1 /* 2 * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README 3 */ 4 5 /* 6 * Written by Anatoly P. Pinchuk pap@namesys.botik.ru 7 * Programm System Institute 8 * Pereslavl-Zalessky Russia 9 */ 10 11 /* 12 * This file contains functions dealing with S+tree 13 * 14 * B_IS_IN_TREE 15 * copy_item_head 16 * comp_short_keys 17 * comp_keys 18 * comp_short_le_keys 19 * le_key2cpu_key 20 * comp_le_keys 21 * bin_search 22 * get_lkey 23 * get_rkey 24 * key_in_buffer 25 * decrement_bcount 26 * reiserfs_check_path 27 * pathrelse_and_restore 28 * pathrelse 29 * search_by_key_reada 30 * search_by_key 31 * search_for_position_by_key 32 * comp_items 33 * prepare_for_direct_item 34 * prepare_for_direntry_item 35 * prepare_for_delete_or_cut 36 * calc_deleted_bytes_number 37 * init_tb_struct 38 * padd_item 39 * reiserfs_delete_item 40 * reiserfs_delete_solid_item 41 * reiserfs_delete_object 42 * maybe_indirect_to_direct 43 * indirect_to_direct_roll_back 44 * reiserfs_cut_from_item 45 * truncate_directory 46 * reiserfs_do_truncate 47 * reiserfs_paste_into_item 48 * reiserfs_insert_item 49 */ 50 51 #include <linux/time.h> 52 #include <linux/string.h> 53 #include <linux/pagemap.h> 54 #include <linux/reiserfs_fs.h> 55 #include <linux/buffer_head.h> 56 #include <linux/quotaops.h> 57 58 /* Does the buffer contain a disk block which is in the tree. */ 59 inline int B_IS_IN_TREE(const struct buffer_head *bh) 60 { 61 62 RFALSE(B_LEVEL(bh) > MAX_HEIGHT, 63 "PAP-1010: block (%b) has too big level (%z)", bh, bh); 64 65 return (B_LEVEL(bh) != FREE_LEVEL); 66 } 67 68 // 69 // to gets item head in le form 70 // 71 inline void copy_item_head(struct item_head *to, 72 const struct item_head *from) 73 { 74 memcpy(to, from, IH_SIZE); 75 } 76 77 /* k1 is pointer to on-disk structure which is stored in little-endian 78 form. k2 is pointer to cpu variable. For key of items of the same 79 object this returns 0. 80 Returns: -1 if key1 < key2 81 0 if key1 == key2 82 1 if key1 > key2 */ 83 inline int comp_short_keys(const struct reiserfs_key *le_key, 84 const struct cpu_key *cpu_key) 85 { 86 __u32 n; 87 n = le32_to_cpu(le_key->k_dir_id); 88 if (n < cpu_key->on_disk_key.k_dir_id) 89 return -1; 90 if (n > cpu_key->on_disk_key.k_dir_id) 91 return 1; 92 n = le32_to_cpu(le_key->k_objectid); 93 if (n < cpu_key->on_disk_key.k_objectid) 94 return -1; 95 if (n > cpu_key->on_disk_key.k_objectid) 96 return 1; 97 return 0; 98 } 99 100 /* k1 is pointer to on-disk structure which is stored in little-endian 101 form. k2 is pointer to cpu variable. 102 Compare keys using all 4 key fields. 103 Returns: -1 if key1 < key2 0 104 if key1 = key2 1 if key1 > key2 */ 105 static inline int comp_keys(const struct reiserfs_key *le_key, 106 const struct cpu_key *cpu_key) 107 { 108 int retval; 109 110 retval = comp_short_keys(le_key, cpu_key); 111 if (retval) 112 return retval; 113 if (le_key_k_offset(le_key_version(le_key), le_key) < 114 cpu_key_k_offset(cpu_key)) 115 return -1; 116 if (le_key_k_offset(le_key_version(le_key), le_key) > 117 cpu_key_k_offset(cpu_key)) 118 return 1; 119 120 if (cpu_key->key_length == 3) 121 return 0; 122 123 /* this part is needed only when tail conversion is in progress */ 124 if (le_key_k_type(le_key_version(le_key), le_key) < 125 cpu_key_k_type(cpu_key)) 126 return -1; 127 128 if (le_key_k_type(le_key_version(le_key), le_key) > 129 cpu_key_k_type(cpu_key)) 130 return 1; 131 132 return 0; 133 } 134 135 inline int comp_short_le_keys(const struct reiserfs_key *key1, 136 const struct reiserfs_key *key2) 137 { 138 __u32 *k1_u32, *k2_u32; 139 int key_length = REISERFS_SHORT_KEY_LEN; 140 141 k1_u32 = (__u32 *) key1; 142 k2_u32 = (__u32 *) key2; 143 for (; key_length--; ++k1_u32, ++k2_u32) { 144 if (le32_to_cpu(*k1_u32) < le32_to_cpu(*k2_u32)) 145 return -1; 146 if (le32_to_cpu(*k1_u32) > le32_to_cpu(*k2_u32)) 147 return 1; 148 } 149 return 0; 150 } 151 152 inline void le_key2cpu_key(struct cpu_key *to, const struct reiserfs_key *from) 153 { 154 int version; 155 to->on_disk_key.k_dir_id = le32_to_cpu(from->k_dir_id); 156 to->on_disk_key.k_objectid = le32_to_cpu(from->k_objectid); 157 158 // find out version of the key 159 version = le_key_version(from); 160 to->version = version; 161 to->on_disk_key.k_offset = le_key_k_offset(version, from); 162 to->on_disk_key.k_type = le_key_k_type(version, from); 163 } 164 165 // this does not say which one is bigger, it only returns 1 if keys 166 // are not equal, 0 otherwise 167 inline int comp_le_keys(const struct reiserfs_key *k1, 168 const struct reiserfs_key *k2) 169 { 170 return memcmp(k1, k2, sizeof(struct reiserfs_key)); 171 } 172 173 /************************************************************************** 174 * Binary search toolkit function * 175 * Search for an item in the array by the item key * 176 * Returns: 1 if found, 0 if not found; * 177 * *pos = number of the searched element if found, else the * 178 * number of the first element that is larger than key. * 179 **************************************************************************/ 180 /* For those not familiar with binary search: lbound is the leftmost item that it 181 could be, rbound the rightmost item that it could be. We examine the item 182 halfway between lbound and rbound, and that tells us either that we can increase 183 lbound, or decrease rbound, or that we have found it, or if lbound <= rbound that 184 there are no possible items, and we have not found it. With each examination we 185 cut the number of possible items it could be by one more than half rounded down, 186 or we find it. */ 187 static inline int bin_search(const void *key, /* Key to search for. */ 188 const void *base, /* First item in the array. */ 189 int num, /* Number of items in the array. */ 190 int width, /* Item size in the array. 191 searched. Lest the reader be 192 confused, note that this is crafted 193 as a general function, and when it 194 is applied specifically to the array 195 of item headers in a node, width 196 is actually the item header size not 197 the item size. */ 198 int *pos /* Number of the searched for element. */ 199 ) 200 { 201 int rbound, lbound, j; 202 203 for (j = ((rbound = num - 1) + (lbound = 0)) / 2; 204 lbound <= rbound; j = (rbound + lbound) / 2) 205 switch (comp_keys 206 ((struct reiserfs_key *)((char *)base + j * width), 207 (struct cpu_key *)key)) { 208 case -1: 209 lbound = j + 1; 210 continue; 211 case 1: 212 rbound = j - 1; 213 continue; 214 case 0: 215 *pos = j; 216 return ITEM_FOUND; /* Key found in the array. */ 217 } 218 219 /* bin_search did not find given key, it returns position of key, 220 that is minimal and greater than the given one. */ 221 *pos = lbound; 222 return ITEM_NOT_FOUND; 223 } 224 225 226 /* Minimal possible key. It is never in the tree. */ 227 const struct reiserfs_key MIN_KEY = { 0, 0, {{0, 0},} }; 228 229 /* Maximal possible key. It is never in the tree. */ 230 static const struct reiserfs_key MAX_KEY = { 231 __constant_cpu_to_le32(0xffffffff), 232 __constant_cpu_to_le32(0xffffffff), 233 {{__constant_cpu_to_le32(0xffffffff), 234 __constant_cpu_to_le32(0xffffffff)},} 235 }; 236 237 /* Get delimiting key of the buffer by looking for it in the buffers in the path, starting from the bottom 238 of the path, and going upwards. We must check the path's validity at each step. If the key is not in 239 the path, there is no delimiting key in the tree (buffer is first or last buffer in tree), and in this 240 case we return a special key, either MIN_KEY or MAX_KEY. */ 241 static inline const struct reiserfs_key *get_lkey(const struct treepath *chk_path, 242 const struct super_block *sb) 243 { 244 int position, path_offset = chk_path->path_length; 245 struct buffer_head *parent; 246 247 RFALSE(path_offset < FIRST_PATH_ELEMENT_OFFSET, 248 "PAP-5010: invalid offset in the path"); 249 250 /* While not higher in path than first element. */ 251 while (path_offset-- > FIRST_PATH_ELEMENT_OFFSET) { 252 253 RFALSE(!buffer_uptodate 254 (PATH_OFFSET_PBUFFER(chk_path, path_offset)), 255 "PAP-5020: parent is not uptodate"); 256 257 /* Parent at the path is not in the tree now. */ 258 if (!B_IS_IN_TREE 259 (parent = 260 PATH_OFFSET_PBUFFER(chk_path, path_offset))) 261 return &MAX_KEY; 262 /* Check whether position in the parent is correct. */ 263 if ((position = 264 PATH_OFFSET_POSITION(chk_path, 265 path_offset)) > 266 B_NR_ITEMS(parent)) 267 return &MAX_KEY; 268 /* Check whether parent at the path really points to the child. */ 269 if (B_N_CHILD_NUM(parent, position) != 270 PATH_OFFSET_PBUFFER(chk_path, 271 path_offset + 1)->b_blocknr) 272 return &MAX_KEY; 273 /* Return delimiting key if position in the parent is not equal to zero. */ 274 if (position) 275 return B_N_PDELIM_KEY(parent, position - 1); 276 } 277 /* Return MIN_KEY if we are in the root of the buffer tree. */ 278 if (PATH_OFFSET_PBUFFER(chk_path, FIRST_PATH_ELEMENT_OFFSET)-> 279 b_blocknr == SB_ROOT_BLOCK(sb)) 280 return &MIN_KEY; 281 return &MAX_KEY; 282 } 283 284 /* Get delimiting key of the buffer at the path and its right neighbor. */ 285 inline const struct reiserfs_key *get_rkey(const struct treepath *chk_path, 286 const struct super_block *sb) 287 { 288 int position, path_offset = chk_path->path_length; 289 struct buffer_head *parent; 290 291 RFALSE(path_offset < FIRST_PATH_ELEMENT_OFFSET, 292 "PAP-5030: invalid offset in the path"); 293 294 while (path_offset-- > FIRST_PATH_ELEMENT_OFFSET) { 295 296 RFALSE(!buffer_uptodate 297 (PATH_OFFSET_PBUFFER(chk_path, path_offset)), 298 "PAP-5040: parent is not uptodate"); 299 300 /* Parent at the path is not in the tree now. */ 301 if (!B_IS_IN_TREE 302 (parent = 303 PATH_OFFSET_PBUFFER(chk_path, path_offset))) 304 return &MIN_KEY; 305 /* Check whether position in the parent is correct. */ 306 if ((position = 307 PATH_OFFSET_POSITION(chk_path, 308 path_offset)) > 309 B_NR_ITEMS(parent)) 310 return &MIN_KEY; 311 /* Check whether parent at the path really points to the child. */ 312 if (B_N_CHILD_NUM(parent, position) != 313 PATH_OFFSET_PBUFFER(chk_path, 314 path_offset + 1)->b_blocknr) 315 return &MIN_KEY; 316 /* Return delimiting key if position in the parent is not the last one. */ 317 if (position != B_NR_ITEMS(parent)) 318 return B_N_PDELIM_KEY(parent, position); 319 } 320 /* Return MAX_KEY if we are in the root of the buffer tree. */ 321 if (PATH_OFFSET_PBUFFER(chk_path, FIRST_PATH_ELEMENT_OFFSET)-> 322 b_blocknr == SB_ROOT_BLOCK(sb)) 323 return &MAX_KEY; 324 return &MIN_KEY; 325 } 326 327 /* Check whether a key is contained in the tree rooted from a buffer at a path. */ 328 /* This works by looking at the left and right delimiting keys for the buffer in the last path_element in 329 the path. These delimiting keys are stored at least one level above that buffer in the tree. If the 330 buffer is the first or last node in the tree order then one of the delimiting keys may be absent, and in 331 this case get_lkey and get_rkey return a special key which is MIN_KEY or MAX_KEY. */ 332 static inline int key_in_buffer(struct treepath *chk_path, /* Path which should be checked. */ 333 const struct cpu_key *key, /* Key which should be checked. */ 334 struct super_block *sb 335 ) 336 { 337 338 RFALSE(!key || chk_path->path_length < FIRST_PATH_ELEMENT_OFFSET 339 || chk_path->path_length > MAX_HEIGHT, 340 "PAP-5050: pointer to the key(%p) is NULL or invalid path length(%d)", 341 key, chk_path->path_length); 342 RFALSE(!PATH_PLAST_BUFFER(chk_path)->b_bdev, 343 "PAP-5060: device must not be NODEV"); 344 345 if (comp_keys(get_lkey(chk_path, sb), key) == 1) 346 /* left delimiting key is bigger, that the key we look for */ 347 return 0; 348 /* if ( comp_keys(key, get_rkey(chk_path, sb)) != -1 ) */ 349 if (comp_keys(get_rkey(chk_path, sb), key) != 1) 350 /* key must be less than right delimitiing key */ 351 return 0; 352 return 1; 353 } 354 355 int reiserfs_check_path(struct treepath *p) 356 { 357 RFALSE(p->path_length != ILLEGAL_PATH_ELEMENT_OFFSET, 358 "path not properly relsed"); 359 return 0; 360 } 361 362 /* Drop the reference to each buffer in a path and restore 363 * dirty bits clean when preparing the buffer for the log. 364 * This version should only be called from fix_nodes() */ 365 void pathrelse_and_restore(struct super_block *sb, 366 struct treepath *search_path) 367 { 368 int path_offset = search_path->path_length; 369 370 RFALSE(path_offset < ILLEGAL_PATH_ELEMENT_OFFSET, 371 "clm-4000: invalid path offset"); 372 373 while (path_offset > ILLEGAL_PATH_ELEMENT_OFFSET) { 374 struct buffer_head *bh; 375 bh = PATH_OFFSET_PBUFFER(search_path, path_offset--); 376 reiserfs_restore_prepared_buffer(sb, bh); 377 brelse(bh); 378 } 379 search_path->path_length = ILLEGAL_PATH_ELEMENT_OFFSET; 380 } 381 382 /* Drop the reference to each buffer in a path */ 383 void pathrelse(struct treepath *search_path) 384 { 385 int path_offset = search_path->path_length; 386 387 RFALSE(path_offset < ILLEGAL_PATH_ELEMENT_OFFSET, 388 "PAP-5090: invalid path offset"); 389 390 while (path_offset > ILLEGAL_PATH_ELEMENT_OFFSET) 391 brelse(PATH_OFFSET_PBUFFER(search_path, path_offset--)); 392 393 search_path->path_length = ILLEGAL_PATH_ELEMENT_OFFSET; 394 } 395 396 static int is_leaf(char *buf, int blocksize, struct buffer_head *bh) 397 { 398 struct block_head *blkh; 399 struct item_head *ih; 400 int used_space; 401 int prev_location; 402 int i; 403 int nr; 404 405 blkh = (struct block_head *)buf; 406 if (blkh_level(blkh) != DISK_LEAF_NODE_LEVEL) { 407 reiserfs_warning(NULL, "reiserfs-5080", 408 "this should be caught earlier"); 409 return 0; 410 } 411 412 nr = blkh_nr_item(blkh); 413 if (nr < 1 || nr > ((blocksize - BLKH_SIZE) / (IH_SIZE + MIN_ITEM_LEN))) { 414 /* item number is too big or too small */ 415 reiserfs_warning(NULL, "reiserfs-5081", 416 "nr_item seems wrong: %z", bh); 417 return 0; 418 } 419 ih = (struct item_head *)(buf + BLKH_SIZE) + nr - 1; 420 used_space = BLKH_SIZE + IH_SIZE * nr + (blocksize - ih_location(ih)); 421 if (used_space != blocksize - blkh_free_space(blkh)) { 422 /* free space does not match to calculated amount of use space */ 423 reiserfs_warning(NULL, "reiserfs-5082", 424 "free space seems wrong: %z", bh); 425 return 0; 426 } 427 // FIXME: it is_leaf will hit performance too much - we may have 428 // return 1 here 429 430 /* check tables of item heads */ 431 ih = (struct item_head *)(buf + BLKH_SIZE); 432 prev_location = blocksize; 433 for (i = 0; i < nr; i++, ih++) { 434 if (le_ih_k_type(ih) == TYPE_ANY) { 435 reiserfs_warning(NULL, "reiserfs-5083", 436 "wrong item type for item %h", 437 ih); 438 return 0; 439 } 440 if (ih_location(ih) >= blocksize 441 || ih_location(ih) < IH_SIZE * nr) { 442 reiserfs_warning(NULL, "reiserfs-5084", 443 "item location seems wrong: %h", 444 ih); 445 return 0; 446 } 447 if (ih_item_len(ih) < 1 448 || ih_item_len(ih) > MAX_ITEM_LEN(blocksize)) { 449 reiserfs_warning(NULL, "reiserfs-5085", 450 "item length seems wrong: %h", 451 ih); 452 return 0; 453 } 454 if (prev_location - ih_location(ih) != ih_item_len(ih)) { 455 reiserfs_warning(NULL, "reiserfs-5086", 456 "item location seems wrong " 457 "(second one): %h", ih); 458 return 0; 459 } 460 prev_location = ih_location(ih); 461 } 462 463 // one may imagine much more checks 464 return 1; 465 } 466 467 /* returns 1 if buf looks like an internal node, 0 otherwise */ 468 static int is_internal(char *buf, int blocksize, struct buffer_head *bh) 469 { 470 struct block_head *blkh; 471 int nr; 472 int used_space; 473 474 blkh = (struct block_head *)buf; 475 nr = blkh_level(blkh); 476 if (nr <= DISK_LEAF_NODE_LEVEL || nr > MAX_HEIGHT) { 477 /* this level is not possible for internal nodes */ 478 reiserfs_warning(NULL, "reiserfs-5087", 479 "this should be caught earlier"); 480 return 0; 481 } 482 483 nr = blkh_nr_item(blkh); 484 if (nr > (blocksize - BLKH_SIZE - DC_SIZE) / (KEY_SIZE + DC_SIZE)) { 485 /* for internal which is not root we might check min number of keys */ 486 reiserfs_warning(NULL, "reiserfs-5088", 487 "number of key seems wrong: %z", bh); 488 return 0; 489 } 490 491 used_space = BLKH_SIZE + KEY_SIZE * nr + DC_SIZE * (nr + 1); 492 if (used_space != blocksize - blkh_free_space(blkh)) { 493 reiserfs_warning(NULL, "reiserfs-5089", 494 "free space seems wrong: %z", bh); 495 return 0; 496 } 497 // one may imagine much more checks 498 return 1; 499 } 500 501 // make sure that bh contains formatted node of reiserfs tree of 502 // 'level'-th level 503 static int is_tree_node(struct buffer_head *bh, int level) 504 { 505 if (B_LEVEL(bh) != level) { 506 reiserfs_warning(NULL, "reiserfs-5090", "node level %d does " 507 "not match to the expected one %d", 508 B_LEVEL(bh), level); 509 return 0; 510 } 511 if (level == DISK_LEAF_NODE_LEVEL) 512 return is_leaf(bh->b_data, bh->b_size, bh); 513 514 return is_internal(bh->b_data, bh->b_size, bh); 515 } 516 517 #define SEARCH_BY_KEY_READA 16 518 519 /* 520 * The function is NOT SCHEDULE-SAFE! 521 * It might unlock the write lock if we needed to wait for a block 522 * to be read. Note that in this case it won't recover the lock to avoid 523 * high contention resulting from too much lock requests, especially 524 * the caller (search_by_key) will perform other schedule-unsafe 525 * operations just after calling this function. 526 * 527 * @return true if we have unlocked 528 */ 529 static bool search_by_key_reada(struct super_block *s, 530 struct buffer_head **bh, 531 b_blocknr_t *b, int num) 532 { 533 int i, j; 534 bool unlocked = false; 535 536 for (i = 0; i < num; i++) { 537 bh[i] = sb_getblk(s, b[i]); 538 } 539 /* 540 * We are going to read some blocks on which we 541 * have a reference. It's safe, though we might be 542 * reading blocks concurrently changed if we release 543 * the lock. But it's still fine because we check later 544 * if the tree changed 545 */ 546 for (j = 0; j < i; j++) { 547 /* 548 * note, this needs attention if we are getting rid of the BKL 549 * you have to make sure the prepared bit isn't set on this buffer 550 */ 551 if (!buffer_uptodate(bh[j])) { 552 if (!unlocked) { 553 reiserfs_write_unlock(s); 554 unlocked = true; 555 } 556 ll_rw_block(READA, 1, bh + j); 557 } 558 brelse(bh[j]); 559 } 560 return unlocked; 561 } 562 563 /************************************************************************** 564 * Algorithm SearchByKey * 565 * look for item in the Disk S+Tree by its key * 566 * Input: sb - super block * 567 * key - pointer to the key to search * 568 * Output: ITEM_FOUND, ITEM_NOT_FOUND or IO_ERROR * 569 * search_path - path from the root to the needed leaf * 570 **************************************************************************/ 571 572 /* This function fills up the path from the root to the leaf as it 573 descends the tree looking for the key. It uses reiserfs_bread to 574 try to find buffers in the cache given their block number. If it 575 does not find them in the cache it reads them from disk. For each 576 node search_by_key finds using reiserfs_bread it then uses 577 bin_search to look through that node. bin_search will find the 578 position of the block_number of the next node if it is looking 579 through an internal node. If it is looking through a leaf node 580 bin_search will find the position of the item which has key either 581 equal to given key, or which is the maximal key less than the given 582 key. search_by_key returns a path that must be checked for the 583 correctness of the top of the path but need not be checked for the 584 correctness of the bottom of the path */ 585 /* The function is NOT SCHEDULE-SAFE! */ 586 int search_by_key(struct super_block *sb, const struct cpu_key *key, /* Key to search. */ 587 struct treepath *search_path,/* This structure was 588 allocated and initialized 589 by the calling 590 function. It is filled up 591 by this function. */ 592 int stop_level /* How far down the tree to search. To 593 stop at leaf level - set to 594 DISK_LEAF_NODE_LEVEL */ 595 ) 596 { 597 b_blocknr_t block_number; 598 int expected_level; 599 struct buffer_head *bh; 600 struct path_element *last_element; 601 int node_level, retval; 602 int right_neighbor_of_leaf_node; 603 int fs_gen; 604 struct buffer_head *reada_bh[SEARCH_BY_KEY_READA]; 605 b_blocknr_t reada_blocks[SEARCH_BY_KEY_READA]; 606 int reada_count = 0; 607 608 #ifdef CONFIG_REISERFS_CHECK 609 int repeat_counter = 0; 610 #endif 611 612 PROC_INFO_INC(sb, search_by_key); 613 614 /* As we add each node to a path we increase its count. This means that 615 we must be careful to release all nodes in a path before we either 616 discard the path struct or re-use the path struct, as we do here. */ 617 618 pathrelse(search_path); 619 620 right_neighbor_of_leaf_node = 0; 621 622 /* With each iteration of this loop we search through the items in the 623 current node, and calculate the next current node(next path element) 624 for the next iteration of this loop.. */ 625 block_number = SB_ROOT_BLOCK(sb); 626 expected_level = -1; 627 while (1) { 628 629 #ifdef CONFIG_REISERFS_CHECK 630 if (!(++repeat_counter % 50000)) 631 reiserfs_warning(sb, "PAP-5100", 632 "%s: there were %d iterations of " 633 "while loop looking for key %K", 634 current->comm, repeat_counter, 635 key); 636 #endif 637 638 /* prep path to have another element added to it. */ 639 last_element = 640 PATH_OFFSET_PELEMENT(search_path, 641 ++search_path->path_length); 642 fs_gen = get_generation(sb); 643 644 /* Read the next tree node, and set the last element in the path to 645 have a pointer to it. */ 646 if ((bh = last_element->pe_buffer = 647 sb_getblk(sb, block_number))) { 648 bool unlocked = false; 649 650 if (!buffer_uptodate(bh) && reada_count > 1) 651 /* may unlock the write lock */ 652 unlocked = search_by_key_reada(sb, reada_bh, 653 reada_blocks, reada_count); 654 /* 655 * If we haven't already unlocked the write lock, 656 * then we need to do that here before reading 657 * the current block 658 */ 659 if (!buffer_uptodate(bh) && !unlocked) { 660 reiserfs_write_unlock(sb); 661 unlocked = true; 662 } 663 ll_rw_block(READ, 1, &bh); 664 wait_on_buffer(bh); 665 666 if (unlocked) 667 reiserfs_write_lock(sb); 668 if (!buffer_uptodate(bh)) 669 goto io_error; 670 } else { 671 io_error: 672 search_path->path_length--; 673 pathrelse(search_path); 674 return IO_ERROR; 675 } 676 reada_count = 0; 677 if (expected_level == -1) 678 expected_level = SB_TREE_HEIGHT(sb); 679 expected_level--; 680 681 /* It is possible that schedule occurred. We must check whether the key 682 to search is still in the tree rooted from the current buffer. If 683 not then repeat search from the root. */ 684 if (fs_changed(fs_gen, sb) && 685 (!B_IS_IN_TREE(bh) || 686 B_LEVEL(bh) != expected_level || 687 !key_in_buffer(search_path, key, sb))) { 688 PROC_INFO_INC(sb, search_by_key_fs_changed); 689 PROC_INFO_INC(sb, search_by_key_restarted); 690 PROC_INFO_INC(sb, 691 sbk_restarted[expected_level - 1]); 692 pathrelse(search_path); 693 694 /* Get the root block number so that we can repeat the search 695 starting from the root. */ 696 block_number = SB_ROOT_BLOCK(sb); 697 expected_level = -1; 698 right_neighbor_of_leaf_node = 0; 699 700 /* repeat search from the root */ 701 continue; 702 } 703 704 /* only check that the key is in the buffer if key is not 705 equal to the MAX_KEY. Latter case is only possible in 706 "finish_unfinished()" processing during mount. */ 707 RFALSE(comp_keys(&MAX_KEY, key) && 708 !key_in_buffer(search_path, key, sb), 709 "PAP-5130: key is not in the buffer"); 710 #ifdef CONFIG_REISERFS_CHECK 711 if (REISERFS_SB(sb)->cur_tb) { 712 print_cur_tb("5140"); 713 reiserfs_panic(sb, "PAP-5140", 714 "schedule occurred in do_balance!"); 715 } 716 #endif 717 718 // make sure, that the node contents look like a node of 719 // certain level 720 if (!is_tree_node(bh, expected_level)) { 721 reiserfs_error(sb, "vs-5150", 722 "invalid format found in block %ld. " 723 "Fsck?", bh->b_blocknr); 724 pathrelse(search_path); 725 return IO_ERROR; 726 } 727 728 /* ok, we have acquired next formatted node in the tree */ 729 node_level = B_LEVEL(bh); 730 731 PROC_INFO_BH_STAT(sb, bh, node_level - 1); 732 733 RFALSE(node_level < stop_level, 734 "vs-5152: tree level (%d) is less than stop level (%d)", 735 node_level, stop_level); 736 737 retval = bin_search(key, B_N_PITEM_HEAD(bh, 0), 738 B_NR_ITEMS(bh), 739 (node_level == 740 DISK_LEAF_NODE_LEVEL) ? IH_SIZE : 741 KEY_SIZE, 742 &(last_element->pe_position)); 743 if (node_level == stop_level) { 744 return retval; 745 } 746 747 /* we are not in the stop level */ 748 if (retval == ITEM_FOUND) 749 /* item has been found, so we choose the pointer which is to the right of the found one */ 750 last_element->pe_position++; 751 752 /* if item was not found we choose the position which is to 753 the left of the found item. This requires no code, 754 bin_search did it already. */ 755 756 /* So we have chosen a position in the current node which is 757 an internal node. Now we calculate child block number by 758 position in the node. */ 759 block_number = 760 B_N_CHILD_NUM(bh, last_element->pe_position); 761 762 /* if we are going to read leaf nodes, try for read ahead as well */ 763 if ((search_path->reada & PATH_READA) && 764 node_level == DISK_LEAF_NODE_LEVEL + 1) { 765 int pos = last_element->pe_position; 766 int limit = B_NR_ITEMS(bh); 767 struct reiserfs_key *le_key; 768 769 if (search_path->reada & PATH_READA_BACK) 770 limit = 0; 771 while (reada_count < SEARCH_BY_KEY_READA) { 772 if (pos == limit) 773 break; 774 reada_blocks[reada_count++] = 775 B_N_CHILD_NUM(bh, pos); 776 if (search_path->reada & PATH_READA_BACK) 777 pos--; 778 else 779 pos++; 780 781 /* 782 * check to make sure we're in the same object 783 */ 784 le_key = B_N_PDELIM_KEY(bh, pos); 785 if (le32_to_cpu(le_key->k_objectid) != 786 key->on_disk_key.k_objectid) { 787 break; 788 } 789 } 790 } 791 } 792 } 793 794 /* Form the path to an item and position in this item which contains 795 file byte defined by key. If there is no such item 796 corresponding to the key, we point the path to the item with 797 maximal key less than key, and *pos_in_item is set to one 798 past the last entry/byte in the item. If searching for entry in a 799 directory item, and it is not found, *pos_in_item is set to one 800 entry more than the entry with maximal key which is less than the 801 sought key. 802 803 Note that if there is no entry in this same node which is one more, 804 then we point to an imaginary entry. for direct items, the 805 position is in units of bytes, for indirect items the position is 806 in units of blocknr entries, for directory items the position is in 807 units of directory entries. */ 808 809 /* The function is NOT SCHEDULE-SAFE! */ 810 int search_for_position_by_key(struct super_block *sb, /* Pointer to the super block. */ 811 const struct cpu_key *p_cpu_key, /* Key to search (cpu variable) */ 812 struct treepath *search_path /* Filled up by this function. */ 813 ) 814 { 815 struct item_head *p_le_ih; /* pointer to on-disk structure */ 816 int blk_size; 817 loff_t item_offset, offset; 818 struct reiserfs_dir_entry de; 819 int retval; 820 821 /* If searching for directory entry. */ 822 if (is_direntry_cpu_key(p_cpu_key)) 823 return search_by_entry_key(sb, p_cpu_key, search_path, 824 &de); 825 826 /* If not searching for directory entry. */ 827 828 /* If item is found. */ 829 retval = search_item(sb, p_cpu_key, search_path); 830 if (retval == IO_ERROR) 831 return retval; 832 if (retval == ITEM_FOUND) { 833 834 RFALSE(!ih_item_len 835 (B_N_PITEM_HEAD 836 (PATH_PLAST_BUFFER(search_path), 837 PATH_LAST_POSITION(search_path))), 838 "PAP-5165: item length equals zero"); 839 840 pos_in_item(search_path) = 0; 841 return POSITION_FOUND; 842 } 843 844 RFALSE(!PATH_LAST_POSITION(search_path), 845 "PAP-5170: position equals zero"); 846 847 /* Item is not found. Set path to the previous item. */ 848 p_le_ih = 849 B_N_PITEM_HEAD(PATH_PLAST_BUFFER(search_path), 850 --PATH_LAST_POSITION(search_path)); 851 blk_size = sb->s_blocksize; 852 853 if (comp_short_keys(&(p_le_ih->ih_key), p_cpu_key)) { 854 return FILE_NOT_FOUND; 855 } 856 // FIXME: quite ugly this far 857 858 item_offset = le_ih_k_offset(p_le_ih); 859 offset = cpu_key_k_offset(p_cpu_key); 860 861 /* Needed byte is contained in the item pointed to by the path. */ 862 if (item_offset <= offset && 863 item_offset + op_bytes_number(p_le_ih, blk_size) > offset) { 864 pos_in_item(search_path) = offset - item_offset; 865 if (is_indirect_le_ih(p_le_ih)) { 866 pos_in_item(search_path) /= blk_size; 867 } 868 return POSITION_FOUND; 869 } 870 871 /* Needed byte is not contained in the item pointed to by the 872 path. Set pos_in_item out of the item. */ 873 if (is_indirect_le_ih(p_le_ih)) 874 pos_in_item(search_path) = 875 ih_item_len(p_le_ih) / UNFM_P_SIZE; 876 else 877 pos_in_item(search_path) = ih_item_len(p_le_ih); 878 879 return POSITION_NOT_FOUND; 880 } 881 882 /* Compare given item and item pointed to by the path. */ 883 int comp_items(const struct item_head *stored_ih, const struct treepath *path) 884 { 885 struct buffer_head *bh = PATH_PLAST_BUFFER(path); 886 struct item_head *ih; 887 888 /* Last buffer at the path is not in the tree. */ 889 if (!B_IS_IN_TREE(bh)) 890 return 1; 891 892 /* Last path position is invalid. */ 893 if (PATH_LAST_POSITION(path) >= B_NR_ITEMS(bh)) 894 return 1; 895 896 /* we need only to know, whether it is the same item */ 897 ih = get_ih(path); 898 return memcmp(stored_ih, ih, IH_SIZE); 899 } 900 901 /* unformatted nodes are not logged anymore, ever. This is safe 902 ** now 903 */ 904 #define held_by_others(bh) (atomic_read(&(bh)->b_count) > 1) 905 906 // block can not be forgotten as it is in I/O or held by someone 907 #define block_in_use(bh) (buffer_locked(bh) || (held_by_others(bh))) 908 909 // prepare for delete or cut of direct item 910 static inline int prepare_for_direct_item(struct treepath *path, 911 struct item_head *le_ih, 912 struct inode *inode, 913 loff_t new_file_length, int *cut_size) 914 { 915 loff_t round_len; 916 917 if (new_file_length == max_reiserfs_offset(inode)) { 918 /* item has to be deleted */ 919 *cut_size = -(IH_SIZE + ih_item_len(le_ih)); 920 return M_DELETE; 921 } 922 // new file gets truncated 923 if (get_inode_item_key_version(inode) == KEY_FORMAT_3_6) { 924 // 925 round_len = ROUND_UP(new_file_length); 926 /* this was new_file_length < le_ih ... */ 927 if (round_len < le_ih_k_offset(le_ih)) { 928 *cut_size = -(IH_SIZE + ih_item_len(le_ih)); 929 return M_DELETE; /* Delete this item. */ 930 } 931 /* Calculate first position and size for cutting from item. */ 932 pos_in_item(path) = round_len - (le_ih_k_offset(le_ih) - 1); 933 *cut_size = -(ih_item_len(le_ih) - pos_in_item(path)); 934 935 return M_CUT; /* Cut from this item. */ 936 } 937 938 // old file: items may have any length 939 940 if (new_file_length < le_ih_k_offset(le_ih)) { 941 *cut_size = -(IH_SIZE + ih_item_len(le_ih)); 942 return M_DELETE; /* Delete this item. */ 943 } 944 /* Calculate first position and size for cutting from item. */ 945 *cut_size = -(ih_item_len(le_ih) - 946 (pos_in_item(path) = 947 new_file_length + 1 - le_ih_k_offset(le_ih))); 948 return M_CUT; /* Cut from this item. */ 949 } 950 951 static inline int prepare_for_direntry_item(struct treepath *path, 952 struct item_head *le_ih, 953 struct inode *inode, 954 loff_t new_file_length, 955 int *cut_size) 956 { 957 if (le_ih_k_offset(le_ih) == DOT_OFFSET && 958 new_file_length == max_reiserfs_offset(inode)) { 959 RFALSE(ih_entry_count(le_ih) != 2, 960 "PAP-5220: incorrect empty directory item (%h)", le_ih); 961 *cut_size = -(IH_SIZE + ih_item_len(le_ih)); 962 return M_DELETE; /* Delete the directory item containing "." and ".." entry. */ 963 } 964 965 if (ih_entry_count(le_ih) == 1) { 966 /* Delete the directory item such as there is one record only 967 in this item */ 968 *cut_size = -(IH_SIZE + ih_item_len(le_ih)); 969 return M_DELETE; 970 } 971 972 /* Cut one record from the directory item. */ 973 *cut_size = 974 -(DEH_SIZE + 975 entry_length(get_last_bh(path), le_ih, pos_in_item(path))); 976 return M_CUT; 977 } 978 979 #define JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD (2 * JOURNAL_PER_BALANCE_CNT + 1) 980 981 /* If the path points to a directory or direct item, calculate mode and the size cut, for balance. 982 If the path points to an indirect item, remove some number of its unformatted nodes. 983 In case of file truncate calculate whether this item must be deleted/truncated or last 984 unformatted node of this item will be converted to a direct item. 985 This function returns a determination of what balance mode the calling function should employ. */ 986 static char prepare_for_delete_or_cut(struct reiserfs_transaction_handle *th, struct inode *inode, struct treepath *path, const struct cpu_key *item_key, int *removed, /* Number of unformatted nodes which were removed 987 from end of the file. */ 988 int *cut_size, unsigned long long new_file_length /* MAX_KEY_OFFSET in case of delete. */ 989 ) 990 { 991 struct super_block *sb = inode->i_sb; 992 struct item_head *p_le_ih = PATH_PITEM_HEAD(path); 993 struct buffer_head *bh = PATH_PLAST_BUFFER(path); 994 995 BUG_ON(!th->t_trans_id); 996 997 /* Stat_data item. */ 998 if (is_statdata_le_ih(p_le_ih)) { 999 1000 RFALSE(new_file_length != max_reiserfs_offset(inode), 1001 "PAP-5210: mode must be M_DELETE"); 1002 1003 *cut_size = -(IH_SIZE + ih_item_len(p_le_ih)); 1004 return M_DELETE; 1005 } 1006 1007 /* Directory item. */ 1008 if (is_direntry_le_ih(p_le_ih)) 1009 return prepare_for_direntry_item(path, p_le_ih, inode, 1010 new_file_length, 1011 cut_size); 1012 1013 /* Direct item. */ 1014 if (is_direct_le_ih(p_le_ih)) 1015 return prepare_for_direct_item(path, p_le_ih, inode, 1016 new_file_length, cut_size); 1017 1018 /* Case of an indirect item. */ 1019 { 1020 int blk_size = sb->s_blocksize; 1021 struct item_head s_ih; 1022 int need_re_search; 1023 int delete = 0; 1024 int result = M_CUT; 1025 int pos = 0; 1026 1027 if ( new_file_length == max_reiserfs_offset (inode) ) { 1028 /* prepare_for_delete_or_cut() is called by 1029 * reiserfs_delete_item() */ 1030 new_file_length = 0; 1031 delete = 1; 1032 } 1033 1034 do { 1035 need_re_search = 0; 1036 *cut_size = 0; 1037 bh = PATH_PLAST_BUFFER(path); 1038 copy_item_head(&s_ih, PATH_PITEM_HEAD(path)); 1039 pos = I_UNFM_NUM(&s_ih); 1040 1041 while (le_ih_k_offset (&s_ih) + (pos - 1) * blk_size > new_file_length) { 1042 __le32 *unfm; 1043 __u32 block; 1044 1045 /* Each unformatted block deletion may involve one additional 1046 * bitmap block into the transaction, thereby the initial 1047 * journal space reservation might not be enough. */ 1048 if (!delete && (*cut_size) != 0 && 1049 reiserfs_transaction_free_space(th) < JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD) 1050 break; 1051 1052 unfm = (__le32 *)B_I_PITEM(bh, &s_ih) + pos - 1; 1053 block = get_block_num(unfm, 0); 1054 1055 if (block != 0) { 1056 reiserfs_prepare_for_journal(sb, bh, 1); 1057 put_block_num(unfm, 0, 0); 1058 journal_mark_dirty(th, sb, bh); 1059 reiserfs_free_block(th, inode, block, 1); 1060 } 1061 1062 reiserfs_write_unlock(sb); 1063 cond_resched(); 1064 reiserfs_write_lock(sb); 1065 1066 if (item_moved (&s_ih, path)) { 1067 need_re_search = 1; 1068 break; 1069 } 1070 1071 pos --; 1072 (*removed)++; 1073 (*cut_size) -= UNFM_P_SIZE; 1074 1075 if (pos == 0) { 1076 (*cut_size) -= IH_SIZE; 1077 result = M_DELETE; 1078 break; 1079 } 1080 } 1081 /* a trick. If the buffer has been logged, this will do nothing. If 1082 ** we've broken the loop without logging it, it will restore the 1083 ** buffer */ 1084 reiserfs_restore_prepared_buffer(sb, bh); 1085 } while (need_re_search && 1086 search_for_position_by_key(sb, item_key, path) == POSITION_FOUND); 1087 pos_in_item(path) = pos * UNFM_P_SIZE; 1088 1089 if (*cut_size == 0) { 1090 /* Nothing were cut. maybe convert last unformatted node to the 1091 * direct item? */ 1092 result = M_CONVERT; 1093 } 1094 return result; 1095 } 1096 } 1097 1098 /* Calculate number of bytes which will be deleted or cut during balance */ 1099 static int calc_deleted_bytes_number(struct tree_balance *tb, char mode) 1100 { 1101 int del_size; 1102 struct item_head *p_le_ih = PATH_PITEM_HEAD(tb->tb_path); 1103 1104 if (is_statdata_le_ih(p_le_ih)) 1105 return 0; 1106 1107 del_size = 1108 (mode == 1109 M_DELETE) ? ih_item_len(p_le_ih) : -tb->insert_size[0]; 1110 if (is_direntry_le_ih(p_le_ih)) { 1111 /* return EMPTY_DIR_SIZE; We delete emty directoris only. 1112 * we can't use EMPTY_DIR_SIZE, as old format dirs have a different 1113 * empty size. ick. FIXME, is this right? */ 1114 return del_size; 1115 } 1116 1117 if (is_indirect_le_ih(p_le_ih)) 1118 del_size = (del_size / UNFM_P_SIZE) * 1119 (PATH_PLAST_BUFFER(tb->tb_path)->b_size); 1120 return del_size; 1121 } 1122 1123 static void init_tb_struct(struct reiserfs_transaction_handle *th, 1124 struct tree_balance *tb, 1125 struct super_block *sb, 1126 struct treepath *path, int size) 1127 { 1128 1129 BUG_ON(!th->t_trans_id); 1130 1131 memset(tb, '\0', sizeof(struct tree_balance)); 1132 tb->transaction_handle = th; 1133 tb->tb_sb = sb; 1134 tb->tb_path = path; 1135 PATH_OFFSET_PBUFFER(path, ILLEGAL_PATH_ELEMENT_OFFSET) = NULL; 1136 PATH_OFFSET_POSITION(path, ILLEGAL_PATH_ELEMENT_OFFSET) = 0; 1137 tb->insert_size[0] = size; 1138 } 1139 1140 void padd_item(char *item, int total_length, int length) 1141 { 1142 int i; 1143 1144 for (i = total_length; i > length;) 1145 item[--i] = 0; 1146 } 1147 1148 #ifdef REISERQUOTA_DEBUG 1149 char key2type(struct reiserfs_key *ih) 1150 { 1151 if (is_direntry_le_key(2, ih)) 1152 return 'd'; 1153 if (is_direct_le_key(2, ih)) 1154 return 'D'; 1155 if (is_indirect_le_key(2, ih)) 1156 return 'i'; 1157 if (is_statdata_le_key(2, ih)) 1158 return 's'; 1159 return 'u'; 1160 } 1161 1162 char head2type(struct item_head *ih) 1163 { 1164 if (is_direntry_le_ih(ih)) 1165 return 'd'; 1166 if (is_direct_le_ih(ih)) 1167 return 'D'; 1168 if (is_indirect_le_ih(ih)) 1169 return 'i'; 1170 if (is_statdata_le_ih(ih)) 1171 return 's'; 1172 return 'u'; 1173 } 1174 #endif 1175 1176 /* Delete object item. 1177 * th - active transaction handle 1178 * path - path to the deleted item 1179 * item_key - key to search for the deleted item 1180 * indode - used for updating i_blocks and quotas 1181 * un_bh - NULL or unformatted node pointer 1182 */ 1183 int reiserfs_delete_item(struct reiserfs_transaction_handle *th, 1184 struct treepath *path, const struct cpu_key *item_key, 1185 struct inode *inode, struct buffer_head *un_bh) 1186 { 1187 struct super_block *sb = inode->i_sb; 1188 struct tree_balance s_del_balance; 1189 struct item_head s_ih; 1190 struct item_head *q_ih; 1191 int quota_cut_bytes; 1192 int ret_value, del_size, removed; 1193 1194 #ifdef CONFIG_REISERFS_CHECK 1195 char mode; 1196 int iter = 0; 1197 #endif 1198 1199 BUG_ON(!th->t_trans_id); 1200 1201 init_tb_struct(th, &s_del_balance, sb, path, 1202 0 /*size is unknown */ ); 1203 1204 while (1) { 1205 removed = 0; 1206 1207 #ifdef CONFIG_REISERFS_CHECK 1208 iter++; 1209 mode = 1210 #endif 1211 prepare_for_delete_or_cut(th, inode, path, 1212 item_key, &removed, 1213 &del_size, 1214 max_reiserfs_offset(inode)); 1215 1216 RFALSE(mode != M_DELETE, "PAP-5320: mode must be M_DELETE"); 1217 1218 copy_item_head(&s_ih, PATH_PITEM_HEAD(path)); 1219 s_del_balance.insert_size[0] = del_size; 1220 1221 ret_value = fix_nodes(M_DELETE, &s_del_balance, NULL, NULL); 1222 if (ret_value != REPEAT_SEARCH) 1223 break; 1224 1225 PROC_INFO_INC(sb, delete_item_restarted); 1226 1227 // file system changed, repeat search 1228 ret_value = 1229 search_for_position_by_key(sb, item_key, path); 1230 if (ret_value == IO_ERROR) 1231 break; 1232 if (ret_value == FILE_NOT_FOUND) { 1233 reiserfs_warning(sb, "vs-5340", 1234 "no items of the file %K found", 1235 item_key); 1236 break; 1237 } 1238 } /* while (1) */ 1239 1240 if (ret_value != CARRY_ON) { 1241 unfix_nodes(&s_del_balance); 1242 return 0; 1243 } 1244 // reiserfs_delete_item returns item length when success 1245 ret_value = calc_deleted_bytes_number(&s_del_balance, M_DELETE); 1246 q_ih = get_ih(path); 1247 quota_cut_bytes = ih_item_len(q_ih); 1248 1249 /* hack so the quota code doesn't have to guess if the file 1250 ** has a tail. On tail insert, we allocate quota for 1 unformatted node. 1251 ** We test the offset because the tail might have been 1252 ** split into multiple items, and we only want to decrement for 1253 ** the unfm node once 1254 */ 1255 if (!S_ISLNK(inode->i_mode) && is_direct_le_ih(q_ih)) { 1256 if ((le_ih_k_offset(q_ih) & (sb->s_blocksize - 1)) == 1) { 1257 quota_cut_bytes = sb->s_blocksize + UNFM_P_SIZE; 1258 } else { 1259 quota_cut_bytes = 0; 1260 } 1261 } 1262 1263 if (un_bh) { 1264 int off; 1265 char *data; 1266 1267 /* We are in direct2indirect conversion, so move tail contents 1268 to the unformatted node */ 1269 /* note, we do the copy before preparing the buffer because we 1270 ** don't care about the contents of the unformatted node yet. 1271 ** the only thing we really care about is the direct item's data 1272 ** is in the unformatted node. 1273 ** 1274 ** Otherwise, we would have to call reiserfs_prepare_for_journal on 1275 ** the unformatted node, which might schedule, meaning we'd have to 1276 ** loop all the way back up to the start of the while loop. 1277 ** 1278 ** The unformatted node must be dirtied later on. We can't be 1279 ** sure here if the entire tail has been deleted yet. 1280 ** 1281 ** un_bh is from the page cache (all unformatted nodes are 1282 ** from the page cache) and might be a highmem page. So, we 1283 ** can't use un_bh->b_data. 1284 ** -clm 1285 */ 1286 1287 data = kmap_atomic(un_bh->b_page, KM_USER0); 1288 off = ((le_ih_k_offset(&s_ih) - 1) & (PAGE_CACHE_SIZE - 1)); 1289 memcpy(data + off, 1290 B_I_PITEM(PATH_PLAST_BUFFER(path), &s_ih), 1291 ret_value); 1292 kunmap_atomic(data, KM_USER0); 1293 } 1294 /* Perform balancing after all resources have been collected at once. */ 1295 do_balance(&s_del_balance, NULL, NULL, M_DELETE); 1296 1297 #ifdef REISERQUOTA_DEBUG 1298 reiserfs_debug(sb, REISERFS_DEBUG_CODE, 1299 "reiserquota delete_item(): freeing %u, id=%u type=%c", 1300 quota_cut_bytes, inode->i_uid, head2type(&s_ih)); 1301 #endif 1302 vfs_dq_free_space_nodirty(inode, quota_cut_bytes); 1303 1304 /* Return deleted body length */ 1305 return ret_value; 1306 } 1307 1308 /* Summary Of Mechanisms For Handling Collisions Between Processes: 1309 1310 deletion of the body of the object is performed by iput(), with the 1311 result that if multiple processes are operating on a file, the 1312 deletion of the body of the file is deferred until the last process 1313 that has an open inode performs its iput(). 1314 1315 writes and truncates are protected from collisions by use of 1316 semaphores. 1317 1318 creates, linking, and mknod are protected from collisions with other 1319 processes by making the reiserfs_add_entry() the last step in the 1320 creation, and then rolling back all changes if there was a collision. 1321 - Hans 1322 */ 1323 1324 /* this deletes item which never gets split */ 1325 void reiserfs_delete_solid_item(struct reiserfs_transaction_handle *th, 1326 struct inode *inode, struct reiserfs_key *key) 1327 { 1328 struct tree_balance tb; 1329 INITIALIZE_PATH(path); 1330 int item_len = 0; 1331 int tb_init = 0; 1332 struct cpu_key cpu_key; 1333 int retval; 1334 int quota_cut_bytes = 0; 1335 1336 BUG_ON(!th->t_trans_id); 1337 1338 le_key2cpu_key(&cpu_key, key); 1339 1340 while (1) { 1341 retval = search_item(th->t_super, &cpu_key, &path); 1342 if (retval == IO_ERROR) { 1343 reiserfs_error(th->t_super, "vs-5350", 1344 "i/o failure occurred trying " 1345 "to delete %K", &cpu_key); 1346 break; 1347 } 1348 if (retval != ITEM_FOUND) { 1349 pathrelse(&path); 1350 // No need for a warning, if there is just no free space to insert '..' item into the newly-created subdir 1351 if (! 1352 ((unsigned long long) 1353 GET_HASH_VALUE(le_key_k_offset 1354 (le_key_version(key), key)) == 0 1355 && (unsigned long long) 1356 GET_GENERATION_NUMBER(le_key_k_offset 1357 (le_key_version(key), 1358 key)) == 1)) 1359 reiserfs_warning(th->t_super, "vs-5355", 1360 "%k not found", key); 1361 break; 1362 } 1363 if (!tb_init) { 1364 tb_init = 1; 1365 item_len = ih_item_len(PATH_PITEM_HEAD(&path)); 1366 init_tb_struct(th, &tb, th->t_super, &path, 1367 -(IH_SIZE + item_len)); 1368 } 1369 quota_cut_bytes = ih_item_len(PATH_PITEM_HEAD(&path)); 1370 1371 retval = fix_nodes(M_DELETE, &tb, NULL, NULL); 1372 if (retval == REPEAT_SEARCH) { 1373 PROC_INFO_INC(th->t_super, delete_solid_item_restarted); 1374 continue; 1375 } 1376 1377 if (retval == CARRY_ON) { 1378 do_balance(&tb, NULL, NULL, M_DELETE); 1379 if (inode) { /* Should we count quota for item? (we don't count quotas for save-links) */ 1380 #ifdef REISERQUOTA_DEBUG 1381 reiserfs_debug(th->t_super, REISERFS_DEBUG_CODE, 1382 "reiserquota delete_solid_item(): freeing %u id=%u type=%c", 1383 quota_cut_bytes, inode->i_uid, 1384 key2type(key)); 1385 #endif 1386 vfs_dq_free_space_nodirty(inode, 1387 quota_cut_bytes); 1388 } 1389 break; 1390 } 1391 // IO_ERROR, NO_DISK_SPACE, etc 1392 reiserfs_warning(th->t_super, "vs-5360", 1393 "could not delete %K due to fix_nodes failure", 1394 &cpu_key); 1395 unfix_nodes(&tb); 1396 break; 1397 } 1398 1399 reiserfs_check_path(&path); 1400 } 1401 1402 int reiserfs_delete_object(struct reiserfs_transaction_handle *th, 1403 struct inode *inode) 1404 { 1405 int err; 1406 inode->i_size = 0; 1407 BUG_ON(!th->t_trans_id); 1408 1409 /* for directory this deletes item containing "." and ".." */ 1410 err = 1411 reiserfs_do_truncate(th, inode, NULL, 0 /*no timestamp updates */ ); 1412 if (err) 1413 return err; 1414 1415 #if defined( USE_INODE_GENERATION_COUNTER ) 1416 if (!old_format_only(th->t_super)) { 1417 __le32 *inode_generation; 1418 1419 inode_generation = 1420 &REISERFS_SB(th->t_super)->s_rs->s_inode_generation; 1421 le32_add_cpu(inode_generation, 1); 1422 } 1423 /* USE_INODE_GENERATION_COUNTER */ 1424 #endif 1425 reiserfs_delete_solid_item(th, inode, INODE_PKEY(inode)); 1426 1427 return err; 1428 } 1429 1430 static void unmap_buffers(struct page *page, loff_t pos) 1431 { 1432 struct buffer_head *bh; 1433 struct buffer_head *head; 1434 struct buffer_head *next; 1435 unsigned long tail_index; 1436 unsigned long cur_index; 1437 1438 if (page) { 1439 if (page_has_buffers(page)) { 1440 tail_index = pos & (PAGE_CACHE_SIZE - 1); 1441 cur_index = 0; 1442 head = page_buffers(page); 1443 bh = head; 1444 do { 1445 next = bh->b_this_page; 1446 1447 /* we want to unmap the buffers that contain the tail, and 1448 ** all the buffers after it (since the tail must be at the 1449 ** end of the file). We don't want to unmap file data 1450 ** before the tail, since it might be dirty and waiting to 1451 ** reach disk 1452 */ 1453 cur_index += bh->b_size; 1454 if (cur_index > tail_index) { 1455 reiserfs_unmap_buffer(bh); 1456 } 1457 bh = next; 1458 } while (bh != head); 1459 } 1460 } 1461 } 1462 1463 static int maybe_indirect_to_direct(struct reiserfs_transaction_handle *th, 1464 struct inode *inode, 1465 struct page *page, 1466 struct treepath *path, 1467 const struct cpu_key *item_key, 1468 loff_t new_file_size, char *mode) 1469 { 1470 struct super_block *sb = inode->i_sb; 1471 int block_size = sb->s_blocksize; 1472 int cut_bytes; 1473 BUG_ON(!th->t_trans_id); 1474 BUG_ON(new_file_size != inode->i_size); 1475 1476 /* the page being sent in could be NULL if there was an i/o error 1477 ** reading in the last block. The user will hit problems trying to 1478 ** read the file, but for now we just skip the indirect2direct 1479 */ 1480 if (atomic_read(&inode->i_count) > 1 || 1481 !tail_has_to_be_packed(inode) || 1482 !page || (REISERFS_I(inode)->i_flags & i_nopack_mask)) { 1483 /* leave tail in an unformatted node */ 1484 *mode = M_SKIP_BALANCING; 1485 cut_bytes = 1486 block_size - (new_file_size & (block_size - 1)); 1487 pathrelse(path); 1488 return cut_bytes; 1489 } 1490 /* Perform the conversion to a direct_item. */ 1491 /* return indirect_to_direct(inode, path, item_key, 1492 new_file_size, mode); */ 1493 return indirect2direct(th, inode, page, path, item_key, 1494 new_file_size, mode); 1495 } 1496 1497 /* we did indirect_to_direct conversion. And we have inserted direct 1498 item successesfully, but there were no disk space to cut unfm 1499 pointer being converted. Therefore we have to delete inserted 1500 direct item(s) */ 1501 static void indirect_to_direct_roll_back(struct reiserfs_transaction_handle *th, 1502 struct inode *inode, struct treepath *path) 1503 { 1504 struct cpu_key tail_key; 1505 int tail_len; 1506 int removed; 1507 BUG_ON(!th->t_trans_id); 1508 1509 make_cpu_key(&tail_key, inode, inode->i_size + 1, TYPE_DIRECT, 4); // !!!! 1510 tail_key.key_length = 4; 1511 1512 tail_len = 1513 (cpu_key_k_offset(&tail_key) & (inode->i_sb->s_blocksize - 1)) - 1; 1514 while (tail_len) { 1515 /* look for the last byte of the tail */ 1516 if (search_for_position_by_key(inode->i_sb, &tail_key, path) == 1517 POSITION_NOT_FOUND) 1518 reiserfs_panic(inode->i_sb, "vs-5615", 1519 "found invalid item"); 1520 RFALSE(path->pos_in_item != 1521 ih_item_len(PATH_PITEM_HEAD(path)) - 1, 1522 "vs-5616: appended bytes found"); 1523 PATH_LAST_POSITION(path)--; 1524 1525 removed = 1526 reiserfs_delete_item(th, path, &tail_key, inode, 1527 NULL /*unbh not needed */ ); 1528 RFALSE(removed <= 0 1529 || removed > tail_len, 1530 "vs-5617: there was tail %d bytes, removed item length %d bytes", 1531 tail_len, removed); 1532 tail_len -= removed; 1533 set_cpu_key_k_offset(&tail_key, 1534 cpu_key_k_offset(&tail_key) - removed); 1535 } 1536 reiserfs_warning(inode->i_sb, "reiserfs-5091", "indirect_to_direct " 1537 "conversion has been rolled back due to " 1538 "lack of disk space"); 1539 //mark_file_without_tail (inode); 1540 mark_inode_dirty(inode); 1541 } 1542 1543 /* (Truncate or cut entry) or delete object item. Returns < 0 on failure */ 1544 int reiserfs_cut_from_item(struct reiserfs_transaction_handle *th, 1545 struct treepath *path, 1546 struct cpu_key *item_key, 1547 struct inode *inode, 1548 struct page *page, loff_t new_file_size) 1549 { 1550 struct super_block *sb = inode->i_sb; 1551 /* Every function which is going to call do_balance must first 1552 create a tree_balance structure. Then it must fill up this 1553 structure by using the init_tb_struct and fix_nodes functions. 1554 After that we can make tree balancing. */ 1555 struct tree_balance s_cut_balance; 1556 struct item_head *p_le_ih; 1557 int cut_size = 0, /* Amount to be cut. */ 1558 ret_value = CARRY_ON, removed = 0, /* Number of the removed unformatted nodes. */ 1559 is_inode_locked = 0; 1560 char mode; /* Mode of the balance. */ 1561 int retval2 = -1; 1562 int quota_cut_bytes; 1563 loff_t tail_pos = 0; 1564 1565 BUG_ON(!th->t_trans_id); 1566 1567 init_tb_struct(th, &s_cut_balance, inode->i_sb, path, 1568 cut_size); 1569 1570 /* Repeat this loop until we either cut the item without needing 1571 to balance, or we fix_nodes without schedule occurring */ 1572 while (1) { 1573 /* Determine the balance mode, position of the first byte to 1574 be cut, and size to be cut. In case of the indirect item 1575 free unformatted nodes which are pointed to by the cut 1576 pointers. */ 1577 1578 mode = 1579 prepare_for_delete_or_cut(th, inode, path, 1580 item_key, &removed, 1581 &cut_size, new_file_size); 1582 if (mode == M_CONVERT) { 1583 /* convert last unformatted node to direct item or leave 1584 tail in the unformatted node */ 1585 RFALSE(ret_value != CARRY_ON, 1586 "PAP-5570: can not convert twice"); 1587 1588 ret_value = 1589 maybe_indirect_to_direct(th, inode, page, 1590 path, item_key, 1591 new_file_size, &mode); 1592 if (mode == M_SKIP_BALANCING) 1593 /* tail has been left in the unformatted node */ 1594 return ret_value; 1595 1596 is_inode_locked = 1; 1597 1598 /* removing of last unformatted node will change value we 1599 have to return to truncate. Save it */ 1600 retval2 = ret_value; 1601 /*retval2 = sb->s_blocksize - (new_file_size & (sb->s_blocksize - 1)); */ 1602 1603 /* So, we have performed the first part of the conversion: 1604 inserting the new direct item. Now we are removing the 1605 last unformatted node pointer. Set key to search for 1606 it. */ 1607 set_cpu_key_k_type(item_key, TYPE_INDIRECT); 1608 item_key->key_length = 4; 1609 new_file_size -= 1610 (new_file_size & (sb->s_blocksize - 1)); 1611 tail_pos = new_file_size; 1612 set_cpu_key_k_offset(item_key, new_file_size + 1); 1613 if (search_for_position_by_key 1614 (sb, item_key, 1615 path) == POSITION_NOT_FOUND) { 1616 print_block(PATH_PLAST_BUFFER(path), 3, 1617 PATH_LAST_POSITION(path) - 1, 1618 PATH_LAST_POSITION(path) + 1); 1619 reiserfs_panic(sb, "PAP-5580", "item to " 1620 "convert does not exist (%K)", 1621 item_key); 1622 } 1623 continue; 1624 } 1625 if (cut_size == 0) { 1626 pathrelse(path); 1627 return 0; 1628 } 1629 1630 s_cut_balance.insert_size[0] = cut_size; 1631 1632 ret_value = fix_nodes(mode, &s_cut_balance, NULL, NULL); 1633 if (ret_value != REPEAT_SEARCH) 1634 break; 1635 1636 PROC_INFO_INC(sb, cut_from_item_restarted); 1637 1638 ret_value = 1639 search_for_position_by_key(sb, item_key, path); 1640 if (ret_value == POSITION_FOUND) 1641 continue; 1642 1643 reiserfs_warning(sb, "PAP-5610", "item %K not found", 1644 item_key); 1645 unfix_nodes(&s_cut_balance); 1646 return (ret_value == IO_ERROR) ? -EIO : -ENOENT; 1647 } /* while */ 1648 1649 // check fix_nodes results (IO_ERROR or NO_DISK_SPACE) 1650 if (ret_value != CARRY_ON) { 1651 if (is_inode_locked) { 1652 // FIXME: this seems to be not needed: we are always able 1653 // to cut item 1654 indirect_to_direct_roll_back(th, inode, path); 1655 } 1656 if (ret_value == NO_DISK_SPACE) 1657 reiserfs_warning(sb, "reiserfs-5092", 1658 "NO_DISK_SPACE"); 1659 unfix_nodes(&s_cut_balance); 1660 return -EIO; 1661 } 1662 1663 /* go ahead and perform balancing */ 1664 1665 RFALSE(mode == M_PASTE || mode == M_INSERT, "invalid mode"); 1666 1667 /* Calculate number of bytes that need to be cut from the item. */ 1668 quota_cut_bytes = 1669 (mode == 1670 M_DELETE) ? ih_item_len(get_ih(path)) : -s_cut_balance. 1671 insert_size[0]; 1672 if (retval2 == -1) 1673 ret_value = calc_deleted_bytes_number(&s_cut_balance, mode); 1674 else 1675 ret_value = retval2; 1676 1677 /* For direct items, we only change the quota when deleting the last 1678 ** item. 1679 */ 1680 p_le_ih = PATH_PITEM_HEAD(s_cut_balance.tb_path); 1681 if (!S_ISLNK(inode->i_mode) && is_direct_le_ih(p_le_ih)) { 1682 if (mode == M_DELETE && 1683 (le_ih_k_offset(p_le_ih) & (sb->s_blocksize - 1)) == 1684 1) { 1685 // FIXME: this is to keep 3.5 happy 1686 REISERFS_I(inode)->i_first_direct_byte = U32_MAX; 1687 quota_cut_bytes = sb->s_blocksize + UNFM_P_SIZE; 1688 } else { 1689 quota_cut_bytes = 0; 1690 } 1691 } 1692 #ifdef CONFIG_REISERFS_CHECK 1693 if (is_inode_locked) { 1694 struct item_head *le_ih = 1695 PATH_PITEM_HEAD(s_cut_balance.tb_path); 1696 /* we are going to complete indirect2direct conversion. Make 1697 sure, that we exactly remove last unformatted node pointer 1698 of the item */ 1699 if (!is_indirect_le_ih(le_ih)) 1700 reiserfs_panic(sb, "vs-5652", 1701 "item must be indirect %h", le_ih); 1702 1703 if (mode == M_DELETE && ih_item_len(le_ih) != UNFM_P_SIZE) 1704 reiserfs_panic(sb, "vs-5653", "completing " 1705 "indirect2direct conversion indirect " 1706 "item %h being deleted must be of " 1707 "4 byte long", le_ih); 1708 1709 if (mode == M_CUT 1710 && s_cut_balance.insert_size[0] != -UNFM_P_SIZE) { 1711 reiserfs_panic(sb, "vs-5654", "can not complete " 1712 "indirect2direct conversion of %h " 1713 "(CUT, insert_size==%d)", 1714 le_ih, s_cut_balance.insert_size[0]); 1715 } 1716 /* it would be useful to make sure, that right neighboring 1717 item is direct item of this file */ 1718 } 1719 #endif 1720 1721 do_balance(&s_cut_balance, NULL, NULL, mode); 1722 if (is_inode_locked) { 1723 /* we've done an indirect->direct conversion. when the data block 1724 ** was freed, it was removed from the list of blocks that must 1725 ** be flushed before the transaction commits, make sure to 1726 ** unmap and invalidate it 1727 */ 1728 unmap_buffers(page, tail_pos); 1729 REISERFS_I(inode)->i_flags &= ~i_pack_on_close_mask; 1730 } 1731 #ifdef REISERQUOTA_DEBUG 1732 reiserfs_debug(inode->i_sb, REISERFS_DEBUG_CODE, 1733 "reiserquota cut_from_item(): freeing %u id=%u type=%c", 1734 quota_cut_bytes, inode->i_uid, '?'); 1735 #endif 1736 vfs_dq_free_space_nodirty(inode, quota_cut_bytes); 1737 return ret_value; 1738 } 1739 1740 static void truncate_directory(struct reiserfs_transaction_handle *th, 1741 struct inode *inode) 1742 { 1743 BUG_ON(!th->t_trans_id); 1744 if (inode->i_nlink) 1745 reiserfs_error(inode->i_sb, "vs-5655", "link count != 0"); 1746 1747 set_le_key_k_offset(KEY_FORMAT_3_5, INODE_PKEY(inode), DOT_OFFSET); 1748 set_le_key_k_type(KEY_FORMAT_3_5, INODE_PKEY(inode), TYPE_DIRENTRY); 1749 reiserfs_delete_solid_item(th, inode, INODE_PKEY(inode)); 1750 reiserfs_update_sd(th, inode); 1751 set_le_key_k_offset(KEY_FORMAT_3_5, INODE_PKEY(inode), SD_OFFSET); 1752 set_le_key_k_type(KEY_FORMAT_3_5, INODE_PKEY(inode), TYPE_STAT_DATA); 1753 } 1754 1755 /* Truncate file to the new size. Note, this must be called with a transaction 1756 already started */ 1757 int reiserfs_do_truncate(struct reiserfs_transaction_handle *th, 1758 struct inode *inode, /* ->i_size contains new size */ 1759 struct page *page, /* up to date for last block */ 1760 int update_timestamps /* when it is called by 1761 file_release to convert 1762 the tail - no timestamps 1763 should be updated */ 1764 ) 1765 { 1766 INITIALIZE_PATH(s_search_path); /* Path to the current object item. */ 1767 struct item_head *p_le_ih; /* Pointer to an item header. */ 1768 struct cpu_key s_item_key; /* Key to search for a previous file item. */ 1769 loff_t file_size, /* Old file size. */ 1770 new_file_size; /* New file size. */ 1771 int deleted; /* Number of deleted or truncated bytes. */ 1772 int retval; 1773 int err = 0; 1774 1775 BUG_ON(!th->t_trans_id); 1776 if (! 1777 (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) 1778 || S_ISLNK(inode->i_mode))) 1779 return 0; 1780 1781 if (S_ISDIR(inode->i_mode)) { 1782 // deletion of directory - no need to update timestamps 1783 truncate_directory(th, inode); 1784 return 0; 1785 } 1786 1787 /* Get new file size. */ 1788 new_file_size = inode->i_size; 1789 1790 // FIXME: note, that key type is unimportant here 1791 make_cpu_key(&s_item_key, inode, max_reiserfs_offset(inode), 1792 TYPE_DIRECT, 3); 1793 1794 retval = 1795 search_for_position_by_key(inode->i_sb, &s_item_key, 1796 &s_search_path); 1797 if (retval == IO_ERROR) { 1798 reiserfs_error(inode->i_sb, "vs-5657", 1799 "i/o failure occurred trying to truncate %K", 1800 &s_item_key); 1801 err = -EIO; 1802 goto out; 1803 } 1804 if (retval == POSITION_FOUND || retval == FILE_NOT_FOUND) { 1805 reiserfs_error(inode->i_sb, "PAP-5660", 1806 "wrong result %d of search for %K", retval, 1807 &s_item_key); 1808 1809 err = -EIO; 1810 goto out; 1811 } 1812 1813 s_search_path.pos_in_item--; 1814 1815 /* Get real file size (total length of all file items) */ 1816 p_le_ih = PATH_PITEM_HEAD(&s_search_path); 1817 if (is_statdata_le_ih(p_le_ih)) 1818 file_size = 0; 1819 else { 1820 loff_t offset = le_ih_k_offset(p_le_ih); 1821 int bytes = 1822 op_bytes_number(p_le_ih, inode->i_sb->s_blocksize); 1823 1824 /* this may mismatch with real file size: if last direct item 1825 had no padding zeros and last unformatted node had no free 1826 space, this file would have this file size */ 1827 file_size = offset + bytes - 1; 1828 } 1829 /* 1830 * are we doing a full truncate or delete, if so 1831 * kick in the reada code 1832 */ 1833 if (new_file_size == 0) 1834 s_search_path.reada = PATH_READA | PATH_READA_BACK; 1835 1836 if (file_size == 0 || file_size < new_file_size) { 1837 goto update_and_out; 1838 } 1839 1840 /* Update key to search for the last file item. */ 1841 set_cpu_key_k_offset(&s_item_key, file_size); 1842 1843 do { 1844 /* Cut or delete file item. */ 1845 deleted = 1846 reiserfs_cut_from_item(th, &s_search_path, &s_item_key, 1847 inode, page, new_file_size); 1848 if (deleted < 0) { 1849 reiserfs_warning(inode->i_sb, "vs-5665", 1850 "reiserfs_cut_from_item failed"); 1851 reiserfs_check_path(&s_search_path); 1852 return 0; 1853 } 1854 1855 RFALSE(deleted > file_size, 1856 "PAP-5670: reiserfs_cut_from_item: too many bytes deleted: deleted %d, file_size %lu, item_key %K", 1857 deleted, file_size, &s_item_key); 1858 1859 /* Change key to search the last file item. */ 1860 file_size -= deleted; 1861 1862 set_cpu_key_k_offset(&s_item_key, file_size); 1863 1864 /* While there are bytes to truncate and previous file item is presented in the tree. */ 1865 1866 /* 1867 ** This loop could take a really long time, and could log 1868 ** many more blocks than a transaction can hold. So, we do a polite 1869 ** journal end here, and if the transaction needs ending, we make 1870 ** sure the file is consistent before ending the current trans 1871 ** and starting a new one 1872 */ 1873 if (journal_transaction_should_end(th, 0) || 1874 reiserfs_transaction_free_space(th) <= JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD) { 1875 int orig_len_alloc = th->t_blocks_allocated; 1876 pathrelse(&s_search_path); 1877 1878 if (update_timestamps) { 1879 inode->i_mtime = CURRENT_TIME_SEC; 1880 inode->i_ctime = CURRENT_TIME_SEC; 1881 } 1882 reiserfs_update_sd(th, inode); 1883 1884 err = journal_end(th, inode->i_sb, orig_len_alloc); 1885 if (err) 1886 goto out; 1887 err = journal_begin(th, inode->i_sb, 1888 JOURNAL_FOR_FREE_BLOCK_AND_UPDATE_SD + JOURNAL_PER_BALANCE_CNT * 4) ; 1889 if (err) 1890 goto out; 1891 reiserfs_update_inode_transaction(inode); 1892 } 1893 } while (file_size > ROUND_UP(new_file_size) && 1894 search_for_position_by_key(inode->i_sb, &s_item_key, 1895 &s_search_path) == POSITION_FOUND); 1896 1897 RFALSE(file_size > ROUND_UP(new_file_size), 1898 "PAP-5680: truncate did not finish: new_file_size %Ld, current %Ld, oid %d", 1899 new_file_size, file_size, s_item_key.on_disk_key.k_objectid); 1900 1901 update_and_out: 1902 if (update_timestamps) { 1903 // this is truncate, not file closing 1904 inode->i_mtime = CURRENT_TIME_SEC; 1905 inode->i_ctime = CURRENT_TIME_SEC; 1906 } 1907 reiserfs_update_sd(th, inode); 1908 1909 out: 1910 pathrelse(&s_search_path); 1911 return err; 1912 } 1913 1914 #ifdef CONFIG_REISERFS_CHECK 1915 // this makes sure, that we __append__, not overwrite or add holes 1916 static void check_research_for_paste(struct treepath *path, 1917 const struct cpu_key *key) 1918 { 1919 struct item_head *found_ih = get_ih(path); 1920 1921 if (is_direct_le_ih(found_ih)) { 1922 if (le_ih_k_offset(found_ih) + 1923 op_bytes_number(found_ih, 1924 get_last_bh(path)->b_size) != 1925 cpu_key_k_offset(key) 1926 || op_bytes_number(found_ih, 1927 get_last_bh(path)->b_size) != 1928 pos_in_item(path)) 1929 reiserfs_panic(NULL, "PAP-5720", "found direct item " 1930 "%h or position (%d) does not match " 1931 "to key %K", found_ih, 1932 pos_in_item(path), key); 1933 } 1934 if (is_indirect_le_ih(found_ih)) { 1935 if (le_ih_k_offset(found_ih) + 1936 op_bytes_number(found_ih, 1937 get_last_bh(path)->b_size) != 1938 cpu_key_k_offset(key) 1939 || I_UNFM_NUM(found_ih) != pos_in_item(path) 1940 || get_ih_free_space(found_ih) != 0) 1941 reiserfs_panic(NULL, "PAP-5730", "found indirect " 1942 "item (%h) or position (%d) does not " 1943 "match to key (%K)", 1944 found_ih, pos_in_item(path), key); 1945 } 1946 } 1947 #endif /* config reiserfs check */ 1948 1949 /* Paste bytes to the existing item. Returns bytes number pasted into the item. */ 1950 int reiserfs_paste_into_item(struct reiserfs_transaction_handle *th, struct treepath *search_path, /* Path to the pasted item. */ 1951 const struct cpu_key *key, /* Key to search for the needed item. */ 1952 struct inode *inode, /* Inode item belongs to */ 1953 const char *body, /* Pointer to the bytes to paste. */ 1954 int pasted_size) 1955 { /* Size of pasted bytes. */ 1956 struct tree_balance s_paste_balance; 1957 int retval; 1958 int fs_gen; 1959 1960 BUG_ON(!th->t_trans_id); 1961 1962 fs_gen = get_generation(inode->i_sb); 1963 1964 #ifdef REISERQUOTA_DEBUG 1965 reiserfs_debug(inode->i_sb, REISERFS_DEBUG_CODE, 1966 "reiserquota paste_into_item(): allocating %u id=%u type=%c", 1967 pasted_size, inode->i_uid, 1968 key2type(&(key->on_disk_key))); 1969 #endif 1970 1971 if (vfs_dq_alloc_space_nodirty(inode, pasted_size)) { 1972 pathrelse(search_path); 1973 return -EDQUOT; 1974 } 1975 init_tb_struct(th, &s_paste_balance, th->t_super, search_path, 1976 pasted_size); 1977 #ifdef DISPLACE_NEW_PACKING_LOCALITIES 1978 s_paste_balance.key = key->on_disk_key; 1979 #endif 1980 1981 /* DQUOT_* can schedule, must check before the fix_nodes */ 1982 if (fs_changed(fs_gen, inode->i_sb)) { 1983 goto search_again; 1984 } 1985 1986 while ((retval = 1987 fix_nodes(M_PASTE, &s_paste_balance, NULL, 1988 body)) == REPEAT_SEARCH) { 1989 search_again: 1990 /* file system changed while we were in the fix_nodes */ 1991 PROC_INFO_INC(th->t_super, paste_into_item_restarted); 1992 retval = 1993 search_for_position_by_key(th->t_super, key, 1994 search_path); 1995 if (retval == IO_ERROR) { 1996 retval = -EIO; 1997 goto error_out; 1998 } 1999 if (retval == POSITION_FOUND) { 2000 reiserfs_warning(inode->i_sb, "PAP-5710", 2001 "entry or pasted byte (%K) exists", 2002 key); 2003 retval = -EEXIST; 2004 goto error_out; 2005 } 2006 #ifdef CONFIG_REISERFS_CHECK 2007 check_research_for_paste(search_path, key); 2008 #endif 2009 } 2010 2011 /* Perform balancing after all resources are collected by fix_nodes, and 2012 accessing them will not risk triggering schedule. */ 2013 if (retval == CARRY_ON) { 2014 do_balance(&s_paste_balance, NULL /*ih */ , body, M_PASTE); 2015 return 0; 2016 } 2017 retval = (retval == NO_DISK_SPACE) ? -ENOSPC : -EIO; 2018 error_out: 2019 /* this also releases the path */ 2020 unfix_nodes(&s_paste_balance); 2021 #ifdef REISERQUOTA_DEBUG 2022 reiserfs_debug(inode->i_sb, REISERFS_DEBUG_CODE, 2023 "reiserquota paste_into_item(): freeing %u id=%u type=%c", 2024 pasted_size, inode->i_uid, 2025 key2type(&(key->on_disk_key))); 2026 #endif 2027 vfs_dq_free_space_nodirty(inode, pasted_size); 2028 return retval; 2029 } 2030 2031 /* Insert new item into the buffer at the path. 2032 * th - active transaction handle 2033 * path - path to the inserted item 2034 * ih - pointer to the item header to insert 2035 * body - pointer to the bytes to insert 2036 */ 2037 int reiserfs_insert_item(struct reiserfs_transaction_handle *th, 2038 struct treepath *path, const struct cpu_key *key, 2039 struct item_head *ih, struct inode *inode, 2040 const char *body) 2041 { 2042 struct tree_balance s_ins_balance; 2043 int retval; 2044 int fs_gen = 0; 2045 int quota_bytes = 0; 2046 2047 BUG_ON(!th->t_trans_id); 2048 2049 if (inode) { /* Do we count quotas for item? */ 2050 fs_gen = get_generation(inode->i_sb); 2051 quota_bytes = ih_item_len(ih); 2052 2053 /* hack so the quota code doesn't have to guess if the file has 2054 ** a tail, links are always tails, so there's no guessing needed 2055 */ 2056 if (!S_ISLNK(inode->i_mode) && is_direct_le_ih(ih)) 2057 quota_bytes = inode->i_sb->s_blocksize + UNFM_P_SIZE; 2058 #ifdef REISERQUOTA_DEBUG 2059 reiserfs_debug(inode->i_sb, REISERFS_DEBUG_CODE, 2060 "reiserquota insert_item(): allocating %u id=%u type=%c", 2061 quota_bytes, inode->i_uid, head2type(ih)); 2062 #endif 2063 /* We can't dirty inode here. It would be immediately written but 2064 * appropriate stat item isn't inserted yet... */ 2065 if (vfs_dq_alloc_space_nodirty(inode, quota_bytes)) { 2066 pathrelse(path); 2067 return -EDQUOT; 2068 } 2069 } 2070 init_tb_struct(th, &s_ins_balance, th->t_super, path, 2071 IH_SIZE + ih_item_len(ih)); 2072 #ifdef DISPLACE_NEW_PACKING_LOCALITIES 2073 s_ins_balance.key = key->on_disk_key; 2074 #endif 2075 /* DQUOT_* can schedule, must check to be sure calling fix_nodes is safe */ 2076 if (inode && fs_changed(fs_gen, inode->i_sb)) { 2077 goto search_again; 2078 } 2079 2080 while ((retval = 2081 fix_nodes(M_INSERT, &s_ins_balance, ih, 2082 body)) == REPEAT_SEARCH) { 2083 search_again: 2084 /* file system changed while we were in the fix_nodes */ 2085 PROC_INFO_INC(th->t_super, insert_item_restarted); 2086 retval = search_item(th->t_super, key, path); 2087 if (retval == IO_ERROR) { 2088 retval = -EIO; 2089 goto error_out; 2090 } 2091 if (retval == ITEM_FOUND) { 2092 reiserfs_warning(th->t_super, "PAP-5760", 2093 "key %K already exists in the tree", 2094 key); 2095 retval = -EEXIST; 2096 goto error_out; 2097 } 2098 } 2099 2100 /* make balancing after all resources will be collected at a time */ 2101 if (retval == CARRY_ON) { 2102 do_balance(&s_ins_balance, ih, body, M_INSERT); 2103 return 0; 2104 } 2105 2106 retval = (retval == NO_DISK_SPACE) ? -ENOSPC : -EIO; 2107 error_out: 2108 /* also releases the path */ 2109 unfix_nodes(&s_ins_balance); 2110 #ifdef REISERQUOTA_DEBUG 2111 reiserfs_debug(th->t_super, REISERFS_DEBUG_CODE, 2112 "reiserquota insert_item(): freeing %u id=%u type=%c", 2113 quota_bytes, inode->i_uid, head2type(ih)); 2114 #endif 2115 if (inode) 2116 vfs_dq_free_space_nodirty(inode, quota_bytes); 2117 return retval; 2118 } 2119