rehash.c revision 5797cb017cb8accbab8c69aaf83269c5d2d72e4e
1/* 2 * rehash.c --- rebuild hash tree directories 3 * 4 * Copyright (C) 2002 Theodore Ts'o 5 * 6 * %Begin-Header% 7 * This file may be redistributed under the terms of the GNU Public 8 * License. 9 * %End-Header% 10 * 11 * This algorithm is designed for simplicity of implementation and to 12 * pack the directory as much as possible. It however requires twice 13 * as much memory as the size of the directory. The maximum size 14 * directory supported using a 4k blocksize is roughly a gigabyte, and 15 * so there may very well be problems with machines that don't have 16 * virtual memory, and obscenely large directories. 17 * 18 * An alternate algorithm which is much more disk intensive could be 19 * written, and probably will need to be written in the future. The 20 * design goals of such an algorithm are: (a) use (roughly) constant 21 * amounts of memory, no matter how large the directory, (b) the 22 * directory must be safe at all times, even if e2fsck is interrupted 23 * in the middle, (c) we must use minimal amounts of extra disk 24 * blocks. This pretty much requires an incremental approach, where 25 * we are reading from one part of the directory, and inserting into 26 * the front half. So the algorithm will have to keep track of a 27 * moving block boundary between the new tree and the old tree, and 28 * files will need to be moved from the old directory and inserted 29 * into the new tree. If the new directory requires space which isn't 30 * yet available, blocks from the beginning part of the old directory 31 * may need to be moved to the end of the directory to make room for 32 * the new tree: 33 * 34 * -------------------------------------------------------- 35 * | new tree | | old tree | 36 * -------------------------------------------------------- 37 * ^ ptr ^ptr 38 * tail new head old 39 * 40 * This is going to be a pain in the tuckus to implement, and will 41 * require a lot more disk accesses. So I'm going to skip it for now; 42 * it's only really going to be an issue for really, really big 43 * filesystems (when we reach the level of tens of millions of files 44 * in a single directory). It will probably be easier to simply 45 * require that e2fsck use VM first. 46 */ 47 48#include "config.h" 49#include <string.h> 50#include <ctype.h> 51#include <errno.h> 52#include "e2fsck.h" 53#include "problem.h" 54 55struct fill_dir_struct { 56 char *buf; 57 struct ext2_inode *inode; 58 errcode_t err; 59 e2fsck_t ctx; 60 struct hash_entry *harray; 61 int max_array, num_array; 62 unsigned int dir_size; 63 int compress; 64 ino_t parent; 65}; 66 67struct hash_entry { 68 ext2_dirhash_t hash; 69 ext2_dirhash_t minor_hash; 70 ino_t ino; 71 struct ext2_dir_entry *dir; 72}; 73 74struct out_dir { 75 int num; 76 int max; 77 char *buf; 78 ext2_dirhash_t *hashes; 79}; 80 81static int fill_dir_block(ext2_filsys fs, 82 blk64_t *block_nr, 83 e2_blkcnt_t blockcnt, 84 blk64_t ref_block EXT2FS_ATTR((unused)), 85 int ref_offset EXT2FS_ATTR((unused)), 86 void *priv_data) 87{ 88 struct fill_dir_struct *fd = (struct fill_dir_struct *) priv_data; 89 struct hash_entry *new_array, *ent; 90 struct ext2_dir_entry *dirent; 91 char *dir; 92 unsigned int offset, dir_offset, rec_len; 93 int hash_alg; 94 95 if (blockcnt < 0) 96 return 0; 97 98 offset = blockcnt * fs->blocksize; 99 if (offset + fs->blocksize > fd->inode->i_size) { 100 fd->err = EXT2_ET_DIR_CORRUPTED; 101 return BLOCK_ABORT; 102 } 103 dir = (fd->buf+offset); 104 if (HOLE_BLKADDR(*block_nr)) { 105 memset(dir, 0, fs->blocksize); 106 dirent = (struct ext2_dir_entry *) dir; 107 (void) ext2fs_set_rec_len(fs, fs->blocksize, dirent); 108 } else { 109 fd->err = ext2fs_read_dir_block3(fs, *block_nr, dir, 0); 110 if (fd->err) 111 return BLOCK_ABORT; 112 } 113 hash_alg = fs->super->s_def_hash_version; 114 if ((hash_alg <= EXT2_HASH_TEA) && 115 (fs->super->s_flags & EXT2_FLAGS_UNSIGNED_HASH)) 116 hash_alg += 3; 117 /* While the directory block is "hot", index it. */ 118 dir_offset = 0; 119 while (dir_offset < fs->blocksize) { 120 dirent = (struct ext2_dir_entry *) (dir + dir_offset); 121 (void) ext2fs_get_rec_len(fs, dirent, &rec_len); 122 if (((dir_offset + rec_len) > fs->blocksize) || 123 (rec_len < 8) || 124 ((rec_len % 4) != 0) || 125 (((dirent->name_len & 0xFF)+8U) > rec_len)) { 126 fd->err = EXT2_ET_DIR_CORRUPTED; 127 return BLOCK_ABORT; 128 } 129 dir_offset += rec_len; 130 if (dirent->inode == 0) 131 continue; 132 if (!fd->compress && ((dirent->name_len&0xFF) == 1) && 133 (dirent->name[0] == '.')) 134 continue; 135 if (!fd->compress && ((dirent->name_len&0xFF) == 2) && 136 (dirent->name[0] == '.') && (dirent->name[1] == '.')) { 137 fd->parent = dirent->inode; 138 continue; 139 } 140 if (fd->num_array >= fd->max_array) { 141 new_array = realloc(fd->harray, 142 sizeof(struct hash_entry) * (fd->max_array+500)); 143 if (!new_array) { 144 fd->err = ENOMEM; 145 return BLOCK_ABORT; 146 } 147 fd->harray = new_array; 148 fd->max_array += 500; 149 } 150 ent = fd->harray + fd->num_array++; 151 ent->dir = dirent; 152 fd->dir_size += EXT2_DIR_REC_LEN(dirent->name_len & 0xFF); 153 ent->ino = dirent->inode; 154 if (fd->compress) 155 ent->hash = ent->minor_hash = 0; 156 else { 157 fd->err = ext2fs_dirhash(hash_alg, dirent->name, 158 dirent->name_len & 0xFF, 159 fs->super->s_hash_seed, 160 &ent->hash, &ent->minor_hash); 161 if (fd->err) 162 return BLOCK_ABORT; 163 } 164 } 165 166 return 0; 167} 168 169/* Used for sorting the hash entry */ 170static EXT2_QSORT_TYPE ino_cmp(const void *a, const void *b) 171{ 172 const struct hash_entry *he_a = (const struct hash_entry *) a; 173 const struct hash_entry *he_b = (const struct hash_entry *) b; 174 175 return (he_a->ino - he_b->ino); 176} 177 178/* Used for sorting the hash entry */ 179static EXT2_QSORT_TYPE name_cmp(const void *a, const void *b) 180{ 181 const struct hash_entry *he_a = (const struct hash_entry *) a; 182 const struct hash_entry *he_b = (const struct hash_entry *) b; 183 int ret; 184 int min_len; 185 186 min_len = he_a->dir->name_len; 187 if (min_len > he_b->dir->name_len) 188 min_len = he_b->dir->name_len; 189 190 ret = strncmp(he_a->dir->name, he_b->dir->name, min_len); 191 if (ret == 0) { 192 if (he_a->dir->name_len > he_b->dir->name_len) 193 ret = 1; 194 else if (he_a->dir->name_len < he_b->dir->name_len) 195 ret = -1; 196 else 197 ret = he_b->dir->inode - he_a->dir->inode; 198 } 199 return ret; 200} 201 202/* Used for sorting the hash entry */ 203static EXT2_QSORT_TYPE hash_cmp(const void *a, const void *b) 204{ 205 const struct hash_entry *he_a = (const struct hash_entry *) a; 206 const struct hash_entry *he_b = (const struct hash_entry *) b; 207 int ret; 208 209 if (he_a->hash > he_b->hash) 210 ret = 1; 211 else if (he_a->hash < he_b->hash) 212 ret = -1; 213 else { 214 if (he_a->minor_hash > he_b->minor_hash) 215 ret = 1; 216 else if (he_a->minor_hash < he_b->minor_hash) 217 ret = -1; 218 else 219 ret = name_cmp(a, b); 220 } 221 return ret; 222} 223 224static errcode_t alloc_size_dir(ext2_filsys fs, struct out_dir *outdir, 225 int blocks) 226{ 227 void *new_mem; 228 229 if (outdir->max) { 230 new_mem = realloc(outdir->buf, blocks * fs->blocksize); 231 if (!new_mem) 232 return ENOMEM; 233 outdir->buf = new_mem; 234 new_mem = realloc(outdir->hashes, 235 blocks * sizeof(ext2_dirhash_t)); 236 if (!new_mem) 237 return ENOMEM; 238 outdir->hashes = new_mem; 239 } else { 240 outdir->buf = malloc(blocks * fs->blocksize); 241 outdir->hashes = malloc(blocks * sizeof(ext2_dirhash_t)); 242 outdir->num = 0; 243 } 244 outdir->max = blocks; 245 return 0; 246} 247 248static void free_out_dir(struct out_dir *outdir) 249{ 250 free(outdir->buf); 251 free(outdir->hashes); 252 outdir->max = 0; 253 outdir->num =0; 254} 255 256static errcode_t get_next_block(ext2_filsys fs, struct out_dir *outdir, 257 char ** ret) 258{ 259 errcode_t retval; 260 261 if (outdir->num >= outdir->max) { 262 retval = alloc_size_dir(fs, outdir, outdir->max + 50); 263 if (retval) 264 return retval; 265 } 266 *ret = outdir->buf + (outdir->num++ * fs->blocksize); 267 memset(*ret, 0, fs->blocksize); 268 return 0; 269} 270 271/* 272 * This function is used to make a unique filename. We do this by 273 * appending ~0, and then incrementing the number. However, we cannot 274 * expand the length of the filename beyond the padding available in 275 * the directory entry. 276 */ 277static void mutate_name(char *str, __u16 *len) 278{ 279 int i; 280 __u16 l = *len & 0xFF, h = *len & 0xff00; 281 282 /* 283 * First check to see if it looks the name has been mutated 284 * already 285 */ 286 for (i = l-1; i > 0; i--) { 287 if (!isdigit(str[i])) 288 break; 289 } 290 if ((i == l-1) || (str[i] != '~')) { 291 if (((l-1) & 3) < 2) 292 l += 2; 293 else 294 l = (l+3) & ~3; 295 str[l-2] = '~'; 296 str[l-1] = '0'; 297 *len = l | h; 298 return; 299 } 300 for (i = l-1; i >= 0; i--) { 301 if (isdigit(str[i])) { 302 if (str[i] == '9') 303 str[i] = '0'; 304 else { 305 str[i]++; 306 return; 307 } 308 continue; 309 } 310 if (i == 1) { 311 if (str[0] == 'z') 312 str[0] = 'A'; 313 else if (str[0] == 'Z') { 314 str[0] = '~'; 315 str[1] = '0'; 316 } else 317 str[0]++; 318 } else if (i > 0) { 319 str[i] = '1'; 320 str[i-1] = '~'; 321 } else { 322 if (str[0] == '~') 323 str[0] = 'a'; 324 else 325 str[0]++; 326 } 327 break; 328 } 329} 330 331static int duplicate_search_and_fix(e2fsck_t ctx, ext2_filsys fs, 332 ext2_ino_t ino, 333 struct fill_dir_struct *fd) 334{ 335 struct problem_context pctx; 336 struct hash_entry *ent, *prev; 337 int i, j; 338 int fixed = 0; 339 char new_name[256]; 340 __u16 new_len; 341 int hash_alg; 342 343 clear_problem_context(&pctx); 344 pctx.ino = ino; 345 346 hash_alg = fs->super->s_def_hash_version; 347 if ((hash_alg <= EXT2_HASH_TEA) && 348 (fs->super->s_flags & EXT2_FLAGS_UNSIGNED_HASH)) 349 hash_alg += 3; 350 351 for (i=1; i < fd->num_array; i++) { 352 ent = fd->harray + i; 353 prev = ent - 1; 354 if (!ent->dir->inode || 355 ((ent->dir->name_len & 0xFF) != 356 (prev->dir->name_len & 0xFF)) || 357 (strncmp(ent->dir->name, prev->dir->name, 358 ent->dir->name_len & 0xFF))) 359 continue; 360 pctx.dirent = ent->dir; 361 if ((ent->dir->inode == prev->dir->inode) && 362 fix_problem(ctx, PR_2_DUPLICATE_DIRENT, &pctx)) { 363 e2fsck_adjust_inode_count(ctx, ent->dir->inode, -1); 364 ent->dir->inode = 0; 365 fixed++; 366 continue; 367 } 368 memcpy(new_name, ent->dir->name, ent->dir->name_len & 0xFF); 369 new_len = ent->dir->name_len; 370 mutate_name(new_name, &new_len); 371 for (j=0; j < fd->num_array; j++) { 372 if ((i==j) || 373 ((new_len & 0xFF) != 374 (fd->harray[j].dir->name_len & 0xFF)) || 375 (strncmp(new_name, fd->harray[j].dir->name, 376 new_len & 0xFF))) 377 continue; 378 mutate_name(new_name, &new_len); 379 380 j = -1; 381 } 382 new_name[new_len & 0xFF] = 0; 383 pctx.str = new_name; 384 if (fix_problem(ctx, PR_2_NON_UNIQUE_FILE, &pctx)) { 385 memcpy(ent->dir->name, new_name, new_len & 0xFF); 386 ent->dir->name_len = new_len; 387 ext2fs_dirhash(hash_alg, ent->dir->name, 388 ent->dir->name_len & 0xFF, 389 fs->super->s_hash_seed, 390 &ent->hash, &ent->minor_hash); 391 fixed++; 392 } 393 } 394 return fixed; 395} 396 397 398static errcode_t copy_dir_entries(e2fsck_t ctx, 399 struct fill_dir_struct *fd, 400 struct out_dir *outdir) 401{ 402 ext2_filsys fs = ctx->fs; 403 errcode_t retval; 404 char *block_start; 405 struct hash_entry *ent; 406 struct ext2_dir_entry *dirent; 407 unsigned int rec_len, prev_rec_len, left, slack, offset; 408 int i; 409 ext2_dirhash_t prev_hash; 410 411 if (ctx->htree_slack_percentage == 255) { 412 profile_get_uint(ctx->profile, "options", 413 "indexed_dir_slack_percentage", 414 0, 20, 415 &ctx->htree_slack_percentage); 416 if (ctx->htree_slack_percentage > 100) 417 ctx->htree_slack_percentage = 20; 418 } 419 420 outdir->max = 0; 421 retval = alloc_size_dir(fs, outdir, 422 (fd->dir_size / fs->blocksize) + 2); 423 if (retval) 424 return retval; 425 outdir->num = fd->compress ? 0 : 1; 426 offset = 0; 427 outdir->hashes[0] = 0; 428 prev_hash = 1; 429 if ((retval = get_next_block(fs, outdir, &block_start))) 430 return retval; 431 dirent = (struct ext2_dir_entry *) block_start; 432 prev_rec_len = 0; 433 rec_len = 0; 434 left = fs->blocksize; 435 slack = fd->compress ? 12 : 436 (fs->blocksize * ctx->htree_slack_percentage)/100; 437 if (slack < 12) 438 slack = 12; 439 for (i = 0; i < fd->num_array; i++) { 440 ent = fd->harray + i; 441 if (ent->dir->inode == 0) 442 continue; 443 rec_len = EXT2_DIR_REC_LEN(ent->dir->name_len & 0xFF); 444 if (rec_len > left) { 445 if (left) { 446 left += prev_rec_len; 447 retval = ext2fs_set_rec_len(fs, left, dirent); 448 if (retval) 449 return retval; 450 } 451 if ((retval = get_next_block(fs, outdir, 452 &block_start))) 453 return retval; 454 offset = 0; 455 } 456 left = fs->blocksize - offset; 457 dirent = (struct ext2_dir_entry *) (block_start + offset); 458 if (offset == 0) { 459 if (ent->hash == prev_hash) 460 outdir->hashes[outdir->num-1] = ent->hash | 1; 461 else 462 outdir->hashes[outdir->num-1] = ent->hash; 463 } 464 dirent->inode = ent->dir->inode; 465 dirent->name_len = ent->dir->name_len; 466 retval = ext2fs_set_rec_len(fs, rec_len, dirent); 467 if (retval) 468 return retval; 469 prev_rec_len = rec_len; 470 memcpy(dirent->name, ent->dir->name, dirent->name_len & 0xFF); 471 offset += rec_len; 472 left -= rec_len; 473 if (left < slack) { 474 prev_rec_len += left; 475 retval = ext2fs_set_rec_len(fs, prev_rec_len, dirent); 476 if (retval) 477 return retval; 478 offset += left; 479 left = 0; 480 } 481 prev_hash = ent->hash; 482 } 483 if (left) 484 retval = ext2fs_set_rec_len(fs, rec_len + left, dirent); 485 486 return retval; 487} 488 489 490static struct ext2_dx_root_info *set_root_node(ext2_filsys fs, char *buf, 491 ext2_ino_t ino, ext2_ino_t parent) 492{ 493 struct ext2_dir_entry *dir; 494 struct ext2_dx_root_info *root; 495 struct ext2_dx_countlimit *limits; 496 int filetype = 0; 497 498 if (fs->super->s_feature_incompat & EXT2_FEATURE_INCOMPAT_FILETYPE) 499 filetype = EXT2_FT_DIR << 8; 500 501 memset(buf, 0, fs->blocksize); 502 dir = (struct ext2_dir_entry *) buf; 503 dir->inode = ino; 504 dir->name[0] = '.'; 505 dir->name_len = 1 | filetype; 506 dir->rec_len = 12; 507 dir = (struct ext2_dir_entry *) (buf + 12); 508 dir->inode = parent; 509 dir->name[0] = '.'; 510 dir->name[1] = '.'; 511 dir->name_len = 2 | filetype; 512 dir->rec_len = fs->blocksize - 12; 513 514 root = (struct ext2_dx_root_info *) (buf+24); 515 root->reserved_zero = 0; 516 root->hash_version = fs->super->s_def_hash_version; 517 root->info_length = 8; 518 root->indirect_levels = 0; 519 root->unused_flags = 0; 520 521 limits = (struct ext2_dx_countlimit *) (buf+32); 522 limits->limit = (fs->blocksize - 32) / sizeof(struct ext2_dx_entry); 523 limits->count = 0; 524 525 return root; 526} 527 528 529static struct ext2_dx_entry *set_int_node(ext2_filsys fs, char *buf) 530{ 531 struct ext2_dir_entry *dir; 532 struct ext2_dx_countlimit *limits; 533 534 memset(buf, 0, fs->blocksize); 535 dir = (struct ext2_dir_entry *) buf; 536 dir->inode = 0; 537 (void) ext2fs_set_rec_len(fs, fs->blocksize, dir); 538 539 limits = (struct ext2_dx_countlimit *) (buf+8); 540 limits->limit = (fs->blocksize - 8) / sizeof(struct ext2_dx_entry); 541 limits->count = 0; 542 543 return (struct ext2_dx_entry *) limits; 544} 545 546/* 547 * This function takes the leaf nodes which have been written in 548 * outdir, and populates the root node and any necessary interior nodes. 549 */ 550static errcode_t calculate_tree(ext2_filsys fs, 551 struct out_dir *outdir, 552 ext2_ino_t ino, 553 ext2_ino_t parent) 554{ 555 struct ext2_dx_root_info *root_info; 556 struct ext2_dx_entry *root, *dx_ent = 0; 557 struct ext2_dx_countlimit *root_limit, *limit; 558 errcode_t retval; 559 char * block_start; 560 int i, c1, c2, nblks; 561 int limit_offset, root_offset; 562 563 root_info = set_root_node(fs, outdir->buf, ino, parent); 564 root_offset = limit_offset = ((char *) root_info - outdir->buf) + 565 root_info->info_length; 566 root_limit = (struct ext2_dx_countlimit *) (outdir->buf + limit_offset); 567 c1 = root_limit->limit; 568 nblks = outdir->num; 569 570 /* Write out the pointer blocks */ 571 if (nblks-1 <= c1) { 572 /* Just write out the root block, and we're done */ 573 root = (struct ext2_dx_entry *) (outdir->buf + root_offset); 574 for (i=1; i < nblks; i++) { 575 root->block = ext2fs_cpu_to_le32(i); 576 if (i != 1) 577 root->hash = 578 ext2fs_cpu_to_le32(outdir->hashes[i]); 579 root++; 580 c1--; 581 } 582 } else { 583 c2 = 0; 584 limit = 0; 585 root_info->indirect_levels = 1; 586 for (i=1; i < nblks; i++) { 587 if (c1 == 0) 588 return ENOSPC; 589 if (c2 == 0) { 590 if (limit) 591 limit->limit = limit->count = 592 ext2fs_cpu_to_le16(limit->limit); 593 root = (struct ext2_dx_entry *) 594 (outdir->buf + root_offset); 595 root->block = ext2fs_cpu_to_le32(outdir->num); 596 if (i != 1) 597 root->hash = 598 ext2fs_cpu_to_le32(outdir->hashes[i]); 599 if ((retval = get_next_block(fs, outdir, 600 &block_start))) 601 return retval; 602 dx_ent = set_int_node(fs, block_start); 603 limit = (struct ext2_dx_countlimit *) dx_ent; 604 c2 = limit->limit; 605 root_offset += sizeof(struct ext2_dx_entry); 606 c1--; 607 } 608 dx_ent->block = ext2fs_cpu_to_le32(i); 609 if (c2 != limit->limit) 610 dx_ent->hash = 611 ext2fs_cpu_to_le32(outdir->hashes[i]); 612 dx_ent++; 613 c2--; 614 } 615 limit->count = ext2fs_cpu_to_le16(limit->limit - c2); 616 limit->limit = ext2fs_cpu_to_le16(limit->limit); 617 } 618 root_limit = (struct ext2_dx_countlimit *) (outdir->buf + limit_offset); 619 root_limit->count = ext2fs_cpu_to_le16(root_limit->limit - c1); 620 root_limit->limit = ext2fs_cpu_to_le16(root_limit->limit); 621 622 return 0; 623} 624 625struct write_dir_struct { 626 struct out_dir *outdir; 627 errcode_t err; 628 e2fsck_t ctx; 629 blk64_t cleared; 630}; 631 632/* 633 * Helper function which writes out a directory block. 634 */ 635static int write_dir_block(ext2_filsys fs, 636 blk64_t *block_nr, 637 e2_blkcnt_t blockcnt, 638 blk64_t ref_block EXT2FS_ATTR((unused)), 639 int ref_offset EXT2FS_ATTR((unused)), 640 void *priv_data) 641{ 642 struct write_dir_struct *wd = (struct write_dir_struct *) priv_data; 643 blk64_t blk; 644 char *dir; 645 646 if (*block_nr == 0) 647 return 0; 648 if (blockcnt >= wd->outdir->num) { 649 e2fsck_read_bitmaps(wd->ctx); 650 blk = *block_nr; 651 /* 652 * In theory, we only release blocks from the end of the 653 * directory file, so it's fine to clobber a whole cluster at 654 * once. 655 */ 656 if (blk % EXT2FS_CLUSTER_RATIO(fs) == 0) { 657 ext2fs_unmark_block_bitmap2(wd->ctx->block_found_map, 658 blk); 659 ext2fs_block_alloc_stats2(fs, blk, -1); 660 wd->cleared++; 661 } 662 *block_nr = 0; 663 return BLOCK_CHANGED; 664 } 665 if (blockcnt < 0) 666 return 0; 667 668 dir = wd->outdir->buf + (blockcnt * fs->blocksize); 669 wd->err = ext2fs_write_dir_block3(fs, *block_nr, dir, 0); 670 if (wd->err) 671 return BLOCK_ABORT; 672 return 0; 673} 674 675static errcode_t write_directory(e2fsck_t ctx, ext2_filsys fs, 676 struct out_dir *outdir, 677 ext2_ino_t ino, int compress) 678{ 679 struct write_dir_struct wd; 680 errcode_t retval; 681 struct ext2_inode inode; 682 683 retval = e2fsck_expand_directory(ctx, ino, -1, outdir->num); 684 if (retval) 685 return retval; 686 687 wd.outdir = outdir; 688 wd.err = 0; 689 wd.ctx = ctx; 690 wd.cleared = 0; 691 692 retval = ext2fs_block_iterate3(fs, ino, 0, 0, 693 write_dir_block, &wd); 694 if (retval) 695 return retval; 696 if (wd.err) 697 return wd.err; 698 699 e2fsck_read_inode(ctx, ino, &inode, "rehash_dir"); 700 if (compress) 701 inode.i_flags &= ~EXT2_INDEX_FL; 702 else 703 inode.i_flags |= EXT2_INDEX_FL; 704 inode.i_size = outdir->num * fs->blocksize; 705 ext2fs_iblk_sub_blocks(fs, &inode, wd.cleared); 706 e2fsck_write_inode(ctx, ino, &inode, "rehash_dir"); 707 708 return 0; 709} 710 711errcode_t e2fsck_rehash_dir(e2fsck_t ctx, ext2_ino_t ino) 712{ 713 ext2_filsys fs = ctx->fs; 714 errcode_t retval; 715 struct ext2_inode inode; 716 char *dir_buf = 0; 717 struct fill_dir_struct fd; 718 struct out_dir outdir; 719 720 outdir.max = outdir.num = 0; 721 outdir.buf = 0; 722 outdir.hashes = 0; 723 e2fsck_read_inode(ctx, ino, &inode, "rehash_dir"); 724 725 retval = ENOMEM; 726 fd.harray = 0; 727 dir_buf = malloc(inode.i_size); 728 if (!dir_buf) 729 goto errout; 730 731 fd.max_array = inode.i_size / 32; 732 fd.num_array = 0; 733 fd.harray = malloc(fd.max_array * sizeof(struct hash_entry)); 734 if (!fd.harray) 735 goto errout; 736 737 fd.ctx = ctx; 738 fd.buf = dir_buf; 739 fd.inode = &inode; 740 fd.err = 0; 741 fd.dir_size = 0; 742 fd.compress = 0; 743 if (!(fs->super->s_feature_compat & EXT2_FEATURE_COMPAT_DIR_INDEX) || 744 (inode.i_size / fs->blocksize) < 2) 745 fd.compress = 1; 746 fd.parent = 0; 747 748retry_nohash: 749 /* Read in the entire directory into memory */ 750 retval = ext2fs_block_iterate3(fs, ino, 0, 0, 751 fill_dir_block, &fd); 752 if (fd.err) { 753 retval = fd.err; 754 goto errout; 755 } 756 757 /* 758 * If the entries read are less than a block, then don't index 759 * the directory 760 */ 761 if (!fd.compress && (fd.dir_size < (fs->blocksize - 24))) { 762 fd.compress = 1; 763 fd.dir_size = 0; 764 fd.num_array = 0; 765 goto retry_nohash; 766 } 767 768#if 0 769 printf("%d entries (%d bytes) found in inode %d\n", 770 fd.num_array, fd.dir_size, ino); 771#endif 772 773 /* Sort the list */ 774resort: 775 if (fd.compress) 776 qsort(fd.harray+2, fd.num_array-2, sizeof(struct hash_entry), 777 hash_cmp); 778 else 779 qsort(fd.harray, fd.num_array, sizeof(struct hash_entry), 780 hash_cmp); 781 782 /* 783 * Look for duplicates 784 */ 785 if (duplicate_search_and_fix(ctx, fs, ino, &fd)) 786 goto resort; 787 788 if (ctx->options & E2F_OPT_NO) { 789 retval = 0; 790 goto errout; 791 } 792 793 /* Sort non-hashed directories by inode number */ 794 if (fd.compress) 795 qsort(fd.harray+2, fd.num_array-2, 796 sizeof(struct hash_entry), ino_cmp); 797 798 /* 799 * Copy the directory entries. In a htree directory these 800 * will become the leaf nodes. 801 */ 802 retval = copy_dir_entries(ctx, &fd, &outdir); 803 if (retval) 804 goto errout; 805 806 free(dir_buf); dir_buf = 0; 807 808 if (!fd.compress) { 809 /* Calculate the interior nodes */ 810 retval = calculate_tree(fs, &outdir, ino, fd.parent); 811 if (retval) 812 goto errout; 813 } 814 815 retval = write_directory(ctx, fs, &outdir, ino, fd.compress); 816 if (retval) 817 goto errout; 818 819errout: 820 free(dir_buf); 821 free(fd.harray); 822 823 free_out_dir(&outdir); 824 return retval; 825} 826 827void e2fsck_rehash_directories(e2fsck_t ctx) 828{ 829 struct problem_context pctx; 830#ifdef RESOURCE_TRACK 831 struct resource_track rtrack; 832#endif 833 struct dir_info *dir; 834 ext2_u32_iterate iter; 835 struct dir_info_iter * dirinfo_iter = 0; 836 ext2_ino_t ino; 837 errcode_t retval; 838 int cur, max, all_dirs, first = 1; 839 840 init_resource_track(&rtrack, ctx->fs->io); 841 all_dirs = ctx->options & E2F_OPT_COMPRESS_DIRS; 842 843 if (!ctx->dirs_to_hash && !all_dirs) 844 return; 845 846 e2fsck_get_lost_and_found(ctx, 0); 847 848 clear_problem_context(&pctx); 849 850 cur = 0; 851 if (all_dirs) { 852 dirinfo_iter = e2fsck_dir_info_iter_begin(ctx); 853 max = e2fsck_get_num_dirinfo(ctx); 854 } else { 855 retval = ext2fs_u32_list_iterate_begin(ctx->dirs_to_hash, 856 &iter); 857 if (retval) { 858 pctx.errcode = retval; 859 fix_problem(ctx, PR_3A_OPTIMIZE_ITER, &pctx); 860 return; 861 } 862 max = ext2fs_u32_list_count(ctx->dirs_to_hash); 863 } 864 while (1) { 865 if (all_dirs) { 866 if ((dir = e2fsck_dir_info_iter(ctx, 867 dirinfo_iter)) == 0) 868 break; 869 ino = dir->ino; 870 } else { 871 if (!ext2fs_u32_list_iterate(iter, &ino)) 872 break; 873 } 874 if (ino == ctx->lost_and_found) 875 continue; 876 pctx.dir = ino; 877 if (first) { 878 fix_problem(ctx, PR_3A_PASS_HEADER, &pctx); 879 first = 0; 880 } 881#if 0 882 fix_problem(ctx, PR_3A_OPTIMIZE_DIR, &pctx); 883#endif 884 pctx.errcode = e2fsck_rehash_dir(ctx, ino); 885 if (pctx.errcode) { 886 end_problem_latch(ctx, PR_LATCH_OPTIMIZE_DIR); 887 fix_problem(ctx, PR_3A_OPTIMIZE_DIR_ERR, &pctx); 888 } 889 if (ctx->progress && !ctx->progress_fd) 890 e2fsck_simple_progress(ctx, "Rebuilding directory", 891 100.0 * (float) (++cur) / (float) max, ino); 892 } 893 end_problem_latch(ctx, PR_LATCH_OPTIMIZE_DIR); 894 if (all_dirs) 895 e2fsck_dir_info_iter_end(ctx, dirinfo_iter); 896 else 897 ext2fs_u32_list_iterate_end(iter); 898 899 if (ctx->dirs_to_hash) 900 ext2fs_u32_list_free(ctx->dirs_to_hash); 901 ctx->dirs_to_hash = 0; 902 903 print_resource_track(ctx, "Pass 3A", &rtrack, ctx->fs->io); 904} 905