Mike Frysinger | 51a43b4 | 2005-09-24 07:11:16 +0000 | [diff] [blame^] | 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 <string.h> |
| 49 | #include <ctype.h> |
| 50 | #include <errno.h> |
| 51 | #include "e2fsck.h" |
| 52 | #include "problem.h" |
| 53 | |
| 54 | struct fill_dir_struct { |
| 55 | char *buf; |
| 56 | struct ext2_inode *inode; |
| 57 | int err; |
| 58 | e2fsck_t ctx; |
| 59 | struct hash_entry *harray; |
| 60 | int max_array, num_array; |
| 61 | int dir_size; |
| 62 | int compress; |
| 63 | ino_t parent; |
| 64 | }; |
| 65 | |
| 66 | struct hash_entry { |
| 67 | ext2_dirhash_t hash; |
| 68 | ext2_dirhash_t minor_hash; |
| 69 | struct ext2_dir_entry *dir; |
| 70 | }; |
| 71 | |
| 72 | struct out_dir { |
| 73 | int num; |
| 74 | int max; |
| 75 | char *buf; |
| 76 | ext2_dirhash_t *hashes; |
| 77 | }; |
| 78 | |
| 79 | static int fill_dir_block(ext2_filsys fs, |
| 80 | blk_t *block_nr, |
| 81 | e2_blkcnt_t blockcnt, |
| 82 | blk_t ref_block EXT2FS_ATTR((unused)), |
| 83 | int ref_offset EXT2FS_ATTR((unused)), |
| 84 | void *priv_data) |
| 85 | { |
| 86 | struct fill_dir_struct *fd = (struct fill_dir_struct *) priv_data; |
| 87 | struct hash_entry *new_array, *ent; |
| 88 | struct ext2_dir_entry *dirent; |
| 89 | char *dir; |
| 90 | unsigned int offset, dir_offset; |
| 91 | |
| 92 | if (blockcnt < 0) |
| 93 | return 0; |
| 94 | |
| 95 | offset = blockcnt * fs->blocksize; |
| 96 | if (offset + fs->blocksize > fd->inode->i_size) { |
| 97 | fd->err = EXT2_ET_DIR_CORRUPTED; |
| 98 | return BLOCK_ABORT; |
| 99 | } |
| 100 | dir = (fd->buf+offset); |
| 101 | if (HOLE_BLKADDR(*block_nr)) { |
| 102 | memset(dir, 0, fs->blocksize); |
| 103 | dirent = (struct ext2_dir_entry *) dir; |
| 104 | dirent->rec_len = fs->blocksize; |
| 105 | } else { |
| 106 | fd->err = ext2fs_read_dir_block(fs, *block_nr, dir); |
| 107 | if (fd->err) |
| 108 | return BLOCK_ABORT; |
| 109 | } |
| 110 | /* While the directory block is "hot", index it. */ |
| 111 | dir_offset = 0; |
| 112 | while (dir_offset < fs->blocksize) { |
| 113 | dirent = (struct ext2_dir_entry *) (dir + dir_offset); |
| 114 | if (((dir_offset + dirent->rec_len) > fs->blocksize) || |
| 115 | (dirent->rec_len < 8) || |
| 116 | ((dirent->rec_len % 4) != 0) || |
| 117 | (((dirent->name_len & 0xFF)+8) > dirent->rec_len)) { |
| 118 | fd->err = EXT2_ET_DIR_CORRUPTED; |
| 119 | return BLOCK_ABORT; |
| 120 | } |
| 121 | dir_offset += dirent->rec_len; |
| 122 | if (dirent->inode == 0) |
| 123 | continue; |
| 124 | if (!fd->compress && ((dirent->name_len&0xFF) == 1) && |
| 125 | (dirent->name[0] == '.')) |
| 126 | continue; |
| 127 | if (!fd->compress && ((dirent->name_len&0xFF) == 2) && |
| 128 | (dirent->name[0] == '.') && (dirent->name[1] == '.')) { |
| 129 | fd->parent = dirent->inode; |
| 130 | continue; |
| 131 | } |
| 132 | if (fd->num_array >= fd->max_array) { |
| 133 | new_array = realloc(fd->harray, |
| 134 | sizeof(struct hash_entry) * (fd->max_array+500)); |
| 135 | if (!new_array) { |
| 136 | fd->err = ENOMEM; |
| 137 | return BLOCK_ABORT; |
| 138 | } |
| 139 | fd->harray = new_array; |
| 140 | fd->max_array += 500; |
| 141 | } |
| 142 | ent = fd->harray + fd->num_array++; |
| 143 | ent->dir = dirent; |
| 144 | fd->dir_size += EXT2_DIR_REC_LEN(dirent->name_len & 0xFF); |
| 145 | if (fd->compress) |
| 146 | ent->hash = ent->minor_hash = 0; |
| 147 | else { |
| 148 | fd->err = ext2fs_dirhash(fs->super->s_def_hash_version, |
| 149 | dirent->name, |
| 150 | dirent->name_len & 0xFF, |
| 151 | fs->super->s_hash_seed, |
| 152 | &ent->hash, &ent->minor_hash); |
| 153 | if (fd->err) |
| 154 | return BLOCK_ABORT; |
| 155 | } |
| 156 | } |
| 157 | |
| 158 | return 0; |
| 159 | } |
| 160 | |
| 161 | /* Used for sorting the hash entry */ |
| 162 | static EXT2_QSORT_TYPE name_cmp(const void *a, const void *b) |
| 163 | { |
| 164 | const struct hash_entry *he_a = (const struct hash_entry *) a; |
| 165 | const struct hash_entry *he_b = (const struct hash_entry *) b; |
| 166 | int ret; |
| 167 | int min_len; |
| 168 | |
| 169 | min_len = he_a->dir->name_len; |
| 170 | if (min_len > he_b->dir->name_len) |
| 171 | min_len = he_b->dir->name_len; |
| 172 | |
| 173 | ret = strncmp(he_a->dir->name, he_b->dir->name, min_len); |
| 174 | if (ret == 0) { |
| 175 | if (he_a->dir->name_len > he_b->dir->name_len) |
| 176 | ret = 1; |
| 177 | else if (he_a->dir->name_len < he_b->dir->name_len) |
| 178 | ret = -1; |
| 179 | else |
| 180 | ret = he_b->dir->inode - he_a->dir->inode; |
| 181 | } |
| 182 | return ret; |
| 183 | } |
| 184 | |
| 185 | /* Used for sorting the hash entry */ |
| 186 | static EXT2_QSORT_TYPE hash_cmp(const void *a, const void *b) |
| 187 | { |
| 188 | const struct hash_entry *he_a = (const struct hash_entry *) a; |
| 189 | const struct hash_entry *he_b = (const struct hash_entry *) b; |
| 190 | int ret; |
| 191 | |
| 192 | if (he_a->hash > he_b->hash) |
| 193 | ret = 1; |
| 194 | else if (he_a->hash < he_b->hash) |
| 195 | ret = -1; |
| 196 | else { |
| 197 | if (he_a->minor_hash > he_b->minor_hash) |
| 198 | ret = 1; |
| 199 | else if (he_a->minor_hash < he_b->minor_hash) |
| 200 | ret = -1; |
| 201 | else |
| 202 | ret = name_cmp(a, b); |
| 203 | } |
| 204 | return ret; |
| 205 | } |
| 206 | |
| 207 | static errcode_t alloc_size_dir(ext2_filsys fs, struct out_dir *outdir, |
| 208 | int blocks) |
| 209 | { |
| 210 | void *new_mem; |
| 211 | |
| 212 | if (outdir->max) { |
| 213 | new_mem = realloc(outdir->buf, blocks * fs->blocksize); |
| 214 | if (!new_mem) |
| 215 | return ENOMEM; |
| 216 | outdir->buf = new_mem; |
| 217 | new_mem = realloc(outdir->hashes, |
| 218 | blocks * sizeof(ext2_dirhash_t)); |
| 219 | if (!new_mem) |
| 220 | return ENOMEM; |
| 221 | outdir->hashes = new_mem; |
| 222 | } else { |
| 223 | outdir->buf = malloc(blocks * fs->blocksize); |
| 224 | outdir->hashes = malloc(blocks * sizeof(ext2_dirhash_t)); |
| 225 | outdir->num = 0; |
| 226 | } |
| 227 | outdir->max = blocks; |
| 228 | return 0; |
| 229 | } |
| 230 | |
| 231 | static void free_out_dir(struct out_dir *outdir) |
| 232 | { |
| 233 | if (outdir->buf) |
| 234 | free(outdir->buf); |
| 235 | if (outdir->hashes) |
| 236 | free(outdir->hashes); |
| 237 | outdir->max = 0; |
| 238 | outdir->num =0; |
| 239 | } |
| 240 | |
| 241 | static errcode_t get_next_block(ext2_filsys fs, struct out_dir *outdir, |
| 242 | char ** ret) |
| 243 | { |
| 244 | errcode_t retval; |
| 245 | |
| 246 | if (outdir->num >= outdir->max) { |
| 247 | retval = alloc_size_dir(fs, outdir, outdir->max + 50); |
| 248 | if (retval) |
| 249 | return retval; |
| 250 | } |
| 251 | *ret = outdir->buf + (outdir->num++ * fs->blocksize); |
| 252 | memset(*ret, 0, fs->blocksize); |
| 253 | return 0; |
| 254 | } |
| 255 | |
| 256 | /* |
| 257 | * This function is used to make a unique filename. We do this by |
| 258 | * appending ~0, and then incrementing the number. However, we cannot |
| 259 | * expand the length of the filename beyond the padding available in |
| 260 | * the directory entry. |
| 261 | */ |
| 262 | static void mutate_name(char *str, __u16 *len) |
| 263 | { |
| 264 | int i; |
| 265 | __u16 l = *len & 0xFF, h = *len & 0xff00; |
| 266 | |
| 267 | /* |
| 268 | * First check to see if it looks the name has been mutated |
| 269 | * already |
| 270 | */ |
| 271 | for (i = l-1; i > 0; i--) { |
| 272 | if (!isdigit(str[i])) |
| 273 | break; |
| 274 | } |
| 275 | if ((i == l-1) || (str[i] != '~')) { |
| 276 | if (((l-1) & 3) < 2) |
| 277 | l += 2; |
| 278 | else |
| 279 | l = (l+3) & ~3; |
| 280 | str[l-2] = '~'; |
| 281 | str[l-1] = '0'; |
| 282 | *len = l | h; |
| 283 | return; |
| 284 | } |
| 285 | for (i = l-1; i >= 0; i--) { |
| 286 | if (isdigit(str[i])) { |
| 287 | if (str[i] == '9') |
| 288 | str[i] = '0'; |
| 289 | else { |
| 290 | str[i]++; |
| 291 | return; |
| 292 | } |
| 293 | continue; |
| 294 | } |
| 295 | if (i == 1) { |
| 296 | if (str[0] == 'z') |
| 297 | str[0] = 'A'; |
| 298 | else if (str[0] == 'Z') { |
| 299 | str[0] = '~'; |
| 300 | str[1] = '0'; |
| 301 | } else |
| 302 | str[0]++; |
| 303 | } else if (i > 0) { |
| 304 | str[i] = '1'; |
| 305 | str[i-1] = '~'; |
| 306 | } else { |
| 307 | if (str[0] == '~') |
| 308 | str[0] = 'a'; |
| 309 | else |
| 310 | str[0]++; |
| 311 | } |
| 312 | break; |
| 313 | } |
| 314 | } |
| 315 | |
| 316 | static int duplicate_search_and_fix(e2fsck_t ctx, ext2_filsys fs, |
| 317 | ext2_ino_t ino, |
| 318 | struct fill_dir_struct *fd) |
| 319 | { |
| 320 | struct problem_context pctx; |
| 321 | struct hash_entry *ent, *prev; |
| 322 | int i, j; |
| 323 | int fixed = 0; |
| 324 | char new_name[256]; |
| 325 | __u16 new_len; |
| 326 | |
| 327 | clear_problem_context(&pctx); |
| 328 | pctx.ino = ino; |
| 329 | |
| 330 | for (i=1; i < fd->num_array; i++) { |
| 331 | ent = fd->harray + i; |
| 332 | prev = ent - 1; |
| 333 | if (!ent->dir->inode || |
| 334 | ((ent->dir->name_len & 0xFF) != |
| 335 | (prev->dir->name_len & 0xFF)) || |
| 336 | (strncmp(ent->dir->name, prev->dir->name, |
| 337 | ent->dir->name_len & 0xFF))) |
| 338 | continue; |
| 339 | pctx.dirent = ent->dir; |
| 340 | if ((ent->dir->inode == prev->dir->inode) && |
| 341 | fix_problem(ctx, PR_2_DUPLICATE_DIRENT, &pctx)) { |
| 342 | e2fsck_adjust_inode_count(ctx, ent->dir->inode, -1); |
| 343 | ent->dir->inode = 0; |
| 344 | fixed++; |
| 345 | continue; |
| 346 | } |
| 347 | memcpy(new_name, ent->dir->name, ent->dir->name_len & 0xFF); |
| 348 | new_len = ent->dir->name_len; |
| 349 | mutate_name(new_name, &new_len); |
| 350 | for (j=0; j < fd->num_array; j++) { |
| 351 | if ((i==j) || |
| 352 | ((ent->dir->name_len & 0xFF) != |
| 353 | (fd->harray[j].dir->name_len & 0xFF)) || |
| 354 | (strncmp(new_name, fd->harray[j].dir->name, |
| 355 | new_len & 0xFF))) |
| 356 | continue; |
| 357 | mutate_name(new_name, &new_len); |
| 358 | |
| 359 | j = -1; |
| 360 | } |
| 361 | new_name[new_len & 0xFF] = 0; |
| 362 | pctx.str = new_name; |
| 363 | if (fix_problem(ctx, PR_2_NON_UNIQUE_FILE, &pctx)) { |
| 364 | memcpy(ent->dir->name, new_name, new_len & 0xFF); |
| 365 | ent->dir->name_len = new_len; |
| 366 | ext2fs_dirhash(fs->super->s_def_hash_version, |
| 367 | ent->dir->name, |
| 368 | ent->dir->name_len & 0xFF, |
| 369 | fs->super->s_hash_seed, |
| 370 | &ent->hash, &ent->minor_hash); |
| 371 | fixed++; |
| 372 | } |
| 373 | } |
| 374 | return fixed; |
| 375 | } |
| 376 | |
| 377 | |
| 378 | static errcode_t copy_dir_entries(ext2_filsys fs, |
| 379 | struct fill_dir_struct *fd, |
| 380 | struct out_dir *outdir) |
| 381 | { |
| 382 | errcode_t retval; |
| 383 | char *block_start; |
| 384 | struct hash_entry *ent; |
| 385 | struct ext2_dir_entry *dirent; |
| 386 | int i, rec_len, left; |
| 387 | ext2_dirhash_t prev_hash; |
| 388 | int offset; |
| 389 | |
| 390 | outdir->max = 0; |
| 391 | retval = alloc_size_dir(fs, outdir, |
| 392 | (fd->dir_size / fs->blocksize) + 2); |
| 393 | if (retval) |
| 394 | return retval; |
| 395 | outdir->num = fd->compress ? 0 : 1; |
| 396 | offset = 0; |
| 397 | outdir->hashes[0] = 0; |
| 398 | prev_hash = 1; |
| 399 | if ((retval = get_next_block(fs, outdir, &block_start))) |
| 400 | return retval; |
| 401 | dirent = (struct ext2_dir_entry *) block_start; |
| 402 | left = fs->blocksize; |
| 403 | for (i=0; i < fd->num_array; i++) { |
| 404 | ent = fd->harray + i; |
| 405 | if (ent->dir->inode == 0) |
| 406 | continue; |
| 407 | rec_len = EXT2_DIR_REC_LEN(ent->dir->name_len & 0xFF); |
| 408 | if (rec_len > left) { |
| 409 | if (left) |
| 410 | dirent->rec_len += left; |
| 411 | if ((retval = get_next_block(fs, outdir, |
| 412 | &block_start))) |
| 413 | return retval; |
| 414 | offset = 0; |
| 415 | } |
| 416 | left = fs->blocksize - offset; |
| 417 | dirent = (struct ext2_dir_entry *) (block_start + offset); |
| 418 | if (offset == 0) { |
| 419 | if (ent->hash == prev_hash) |
| 420 | outdir->hashes[outdir->num-1] = ent->hash | 1; |
| 421 | else |
| 422 | outdir->hashes[outdir->num-1] = ent->hash; |
| 423 | } |
| 424 | dirent->inode = ent->dir->inode; |
| 425 | dirent->name_len = ent->dir->name_len; |
| 426 | dirent->rec_len = rec_len; |
| 427 | memcpy(dirent->name, ent->dir->name, dirent->name_len & 0xFF); |
| 428 | offset += rec_len; |
| 429 | left -= rec_len; |
| 430 | if (left < 12) { |
| 431 | dirent->rec_len += left; |
| 432 | offset += left; |
| 433 | left = 0; |
| 434 | } |
| 435 | prev_hash = ent->hash; |
| 436 | } |
| 437 | if (left) |
| 438 | dirent->rec_len += left; |
| 439 | |
| 440 | return 0; |
| 441 | } |
| 442 | |
| 443 | |
| 444 | static struct ext2_dx_root_info *set_root_node(ext2_filsys fs, char *buf, |
| 445 | ext2_ino_t ino, ext2_ino_t parent) |
| 446 | { |
| 447 | struct ext2_dir_entry *dir; |
| 448 | struct ext2_dx_root_info *root; |
| 449 | struct ext2_dx_countlimit *limits; |
| 450 | int filetype = 0; |
| 451 | |
| 452 | if (fs->super->s_feature_incompat & EXT2_FEATURE_INCOMPAT_FILETYPE) |
| 453 | filetype = EXT2_FT_DIR << 8; |
| 454 | |
| 455 | memset(buf, 0, fs->blocksize); |
| 456 | dir = (struct ext2_dir_entry *) buf; |
| 457 | dir->inode = ino; |
| 458 | dir->name[0] = '.'; |
| 459 | dir->name_len = 1 | filetype; |
| 460 | dir->rec_len = 12; |
| 461 | dir = (struct ext2_dir_entry *) (buf + 12); |
| 462 | dir->inode = parent; |
| 463 | dir->name[0] = '.'; |
| 464 | dir->name[1] = '.'; |
| 465 | dir->name_len = 2 | filetype; |
| 466 | dir->rec_len = fs->blocksize - 12; |
| 467 | |
| 468 | root = (struct ext2_dx_root_info *) (buf+24); |
| 469 | root->reserved_zero = 0; |
| 470 | root->hash_version = fs->super->s_def_hash_version; |
| 471 | root->info_length = 8; |
| 472 | root->indirect_levels = 0; |
| 473 | root->unused_flags = 0; |
| 474 | |
| 475 | limits = (struct ext2_dx_countlimit *) (buf+32); |
| 476 | limits->limit = (fs->blocksize - 32) / sizeof(struct ext2_dx_entry); |
| 477 | limits->count = 0; |
| 478 | |
| 479 | return root; |
| 480 | } |
| 481 | |
| 482 | |
| 483 | static struct ext2_dx_entry *set_int_node(ext2_filsys fs, char *buf) |
| 484 | { |
| 485 | struct ext2_dir_entry *dir; |
| 486 | struct ext2_dx_countlimit *limits; |
| 487 | |
| 488 | memset(buf, 0, fs->blocksize); |
| 489 | dir = (struct ext2_dir_entry *) buf; |
| 490 | dir->inode = 0; |
| 491 | dir->rec_len = fs->blocksize; |
| 492 | |
| 493 | limits = (struct ext2_dx_countlimit *) (buf+8); |
| 494 | limits->limit = (fs->blocksize - 8) / sizeof(struct ext2_dx_entry); |
| 495 | limits->count = 0; |
| 496 | |
| 497 | return (struct ext2_dx_entry *) limits; |
| 498 | } |
| 499 | |
| 500 | /* |
| 501 | * This function takes the leaf nodes which have been written in |
| 502 | * outdir, and populates the root node and any necessary interior nodes. |
| 503 | */ |
| 504 | static errcode_t calculate_tree(ext2_filsys fs, |
| 505 | struct out_dir *outdir, |
| 506 | ext2_ino_t ino, |
| 507 | ext2_ino_t parent) |
| 508 | { |
| 509 | struct ext2_dx_root_info *root_info; |
| 510 | struct ext2_dx_entry *root, *dx_ent = 0; |
| 511 | struct ext2_dx_countlimit *root_limit, *limit; |
| 512 | errcode_t retval; |
| 513 | char * block_start; |
| 514 | int i, c1, c2, nblks; |
| 515 | int limit_offset, root_offset; |
| 516 | |
| 517 | root_info = set_root_node(fs, outdir->buf, ino, parent); |
| 518 | root_offset = limit_offset = ((char *) root_info - outdir->buf) + |
| 519 | root_info->info_length; |
| 520 | root_limit = (struct ext2_dx_countlimit *) (outdir->buf + limit_offset); |
| 521 | c1 = root_limit->limit; |
| 522 | nblks = outdir->num; |
| 523 | |
| 524 | /* Write out the pointer blocks */ |
| 525 | if (nblks-1 <= c1) { |
| 526 | /* Just write out the root block, and we're done */ |
| 527 | root = (struct ext2_dx_entry *) (outdir->buf + root_offset); |
| 528 | for (i=1; i < nblks; i++) { |
| 529 | root->block = ext2fs_cpu_to_le32(i); |
| 530 | if (i != 1) |
| 531 | root->hash = |
| 532 | ext2fs_cpu_to_le32(outdir->hashes[i]); |
| 533 | root++; |
| 534 | c1--; |
| 535 | } |
| 536 | } else { |
| 537 | c2 = 0; |
| 538 | limit = 0; |
| 539 | root_info->indirect_levels = 1; |
| 540 | for (i=1; i < nblks; i++) { |
| 541 | if (c1 == 0) |
| 542 | return ENOSPC; |
| 543 | if (c2 == 0) { |
| 544 | if (limit) |
| 545 | limit->limit = limit->count = |
| 546 | ext2fs_cpu_to_le16(limit->limit); |
| 547 | root = (struct ext2_dx_entry *) |
| 548 | (outdir->buf + root_offset); |
| 549 | root->block = ext2fs_cpu_to_le32(outdir->num); |
| 550 | if (i != 1) |
| 551 | root->hash = |
| 552 | ext2fs_cpu_to_le32(outdir->hashes[i]); |
| 553 | if ((retval = get_next_block(fs, outdir, |
| 554 | &block_start))) |
| 555 | return retval; |
| 556 | dx_ent = set_int_node(fs, block_start); |
| 557 | limit = (struct ext2_dx_countlimit *) dx_ent; |
| 558 | c2 = limit->limit; |
| 559 | root_offset += sizeof(struct ext2_dx_entry); |
| 560 | c1--; |
| 561 | } |
| 562 | dx_ent->block = ext2fs_cpu_to_le32(i); |
| 563 | if (c2 != limit->limit) |
| 564 | dx_ent->hash = |
| 565 | ext2fs_cpu_to_le32(outdir->hashes[i]); |
| 566 | dx_ent++; |
| 567 | c2--; |
| 568 | } |
| 569 | limit->count = ext2fs_cpu_to_le16(limit->limit - c2); |
| 570 | limit->limit = ext2fs_cpu_to_le16(limit->limit); |
| 571 | } |
| 572 | root_limit = (struct ext2_dx_countlimit *) (outdir->buf + limit_offset); |
| 573 | root_limit->count = ext2fs_cpu_to_le16(root_limit->limit - c1); |
| 574 | root_limit->limit = ext2fs_cpu_to_le16(root_limit->limit); |
| 575 | |
| 576 | return 0; |
| 577 | } |
| 578 | |
| 579 | struct write_dir_struct { |
| 580 | struct out_dir *outdir; |
| 581 | errcode_t err; |
| 582 | e2fsck_t ctx; |
| 583 | int cleared; |
| 584 | }; |
| 585 | |
| 586 | /* |
| 587 | * Helper function which writes out a directory block. |
| 588 | */ |
| 589 | static int write_dir_block(ext2_filsys fs, |
| 590 | blk_t *block_nr, |
| 591 | e2_blkcnt_t blockcnt, |
| 592 | blk_t ref_block EXT2FS_ATTR((unused)), |
| 593 | int ref_offset EXT2FS_ATTR((unused)), |
| 594 | void *priv_data) |
| 595 | { |
| 596 | struct write_dir_struct *wd = (struct write_dir_struct *) priv_data; |
| 597 | blk_t blk; |
| 598 | char *dir; |
| 599 | |
| 600 | if (*block_nr == 0) |
| 601 | return 0; |
| 602 | if (blockcnt >= wd->outdir->num) { |
| 603 | e2fsck_read_bitmaps(wd->ctx); |
| 604 | blk = *block_nr; |
| 605 | ext2fs_unmark_block_bitmap(wd->ctx->block_found_map, blk); |
| 606 | ext2fs_block_alloc_stats(fs, blk, -1); |
| 607 | *block_nr = 0; |
| 608 | wd->cleared++; |
| 609 | return BLOCK_CHANGED; |
| 610 | } |
| 611 | if (blockcnt < 0) |
| 612 | return 0; |
| 613 | |
| 614 | dir = wd->outdir->buf + (blockcnt * fs->blocksize); |
| 615 | wd->err = ext2fs_write_dir_block(fs, *block_nr, dir); |
| 616 | if (wd->err) |
| 617 | return BLOCK_ABORT; |
| 618 | return 0; |
| 619 | } |
| 620 | |
| 621 | static errcode_t write_directory(e2fsck_t ctx, ext2_filsys fs, |
| 622 | struct out_dir *outdir, |
| 623 | ext2_ino_t ino, int compress) |
| 624 | { |
| 625 | struct write_dir_struct wd; |
| 626 | errcode_t retval; |
| 627 | struct ext2_inode inode; |
| 628 | |
| 629 | retval = e2fsck_expand_directory(ctx, ino, -1, outdir->num); |
| 630 | if (retval) |
| 631 | return retval; |
| 632 | |
| 633 | wd.outdir = outdir; |
| 634 | wd.err = 0; |
| 635 | wd.ctx = ctx; |
| 636 | wd.cleared = 0; |
| 637 | |
| 638 | retval = ext2fs_block_iterate2(fs, ino, 0, 0, |
| 639 | write_dir_block, &wd); |
| 640 | if (retval) |
| 641 | return retval; |
| 642 | if (wd.err) |
| 643 | return wd.err; |
| 644 | |
| 645 | e2fsck_read_inode(ctx, ino, &inode, "rehash_dir"); |
| 646 | if (compress) |
| 647 | inode.i_flags &= ~EXT2_INDEX_FL; |
| 648 | else |
| 649 | inode.i_flags |= EXT2_INDEX_FL; |
| 650 | inode.i_size = outdir->num * fs->blocksize; |
| 651 | inode.i_blocks -= (fs->blocksize / 512) * wd.cleared; |
| 652 | e2fsck_write_inode(ctx, ino, &inode, "rehash_dir"); |
| 653 | |
| 654 | return 0; |
| 655 | } |
| 656 | |
| 657 | errcode_t e2fsck_rehash_dir(e2fsck_t ctx, ext2_ino_t ino) |
| 658 | { |
| 659 | ext2_filsys fs = ctx->fs; |
| 660 | errcode_t retval; |
| 661 | struct ext2_inode inode; |
| 662 | char *dir_buf = 0; |
| 663 | struct fill_dir_struct fd; |
| 664 | struct out_dir outdir; |
| 665 | |
| 666 | outdir.max = outdir.num = 0; |
| 667 | outdir.buf = 0; |
| 668 | outdir.hashes = 0; |
| 669 | e2fsck_read_inode(ctx, ino, &inode, "rehash_dir"); |
| 670 | |
| 671 | retval = ENOMEM; |
| 672 | fd.harray = 0; |
| 673 | dir_buf = malloc(inode.i_size); |
| 674 | if (!dir_buf) |
| 675 | goto errout; |
| 676 | |
| 677 | fd.max_array = inode.i_size / 32; |
| 678 | fd.num_array = 0; |
| 679 | fd.harray = malloc(fd.max_array * sizeof(struct hash_entry)); |
| 680 | if (!fd.harray) |
| 681 | goto errout; |
| 682 | |
| 683 | fd.ctx = ctx; |
| 684 | fd.buf = dir_buf; |
| 685 | fd.inode = &inode; |
| 686 | fd.err = 0; |
| 687 | fd.dir_size = 0; |
| 688 | fd.compress = 0; |
| 689 | if (!(fs->super->s_feature_compat & EXT2_FEATURE_COMPAT_DIR_INDEX) || |
| 690 | (inode.i_size / fs->blocksize) < 2) |
| 691 | fd.compress = 1; |
| 692 | fd.parent = 0; |
| 693 | |
| 694 | /* Read in the entire directory into memory */ |
| 695 | retval = ext2fs_block_iterate2(fs, ino, 0, 0, |
| 696 | fill_dir_block, &fd); |
| 697 | if (fd.err) { |
| 698 | retval = fd.err; |
| 699 | goto errout; |
| 700 | } |
| 701 | |
| 702 | #if 0 |
| 703 | printf("%d entries (%d bytes) found in inode %d\n", |
| 704 | fd.num_array, fd.dir_size, ino); |
| 705 | #endif |
| 706 | |
| 707 | /* Sort the list */ |
| 708 | resort: |
| 709 | if (fd.compress) |
| 710 | qsort(fd.harray+2, fd.num_array-2, |
| 711 | sizeof(struct hash_entry), name_cmp); |
| 712 | else |
| 713 | qsort(fd.harray, fd.num_array, |
| 714 | sizeof(struct hash_entry), hash_cmp); |
| 715 | |
| 716 | /* |
| 717 | * Look for duplicates |
| 718 | */ |
| 719 | if (duplicate_search_and_fix(ctx, fs, ino, &fd)) |
| 720 | goto resort; |
| 721 | |
| 722 | if (ctx->options & E2F_OPT_NO) { |
| 723 | retval = 0; |
| 724 | goto errout; |
| 725 | } |
| 726 | |
| 727 | /* |
| 728 | * Copy the directory entries. In a htree directory these |
| 729 | * will become the leaf nodes. |
| 730 | */ |
| 731 | retval = copy_dir_entries(fs, &fd, &outdir); |
| 732 | if (retval) |
| 733 | goto errout; |
| 734 | |
| 735 | free(dir_buf); dir_buf = 0; |
| 736 | |
| 737 | if (!fd.compress) { |
| 738 | /* Calculate the interior nodes */ |
| 739 | retval = calculate_tree(fs, &outdir, ino, fd.parent); |
| 740 | if (retval) |
| 741 | goto errout; |
| 742 | } |
| 743 | |
| 744 | retval = write_directory(ctx, fs, &outdir, ino, fd.compress); |
| 745 | if (retval) |
| 746 | goto errout; |
| 747 | |
| 748 | errout: |
| 749 | if (dir_buf) |
| 750 | free(dir_buf); |
| 751 | if (fd.harray) |
| 752 | free(fd.harray); |
| 753 | |
| 754 | free_out_dir(&outdir); |
| 755 | return retval; |
| 756 | } |
| 757 | |
| 758 | void e2fsck_rehash_directories(e2fsck_t ctx) |
| 759 | { |
| 760 | struct problem_context pctx; |
| 761 | #ifdef RESOURCE_TRACK |
| 762 | struct resource_track rtrack; |
| 763 | #endif |
| 764 | struct dir_info *dir; |
| 765 | ext2_u32_iterate iter; |
| 766 | ext2_ino_t ino; |
| 767 | errcode_t retval; |
| 768 | int i, cur, max, all_dirs, dir_index, first = 1; |
| 769 | |
| 770 | #ifdef RESOURCE_TRACK |
| 771 | init_resource_track(&rtrack); |
| 772 | #endif |
| 773 | |
| 774 | all_dirs = ctx->options & E2F_OPT_COMPRESS_DIRS; |
| 775 | |
| 776 | if (!ctx->dirs_to_hash && !all_dirs) |
| 777 | return; |
| 778 | |
| 779 | e2fsck_get_lost_and_found(ctx, 0); |
| 780 | |
| 781 | clear_problem_context(&pctx); |
| 782 | |
| 783 | dir_index = ctx->fs->super->s_feature_compat & EXT2_FEATURE_COMPAT_DIR_INDEX; |
| 784 | cur = 0; |
| 785 | if (all_dirs) { |
| 786 | i = 0; |
| 787 | max = e2fsck_get_num_dirinfo(ctx); |
| 788 | } else { |
| 789 | retval = ext2fs_u32_list_iterate_begin(ctx->dirs_to_hash, |
| 790 | &iter); |
| 791 | if (retval) { |
| 792 | pctx.errcode = retval; |
| 793 | fix_problem(ctx, PR_3A_OPTIMIZE_ITER, &pctx); |
| 794 | return; |
| 795 | } |
| 796 | max = ext2fs_u32_list_count(ctx->dirs_to_hash); |
| 797 | } |
| 798 | while (1) { |
| 799 | if (all_dirs) { |
| 800 | if ((dir = e2fsck_dir_info_iter(ctx, &i)) == 0) |
| 801 | break; |
| 802 | ino = dir->ino; |
| 803 | } else { |
| 804 | if (!ext2fs_u32_list_iterate(iter, &ino)) |
| 805 | break; |
| 806 | } |
| 807 | if (ino == ctx->lost_and_found) |
| 808 | continue; |
| 809 | pctx.dir = ino; |
| 810 | if (first) { |
| 811 | fix_problem(ctx, PR_3A_PASS_HEADER, &pctx); |
| 812 | first = 0; |
| 813 | } |
| 814 | #if 0 |
| 815 | fix_problem(ctx, PR_3A_OPTIMIZE_DIR, &pctx); |
| 816 | #endif |
| 817 | pctx.errcode = e2fsck_rehash_dir(ctx, ino); |
| 818 | if (pctx.errcode) { |
| 819 | end_problem_latch(ctx, PR_LATCH_OPTIMIZE_DIR); |
| 820 | fix_problem(ctx, PR_3A_OPTIMIZE_DIR_ERR, &pctx); |
| 821 | } |
| 822 | if (ctx->progress && !ctx->progress_fd) |
| 823 | e2fsck_simple_progress(ctx, "Rebuilding directory", |
| 824 | 100.0 * (float) (++cur) / (float) max, ino); |
| 825 | } |
| 826 | end_problem_latch(ctx, PR_LATCH_OPTIMIZE_DIR); |
| 827 | if (!all_dirs) |
| 828 | ext2fs_u32_list_iterate_end(iter); |
| 829 | |
| 830 | if (ctx->dirs_to_hash) |
| 831 | ext2fs_u32_list_free(ctx->dirs_to_hash); |
| 832 | ctx->dirs_to_hash = 0; |
| 833 | |
| 834 | #ifdef RESOURCE_TRACK |
| 835 | if (ctx->options & E2F_OPT_TIME2) { |
| 836 | e2fsck_clear_progbar(ctx); |
| 837 | print_resource_track("Pass 3A", &rtrack); |
| 838 | } |
| 839 | #endif |
| 840 | } |