1 /*- 2 * SPDX-License-Identifier: BSD-4-Clause 3 * 4 * Copyright (C) 1994, 1995, 1997 Wolfgang Solfrank. 5 * Copyright (C) 1994, 1995, 1997 TooLs GmbH. 6 * All rights reserved. 7 * Original code by Paul Popelka (paulp@uts.amdahl.com) (see below). 8 * 9 * Redistribution and use in source and binary forms, with or without 10 * modification, are permitted provided that the following conditions 11 * are met: 12 * 1. Redistributions of source code must retain the above copyright 13 * notice, this list of conditions and the following disclaimer. 14 * 2. Redistributions in binary form must reproduce the above copyright 15 * notice, this list of conditions and the following disclaimer in the 16 * documentation and/or other materials provided with the distribution. 17 * 3. All advertising materials mentioning features or use of this software 18 * must display the following acknowledgement: 19 * This product includes software developed by TooLs GmbH. 20 * 4. The name of TooLs GmbH may not be used to endorse or promote products 21 * derived from this software without specific prior written permission. 22 * 23 * THIS SOFTWARE IS PROVIDED BY TOOLS GMBH ``AS IS'' AND ANY EXPRESS OR 24 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 25 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 26 * IN NO EVENT SHALL TOOLS GMBH BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 27 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 28 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; 29 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, 30 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR 31 * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF 32 * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 33 */ 34 /*- 35 * Written by Paul Popelka (paulp@uts.amdahl.com) 36 * 37 * You can do anything you want with this software, just don't say you wrote 38 * it, and don't remove this notice. 39 * 40 * This software is provided "as is". 41 * 42 * The author supplies this software to be publicly redistributed on the 43 * understanding that the author is not responsible for the correct 44 * functioning of this software in any circumstances and is not liable for 45 * any damages caused by this software. 46 * 47 * October 1992 48 */ 49 50 #include <sys/cdefs.h> 51 __FBSDID("$FreeBSD$"); 52 53 #include <sys/param.h> 54 #include <sys/endian.h> 55 56 #include <dirent.h> 57 #include <stdio.h> 58 #include <string.h> 59 60 #include <vfs/msdosfs/bpb.h> 61 #include "msdos/direntry.h" 62 #include <vfs/msdosfs/msdosfsmount.h> 63 64 #include "makefs.h" 65 #include "msdos.h" 66 67 static int char8ucs2str(const uint8_t *in, int n, uint16_t *out, int m); 68 static void ucs2pad(uint16_t *buf, int len, int size); 69 static int char8match(uint16_t *w1, uint16_t *w2, int n); 70 71 static const u_char unix2dos[256] = { 72 0, 0, 0, 0, 0, 0, 0, 0, /* 00-07 */ 73 0, 0, 0, 0, 0, 0, 0, 0, /* 08-0f */ 74 0, 0, 0, 0, 0, 0, 0, 0, /* 10-17 */ 75 0, 0, 0, 0, 0, 0, 0, 0, /* 18-1f */ 76 0, '!', 0, '#', '$', '%', '&', '\'', /* 20-27 */ 77 '(', ')', 0, '+', 0, '-', 0, 0, /* 28-2f */ 78 '0', '1', '2', '3', '4', '5', '6', '7', /* 30-37 */ 79 '8', '9', 0, 0, 0, 0, 0, 0, /* 38-3f */ 80 '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G', /* 40-47 */ 81 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', /* 48-4f */ 82 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', /* 50-57 */ 83 'X', 'Y', 'Z', 0, 0, 0, '^', '_', /* 58-5f */ 84 '`', 'A', 'B', 'C', 'D', 'E', 'F', 'G', /* 60-67 */ 85 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', /* 68-6f */ 86 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', /* 70-77 */ 87 'X', 'Y', 'Z', '{', 0, '}', '~', 0, /* 78-7f */ 88 0, 0, 0, 0, 0, 0, 0, 0, /* 80-87 */ 89 0, 0, 0, 0, 0, 0, 0, 0, /* 88-8f */ 90 0, 0, 0, 0, 0, 0, 0, 0, /* 90-97 */ 91 0, 0, 0, 0, 0, 0, 0, 0, /* 98-9f */ 92 0, 0xad, 0xbd, 0x9c, 0xcf, 0xbe, 0xdd, 0xf5, /* a0-a7 */ 93 0xf9, 0xb8, 0xa6, 0xae, 0xaa, 0xf0, 0xa9, 0xee, /* a8-af */ 94 0xf8, 0xf1, 0xfd, 0xfc, 0xef, 0xe6, 0xf4, 0xfa, /* b0-b7 */ 95 0xf7, 0xfb, 0xa7, 0xaf, 0xac, 0xab, 0xf3, 0xa8, /* b8-bf */ 96 0xb7, 0xb5, 0xb6, 0xc7, 0x8e, 0x8f, 0x92, 0x80, /* c0-c7 */ 97 0xd4, 0x90, 0xd2, 0xd3, 0xde, 0xd6, 0xd7, 0xd8, /* c8-cf */ 98 0xd1, 0xa5, 0xe3, 0xe0, 0xe2, 0xe5, 0x99, 0x9e, /* d0-d7 */ 99 0x9d, 0xeb, 0xe9, 0xea, 0x9a, 0xed, 0xe8, 0xe1, /* d8-df */ 100 0xb7, 0xb5, 0xb6, 0xc7, 0x8e, 0x8f, 0x92, 0x80, /* e0-e7 */ 101 0xd4, 0x90, 0xd2, 0xd3, 0xde, 0xd6, 0xd7, 0xd8, /* e8-ef */ 102 0xd1, 0xa5, 0xe3, 0xe0, 0xe2, 0xe5, 0x99, 0xf6, /* f0-f7 */ 103 0x9d, 0xeb, 0xe9, 0xea, 0x9a, 0xed, 0xe8, 0x98, /* f8-ff */ 104 }; 105 106 static const u_char u2l[256] = { 107 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, /* 00-07 */ 108 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, /* 08-0f */ 109 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, /* 10-17 */ 110 0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f, /* 18-1f */ 111 ' ', '!', '"', '#', '$', '%', '&', '\'', /* 20-27 */ 112 '(', ')', '*', '+', ',', '-', '.', '/', /* 28-2f */ 113 '0', '1', '2', '3', '4', '5', '6', '7', /* 30-37 */ 114 '8', '9', ':', ';', '<', '=', '>', '?', /* 38-3f */ 115 '@', 'a', 'b', 'c', 'd', 'e', 'f', 'g', /* 40-47 */ 116 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', /* 48-4f */ 117 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', /* 50-57 */ 118 'x', 'y', 'z', '[', '\\', ']', '^', '_', /* 58-5f */ 119 '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g', /* 60-67 */ 120 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', /* 68-6f */ 121 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', /* 70-77 */ 122 'x', 'y', 'z', '{', '|', '}', '~', 0x7f, /* 78-7f */ 123 0x80, 0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, /* 80-87 */ 124 0x88, 0x89, 0x8a, 0x8b, 0x8c, 0x8d, 0x8e, 0x8f, /* 88-8f */ 125 0x90, 0x91, 0x92, 0x93, 0x94, 0x95, 0x96, 0x97, /* 90-97 */ 126 0x98, 0x99, 0x9a, 0x9b, 0x9c, 0x9d, 0x9e, 0x9f, /* 98-9f */ 127 0xa0, 0xa1, 0xa2, 0xa3, 0xa4, 0xa5, 0xa6, 0xa7, /* a0-a7 */ 128 0xa8, 0xa9, 0xaa, 0xab, 0xac, 0xad, 0xae, 0xaf, /* a8-af */ 129 0xb0, 0xb1, 0xb2, 0xb3, 0xb4, 0xb5, 0xb6, 0xb7, /* b0-b7 */ 130 0xb8, 0xb9, 0xba, 0xbb, 0xbc, 0xbd, 0xbe, 0xbf, /* b8-bf */ 131 0xe0, 0xe1, 0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, /* c0-c7 */ 132 0xe8, 0xe9, 0xea, 0xeb, 0xec, 0xed, 0xee, 0xef, /* c8-cf */ 133 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xd7, /* d0-d7 */ 134 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xdf, /* d8-df */ 135 0xe0, 0xe1, 0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, /* e0-e7 */ 136 0xe8, 0xe9, 0xea, 0xeb, 0xec, 0xed, 0xee, 0xef, /* e8-ef */ 137 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, /* f0-f7 */ 138 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff, /* f8-ff */ 139 }; 140 141 /* 142 * Determine the number of slots necessary for Win95 names 143 */ 144 int 145 winSlotCnt(const u_char *un, size_t unlen) 146 { 147 const u_char *cp; 148 149 /* 150 * Drop trailing blanks and dots 151 */ 152 for (cp = un + unlen; unlen > 0; unlen--) 153 if (*--cp != ' ' && *cp != '.') 154 break; 155 156 return howmany(unlen, WIN_CHARS); 157 } 158 159 /* 160 * Compare our filename to the one in the Win95 entry 161 * Returns the checksum or -1 if no match 162 */ 163 int 164 winChkName(const u_char *un, size_t unlen, struct winentry *wep, int chksum) 165 { 166 uint16_t wn[WIN_MAXLEN], *p; 167 uint16_t buf[WIN_CHARS]; 168 int i, len; 169 170 /* 171 * First compare checksums 172 */ 173 if (wep->weCnt & WIN_LAST) 174 chksum = wep->weChksum; 175 else if (chksum != wep->weChksum) 176 chksum = -1; 177 if (chksum == -1) 178 return -1; 179 180 /* 181 * Offset of this entry 182 */ 183 i = ((wep->weCnt & WIN_CNT) - 1) * WIN_CHARS; 184 185 /* 186 * Translate UNIX name to ucs-2 187 */ 188 len = char8ucs2str(un, unlen, wn, WIN_MAXLEN); 189 ucs2pad(wn, len, WIN_MAXLEN); 190 191 if (i >= len + 1) 192 return -1; 193 if ((wep->weCnt & WIN_LAST) && (len - i > WIN_CHARS)) 194 return -1; 195 196 /* 197 * Fetch name segment from directory entry 198 */ 199 p = &buf[0]; 200 memcpy(p, wep->wePart1, sizeof(wep->wePart1)); 201 p += sizeof(wep->wePart1) / sizeof(*p); 202 memcpy(p, wep->wePart2, sizeof(wep->wePart2)); 203 p += sizeof(wep->wePart2) / sizeof(*p); 204 memcpy(p, wep->wePart3, sizeof(wep->wePart3)); 205 206 /* 207 * And compare name segment 208 */ 209 if (!(char8match(&wn[i], buf, WIN_CHARS))) 210 return -1; 211 212 return chksum; 213 } 214 215 /* 216 * Compute the checksum of a DOS filename for Win95 use 217 */ 218 uint8_t 219 winChksum(uint8_t *name) 220 { 221 int i; 222 uint8_t s; 223 224 for (s = 0, i = 11; --i >= 0; s += *name++) 225 s = (s << 7) | (s >> 1); 226 return s; 227 } 228 229 /* 230 * Create a Win95 long name directory entry 231 * Note: assumes that the filename is valid, 232 * i.e. doesn't consist solely of blanks and dots 233 */ 234 int 235 unix2winfn(const u_char *un, size_t unlen, struct winentry *wep, int cnt, 236 int chksum) 237 { 238 uint16_t wn[WIN_MAXLEN], *p; 239 int i, len; 240 const u_char *cp; 241 242 /* 243 * Drop trailing blanks and dots 244 */ 245 for (cp = un + unlen; unlen > 0; unlen--) 246 if (*--cp != ' ' && *cp != '.') 247 break; 248 249 /* 250 * Offset of this entry 251 */ 252 i = (cnt - 1) * WIN_CHARS; 253 254 /* 255 * Translate UNIX name to ucs-2 256 */ 257 len = char8ucs2str(un, unlen, wn, WIN_MAXLEN); 258 ucs2pad(wn, len, WIN_MAXLEN); 259 260 /* 261 * Initialize winentry to some useful default 262 */ 263 memset(wep, 0xff, sizeof(*wep)); 264 wep->weCnt = cnt; 265 wep->weAttributes = ATTR_WIN95; 266 wep->weReserved1 = 0; 267 wep->weChksum = chksum; 268 wep->weReserved2 = 0; 269 270 /* 271 * Store name segment into directory entry 272 */ 273 p = &wn[i]; 274 memcpy(wep->wePart1, p, sizeof(wep->wePart1)); 275 p += sizeof(wep->wePart1) / sizeof(*p); 276 memcpy(wep->wePart2, p, sizeof(wep->wePart2)); 277 p += sizeof(wep->wePart2) / sizeof(*p); 278 memcpy(wep->wePart3, p, sizeof(wep->wePart3)); 279 280 if (len > i + WIN_CHARS) 281 return 1; 282 283 wep->weCnt |= WIN_LAST; 284 return 0; 285 } 286 287 /* 288 * Convert a unix filename to a DOS filename according to Win95 rules. 289 * If applicable and gen is not 0, it is inserted into the converted 290 * filename as a generation number. 291 * Returns 292 * 0 if name couldn't be converted 293 * 1 if the converted name is the same as the original 294 * (no long filename entry necessary for Win95) 295 * 2 if conversion was successful 296 * 3 if conversion was successful and generation number was inserted 297 */ 298 int 299 unix2dosfn(const u_char *un, u_char dn[12], size_t unlen, u_int gen) 300 { 301 int i, j, l; 302 int conv = 1; 303 const u_char *cp, *dp, *dp1; 304 u_char gentext[6], *wcp; 305 int shortlen; 306 307 /* 308 * Fill the dos filename string with blanks. These are DOS's pad 309 * characters. 310 */ 311 for (i = 0; i < 11; i++) 312 dn[i] = ' '; 313 dn[11] = 0; 314 315 /* 316 * The filenames "." and ".." are handled specially, since they 317 * don't follow dos filename rules. 318 */ 319 if (un[0] == '.' && unlen == 1) { 320 dn[0] = '.'; 321 return gen <= 1; 322 } 323 if (un[0] == '.' && un[1] == '.' && unlen == 2) { 324 dn[0] = '.'; 325 dn[1] = '.'; 326 return gen <= 1; 327 } 328 329 /* 330 * Filenames with only blanks and dots are not allowed! 331 */ 332 for (cp = un, i = unlen; --i >= 0; cp++) 333 if (*cp != ' ' && *cp != '.') 334 break; 335 if (i < 0) 336 return 0; 337 338 /* 339 * Now find the extension 340 * Note: dot as first char doesn't start extension 341 * and trailing dots and blanks are ignored 342 */ 343 dp = dp1 = 0; 344 for (cp = un + 1, i = unlen - 1; --i >= 0;) { 345 switch (*cp++) { 346 case '.': 347 if (!dp1) 348 dp1 = cp; 349 break; 350 case ' ': 351 break; 352 default: 353 if (dp1) 354 dp = dp1; 355 dp1 = 0; 356 break; 357 } 358 } 359 360 /* 361 * Now convert it 362 */ 363 if (dp) { 364 if (dp1) 365 l = dp1 - dp; 366 else 367 l = unlen - (dp - un); 368 for (i = 0, j = 8; i < l && j < 11; i++, j++) { 369 if (dp[i] != (dn[j] = unix2dos[dp[i]]) 370 && conv != 3) 371 conv = 2; 372 if (!dn[j]) { 373 conv = 3; 374 dn[j--] = ' '; 375 } 376 } 377 if (i < l) 378 conv = 3; 379 dp--; 380 } else { 381 for (dp = cp; *--dp == ' ' || *dp == '.';); 382 dp++; 383 } 384 385 shortlen = (dp - un) <= 8; 386 387 /* 388 * Now convert the rest of the name 389 */ 390 for (i = j = 0; un < dp && j < 8; i++, j++, un++) { 391 if ((*un == ' ') && shortlen) 392 dn[j] = ' '; 393 else 394 dn[j] = unix2dos[*un]; 395 if ((*un != dn[j]) 396 && conv != 3) 397 conv = 2; 398 if (!dn[j]) { 399 conv = 3; 400 dn[j--] = ' '; 401 } 402 } 403 if (un < dp) 404 conv = 3; 405 /* 406 * If we didn't have any chars in filename, 407 * generate a default 408 */ 409 if (!j) 410 dn[0] = '_'; 411 412 /* 413 * The first character cannot be E5, 414 * because that means a deleted entry 415 */ 416 if (dn[0] == 0xe5) 417 dn[0] = SLOT_E5; 418 419 /* 420 * If there wasn't any char dropped, 421 * there is no place for generation numbers 422 */ 423 if (conv != 3) { 424 if (gen > 1) 425 return 0; 426 return conv; 427 } 428 429 /* 430 * Now insert the generation number into the filename part 431 */ 432 for (wcp = gentext + sizeof(gentext); wcp > gentext && gen; gen /= 10) 433 *--wcp = gen % 10 + '0'; 434 if (gen) 435 return 0; 436 for (i = 8; dn[--i] == ' ';); 437 i++; 438 if (gentext + sizeof(gentext) - wcp + 1 > 8 - i) 439 i = 8 - (gentext + sizeof(gentext) - wcp + 1); 440 dn[i++] = '~'; 441 while (wcp < gentext + sizeof(gentext)) 442 dn[i++] = *wcp++; 443 return 3; 444 } 445 446 /* 447 * Convert 8bit character string into UCS-2 string 448 * return total number of output chacters 449 */ 450 static int 451 char8ucs2str(const uint8_t *in, int n, uint16_t *out, int m) 452 { 453 uint16_t *p; 454 455 p = out; 456 while (n > 0 && in[0] != 0) { 457 if (m < 1) 458 break; 459 if (p) 460 p[0] = htole16(in[0]); 461 p += 1; 462 m -= 1; 463 in += 1; 464 n -= 1; 465 } 466 467 return p - out; 468 } 469 470 static void 471 ucs2pad(uint16_t *buf, int len, int size) 472 { 473 474 if (len < size-1) 475 buf[len++] = 0x0000; 476 while (len < size) 477 buf[len++] = 0xffff; 478 } 479 480 /* 481 * Compare two 8bit char conversions case-insensitive 482 * 483 * uses the DOS case folding table 484 */ 485 static int 486 char8match(uint16_t *w1, uint16_t *w2, int n) 487 { 488 uint16_t u1, u2; 489 490 while (n > 0) { 491 u1 = le16toh(*w1); 492 u2 = le16toh(*w2); 493 if (u1 == 0 || u2 == 0) 494 return u1 == u2; 495 if (u1 > 255 || u2 > 255) 496 return 0; 497 u1 = u2l[u1 & 0xff]; 498 u2 = u2l[u2 & 0xff]; 499 if (u1 != u2) 500 return 0; 501 ++w1; 502 ++w2; 503 --n; 504 } 505 506 return 1; 507 } 508