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