2 * fs/cifs/cifs_unicode.c
4 * Copyright (c) International Business Machines Corp., 2000,2009
5 * Modified by Steve French (sfrench@us.ibm.com)
7 * This program is free software; you can redistribute it and/or modify
8 * it under the terms of the GNU General Public License as published by
9 * the Free Software Foundation; either version 2 of the License, or
10 * (at your option) any later version.
12 * This program is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
15 * the GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with this program; if not, write to the Free Software
19 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
22 #include <linux/slab.h>
23 #include "cifs_fs_sb.h"
24 #include "cifs_unicode.h"
25 #include "cifs_uniupr.h"
28 #include "cifs_debug.h"
31 * cifs_utf16_bytes - how long will a string be after conversion?
32 * @utf16 - pointer to input string
33 * @maxbytes - don't go past this many bytes of input string
34 * @codepage - destination codepage
36 * Walk a utf16le string and return the number of bytes that the string will
37 * be after being converted to the given charset, not including any null
38 * termination required. Don't walk past maxbytes in the source buffer.
41 cifs_utf16_bytes(const __le16 *from, int maxbytes,
42 const struct nls_table *codepage)
45 int charlen, outlen = 0;
46 int maxwords = maxbytes / 2;
47 char tmp[NLS_MAX_CHARSET_SIZE];
50 for (i = 0; i < maxwords; i++) {
51 ftmp = get_unaligned_le16(&from[i]);
55 charlen = codepage->uni2char(ftmp, tmp, NLS_MAX_CHARSET_SIZE);
65 int cifs_remap(struct cifs_sb_info *cifs_sb)
69 if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SFM_CHR)
70 map_type = SFM_MAP_UNI_RSVD;
71 else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR)
72 map_type = SFU_MAP_UNI_RSVD;
74 map_type = NO_MAP_UNI_RSVD;
79 /* Convert character using the SFU - "Services for Unix" remapping range */
81 convert_sfu_char(const __u16 src_char, char *target)
84 * BB: Cannot handle remapping UNI_SLASH until all the calls to
85 * build_path_from_dentry are modified, as they use slash as
113 /* Convert character using the SFM - "Services for Mac" remapping range */
115 convert_sfm_char(const __u16 src_char, char *target)
147 * cifs_mapchar - convert a host-endian char to proper char in codepage
148 * @target - where converted character should be copied
149 * @src_char - 2 byte host-endian source character
150 * @cp - codepage to which character should be converted
151 * @map_type - How should the 7 NTFS/SMB reserved characters be mapped to UCS2?
153 * This function handles the conversion of a single character. It is the
154 * responsibility of the caller to ensure that the target buffer is large
155 * enough to hold the result of the conversion (at least NLS_MAX_CHARSET_SIZE).
158 cifs_mapchar(char *target, const __u16 src_char, const struct nls_table *cp,
163 if ((maptype == SFM_MAP_UNI_RSVD) && convert_sfm_char(src_char, target))
165 else if ((maptype == SFU_MAP_UNI_RSVD) &&
166 convert_sfu_char(src_char, target))
169 /* if character not one of seven in special remap set */
170 len = cp->uni2char(src_char, target, NLS_MAX_CHARSET_SIZE);
179 * cifs_from_utf16 - convert utf16le string to local charset
180 * @to - destination buffer
181 * @from - source buffer
182 * @tolen - destination buffer size (in bytes)
183 * @fromlen - source buffer size (in bytes)
184 * @codepage - codepage to which characters should be converted
185 * @mapchar - should characters be remapped according to the mapchars option?
187 * Convert a little-endian utf16le string (as sent by the server) to a string
188 * in the provided codepage. The tolen and fromlen parameters are to ensure
189 * that the code doesn't walk off of the end of the buffer (which is always
190 * a danger if the alignment of the source buffer is off). The destination
191 * string is always properly null terminated and fits in the destination
192 * buffer. Returns the length of the destination string in bytes (including
195 * Note that some windows versions actually send multiword UTF-16 characters
196 * instead of straight UTF16-2. The linux nls routines however aren't able to
197 * deal with those characters properly. In the event that we get some of
198 * those characters, they won't be translated properly.
201 cifs_from_utf16(char *to, const __le16 *from, int tolen, int fromlen,
202 const struct nls_table *codepage, int map_type)
204 int i, charlen, safelen;
206 int nullsize = nls_nullsize(codepage);
207 int fromwords = fromlen / 2;
208 char tmp[NLS_MAX_CHARSET_SIZE];
212 * because the chars can be of varying widths, we need to take care
213 * not to overflow the destination buffer when we get close to the
214 * end of it. Until we get to this offset, we don't need to check
215 * for overflow however.
217 safelen = tolen - (NLS_MAX_CHARSET_SIZE + nullsize);
219 for (i = 0; i < fromwords; i++) {
220 ftmp = get_unaligned_le16(&from[i]);
225 * check to see if converting this character might make the
226 * conversion bleed into the null terminator
228 if (outlen >= safelen) {
229 charlen = cifs_mapchar(tmp, ftmp, codepage, map_type);
230 if ((outlen + charlen) > (tolen - nullsize))
234 /* put converted char into 'to' buffer */
235 charlen = cifs_mapchar(&to[outlen], ftmp, codepage, map_type);
239 /* properly null-terminate string */
240 for (i = 0; i < nullsize; i++)
247 * NAME: cifs_strtoUTF16()
249 * FUNCTION: Convert character string to unicode string
253 cifs_strtoUTF16(__le16 *to, const char *from, int len,
254 const struct nls_table *codepage)
258 wchar_t wchar_to; /* needed to quiet sparse */
260 /* special case for utf8 to handle no plane0 chars */
261 if (!strcmp(codepage->charset, "utf8")) {
263 * convert utf8 -> utf16, we assume we have enough space
264 * as caller should have assumed conversion does not overflow
265 * in destination len is length in wchar_t units (16bits)
267 i = utf8s_to_utf16s(from, len, UTF16_LITTLE_ENDIAN,
268 (wchar_t *) to, len);
270 /* if success terminate and exit */
274 * if fails fall back to UCS encoding as this
275 * function should not return negative values
276 * currently can fail only if source contains
277 * invalid encoded characters
281 for (i = 0; len && *from; i++, from += charlen, len -= charlen) {
282 charlen = codepage->char2uni(from, len, &wchar_to);
284 cifs_dbg(VFS, "strtoUTF16: char2uni of 0x%x returned %d\n",
286 /* A question mark */
290 put_unaligned_le16(wchar_to, &to[i]);
294 put_unaligned_le16(0, &to[i]);
299 * cifs_strndup_from_utf16 - copy a string from wire format to the local
301 * @src - source string
302 * @maxlen - don't walk past this many bytes in the source string
303 * @is_unicode - is this a unicode string?
304 * @codepage - destination codepage
306 * Take a string given by the server, convert it to the local codepage and
307 * put it in a new buffer. Returns a pointer to the new string or NULL on
311 cifs_strndup_from_utf16(const char *src, const int maxlen,
312 const bool is_unicode, const struct nls_table *codepage)
318 len = cifs_utf16_bytes((__le16 *) src, maxlen, codepage);
319 len += nls_nullsize(codepage);
320 dst = kmalloc(len, GFP_KERNEL);
323 cifs_from_utf16(dst, (__le16 *) src, len, maxlen, codepage,
326 len = strnlen(src, maxlen);
328 dst = kmalloc(len, GFP_KERNEL);
331 strlcpy(dst, src, len);
337 static __le16 convert_to_sfu_char(char src_char)
343 dest_char = cpu_to_le16(UNI_COLON);
346 dest_char = cpu_to_le16(UNI_ASTERISK);
349 dest_char = cpu_to_le16(UNI_QUESTION);
352 dest_char = cpu_to_le16(UNI_LESSTHAN);
355 dest_char = cpu_to_le16(UNI_GRTRTHAN);
358 dest_char = cpu_to_le16(UNI_PIPE);
367 static __le16 convert_to_sfm_char(char src_char)
373 dest_char = cpu_to_le16(SFM_COLON);
376 dest_char = cpu_to_le16(SFM_ASTERISK);
379 dest_char = cpu_to_le16(SFM_QUESTION);
382 dest_char = cpu_to_le16(SFM_LESSTHAN);
385 dest_char = cpu_to_le16(SFM_GRTRTHAN);
388 dest_char = cpu_to_le16(SFM_PIPE);
398 * Convert 16 bit Unicode pathname to wire format from string in current code
399 * page. Conversion may involve remapping up the six characters that are
400 * only legal in POSIX-like OS (if they are present in the string). Path
401 * names are little endian 16 bit Unicode on the wire
404 cifsConvertToUTF16(__le16 *target, const char *source, int srclen,
405 const struct nls_table *cp, int map_chars)
413 if (map_chars == NO_MAP_UNI_RSVD)
414 return cifs_strtoUTF16(target, source, PATH_MAX, cp);
416 for (i = 0; i < srclen; j++) {
417 src_char = source[i];
420 /* check if end of string */
424 /* see if we must remap this char */
425 if (map_chars == SFU_MAP_UNI_RSVD)
426 dst_char = convert_to_sfu_char(src_char);
427 else if (map_chars == SFM_MAP_UNI_RSVD)
428 dst_char = convert_to_sfm_char(src_char);
432 * FIXME: We can not handle remapping backslash (UNI_SLASH)
433 * until all the calls to build_path_from_dentry are modified,
434 * as they use backslash as separator.
437 charlen = cp->char2uni(source + i, srclen - i, &tmp);
438 dst_char = cpu_to_le16(tmp);
441 * if no match, use question mark, which at least in
442 * some cases serves as wild card
445 dst_char = cpu_to_le16(0x003f);
450 * character may take more than one byte in the source string,
451 * but will take exactly two bytes in the target string
454 put_unaligned(dst_char, &target[j]);
458 put_unaligned(0, &target[j]); /* Null terminate target unicode string */
462 #ifdef CONFIG_CIFS_SMB2
464 * cifs_local_to_utf16_bytes - how long will a string be after conversion?
465 * @from - pointer to input string
466 * @maxbytes - don't go past this many bytes of input string
467 * @codepage - source codepage
469 * Walk a string and return the number of bytes that the string will
470 * be after being converted to the given charset, not including any null
471 * termination required. Don't walk past maxbytes in the source buffer.
475 cifs_local_to_utf16_bytes(const char *from, int len,
476 const struct nls_table *codepage)
482 for (i = 0; len && *from; i++, from += charlen, len -= charlen) {
483 charlen = codepage->char2uni(from, len, &wchar_to);
484 /* Failed conversion defaults to a question mark */
488 return 2 * i; /* UTF16 characters are two bytes */
492 * cifs_strndup_to_utf16 - copy a string to wire format from the local codepage
493 * @src - source string
494 * @maxlen - don't walk past this many bytes in the source string
495 * @utf16_len - the length of the allocated string in bytes (including null)
496 * @cp - source codepage
497 * @remap - map special chars
499 * Take a string convert it from the local codepage to UTF16 and
500 * put it in a new buffer. Returns a pointer to the new string or NULL on
504 cifs_strndup_to_utf16(const char *src, const int maxlen, int *utf16_len,
505 const struct nls_table *cp, int remap)
510 len = cifs_local_to_utf16_bytes(src, maxlen, cp);
512 dst = kmalloc(len, GFP_KERNEL);
517 cifsConvertToUTF16(dst, src, strlen(src), cp, remap);
521 #endif /* CONFIG_CIFS_SMB2 */