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_unicode.h"
24 #include "cifs_uniupr.h"
27 #include "cifs_debug.h"
30 * cifs_utf16_bytes - how long will a string be after conversion?
31 * @utf16 - pointer to input string
32 * @maxbytes - don't go past this many bytes of input string
33 * @codepage - destination codepage
35 * Walk a utf16le string and return the number of bytes that the string will
36 * be after being converted to the given charset, not including any null
37 * termination required. Don't walk past maxbytes in the source buffer.
40 cifs_utf16_bytes(const __le16 *from, int maxbytes,
41 const struct nls_table *codepage)
44 int charlen, outlen = 0;
45 int maxwords = maxbytes / 2;
46 char tmp[NLS_MAX_CHARSET_SIZE];
49 for (i = 0; i < maxwords; i++) {
50 ftmp = get_unaligned_le16(&from[i]);
54 charlen = codepage->uni2char(ftmp, tmp, NLS_MAX_CHARSET_SIZE);
64 /* Convert character using the SFU - "Services for Unix" remapping range */
66 convert_sfu_char(const __u16 src_char, char *target)
69 * BB: Cannot handle remapping UNI_SLASH until all the calls to
70 * build_path_from_dentry are modified, as they use slash as
98 /* Convert character using the SFM - "Services for Mac" remapping range */
100 convert_sfm_char(const __u16 src_char, char *target)
132 * cifs_mapchar - convert a host-endian char to proper char in codepage
133 * @target - where converted character should be copied
134 * @src_char - 2 byte host-endian source character
135 * @cp - codepage to which character should be converted
136 * @map_type - How should the 7 NTFS/SMB reserved characters be mapped to UCS2?
138 * This function handles the conversion of a single character. It is the
139 * responsibility of the caller to ensure that the target buffer is large
140 * enough to hold the result of the conversion (at least NLS_MAX_CHARSET_SIZE).
143 cifs_mapchar(char *target, const __u16 src_char, const struct nls_table *cp,
148 if ((maptype == SFM_MAP_UNI_RSVD) && convert_sfm_char(src_char, target))
150 else if ((maptype == SFU_MAP_UNI_RSVD) &&
151 convert_sfu_char(src_char, target))
154 /* if character not one of seven in special remap set */
155 len = cp->uni2char(src_char, target, NLS_MAX_CHARSET_SIZE);
164 * cifs_from_utf16 - convert utf16le string to local charset
165 * @to - destination buffer
166 * @from - source buffer
167 * @tolen - destination buffer size (in bytes)
168 * @fromlen - source buffer size (in bytes)
169 * @codepage - codepage to which characters should be converted
170 * @mapchar - should characters be remapped according to the mapchars option?
172 * Convert a little-endian utf16le string (as sent by the server) to a string
173 * in the provided codepage. The tolen and fromlen parameters are to ensure
174 * that the code doesn't walk off of the end of the buffer (which is always
175 * a danger if the alignment of the source buffer is off). The destination
176 * string is always properly null terminated and fits in the destination
177 * buffer. Returns the length of the destination string in bytes (including
180 * Note that some windows versions actually send multiword UTF-16 characters
181 * instead of straight UTF16-2. The linux nls routines however aren't able to
182 * deal with those characters properly. In the event that we get some of
183 * those characters, they won't be translated properly.
186 cifs_from_utf16(char *to, const __le16 *from, int tolen, int fromlen,
187 const struct nls_table *codepage, int map_type)
189 int i, charlen, safelen;
191 int nullsize = nls_nullsize(codepage);
192 int fromwords = fromlen / 2;
193 char tmp[NLS_MAX_CHARSET_SIZE];
197 * because the chars can be of varying widths, we need to take care
198 * not to overflow the destination buffer when we get close to the
199 * end of it. Until we get to this offset, we don't need to check
200 * for overflow however.
202 safelen = tolen - (NLS_MAX_CHARSET_SIZE + nullsize);
204 for (i = 0; i < fromwords; i++) {
205 ftmp = get_unaligned_le16(&from[i]);
210 * check to see if converting this character might make the
211 * conversion bleed into the null terminator
213 if (outlen >= safelen) {
214 charlen = cifs_mapchar(tmp, ftmp, codepage, map_type);
215 if ((outlen + charlen) > (tolen - nullsize))
219 /* put converted char into 'to' buffer */
220 charlen = cifs_mapchar(&to[outlen], ftmp, codepage, map_type);
224 /* properly null-terminate string */
225 for (i = 0; i < nullsize; i++)
232 * NAME: cifs_strtoUTF16()
234 * FUNCTION: Convert character string to unicode string
238 cifs_strtoUTF16(__le16 *to, const char *from, int len,
239 const struct nls_table *codepage)
243 wchar_t wchar_to; /* needed to quiet sparse */
245 /* special case for utf8 to handle no plane0 chars */
246 if (!strcmp(codepage->charset, "utf8")) {
248 * convert utf8 -> utf16, we assume we have enough space
249 * as caller should have assumed conversion does not overflow
250 * in destination len is length in wchar_t units (16bits)
252 i = utf8s_to_utf16s(from, len, UTF16_LITTLE_ENDIAN,
253 (wchar_t *) to, len);
255 /* if success terminate and exit */
259 * if fails fall back to UCS encoding as this
260 * function should not return negative values
261 * currently can fail only if source contains
262 * invalid encoded characters
266 for (i = 0; len && *from; i++, from += charlen, len -= charlen) {
267 charlen = codepage->char2uni(from, len, &wchar_to);
269 cifs_dbg(VFS, "strtoUTF16: char2uni of 0x%x returned %d\n",
271 /* A question mark */
275 put_unaligned_le16(wchar_to, &to[i]);
279 put_unaligned_le16(0, &to[i]);
284 * cifs_strndup_from_utf16 - copy a string from wire format to the local
286 * @src - source string
287 * @maxlen - don't walk past this many bytes in the source string
288 * @is_unicode - is this a unicode string?
289 * @codepage - destination codepage
291 * Take a string given by the server, convert it to the local codepage and
292 * put it in a new buffer. Returns a pointer to the new string or NULL on
296 cifs_strndup_from_utf16(const char *src, const int maxlen,
297 const bool is_unicode, const struct nls_table *codepage)
303 len = cifs_utf16_bytes((__le16 *) src, maxlen, codepage);
304 len += nls_nullsize(codepage);
305 dst = kmalloc(len, GFP_KERNEL);
308 cifs_from_utf16(dst, (__le16 *) src, len, maxlen, codepage,
311 len = strnlen(src, maxlen);
313 dst = kmalloc(len, GFP_KERNEL);
316 strlcpy(dst, src, len);
323 * Convert 16 bit Unicode pathname to wire format from string in current code
324 * page. Conversion may involve remapping up the six characters that are
325 * only legal in POSIX-like OS (if they are present in the string). Path
326 * names are little endian 16 bit Unicode on the wire
329 cifsConvertToUTF16(__le16 *target, const char *source, int srclen,
330 const struct nls_table *cp, int mapChars)
339 return cifs_strtoUTF16(target, source, PATH_MAX, cp);
341 for (i = 0; i < srclen; j++) {
342 src_char = source[i];
348 dst_char = cpu_to_le16(UNI_COLON);
351 dst_char = cpu_to_le16(UNI_ASTERISK);
354 dst_char = cpu_to_le16(UNI_QUESTION);
357 dst_char = cpu_to_le16(UNI_LESSTHAN);
360 dst_char = cpu_to_le16(UNI_GRTRTHAN);
363 dst_char = cpu_to_le16(UNI_PIPE);
366 * FIXME: We can not handle remapping backslash (UNI_SLASH)
367 * until all the calls to build_path_from_dentry are modified,
368 * as they use backslash as separator.
371 charlen = cp->char2uni(source + i, srclen - i, &tmp);
372 dst_char = cpu_to_le16(tmp);
375 * if no match, use question mark, which at least in
376 * some cases serves as wild card
379 dst_char = cpu_to_le16(0x003f);
384 * character may take more than one byte in the source string,
385 * but will take exactly two bytes in the target string
388 put_unaligned(dst_char, &target[j]);
392 put_unaligned(0, &target[j]); /* Null terminate target unicode string */
396 #ifdef CONFIG_CIFS_SMB2
398 * cifs_local_to_utf16_bytes - how long will a string be after conversion?
399 * @from - pointer to input string
400 * @maxbytes - don't go past this many bytes of input string
401 * @codepage - source codepage
403 * Walk a string and return the number of bytes that the string will
404 * be after being converted to the given charset, not including any null
405 * termination required. Don't walk past maxbytes in the source buffer.
409 cifs_local_to_utf16_bytes(const char *from, int len,
410 const struct nls_table *codepage)
416 for (i = 0; len && *from; i++, from += charlen, len -= charlen) {
417 charlen = codepage->char2uni(from, len, &wchar_to);
418 /* Failed conversion defaults to a question mark */
422 return 2 * i; /* UTF16 characters are two bytes */
426 * cifs_strndup_to_utf16 - copy a string to wire format from the local codepage
427 * @src - source string
428 * @maxlen - don't walk past this many bytes in the source string
429 * @utf16_len - the length of the allocated string in bytes (including null)
430 * @cp - source codepage
431 * @remap - map special chars
433 * Take a string convert it from the local codepage to UTF16 and
434 * put it in a new buffer. Returns a pointer to the new string or NULL on
438 cifs_strndup_to_utf16(const char *src, const int maxlen, int *utf16_len,
439 const struct nls_table *cp, int remap)
444 len = cifs_local_to_utf16_bytes(src, maxlen, cp);
446 dst = kmalloc(len, GFP_KERNEL);
451 cifsConvertToUTF16(dst, src, strlen(src), cp, remap);
455 #endif /* CONFIG_CIFS_SMB2 */