NLS: update handling of Unicode

This patch (as1239) updates the kernel's treatment of Unicode.  The
character-set conversion routines are well behind the current state of
the Unicode specification: They don't recognize the existence of code
points beyond plane 0 or of surrogate pairs in the UTF-16 encoding.

The old wchar_t 16-bit type is retained because it's still used in
lots of places.  This shouldn't cause any new problems; if a
conversion now results in an invalid 16-bit code then before it must
have yielded an undefined code.

Difficult-to-read names like "utf_mbstowcs" are replaced with more
transparent names like "utf8s_to_utf16s" and the ordering of the
parameters is rationalized (buffer lengths come immediate after the
pointers they refer to, and the inputs precede the outputs).
Fortunately the low-level conversion routines are used in only a few
places; the interfaces to the higher-level uni2char and char2uni
methods have been left unchanged.

Signed-off-by: Alan Stern <[email protected]>
Acked-by: Clemens Ladisch <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>

drivers/usb/core/message.c

@@ -780,14 +780,13 @@ int usb_string(struct usb_device *dev, int index, char *buf, size_t size)
 {
 	unsigned char *tbuf;
 	int err;
-	unsigned int u;
 
 	if (dev->state == USB_STATE_SUSPENDED)
 		return -EHOSTUNREACH;
 	if (size <= 0 || !buf || !index)
 		return -EINVAL;
 	buf[0] = 0;
-	tbuf = kmalloc(256 + 2, GFP_NOIO);
+	tbuf = kmalloc(256, GFP_NOIO);
 	if (!tbuf)
 		return -ENOMEM;
 
@@ -814,12 +813,9 @@ int usb_string(struct usb_device *dev, int index, char *buf, size_t size)
 	if (err < 0)
 		goto errout;
 
-	for (u = 2; u < err; u += 2)
-		le16_to_cpus((u16 *)&tbuf[u]);
-	tbuf[u] = 0;
-	tbuf[u + 1] = 0;
 	size--;		/* leave room for trailing NULL char in output buffer */
-	err = utf8_wcstombs(buf, (u16 *)&tbuf[2], size);
+	err = utf16s_to_utf8s((wchar_t *) &tbuf[2], (err - 2) / 2,
+			UTF16_LITTLE_ENDIAN, buf, size);
 	buf[err] = 0;
 
 	if (tbuf[1] != USB_DT_STRING)

fs/befs/linuxvfs.c

@@ -513,7 +513,7 @@ befs_utf2nls(struct super_block *sb, const char *in,
 {
 	struct nls_table *nls = BEFS_SB(sb)->nls;
 	int i, o;
-	wchar_t uni;
+	unicode_t uni;
 	int unilen, utflen;
 	char *result;
 	/* The utf8->nls conversion won't make the final nls string bigger
@@ -539,16 +539,16 @@ befs_utf2nls(struct super_block *sb, const char *in,
 	for (i = o = 0; i < in_len; i += utflen, o += unilen) {
 
 		/* convert from UTF-8 to Unicode */
-		utflen = utf8_mbtowc(&uni, &in[i], in_len - i);
-		if (utflen < 0) {
+		utflen = utf8_to_utf32(&in[i], in_len - i, &uni);
+		if (utflen < 0)
 			goto conv_err;
-		}
 
 		/* convert from Unicode to nls */
+		if (uni > MAX_WCHAR_T)
+			goto conv_err;
 		unilen = nls->uni2char(uni, &result[o], in_len - o);
-		if (unilen < 0) {
+		if (unilen < 0)
 			goto conv_err;
-		}
 	}
 	result[o] = '\0';
 	*out_len = o;
@@ -619,15 +619,13 @@ befs_nls2utf(struct super_block *sb, const char *in,
 
 		/* convert from nls to unicode */
 		unilen = nls->char2uni(&in[i], in_len - i, &uni);
-		if (unilen < 0) {
+		if (unilen < 0)
 			goto conv_err;
-		}
 
 		/* convert from unicode to UTF-8 */
-		utflen = utf8_wctomb(&result[o], uni, 3);
-		if (utflen <= 0) {
+		utflen = utf32_to_utf8(uni, &result[o], 3);
+		if (utflen <= 0)
 			goto conv_err;
-		}
 	}
 
 	result[o] = '\0';

fs/fat/dir.c

@@ -22,6 +22,19 @@
 #include <asm/uaccess.h>
 #include "fat.h"
 
+/*
+ * Maximum buffer size of short name.
+ * [(MSDOS_NAME + '.') * max one char + nul]
+ * For msdos style, ['.' (hidden) + MSDOS_NAME + '.' + nul]
+ */
+#define FAT_MAX_SHORT_SIZE	((MSDOS_NAME + 1) * NLS_MAX_CHARSET_SIZE + 1)
+/*
+ * Maximum buffer size of unicode chars from slots.
+ * [(max longname slots * 13 (size in a slot) + nul) * sizeof(wchar_t)]
+ */
+#define FAT_MAX_UNI_CHARS	((MSDOS_SLOTS - 1) * 13 + 1)
+#define FAT_MAX_UNI_SIZE	(FAT_MAX_UNI_CHARS * sizeof(wchar_t))
+
 static inline loff_t fat_make_i_pos(struct super_block *sb,
 				    struct buffer_head *bh,
 				    struct msdos_dir_entry *de)
@@ -171,7 +184,8 @@ static inline int fat_uni_to_x8(struct msdos_sb_info *sbi, const wchar_t *uni,
 				unsigned char *buf, int size)
 {
 	if (sbi->options.utf8)
-		return utf8_wcstombs(buf, uni, size);
+		return utf16s_to_utf8s(uni, FAT_MAX_UNI_CHARS,
+				UTF16_HOST_ENDIAN, buf, size);
 	else
 		return uni16_to_x8(buf, uni, size, sbi->options.unicode_xlate,
 				   sbi->nls_io);
@@ -324,19 +338,6 @@ static int fat_parse_long(struct inode *dir, loff_t *pos,
 	return 0;
 }
 
-/*
- * Maximum buffer size of short name.
- * [(MSDOS_NAME + '.') * max one char + nul]
- * For msdos style, ['.' (hidden) + MSDOS_NAME + '.' + nul]
- */
-#define FAT_MAX_SHORT_SIZE	((MSDOS_NAME + 1) * NLS_MAX_CHARSET_SIZE + 1)
-/*
- * Maximum buffer size of unicode chars from slots.
- * [(max longname slots * 13 (size in a slot) + nul) * sizeof(wchar_t)]
- */
-#define FAT_MAX_UNI_CHARS	((MSDOS_SLOTS - 1) * 13 + 1)
-#define FAT_MAX_UNI_SIZE	(FAT_MAX_UNI_CHARS * sizeof(wchar_t))
-
 /*
  * Return values: negative -> error, 0 -> not found, positive -> found,
  * value is the total amount of slots, including the shortname entry.

fs/fat/namei_vfat.c

@@ -502,11 +502,11 @@ xlate_to_uni(const unsigned char *name, int len, unsigned char *outname,
 	if (utf8) {
 		int name_len = strlen(name);
 
-		*outlen = utf8_mbstowcs((wchar_t *)outname, name, PATH_MAX);
+		*outlen = utf8s_to_utf16s(name, PATH_MAX, (wchar_t *) outname);
 
 		/*
 		 * We stripped '.'s before and set len appropriately,
-		 * but utf8_mbstowcs doesn't care about len
+		 * but utf8s_to_utf16s doesn't care about len
 		 */
 		*outlen -= (name_len - len);
 

fs/isofs/joliet.c

@@ -37,37 +37,6 @@ uni16_to_x8(unsigned char *ascii, __be16 *uni, int len, struct nls_table *nls)
 	return (op - ascii);
 }
 
-/* Convert big endian wide character string to utf8 */
-static int
-wcsntombs_be(__u8 *s, const __u8 *pwcs, int inlen, int maxlen)
-{
-	const __u8 *ip;
-	__u8 *op;
-	int size;
-	__u16 c;
-
-	op = s;
-	ip = pwcs;
-	while ((*ip || ip[1]) && (maxlen > 0) && (inlen > 0)) {
-		c = (*ip << 8) | ip[1];
-		if (c > 0x7f) {
-			size = utf8_wctomb(op, c, maxlen);
-			if (size == -1) {
-				/* Ignore character and move on */
-				maxlen--;
-			} else {
-				op += size;
-				maxlen -= size;
-			}
-		} else {
-			*op++ = (__u8) c;
-		}
-		ip += 2;
-		inlen--;
-	}
-	return (op - s);
-}
-
 int
 get_joliet_filename(struct iso_directory_record * de, unsigned char *outname, struct inode * inode)
 {
@@ -79,8 +48,9 @@ get_joliet_filename(struct iso_directory_record * de, unsigned char *outname, st
 	nls = ISOFS_SB(inode->i_sb)->s_nls_iocharset;
 
 	if (utf8) {
-		len = wcsntombs_be(outname, de->name,
-				de->name_len[0] >> 1, PAGE_SIZE);
+		len = utf16s_to_utf8s((const wchar_t *) de->name,
+				de->name_len[0] >> 1, UTF16_BIG_ENDIAN,
+				outname, PAGE_SIZE);
 	} else {
 		len = uni16_to_x8(outname, (__be16 *) de->name,
 				de->name_len[0] >> 1, nls);

fs/ncpfs/ncplib_kernel.c

@@ -1113,11 +1113,13 @@ ncp__io2vol(struct ncp_server *server, unsigned char *vname, unsigned int *vlen,
 
 		if (NCP_IS_FLAG(server, NCP_FLAG_UTF8)) {
 			int k;
+			unicode_t u;
 
-			k = utf8_mbtowc(&ec, iname, iname_end - iname);
-			if (k < 0)
+			k = utf8_to_utf32(iname, iname_end - iname, &u);
+			if (k < 0 || u > MAX_WCHAR_T)
 				return -EINVAL;
 			iname += k;
+			ec = u;
 		} else {
 			if (*iname == NCP_ESC) {
 				int k;
@@ -1214,7 +1216,7 @@ ncp__vol2io(struct ncp_server *server, unsigned char *iname, unsigned int *ilen,
 		if (NCP_IS_FLAG(server, NCP_FLAG_UTF8)) {
 			int k;
 
-			k = utf8_wctomb(iname, ec, iname_end - iname);
+			k = utf32_to_utf8(ec, iname, iname_end - iname);
 			if (k < 0) {
 				err = -ENAMETOOLONG;
 				goto quit;