xref: /openbmc/u-boot/drivers/usb/gadget/usbstring.c (revision c6af2e7d)
1 /*
2  * Copyright (C) 2003 David Brownell
3  *
4  * This program is free software; you can redistribute it and/or modify
5  * it under the terms of the GNU Lesser General Public License as published
6  * by the Free Software Foundation; either version 2.1 of the License, or
7  * (at your option) any later version.
8  *
9  * Ported to U-boot by: Thomas Smits <ts.smits@gmail.com> and
10  *                      Remy Bohmer <linux@bohmer.net>
11  */
12 
13 #include <common.h>
14 #include <asm/errno.h>
15 #include <linux/usb/ch9.h>
16 #include <usbdescriptors.h>
17 #include <linux/usb/gadget.h>
18 
19 #include <asm/unaligned.h>
20 
21 
22 static int utf8_to_utf16le(const char *s, __le16 *cp, unsigned len)
23 {
24 	int	count = 0;
25 	u8	c;
26 	u16	uchar;
27 
28 	/*
29 	 * this insists on correct encodings, though not minimal ones.
30 	 * BUT it currently rejects legit 4-byte UTF-8 code points,
31 	 * which need surrogate pairs.  (Unicode 3.1 can use them.)
32 	 */
33 	while (len != 0 && (c = (u8) *s++) != 0) {
34 		if ((c & 0x80)) {
35 			/*
36 			 * 2-byte sequence:
37 			 * 00000yyyyyxxxxxx = 110yyyyy 10xxxxxx
38 			 */
39 			if ((c & 0xe0) == 0xc0) {
40 				uchar = (c & 0x1f) << 6;
41 
42 				c = (u8) *s++;
43 				if ((c & 0xc0) != 0x80)
44 					goto fail;
45 				c &= 0x3f;
46 				uchar |= c;
47 
48 			/*
49 			 * 3-byte sequence (most CJKV characters):
50 			 * zzzzyyyyyyxxxxxx = 1110zzzz 10yyyyyy 10xxxxxx
51 			 */
52 			} else if ((c & 0xf0) == 0xe0) {
53 				uchar = (c & 0x0f) << 12;
54 
55 				c = (u8) *s++;
56 				if ((c & 0xc0) != 0x80)
57 					goto fail;
58 				c &= 0x3f;
59 				uchar |= c << 6;
60 
61 				c = (u8) *s++;
62 				if ((c & 0xc0) != 0x80)
63 					goto fail;
64 				c &= 0x3f;
65 				uchar |= c;
66 
67 				/* no bogus surrogates */
68 				if (0xd800 <= uchar && uchar <= 0xdfff)
69 					goto fail;
70 
71 			/*
72 			 * 4-byte sequence (surrogate pairs, currently rare):
73 			 * 11101110wwwwzzzzyy + 110111yyyyxxxxxx
74 			 *     = 11110uuu 10uuzzzz 10yyyyyy 10xxxxxx
75 			 * (uuuuu = wwww + 1)
76 			 * FIXME accept the surrogate code points (only)
77 			 */
78 			} else
79 				goto fail;
80 		} else
81 			uchar = c;
82 		put_unaligned_le16(uchar, cp++);
83 		count++;
84 		len--;
85 	}
86 	return count;
87 fail:
88 	return -1;
89 }
90 
91 
92 /**
93  * usb_gadget_get_string - fill out a string descriptor
94  * @table: of c strings encoded using UTF-8
95  * @id: string id, from low byte of wValue in get string descriptor
96  * @buf: at least 256 bytes
97  *
98  * Finds the UTF-8 string matching the ID, and converts it into a
99  * string descriptor in utf16-le.
100  * Returns length of descriptor (always even) or negative errno
101  *
102  * If your driver needs stings in multiple languages, you'll probably
103  * "switch (wIndex) { ... }"  in your ep0 string descriptor logic,
104  * using this routine after choosing which set of UTF-8 strings to use.
105  * Note that US-ASCII is a strict subset of UTF-8; any string bytes with
106  * the eighth bit set will be multibyte UTF-8 characters, not ISO-8859/1
107  * characters (which are also widely used in C strings).
108  */
109 int
110 usb_gadget_get_string(struct usb_gadget_strings *table, int id, u8 *buf)
111 {
112 	struct usb_string	*s;
113 	int			len;
114 
115 	/* descriptor 0 has the language id */
116 	if (id == 0) {
117 		buf[0] = 4;
118 		buf[1] = USB_DT_STRING;
119 		buf[2] = (u8) table->language;
120 		buf[3] = (u8) (table->language >> 8);
121 		return 4;
122 	}
123 	for (s = table->strings; s && s->s; s++)
124 		if (s->id == id)
125 			break;
126 
127 	/* unrecognized: stall. */
128 	if (!s || !s->s)
129 		return -EINVAL;
130 
131 	/* string descriptors have length, tag, then UTF16-LE text */
132 	len = min((size_t) 126, strlen(s->s));
133 	memset(buf + 2, 0, 2 * len);	/* zero all the bytes */
134 	len = utf8_to_utf16le(s->s, (__le16 *)&buf[2], len);
135 	if (len < 0)
136 		return -EINVAL;
137 	buf[0] = (len + 1) * 2;
138 	buf[1] = USB_DT_STRING;
139 	return buf[0];
140 }
141