xref: /openbmc/u-boot/drivers/usb/gadget/usbstring.c (revision dd1033e4)
1 // SPDX-License-Identifier: LGPL-2.1+
2 /*
3  * Copyright (C) 2003 David Brownell
4  *
5  * Ported to U-Boot by: Thomas Smits <ts.smits@gmail.com> and
6  *                      Remy Bohmer <linux@bohmer.net>
7  */
8 
9 #include <common.h>
10 #include <linux/errno.h>
11 #include <linux/usb/ch9.h>
12 #include <linux/usb/gadget.h>
13 
14 #include <asm/unaligned.h>
15 
16 
17 static int utf8_to_utf16le(const char *s, __le16 *cp, unsigned len)
18 {
19 	int	count = 0;
20 	u8	c;
21 	u16	uchar;
22 
23 	/*
24 	 * this insists on correct encodings, though not minimal ones.
25 	 * BUT it currently rejects legit 4-byte UTF-8 code points,
26 	 * which need surrogate pairs.  (Unicode 3.1 can use them.)
27 	 */
28 	while (len != 0 && (c = (u8) *s++) != 0) {
29 		if ((c & 0x80)) {
30 			/*
31 			 * 2-byte sequence:
32 			 * 00000yyyyyxxxxxx = 110yyyyy 10xxxxxx
33 			 */
34 			if ((c & 0xe0) == 0xc0) {
35 				uchar = (c & 0x1f) << 6;
36 
37 				c = (u8) *s++;
38 				if ((c & 0xc0) != 0x80)
39 					goto fail;
40 				c &= 0x3f;
41 				uchar |= c;
42 
43 			/*
44 			 * 3-byte sequence (most CJKV characters):
45 			 * zzzzyyyyyyxxxxxx = 1110zzzz 10yyyyyy 10xxxxxx
46 			 */
47 			} else if ((c & 0xf0) == 0xe0) {
48 				uchar = (c & 0x0f) << 12;
49 
50 				c = (u8) *s++;
51 				if ((c & 0xc0) != 0x80)
52 					goto fail;
53 				c &= 0x3f;
54 				uchar |= c << 6;
55 
56 				c = (u8) *s++;
57 				if ((c & 0xc0) != 0x80)
58 					goto fail;
59 				c &= 0x3f;
60 				uchar |= c;
61 
62 				/* no bogus surrogates */
63 				if (0xd800 <= uchar && uchar <= 0xdfff)
64 					goto fail;
65 
66 			/*
67 			 * 4-byte sequence (surrogate pairs, currently rare):
68 			 * 11101110wwwwzzzzyy + 110111yyyyxxxxxx
69 			 *     = 11110uuu 10uuzzzz 10yyyyyy 10xxxxxx
70 			 * (uuuuu = wwww + 1)
71 			 * FIXME accept the surrogate code points (only)
72 			 */
73 			} else
74 				goto fail;
75 		} else
76 			uchar = c;
77 		put_unaligned_le16(uchar, cp++);
78 		count++;
79 		len--;
80 	}
81 	return count;
82 fail:
83 	return -1;
84 }
85 
86 
87 /**
88  * usb_gadget_get_string - fill out a string descriptor
89  * @table: of c strings encoded using UTF-8
90  * @id: string id, from low byte of wValue in get string descriptor
91  * @buf: at least 256 bytes
92  *
93  * Finds the UTF-8 string matching the ID, and converts it into a
94  * string descriptor in utf16-le.
95  * Returns length of descriptor (always even) or negative errno
96  *
97  * If your driver needs stings in multiple languages, you'll probably
98  * "switch (wIndex) { ... }"  in your ep0 string descriptor logic,
99  * using this routine after choosing which set of UTF-8 strings to use.
100  * Note that US-ASCII is a strict subset of UTF-8; any string bytes with
101  * the eighth bit set will be multibyte UTF-8 characters, not ISO-8859/1
102  * characters (which are also widely used in C strings).
103  */
104 int
105 usb_gadget_get_string(struct usb_gadget_strings *table, int id, u8 *buf)
106 {
107 	struct usb_string	*s;
108 	int			len;
109 
110 	if (!table)
111 		return -EINVAL;
112 
113 	/* descriptor 0 has the language id */
114 	if (id == 0) {
115 		buf[0] = 4;
116 		buf[1] = USB_DT_STRING;
117 		buf[2] = (u8) table->language;
118 		buf[3] = (u8) (table->language >> 8);
119 		return 4;
120 	}
121 	for (s = table->strings; s && s->s; s++)
122 		if (s->id == id)
123 			break;
124 
125 	/* unrecognized: stall. */
126 	if (!s || !s->s)
127 		return -EINVAL;
128 
129 	/* string descriptors have length, tag, then UTF16-LE text */
130 	len = min((size_t) 126, strlen(s->s));
131 	memset(buf + 2, 0, 2 * len);	/* zero all the bytes */
132 	len = utf8_to_utf16le(s->s, (__le16 *)&buf[2], len);
133 	if (len < 0)
134 		return -EINVAL;
135 	buf[0] = (len + 1) * 2;
136 	buf[1] = USB_DT_STRING;
137 	return buf[0];
138 }
139