xref: /rk3399_rockchip-uboot/drivers/usb/gadget/usbstring.c (revision 23cd138503f90ff6af109c0096727ba641942614)
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 <linux/usb/gadget.h>
17 
18 #include <asm/unaligned.h>
19 
20 
21 static int utf8_to_utf16le(const char *s, __le16 *cp, unsigned len)
22 {
23 	int	count = 0;
24 	u8	c;
25 	u16	uchar;
26 
27 	/* this insists on correct encodings, though not minimal ones.
28 	 * BUT it currently rejects legit 4-byte UTF-8 code points,
29 	 * which need surrogate pairs.  (Unicode 3.1 can use them.)
30 	 */
31 	while (len != 0 && (c = (u8) *s++) != 0) {
32 		if ((c & 0x80)) {
33 			// 2-byte sequence:
34 			// 00000yyyyyxxxxxx = 110yyyyy 10xxxxxx
35 			if ((c & 0xe0) == 0xc0) {
36 				uchar = (c & 0x1f) << 6;
37 
38 				c = (u8) *s++;
39 				if ((c & 0xc0) != 0x80)
40 					goto fail;
41 				c &= 0x3f;
42 				uchar |= c;
43 
44 			// 3-byte sequence (most CJKV characters):
45 			// zzzzyyyyyyxxxxxx = 1110zzzz 10yyyyyy 10xxxxxx
46 			} else if ((c & 0xf0) == 0xe0) {
47 				uchar = (c & 0x0f) << 12;
48 
49 				c = (u8) *s++;
50 				if ((c & 0xc0) != 0x80)
51 					goto fail;
52 				c &= 0x3f;
53 				uchar |= c << 6;
54 
55 				c = (u8) *s++;
56 				if ((c & 0xc0) != 0x80)
57 					goto fail;
58 				c &= 0x3f;
59 				uchar |= c;
60 
61 				/* no bogus surrogates */
62 				if (0xd800 <= uchar && uchar <= 0xdfff)
63 					goto fail;
64 
65 			// 4-byte sequence (surrogate pairs, currently rare):
66 			// 11101110wwwwzzzzyy + 110111yyyyxxxxxx
67 			//     = 11110uuu 10uuzzzz 10yyyyyy 10xxxxxx
68 			// (uuuuu = wwww + 1)
69 			// FIXME accept the surrogate code points (only)
70 
71 			} else
72 				goto fail;
73 		} else
74 			uchar = c;
75 		put_unaligned_le16(uchar, cp++);
76 		count++;
77 		len--;
78 	}
79 	return count;
80 fail:
81 	return -1;
82 }
83 
84 
85 /**
86  * usb_gadget_get_string - fill out a string descriptor
87  * @table: of c strings encoded using UTF-8
88  * @id: string id, from low byte of wValue in get string descriptor
89  * @buf: at least 256 bytes
90  *
91  * Finds the UTF-8 string matching the ID, and converts it into a
92  * string descriptor in utf16-le.
93  * Returns length of descriptor (always even) or negative errno
94  *
95  * If your driver needs stings in multiple languages, you'll probably
96  * "switch (wIndex) { ... }"  in your ep0 string descriptor logic,
97  * using this routine after choosing which set of UTF-8 strings to use.
98  * Note that US-ASCII is a strict subset of UTF-8; any string bytes with
99  * the eighth bit set will be multibyte UTF-8 characters, not ISO-8859/1
100  * characters (which are also widely used in C strings).
101  */
102 int
103 usb_gadget_get_string (struct usb_gadget_strings *table, int id, u8 *buf)
104 {
105 	struct usb_string	*s;
106 	int			len;
107 
108 	/* descriptor 0 has the language id */
109 	if (id == 0) {
110 		buf [0] = 4;
111 		buf [1] = USB_DT_STRING;
112 		buf [2] = (u8) table->language;
113 		buf [3] = (u8) (table->language >> 8);
114 		return 4;
115 	}
116 	for (s = table->strings; s && s->s; s++)
117 		if (s->id == id)
118 			break;
119 
120 	/* unrecognized: stall. */
121 	if (!s || !s->s)
122 		return -EINVAL;
123 
124 	/* string descriptors have length, tag, then UTF16-LE text */
125 	len = min ((size_t) 126, strlen (s->s));
126 	memset (buf + 2, 0, 2 * len);	/* zero all the bytes */
127 	len = utf8_to_utf16le(s->s, (__le16 *)&buf[2], len);
128 	if (len < 0)
129 		return -EINVAL;
130 	buf [0] = (len + 1) * 2;
131 	buf [1] = USB_DT_STRING;
132 	return buf [0];
133 }
134 
135