usbstring.c 3.2 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136
  1. /*
  2. * Copyright (C) 2003 David Brownell
  3. *
  4. * SPDX-License-Identifier: LGPL-2.1+
  5. *
  6. * Ported to U-boot by: Thomas Smits <ts.smits@gmail.com> and
  7. * Remy Bohmer <linux@bohmer.net>
  8. */
  9. #include <common.h>
  10. #include <asm/errno.h>
  11. #include <linux/usb/ch9.h>
  12. #include <linux/usb/gadget.h>
  13. #include <asm/unaligned.h>
  14. static int utf8_to_utf16le(const char *s, __le16 *cp, unsigned len)
  15. {
  16. int count = 0;
  17. u8 c;
  18. u16 uchar;
  19. /*
  20. * this insists on correct encodings, though not minimal ones.
  21. * BUT it currently rejects legit 4-byte UTF-8 code points,
  22. * which need surrogate pairs. (Unicode 3.1 can use them.)
  23. */
  24. while (len != 0 && (c = (u8) *s++) != 0) {
  25. if ((c & 0x80)) {
  26. /*
  27. * 2-byte sequence:
  28. * 00000yyyyyxxxxxx = 110yyyyy 10xxxxxx
  29. */
  30. if ((c & 0xe0) == 0xc0) {
  31. uchar = (c & 0x1f) << 6;
  32. c = (u8) *s++;
  33. if ((c & 0xc0) != 0x80)
  34. goto fail;
  35. c &= 0x3f;
  36. uchar |= c;
  37. /*
  38. * 3-byte sequence (most CJKV characters):
  39. * zzzzyyyyyyxxxxxx = 1110zzzz 10yyyyyy 10xxxxxx
  40. */
  41. } else if ((c & 0xf0) == 0xe0) {
  42. uchar = (c & 0x0f) << 12;
  43. c = (u8) *s++;
  44. if ((c & 0xc0) != 0x80)
  45. goto fail;
  46. c &= 0x3f;
  47. uchar |= c << 6;
  48. c = (u8) *s++;
  49. if ((c & 0xc0) != 0x80)
  50. goto fail;
  51. c &= 0x3f;
  52. uchar |= c;
  53. /* no bogus surrogates */
  54. if (0xd800 <= uchar && uchar <= 0xdfff)
  55. goto fail;
  56. /*
  57. * 4-byte sequence (surrogate pairs, currently rare):
  58. * 11101110wwwwzzzzyy + 110111yyyyxxxxxx
  59. * = 11110uuu 10uuzzzz 10yyyyyy 10xxxxxx
  60. * (uuuuu = wwww + 1)
  61. * FIXME accept the surrogate code points (only)
  62. */
  63. } else
  64. goto fail;
  65. } else
  66. uchar = c;
  67. put_unaligned_le16(uchar, cp++);
  68. count++;
  69. len--;
  70. }
  71. return count;
  72. fail:
  73. return -1;
  74. }
  75. /**
  76. * usb_gadget_get_string - fill out a string descriptor
  77. * @table: of c strings encoded using UTF-8
  78. * @id: string id, from low byte of wValue in get string descriptor
  79. * @buf: at least 256 bytes
  80. *
  81. * Finds the UTF-8 string matching the ID, and converts it into a
  82. * string descriptor in utf16-le.
  83. * Returns length of descriptor (always even) or negative errno
  84. *
  85. * If your driver needs stings in multiple languages, you'll probably
  86. * "switch (wIndex) { ... }" in your ep0 string descriptor logic,
  87. * using this routine after choosing which set of UTF-8 strings to use.
  88. * Note that US-ASCII is a strict subset of UTF-8; any string bytes with
  89. * the eighth bit set will be multibyte UTF-8 characters, not ISO-8859/1
  90. * characters (which are also widely used in C strings).
  91. */
  92. int
  93. usb_gadget_get_string(struct usb_gadget_strings *table, int id, u8 *buf)
  94. {
  95. struct usb_string *s;
  96. int len;
  97. /* descriptor 0 has the language id */
  98. if (id == 0) {
  99. buf[0] = 4;
  100. buf[1] = USB_DT_STRING;
  101. buf[2] = (u8) table->language;
  102. buf[3] = (u8) (table->language >> 8);
  103. return 4;
  104. }
  105. for (s = table->strings; s && s->s; s++)
  106. if (s->id == id)
  107. break;
  108. /* unrecognized: stall. */
  109. if (!s || !s->s)
  110. return -EINVAL;
  111. /* string descriptors have length, tag, then UTF16-LE text */
  112. len = min((size_t) 126, strlen(s->s));
  113. memset(buf + 2, 0, 2 * len); /* zero all the bytes */
  114. len = utf8_to_utf16le(s->s, (__le16 *)&buf[2], len);
  115. if (len < 0)
  116. return -EINVAL;
  117. buf[0] = (len + 1) * 2;
  118. buf[1] = USB_DT_STRING;
  119. return buf[0];
  120. }