1 | /* Copyright (C) 1991-2022 Free Software Foundation, Inc. |
2 | This file is part of the GNU C Library. |
3 | |
4 | The GNU C Library is free software; you can redistribute it and/or |
5 | modify it under the terms of the GNU Lesser General Public |
6 | License as published by the Free Software Foundation; either |
7 | version 2.1 of the License, or (at your option) any later version. |
8 | |
9 | The GNU C Library is distributed in the hope that it will be useful, |
10 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
11 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
12 | Lesser General Public License for more details. |
13 | |
14 | You should have received a copy of the GNU Lesser General Public |
15 | License along with the GNU C Library; if not, see |
16 | <https://www.gnu.org/licenses/>. */ |
17 | |
18 | #include <string.h> |
19 | #include <stdlib.h> |
20 | |
21 | #undef strlen |
22 | |
23 | #ifndef STRLEN |
24 | # define STRLEN strlen |
25 | #endif |
26 | |
27 | /* Return the length of the null-terminated string STR. Scan for |
28 | the null terminator quickly by testing four bytes at a time. */ |
29 | size_t |
30 | STRLEN (const char *str) |
31 | { |
32 | const char *char_ptr; |
33 | const unsigned long int *longword_ptr; |
34 | unsigned long int longword, himagic, lomagic; |
35 | |
36 | /* Handle the first few characters by reading one character at a time. |
37 | Do this until CHAR_PTR is aligned on a longword boundary. */ |
38 | for (char_ptr = str; ((unsigned long int) char_ptr |
39 | & (sizeof (longword) - 1)) != 0; |
40 | ++char_ptr) |
41 | if (*char_ptr == '\0') |
42 | return char_ptr - str; |
43 | |
44 | /* All these elucidatory comments refer to 4-byte longwords, |
45 | but the theory applies equally well to 8-byte longwords. */ |
46 | |
47 | longword_ptr = (unsigned long int *) char_ptr; |
48 | |
49 | /* Bits 31, 24, 16, and 8 of this number are zero. Call these bits |
50 | the "holes." Note that there is a hole just to the left of |
51 | each byte, with an extra at the end: |
52 | |
53 | bits: 01111110 11111110 11111110 11111111 |
54 | bytes: AAAAAAAA BBBBBBBB CCCCCCCC DDDDDDDD |
55 | |
56 | The 1-bits make sure that carries propagate to the next 0-bit. |
57 | The 0-bits provide holes for carries to fall into. */ |
58 | himagic = 0x80808080L; |
59 | lomagic = 0x01010101L; |
60 | if (sizeof (longword) > 4) |
61 | { |
62 | /* 64-bit version of the magic. */ |
63 | /* Do the shift in two steps to avoid a warning if long has 32 bits. */ |
64 | himagic = ((himagic << 16) << 16) | himagic; |
65 | lomagic = ((lomagic << 16) << 16) | lomagic; |
66 | } |
67 | if (sizeof (longword) > 8) |
68 | abort (); |
69 | |
70 | /* Instead of the traditional loop which tests each character, |
71 | we will test a longword at a time. The tricky part is testing |
72 | if *any of the four* bytes in the longword in question are zero. */ |
73 | for (;;) |
74 | { |
75 | longword = *longword_ptr++; |
76 | |
77 | if (((longword - lomagic) & ~longword & himagic) != 0) |
78 | { |
79 | /* Which of the bytes was the zero? If none of them were, it was |
80 | a misfire; continue the search. */ |
81 | |
82 | const char *cp = (const char *) (longword_ptr - 1); |
83 | |
84 | if (cp[0] == 0) |
85 | return cp - str; |
86 | if (cp[1] == 0) |
87 | return cp - str + 1; |
88 | if (cp[2] == 0) |
89 | return cp - str + 2; |
90 | if (cp[3] == 0) |
91 | return cp - str + 3; |
92 | if (sizeof (longword) > 4) |
93 | { |
94 | if (cp[4] == 0) |
95 | return cp - str + 4; |
96 | if (cp[5] == 0) |
97 | return cp - str + 5; |
98 | if (cp[6] == 0) |
99 | return cp - str + 6; |
100 | if (cp[7] == 0) |
101 | return cp - str + 7; |
102 | } |
103 | } |
104 | } |
105 | } |
106 | libc_hidden_builtin_def (strlen) |
107 | |