无意间扫到 glibc strlen.c 中代码, 久久不能忘怀. 在一无所知的编程生涯中又记起点点滴滴:

编程可不是儿戏 ❀, 有些难, 也有些不舍. 随轨迹一同重温, 曾经最熟悉的 strlen 手感吧 ~

/* Copyright (C) 1991-2020 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Written by Torbjorn Granlund (tege@sics.se),
with help from Dan Sahlin (dan@sics.se);
commentary by Jim Blandy (jimb@ai.mit.edu).

The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.

The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
https://www.gnu.org/licenses/. */

#include
#include

#undef strlen

#ifndef STRLEN

define STRLEN strlen

#endif

/* Return the length of the null-terminated string STR. Scan for
the null terminator quickly by testing four bytes at a time. */
size_t
STRLEN (const char *str)
{
const char *char_ptr;
const unsigned long int *longword_ptr;
unsigned long int longword, himagic, lomagic;

/* Handle the first few characters by reading one character at a time.
Do this until CHAR_PTR is aligned on a longword boundary. */
for (char_ptr = str; ((unsigned long int) char_ptr
& (sizeof (longword) - )) != ;
++char_ptr)
if (*char_ptr == '\0')
return char_ptr - str;

/* All these elucidatory comments refer to 4-byte longwords,
but the theory applies equally well to 8-byte longwords. */

longword_ptr = (unsigned long int *) char_ptr;

/* Bits 31, 24, 16, and 8 of this number are zero. Call these bits
the "holes." Note that there is a hole just to the left of
each byte, with an extra at the end:

 bits:  01111110 11111110 11111110 11111111  
 bytes: AAAAAAAA BBBBBBBB CCCCCCCC DDDDDDDD

 The 1-bits make sure that carries propagate to the next 0-bit.  
 The 0-bits provide holes for carries to fall into.  \*/  

himagic = 0x80808080L;
lomagic = 0x01010101L;
if (sizeof (longword) > )
{
/* 64-bit version of the magic. */
/* Do the shift in two steps to avoid a warning if long has 32 bits. */
himagic = ((himagic << ) << ) | himagic; lomagic = ((lomagic << ) << ) | lomagic; } if (sizeof (longword) > )
abort ();

/* Instead of the traditional loop which tests each character,
we will test a longword at a time. The tricky part is testing
if *any of the four* bytes in the longword in question are zero. */
for (;;)
{
longword = *longword_ptr++;

  if (((longword - lomagic) & ~longword & himagic) != )  
{  
  /\* Which of the bytes was the zero?  If none of them were, it was  
     a misfire; continue the search.  \*/

  const char \*cp = (const char \*) (longword\_ptr - );

  if (cp\[\] == )  
    return cp - str;

  if (cp\[\] == )  
    return cp - str + ;  
  if (cp\[\] == )  
    return cp - str + ;  
  if (cp\[\] == )  
    return cp - str + ;  
  if (sizeof (longword) > )  
    {  
      if (cp\[\] == )  
    return cp - str + ;  
      if (cp\[\] == )  
    return cp - str + ;  
      if (cp\[\] == )  
    return cp - str + ;  
      if (cp\[\] == )  
    return cp - str + ;  
    }  
}  
}  

}
libc_hidden_builtin_def (strlen)

1. unsigned long int 字节多大 4 字节, 8 字节 ?

unsigned long int longword, himagic, lomagic;

long 具体多长和平台有关, 例如大多数 linux , x84 sizeof (long) = 4, x64 sizeof (long) = 8.

window x86, x64 sizeof (long) = 4.  (2020年05月28日), C 标准保证 sizeof(long) >= sizeof (int)

具体多少字节交给了实现方.

2. ((unsigned long int) char_ptr & (sizeof (longword) - 1)) 位对齐 ?

/* Handle the first few characters by reading one character at a time.
Do this until CHAR_PTR is aligned on a longword boundary. */
for (char_ptr = str; ((unsigned long int) char_ptr
& (sizeof (longword) - )) != ;
++char_ptr)
if (*char_ptr == '\0')
return char_ptr - str;

起始的这些代码的作用是, 让 chart_ptr 按照 sizeof (unsigned long) 字节大小进行位对齐.

这涉及到多数计算机硬件对齐有要求和性能方面的考虑等等(性能是主要因素).

3. himagic = 0x80808080L; lomagic = 0x01010101L; what fuck ?

/* Bits 31, 24, 16, and 8 of this number are zero. Call these bits
the "holes." Note that there is a hole just to the left of
each byte, with an extra at the end:

 bits:  01111110 11111110 11111110 11111111  
 bytes: AAAAAAAA BBBBBBBB CCCCCCCC DDDDDDDD

 The 1-bits make sure that carries propagate to the next 0-bit.  
 The 0-bits provide holes for carries to fall into.  \*/  

himagic = 0x80808080L;
lomagic = 0x01010101L;
if (sizeof (longword) > )
{
/* 64-bit version of the magic. */
/* Do the shift in two steps to avoid a warning if long has 32 bits. */
himagic = ((himagic << ) << ) | himagic; lomagic = ((lomagic << ) << ) | lomagic; } if (sizeof (longword) > )
abort ();

/* Instead of the traditional loop which tests each character,
we will test a longword at a time. The tricky part is testing
if *any of the four* bytes in the longword in question are zero. */
for (;;)
{
longword = *longword_ptr++;

  if (((longword - lomagic) & ~longword & himagic) != )  
{

3.1 (((longword - lomagic) & ~longword & himagic) != 0) ? mmp ?

可能这就是艺术吧. 想到这个想法的, 真是个天才啊! 好巧妙. 哈哈哈.  我们会分两个小点说明下.

首次看, 感觉有点萌. 我这里用个简单的思路来带大家理解这个问题. 上面代码主要围绕

sizeof (unsigned long) 4 字节和 8 字节去处理得到. 我们简单点, 通过处理 1 字节, 类比递归机制.

搞懂这个公式背后的原理 (ˇˍˇ) ～

/**
* himagic : 1000 0000
* lomagic : 0000 0001
* longword : XXXX XXXX
* /
unsigned long himagic = 0x80L;
unsigned long lomagic = 0x01L;

unsigned long longword ;

随后我们仔细分析下面公式

((longword - lomagic) & ~longword & himagic)

( & himagic ) = ( & 1000 0000) 表明最终只在乎最高位.

longword 分三种情况讨论

longword : 1XXX XXXX =< x <=
longword : 0XXX XXXX < x <
longword : x =

第一种 longword = 1XXX XXXX

那么 ~longword = 0YYY YYYY 显然 ~ longword & himagic = 0000 0000 不用继续了.

第二种 longword = 0XXX XXXX 且不为 0, 及不小于 1

显然 (longword - lomagic) = 0ZZZ ZZZ >= 0 且 < 127, 因为 lomagic = 1;

此刻 (longword - lomagic) & himagic = 0ZZZ ZZZZ & 1000 0000 = 0 , 所以也不需要继续了.

第三种 longword = 0000 0000

那么 ~longword & himagic = 1111 1111 & 1000 0000 = 1000 000;

再看 (longword - lomagic) = (0000 0000 - 0000 0001) , 由于无符号数减法是按照

(补码(0000 0000) + 补码(-000 0001)) = (补码(0000 0000) + 补码(~000 0001 + 1))

= (补码(0000 0000) + 补码(1111 1111)) = 1111 1111 (快捷的可以查公式得到最终结果),

因而 此刻最终结果为 1111 1111 & 1000 0000 = 1000 0000 > 0.

综合讨论, 可以根据上面公式巧妙的筛选出值是否为 0.  对于 2字节, 4 字节, 8 字节, 思路完全相似.

3.2 (sizeof (longword) > 4) ? (sizeof (longword) > 8) 为什么不用宏, 大展宏图呗 ?

宏可以做到多平台源码共享, 无法做到多平台二进制共享. glibc 这么通用项目, 可移植性影响因子

可能会很重. (性能是毒酒, 想活的久还是少喝 ~ )

4. libc_hidden_builtin_def (strlen) ? 闹哪样 ~

理解这个东西, 要引入些场外信息  (不同编译参数会不一样, 这里只抽取其中一条分支解法)

// file : glibc-2.31/include/libc-symbols.h

libc_hidden_builtin_def (strlen)

#define libc_hidden_builtin_def(name) libc_hidden_def (name)

define libc_hidden_def(name) hidden_def (name)

/* Define ALIASNAME as a strong alias for NAME. */

define strong_alias(name, aliasname) _strong_alias(name, aliasname)

define _strong_alias(name, aliasname) \

extern __typeof (name) aliasname __attribute__ ((alias (#name))) \
__attribute_copy__ (name);

/* For assembly, we need to do the opposite of what we do in C:
in assembly gcc __REDIRECT stuff is not in place, so functions
are defined by its normal name and we need to create the
__GI_* alias to it, in C __REDIRECT causes the function definition
to use __GI_* name and we need to add alias to the real name.
There is no reason to use hidden_weak over hidden_def in assembly,
but we provide it for consistency with the C usage.
hidden_proto doesn't make sense for assembly but the equivalent
is to call via the HIDDEN_JUMPTARGET macro instead of JUMPTARGET. */

define hidden_def(name) strong_alias (name, __GI_##name)

/* Undefine (also defined in libc-symbols.h). */
#undef __attribute_copy__
#if __GNUC_PREREQ (9, 0)
/* Copies attributes from the declaration or type referenced by
the argument. */

define __attribute_copy__(arg) __attribute__ ((__copy__ (arg)))

#else

define __attribute_copy__(arg)

#endif

利用上面宏定义, 进行展开

libc_hidden_builtin_def (strlen)
|

hidden_def (strlen)
|

strong_alias (strlen, __GI_strlen)
|

_strong_alias (strlen, __GI_strlen)
|

extern __typeof (strlen) __GI_strlen __attribute__ ((alias ("strlen"))) __attribute_copy__ (strlen);
|
extern __typeof (strlen) __GI_strlen __attribute__ ((alias ("strlen"))) __attribute__ ((__copy__ (strlen)));
``

其中 GUN C 扩展语法

　　__typeof (arg) : 获取变量的声明的类型

　　__attribute__ ((__copy__ (arg))) : GCC 9 以上版本 attribute copy 复制特性

　　alias_name __attribute__ ((alias (name))) : 为 name 声明符号别名 alias name.

总结:  libc_hidden_builtin_def (strlen) 意思是基于 strlen 符号, 重新定义一个符号别名 __GI_strlen. 

(补充资料 strong_alias 注释)

strlen 工程代码有很多种, 我们这里选择一个通用 glibc 版本去思考和分析. 有兴趣可以自行查阅更多. 

随口就来  ~ 做人嘛开心最重要 ~ 千锤百炼芮成钢 ~ 哈哈哈

　　错误是难免的, 欢迎指正和交流 ~