本文对分别测试VC，MinGW，GCC 三种编译器，32位和64位模式，共6种情况下，和64位编程有关的预定义宏的值。对跨平台编程具有参考意义。
Agner Fog 在他的《Calling conventions for different C++ compilers and operating systems》提到一些预定义宏。这里摘录如下。
注：下面的内容来自《Calling conventions for different C++ compilers and operating systems》,Last updated 2012-02-29，作者：By Agner Fog. Copenhagen University College .如与原文内容不符，以原文为准。
Most C++ compilers have a predefined macro containing the version number of the compiler. Programmers can use preprocessing directives to check for the existence of these macros in order to detect which compiler the program is compiled on and thereby fix problems with incompatible compilers.
表格23. 编译器版本预定义宏
|
Compiler |
Predefined macro |
|
Borland |
__BORLANDC__ |
|
Codeplay VectorC |
__VECTORC__ |
|
Digital Mars |
__DMC__ |
|
Gnu |
__GNUC__ |
|
Intel |
__INTEL_COMPILER |
|
Microsoft |
_MSC_VER |
|
Pathscale |
__PATHSCALE__ |
|
Symantec |
__SYMANTECC__ |
| Watcom | __WATCOMC__ |
不幸的是，并不是所有编译器都有良好的文档来告诉你哪些宏是用来说明正在编译的硬件平台和操作系统的。以下宏可以被定义也可以不被定义
表 24. 硬件平台预定义宏
| platform |
Predefined macro |
||
| x86 | _M_IX86 | __INTEL__ | __i386__ |
| x86-64 | _M_X64 | __x86_64__ | __amd64 |
| IA64 | __IA64__ | ||
| DEC Alpha | __ALPHA__ | ||
| Motorola Power PC | __POWERPC__ | ||
| Any little endian | __LITTLE_ENDIAN__ | ||
| Any big endian | __BIG_ENDIAN__ | ||
Table 25. 操作系统预定义宏
|
Operating system |
Predefined macro |
|||
|
DOS 16 bit |
__MSDOS__ |
_MSDOS |
|
|
|
Windows 16 bit |
_WIN16 |
|
|
|
|
Windows 32 bit |
_WIN32 |
__WINDOWS__ |
|
|
|
Windows 64 bit |
_WIN64 |
_WIN32 |
|
|
|
Linux 32 bit |
__unix__ |
__linux__ |
|
|
|
Linux 64 bit |
__unix__ |
__linux__ |
__LP64__ |
__amd64 |
|
BSD |
__unix__ |
__BSD__ |
__FREEBSD__ |
|
|
Mac OS |
__APPLE__ |
__DARWIN__ |
__MACH__ |
|
| OS/2 | __OS2__ | |||
下面的代码主要测试和64编程有关的宏
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 |
void test() { int len=sizeof(int)*8; printf("sizeof(int)=%d\n",len); len=sizeof(int *)*8; printf("sizeof(int*)=%d\n",len); #ifdef _MSC_VER printf("_MSC_VER is defined\n"); #endif #ifdef __GNUC__ printf("__GNUC__ is defined\n"); #endif #ifdef __INTEL__ printf("__INTEL__ is defined\n"); #endif #ifdef __i386__ printf("__i386__ is defined\n"); #endif #ifdef __x86_64__ printf("__x86_64__ is defined\n"); #endif #ifdef _WIN32 printf("_WIN32 is defined\n"); #endif #ifdef _WIN64 printf("_WIN64 is defined\n"); #endif #ifdef __linux__ printf("__linux__ is defined\n"); #endif #ifdef __LP64__ printf("__LP64__ is defined\n"); #endif #ifdef __amd64 printf("__amd64 is defined\n"); #endif } int main(int argc, char* argv[]) { test(); return 0; } |
最后给出结果.
1. 在VC2010， 32位模式下编译，输出下面的信息
sizeof(int)=32
sizeof(int*)=32
_MSC_VER is defined
_WIN32 is defined
2. 在MinGW下编译 输出下面的信息
sizeof(int)=32
sizeof(int*)=32
__GNUC__ is defined
__i386__ is defined
_WIN32 is defined
3. 在64位Fedora19, 使用gcc -m32 编译，输出下面的信息
sizeof(int)=32
sizeof(int*)=32
__GNUC__ is defined
__i386__ is defined
__linux__ is defined
4. 在VC2010， 64位模式下编译，输出下面的信息
sizeof(int)=32
sizeof(int*)=64
_MSC_VER is defined
_WIN32 is defined
_WIN64 is defined
5. 在MinGW64下编译 输出下面的信息
sizeof(int)=32
sizeof(int*)=64
__GNUC__ is defined
__x86_64__ is defined
_WIN32 is defined
_WIN64 is defined
__amd64 is defined
6. 在64位Fedora19, 使用gcc -m64 编译，输出下面的信息
sizeof(int)=32
sizeof(int*)=64
__GNUC__ is defined
__x86_64__ is defined
__linux__ is defined
__LP64__ is defined
__amd64 is defined
注：在VC下直接用集成环境编译，在MinGW和Linux下直接使用 gcc来编译。