linux系统编程：setjmp和longjmp函数用法

#include <stdio.h>

#include <setjmp.h>

//jmp_buf：数组，保存栈信息即运行环境

jmp_buf buf;

double Divide(double a, double b)

{

if (b == 0.0)

{

longjmp(buf, 1); // throw

}

else

return a / b;

}

//setjmp保存当前栈信息，成功返回0，当执行到longjmp时，

//恢复栈信息即跳转到setjmp位置重新执行setjmp

//且此次返回值根据longjmp函数参数给定

int main(void)

{

int ret;

ret = setjmp(buf);

if (ret == 0) // try

{

printf("division ...\n");

printf("%f\n", Divide(5.0, 0.0));

}

else if (ret == 1) // catch

{

printf("divisiong by zero\n");

}

return 0;

}

运行结果：

division ...

divisiong by zero

其实上次错误处理模式已经是c++异常处理雏形

上述相当于：

#include <iostream>

using namespace std;

double Divide(double a, double b)

{

if (b == 0.0)

{

throw 1; // throw

}

else

return a / b;

}

int main(void)

{

try // try

{

cout<<"division ..."<<endl;

cout<<Divide(3.0, 0.0)<<endl;

cout<<Divide(5.0, 0.0)<<endl;

}

catch (int) // catch

{

cout<<"divisiong by zero"<<endl;

}

return 0;

}

运行结果同上

转载于:https://www.cnblogs.com/mickole/p/3258603.html