理论上代码一和代码二计算量是一样，但是单经常代码二慢几乎是两倍，为什么？先看以下实验：

代码一：单个循环

1
2
3
4
5
6
7

int i;
for (i = 0; i < exchanges; i++)
{
    // some code
    // some more code
    // even more code
}

代码二：三个循环

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15

int i;
for (i = 0; i < exchanges; i++)
{
    // some code
}

for (i = 0; i < exchanges; i++)
{
    // some more code
}

for (i = 0; i < exchanges; i++)
{
    // even more code
}

测试代码

 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

#include <time.h>
#include <iostream>
using namespace std;
#define TOGETHER 1;
int main(){

    int size = 10000;
    int *data = new int[size];

    clock_t start = clock();

    for (int i = 0; i < 1000000; i++)
    {
        #ifdef TOGETHER
                //10.114s
                for (int c = 0; c < size; c++){
                    data[c] *= 10;
                    data[c] += 7;
                    data[c] &= 15;
                }
        #else
                //20.695s
                for (int c = 0; c < size; c++){
                    data[c] *= 10;
                }
                for (int c = 0; c < size; c++){
                    data[c] += 7;
                }
                for (int c = 0; c < size; c++){
                    data[c] &= 15;
                }
        #endif
    }

    clock_t end = clock();
    cout << (double)(end - start) / CLOCKS_PER_SEC << endl;
}

我的测试平台一：

名称 配置 备注
CPU Intel(R) Core(TM) i3-7100U CPU @ 2.40GHz
RAM 4GB
OS Windows Server 2019 Datacenter
ESXi 6.7.0 Update 3 (Build 14320388)		 建立在Exsi上 的虚拟机
c++ GCC 9.2.0 (Rev2, Built by MSYS2 project)
编译参数 -std=gnu++11 -O2 -Wall -lstdc++fs -lm
测试结果
1个循环 9.671s
3个循环 20.062s

我的测试平台二：

名称 配置 备注
CPU Intel(R) Core(TM) i7-4790K CPU @ 4.00GHz
RAM 16GB
OS Windows 10 企业版 1809 LTSC 物理机
c++ GCC 9.2.0 (Rev2, Built by MSYS2 project)
编译参数 -std=gnu++11 -O2 -Wall -lstdc++fs -lm
测试结果
1个循环 4.762s
3个循环 10.231s

原因

Each additional for-loop would cost the equivalent of two int allocations.

参考

https://stackoverflow.com/questions/13866962/what-is-the-overhead-in-splitting-a-for-loop-into-multiple-for-loops-if-the-tot