Question

我已经在.Net Core 2.2中编写了下一个测试程序，并在3台不同的PC上执行它得到的结果大致相同：第二种执行方法总是执行得更快。

代码：

using System;
using System.Diagnostics;

class Program
{
    static void Main()
    {
        for (int i = 0; i < 10; i++)
        {
            RunTest();
        }
    }

    private static void RunTest()
    {
        const long cnt = 1000000000;

        var time1 = Test1(cnt);
        var time2 = Test2(cnt);

        Console.WriteLine($"Done in: {time1} vs {time2} diff: {time1 - time2}");
    }

    private static TimeSpan Test1(long cnt)
    {
        var v = 1;
        var timer = Stopwatch.StartNew();
        for (long i = cnt; i != 0; i--)
        {
            v = RotateBit(v);
        }
        timer.Stop();

        return timer.Elapsed;
    }

    private static TimeSpan Test2(long cnt)
    {
        var v = 1;
        var timer = Stopwatch.StartNew();
        for (long i = cnt; i != 0; i--)
        {
            v = RotateBit(v);
        }
        timer.Stop();

        return timer.Elapsed;
    }

    private static int RotateBit(int v)
    {
        v = v << 1;

        if ((v & 0b0000_1111) == 0)
        {
            v = 1;
        }

        return v;
    }
}

请注意，Test1和Test2方法是绝对复制。没有第一个Test1或Test2的仪表，首先执行的仪表总是快将近2倍：Done in: 00:00:00.8881566 vs 00:00:00.4706360 diff: 00:00:00.4175206，即使RunTest方法执行了多次。

发生这种奇怪行为的原因可能是什么？

执行该方法的副本始终可以使性能提高2倍

0 个答案: