How Much Slower Are C# Exceptions Than Return Codes?

The abnormal processing mechanism in C#is not significant in the case of stacks, but its performance loss is still worthy of attention.

A recent study shows that it takes about 29914 milliseconds to deal with one million abnormalities, that is, about 33 abnormalities per milliseconds. Nevertheless, in most cases, the use of anomalies to replace the return code is still efficient enough.

However, it is important to note that the exception processing speed is at least 30,000 times slower than the return code. Moreover, these data are only minimal estimates, and additional performance expenses will be generated in the actual application scenarios.

The test was performed on a laptop computer equipped with Intel Core2 Duo T8100 @ 2.1 GHz processor, using the .NET 4.0 framework, published a version, and excluded the debugger (the debugger would significantly affect performance).

The following is the test code used in the experiment:

static void Main(string[] args)
{
    int iterations = 1000000;
    Console.WriteLine("Starting " + iterations.ToString() + " iterations...\n");

    var stopwatch = new Stopwatch();

    // 测试异常
    stopwatch.Reset();
    stopwatch.Start();
    for (int i = 1; i <= iterations; i++)
    {
        try
        {
            // 模拟异常抛出
            throw new Exception();
        }
        catch (Exception)
        {
            // 捕获异常
        }
    }
    stopwatch.Stop();
    Console.WriteLine("Exceptions took: " + stopwatch.ElapsedMilliseconds + "ms");


    // 测试返回码 (此处应添加返回码测试代码)

}

(Note: The original text lacks the return code test code part, which retains the placeholders here)

The above is the detailed content of How Much Slower Are C# Exceptions Than Return Codes?. For more information, please follow other related articles on the PHP Chinese website!