Following the stackoverflow question.

I've encountered with one performance problem.
Let's talk code. I simplified the code as much as I could to reproduce the issue.
Suppose we have a generic class. It has an empty list inside and does something with T in constructor. It has Run method that calls an IEnumerable<T> method on the list, e.g. Any().

public class BaseClass<T>
{
    private List<T> _list = new List<T>();

    public BaseClass()
    {
        Enumerable.Empty<T>();
        // or Enumerable.Repeat(new T(), 10);
        // or even new T();
        // or foreach (var item in _list) {}
    }

    public void Run()
    {
        for (var i = 0; i < 8000000; i++)
        {
            if (_list.Any())
            // or if (_list.Count() > 0)
            // or if (_list.FirstOrDefault() != null)
            // or if (_list.SingleOrDefault() != null)
            // or other IEnumerable<T> method
            {
                return;
            }
        }
    }
}

Then we have a derived class which is empty:

public class DerivedClass : BaseClass<object>
{
}

Let's measure the performance of running ClassBase<T>.Run method from both classes. Accessing from derived type is 4 times slower that from base class. And I can't understand why that happens. Compiled in Release mode, result is the same with warm up. It happens on .NET 4.5 only.

public class Program
{
    public static void Main()
    {
        Measure(new DerivedClass());
        Measure(new BaseClass<object>());
    }

    private static void Measure(BaseClass<object> baseClass)
    {
        var sw = Stopwatch.StartNew();
        baseClass.Run();
        sw.Stop();
        Console.WriteLine(sw.ElapsedMilliseconds);
    }
}

Full listing on gist

I've got an answer on Microsoft Connect

It is related to dictionary lookups in shared generics code. The heuristic in runtime and JIT do not work well for this particular test. We will take a look what can be done about it.

In the meantime, you can workaround it by adding two dummy methods to the BaseClass (do not even need to be called). It will cause the heuristic to work as one would expect.