向下转换指向成员函数的指针.这是合法用法吗?

我将指向成员函数的指针列表存储在数组中。 我想索引到数组中并执行相应的函数。 将有许多数组列出来自不同类的函数(全部派生自 Base(，因此在编译时不知道该类。

我有这个方案在工作，但我对不得不在一个地方使用空指针并不完全满意，但我似乎无法避免它。

根据 C++11 标准，我在 Base 和派生成员函数指针之间的转换是否合法(它正在使用 g++(。我将不胜感激语言律师的建议！

下面是我的代码的精简版，但可运行。

#include <iostream>
using std::cout;
//*************************************
class Base {
public:
  typedef int (Base::*BaseThunk)();
  virtual int execute(BaseThunk x) {return 0;}
};
//*************************************
class Derived : public Base {
public:
  typedef int (Derived::*DerivedThunk)();
  int execute(BaseThunk step) {
    return (this->*step)(); //Is this OK ? step is really a DerivedThunk.
  }
  int f1() { cout<<"1n";return 1;}
  int f2() { cout<<"2n";return 2;}
  int f3() { cout<<"3n";return 3;}
  static DerivedThunk steps[];
};
//Here is an array of pointers to member functions of the Derived class.
Derived::DerivedThunk Derived::steps[] = {&Derived::f1, &Derived::f2, &Derived::f3};
//*************************************
class Intermediate : public Base {
public:
  void f(void *x) { //I am worried about using void pointer here !
    BaseThunk *seq = reinterpret_cast<BaseThunk *>(x);
    Derived d;
    d.execute(seq[2]);
  }
};
//*************************************
int main() {
  Intermediate b;
  b.f(&Derived::steps);
}