对象内存布局 (15)

重复继承

 

下面我们再来看看，发生重复继承的情况。所谓重复继承，也就是某个基类被间接地重复继承了多次。

 

下图是一个继承图，我们重载了父类的f()函数。

 

其类继承的源代码如下所示。其中，每个类都有两个变量，一个是整形（4字节），一个是字符（1字节），而且还有自己的虚函数，自己overwrite父类的虚函数。如子类D中，f()覆盖了超类的函数， f1()和f2() 覆盖了其父类的虚函数，Df()为自己的虚函数。

class B

{

    public:

        int ib;

        char cb;

    public:

        B():ib(0),cb('B') {}

 

        virtual void f() { cout << "B::f()" << endl;}

        virtual void Bf() { cout << "B::Bf()" << endl;}

};

class B1 :  public B

{

    public:

        int ib1;

        char cb1;

    public:

        B1():ib1(11),cb1('1') {}

 

        virtual void f() { cout << "B1::f()" << endl;}

        virtual void f1() { cout << "B1::f1()" << endl;}

        virtual void Bf1() { cout << "B1::Bf1()" << endl;}

 

};

class B2:  public B

{

    public:

        int ib2;

        char cb2;

    public:

        B2():ib2(12),cb2('2') {}

 

        virtual void f() { cout << "B2::f()" << endl;}

        virtual void f2() { cout << "B2::f2()" << endl;}

        virtual void Bf2() { cout << "B2::Bf2()" << endl;}

       

};

 

class D : public B1, public B2

{

    public:

        int id;

        char cd;

    public:

        D():id(100),cd('D') {}

 

        virtual void f() { cout << "D::f()" << endl;}

        virtual void f1() { cout << "D::f1()" << endl;}

        virtual void f2() { cout << "D::f2()" << endl;}

        virtual void Df() { cout << "D::Df()" << endl;}

       

};

我们用来存取子类内存布局的代码如下所示：（在VC++ 2003和G++ 3.4.4下）

    typedef void(*Fun)(void);

    int** pVtab = NULL;

    Fun pFun = NULL;

 

    D d;

    pVtab = (int**)&d;

    cout << "[0] D::B1::_vptr->" << endl;

    pFun = (Fun)pVtab[0][0];

    cout << "     [0] ";    pFun();

    pFun = (Fun)pVtab[0][1];

    cout << "     [1] ";    pFun();

    pFun = (Fun)pVtab[0][2];

    cout << "     [2] ";    pFun();

    pFun = (Fun)pVtab[0][3];

    cout << "     [3] ";    pFun();

    pFun = (Fun)pVtab[0][4];

    cout << "     [4] ";    pFun();

    pFun = (Fun)pVtab[0][5];

    cout << "     [5] 0x" << pFun << endl;

   

    cout << "[1] B::ib = " << (int)pVtab[1] << endl;

    cout << "[2] B::cb = " << (char)pVtab[2] << endl;

    cout << "[3] B1::ib1 = " << (int)pVtab[3] << endl;

    cout << "[4] B1::cb1 = " << (char)pVtab[4] << endl;

 

    cout << "[5] D::B2::_vptr->" << endl;

    pFun = (Fun)pVtab[5][0];

    cout << "     [0] ";    pFun();

    pFun = (Fun)pVtab[5][1];

    cout << "     [1] ";    pFun();

    pFun = (Fun)pVtab[5][2];

    cout << "     [2] ";    pFun();

    pFun = (Fun)pVtab[5][3];

    cout << "     [3] ";    pFun();

    pFun = (Fun)pVtab[5][4];

    cout << "     [4] 0x" << pFun << endl;

 

    cout << "[6] B::ib = " << (int)pVtab[6] << endl;

    cout << "[7] B::cb = " << (char)pVtab[7] << endl;   

    cout << "[8] B2::ib2 = " << (int)pVtab[8] << endl;

    cout << "[9] B2::cb2 = " << (char)pVtab[9] << endl;

 

    cout << "[10] D::id = " << (int)pVtab[10] << endl;

    cout << "[11] D::cd = " << (char)pVtab[11] << endl;

 

程序运行结果如下：

GCC 3.4.4	VC++ 2003
[0] D::B1::_vptr->      [0] D::f()      [1] B::Bf()      [2] D::f1()      [3] B1::Bf1()      [4] D::f2()      [5] 0x1 [1] B::ib = 0 [2] B::cb = B [3] B1::ib1 = 11 [4] B1::cb1 = 1 [5] D::B2::_vptr->      [0] D::f()      [1] B::Bf()      [2] D::f2()      [3] B2::Bf2()      [4] 0x0 [6] B::ib = 0 [7] B::cb = B [8] B2::ib2 = 12 [9] B2::cb2 = 2 [10] D::id = 100 [11] D::cd = D	[0] D::B1::_vptr->      [0] D::f()      [1] B::Bf()      [2] D::f1()      [3] B1::Bf1()      [4] D::Df()      [5] 0x00000000 [1] B::ib = 0 [2] B::cb = B [3] B1::ib1 = 11 [4] B1::cb1 = 1 [5] D::B2::_vptr->      [0] D::f()      [1] B::Bf()      [2] D::f2()      [3] B2::Bf2()      [4] 0x00000000 [6] B::ib = 0 [7] B::cb = B [8] B2::ib2 = 12 [9] B2::cb2 = 2 [10] D::id = 100 [11] D::cd = D

下面是对于子类实例中的虚函数表的图：

 

 

 

我们可以看见，最顶端的父类B其成员变量存在于B1和B2中，并被D给继承下去了。而在D中，其有B1和B2的实例，于是B的成员在D的实例中存在两份，一份是B1继承而来的，另一份是B2继承而来的。所以，如果我们使用以下语句，则会产生二义性编译错误：

D d;

d.ib = 0;               //二义性错误

d.B1::ib = 1;           //正确

d.B2::ib = 2;           //正确

注意，上面例程中的最后两条语句存取的是两个变量。虽然我们消除了二义性的编译错误，但B类在D中还是有两个实例，这种继承造成了数据的重复，我们叫这种继承为重复继承。重复的基类数据成员可能并不是我们想要的。所以，C++引入了虚基类的概念。

 

 

钻石型多重虚拟继承

 

虚拟继承的出现就是为了解决重复继承中多个间接父类的问题的。钻石型的结构是其最经典的结构。也是我们在这里要讨论的结构：

 

上述的“重复继承”只需要把B1和B2继承B的语法中加上virtual 关键，就成了虚拟继承，其继承图如下所示：

 

上图和前面的“重复继承”中的类的内部数据和接口都是完全一样的，只是我们采用了虚拟继承：其省略后的源码如下所示：

 

class B {……};

class B1 : virtual public B{……};

class B2: virtual public B{……};

class D : public B1, public B2{ …… };

 

在查看D之前，我们先看一看单一虚拟继承的情况。下面是一段在VC++2003下的测试程序：（因为VC++和GCC的内存而局上有一些细节上的不同，所以这里只给出VC++的程序，GCC下的程序大家可以根据我给出的程序自己仿照着写一个去试一试）：

 

    int** pVtab = NULL;

    Fun pFun = NULL;

 

    B1 bb1;

 

    pVtab = (int**)&bb1;

    cout << "[0] B1::_vptr->" << endl;

    pFun = (Fun)pVtab[0][0];

    cout << "     [0] ";

    pFun(); //B1::f1();

    cout << "     [1] ";

    pFun = (Fun)pVtab[0][1];

    pFun(); //B1::bf1();

    cout << "     [2] ";

    cout << pVtab[0][2] << endl;

 

    cout << "[1] = 0x";

    cout << (int*)*((int*)(&bb1)+1) <<endl; //B1::ib1

    cout << "[2] B1::ib1 = ";

    cout << (int)*((int*)(&bb1)+2) <<endl; //B1::ib1

    cout << "[3] B1::cb1 = ";

    cout << (char)*((int*)(&bb1)+3) << endl; //B1::cb1

    cout << "[4] = 0x";

    cout << (int*)*((int*)(&bb1)+4) << endl; //NULL

    cout << "[5] B::_vptr->" << endl;

    pFun = (Fun)pVtab[5][0];

    cout << "     [0] ";

    pFun(); //B1::f();

    pFun = (Fun)pVtab[5][1];

    cout << "     [1] ";

    pFun(); //B::Bf();

    cout << "     [2] ";

    cout << "0x" << (Fun)pVtab[5][2] << endl;

 

    cout << "[6] B::ib = ";

    cout << (int)*((int*)(&bb1)+6) <<endl; //B::ib

    cout << "[7] B::cb = ";

 

其运行结果如下（我结出了GCC的和VC++2003的对比）：

 

GCC 3.4.4	VC++ 2003
[0] B1::_vptr ->     [0] : B1::f()     [1] : B1::f1()     [2] : B1::Bf1()     [3] : 0 [1] B1::ib1 : 11 [2] B1::cb1 : 1 [3] B::_vptr ->     [0] : B1::f()     [1] : B::Bf()     [2] : 0 [4] B::ib : 0 [5] B::cb : B [6] NULL : 0	[0] B1::_vptr->      [0] B1::f1()      [1] B1::Bf1()      [2] 0 [1] = 0x00454310 ç该地址取值后是-4 [2] B1::ib1 = 11 [3] B1::cb1 = 1 [4] = 0x00000000 [5] B::_vptr->      [0] B1::f()      [1] B::Bf()      [2] 0x00000000 [6] B::ib = 0 [7] B::cb = B

 

 

这里，大家可以自己对比一下。关于细节上，我会在后面一并再说。

 

下面的测试程序是看子类D的内存布局，同样是VC++ 2003的（因为VC++和GCC的内存布局上有一些细节上的不同，而VC++的相对要清楚很多，所以这里只给出VC++的程序，GCC下的程序大家可以根据我给出的程序自己仿照着写一个去试一试）：

 

    D d;

    pVtab = (int**)&d;

    cout << "[0] D::B1::_vptr->" << endl;

    pFun = (Fun)pVtab[0][0];

    cout << "     [0] ";    pFun(); //D::f1();

    pFun = (Fun)pVtab[0][1];

    cout << "     [1] ";    pFun(); //B1::Bf1();

    pFun = (Fun)pVtab[0][2];

    cout << "     [2] ";    pFun(); //D::Df();

    pFun = (Fun)pVtab[0][3];

    cout << "     [3] ";

    cout << pFun << endl;

 

    //cout << pVtab[4][2] << endl;

    cout << "[1] = 0x";

    cout <<  (int*)((&dd)+1) <<endl; //????

    cout << "[2] B1::ib1 = ";

    cout << *((int*)(&dd)+2) <<endl; //B1::ib1

    cout << "[3] B1::cb1 = ";

    cout << (char)*((int*)(&dd)+3) << endl; //B1::cb1

    //---------------------

    cout << "[4] D::B2::_vptr->" << endl;

    pFun = (Fun)pVtab[4][0];

    cout << "     [0] ";    pFun(); //D::f2();

    pFun = (Fun)pVtab[4][1];

    cout << "     [1] ";    pFun(); //B2::Bf2();

    pFun = (Fun)pVtab[4][2];

    cout << "     [2] ";

    cout << pFun << endl;

   

    cout << "[5] = 0x";

    cout << *((int*)(&dd)+5) << endl; // ???

    cout << "[6] B2::ib2 = ";

    cout << (int)*((int*)(&dd)+6) <<endl; //B2::ib2

    cout << "[7] B2::cb2 = ";

    cout << (char)*((int*)(&dd)+7) << endl; //B2::cb2

    cout << "[8] D::id = ";

    cout << *((int*)(&dd)+8) << endl; //D::id

    cout << "[9] D::cd = ";

    cout << (char)*((int*)(&dd)+9) << endl;//D::cd

    cout << "[10]  = 0x";

    cout << (int*)*((int*)(&dd)+10) << endl;

    //---------------------

    cout << "[11] D::B::_vptr->" << endl;

    pFun = (Fun)pVtab[11][0];

    cout << "     [0] ";    pFun(); //D::f();

    pFun = (Fun)pVtab[11][1];

    cout << "     [1] ";    pFun(); //B::Bf();

    pFun = (Fun)pVtab[11][2];

    cout << "     [2] ";

    cout << pFun << endl;

 

    cout << "[12] B::ib = ";

    cout << *((int*)(&dd)+12) << endl; //B::ib

    cout << "[13] B::cb = ";

    cout << (char)*((int*)(&dd)+13) <<endl;//B::cb

 完整的测试代码：

#include<iostream>
using namespace std;

class B
{
public:
    int ib;
    char cb;
public:
    B():ib(10),cb('B') {}

    virtual void f()
    {
        cout << "B::f()" << endl;
    }
    virtual void Bf()
    {
        cout << "B::Bf()" << endl;
    }
};
class B1 :  public virtual B
{
public:
    int ib1;
    char cb1;
public:
    B1():ib1(11),cb1('1') {}

    virtual void f()
    {
        cout << "B1::f()" << endl;
    }
    virtual void f1()
    {
        cout << "B1::f1()" << endl;
    }
    virtual void Bf1()
    {
        cout << "B1::Bf1()" << endl;
    }

};
class B2:  public virtual B
{
public:
    int ib2;
    char cb2;
public:
    B2():ib2(12),cb2('2') {}

    virtual void f()
    {
        cout << "B2::f()" << endl;
    }
    virtual void f2()
    {
        cout << "B2::f2()" << endl;
    }
    virtual void Bf2()
    {
        cout << "B2::Bf2()" << endl;
    }

};

class D : public B1, public B2
{
public:
    int id;
    char cd;
public:
    D():id(100),cd('D') {}

    virtual void f()
    {
        cout << "D::f()" << endl;
    }
    virtual void f1()
    {
        cout << "D::f1()" << endl;
    }
    virtual void f2()
    {
        cout << "D::f2()" << endl;
    }
    virtual void Df()
    {
        cout << "D::Df()" << endl;
    }

};


typedef void(*Fun)(void);

int main()
{
    int** pVtab = NULL;
    Fun pFun = NULL;

    B1 bb1;

    pVtab = (int**)&bb1;
    //第一个虚函数表的地址
    cout<<"[0] 0x ";
    cout<<(int*)pVtab[0]<<endl;
    cout << "[0] B1::_vptr->" << endl;
    pFun = (Fun)pVtab[0][0];
    cout << "     [0] ";
    pFun(); //B1::f1();
    cout << "     [1] ";
    pFun = (Fun)pVtab[0][1];
    pFun(); //B1::bf1();
    cout << "     [2] ";
    pFun = (Fun)pVtab[0][2];
    pFun(); //B1::bf1();
    cout << "     [3] ";
    cout << pVtab[0][3] << endl;

    cout << "[1] B1::ib1 = ";
    cout << *((int*)(&bb1)+1) <<endl; //B1::ib1
    cout << "[2] B1::cb1 = ";
    cout << (char)*((int*)(&bb1)+2) <<endl; //B1::cb1
    //第二个虚函数表的地址
    cout<< "[3] 0x ";
    cout<<(int*)pVtab[3]<<endl;
    cout << "[3] B::_vptr->" << endl;
    pFun = (Fun)pVtab[3][0];
    cout << "     [0] ";
    pFun(); //B1::f();
    pFun = (Fun)pVtab[3][1];
    cout << "     [1] ";
    pFun(); //B::Bf();
    cout << "     [2] ";
    cout << "0x " << (Fun)pVtab[3][2] << endl;

    cout << "[4] B::ib = ";
    cout << (int)*((int*)(&bb1)+4) <<endl; //B::ib
    cout << "[5] B::cb = ";
    cout << (char)*((int*)(&bb1)+5) <<endl; //B::cb
    cout<<endl;

    D d;

    pVtab = (int**)&d;
     //第一个虚函数表的地址
    cout<<"[0] 0x ";
    cout<<(int*)pVtab[0]<<endl;
    cout << "[0] D::B1::_vptr->" << endl;
    pFun = (Fun)pVtab[0][0];
    cout << "     [0] ";
    pFun(); //D::f1();
    pFun = (Fun)pVtab[0][1];
    cout << "     [1] ";
    pFun(); //B1::Bf1();
    pFun = (Fun)pVtab[0][2];
    cout << "     [2] ";
    pFun(); //D::Df();
    pFun = (Fun)pVtab[0][3];
    cout << "     [3] ";
    pFun();
    pFun = (Fun)pVtab[0][4];
    cout << "     [4] ";
    pFun();
    pFun = (Fun)pVtab[0][5];
    cout << "     [5] ";
    cout << pFun << endl;

    //cout << pVtab[4][2] << endl;
    cout << "[1] = 0x ";
    cout <<  *((int*)((&d)+1)) <<endl; //????

    cout << "[1] B1::ib1 = ";
    cout << *((int*)(&d)+1) <<endl; //B1::ib1
    cout << "[2] B1::cb1 = ";
    cout << (char)*((int*)(&d)+2) << endl; //B1::cb1
    //---------------------
     //第二个虚函数表的地址
    cout<<"[3] 0x ";
    cout<<(int*)pVtab[3]<<endl;
    cout << "[3] D::B2::_vptr->" << endl;
    pFun = (Fun)pVtab[3][0];
    cout << "     [0] ";
    pFun(); //D::f2();
    pFun = (Fun)pVtab[3][1];
    cout << "     [1] ";
    pFun(); //B2::Bf2();
    pFun = (Fun)pVtab[3][2];
    cout << "     [2] ";
    pFun();
    pFun = (Fun)pVtab[3][3];
    cout << "     [3] ";
    cout << pFun << endl;

    cout << "[4] B2::ib2 = ";
    cout << (int)*((int*)(&d)+4) <<endl; //B2::ib2
    cout << "[5] B2::cb2 = ";
    cout << (char)*((int*)(&d)+5) << endl; //B2::cb2

    cout << "[6] D::id = ";
    cout << *((int*)(&d)+6) << endl; //D::id
    cout << "[7] D::cd = ";
    cout << (char)*((int*)(&d)+7) << endl;//D::cd

    cout << "[8]  = 0x ";
    cout << (int*)*((int*)(&d)+8) << endl;
    //---------------------
     //第三个虚函数表的地址
    cout<<"[8] 0x ";
    cout<<(int*)pVtab[8]<<endl;
    cout << "[8] D::B::_vptr->" << endl;
    pFun = (Fun)pVtab[8][0];
    cout << "     [0] ";
    pFun(); //D::f();
    pFun = (Fun)pVtab[8][1];
    cout << "     [1] ";
    pFun(); //B::Bf();
    pFun = (Fun)pVtab[8][2];
    cout << "     [2] ";
    cout << pFun << endl;

    cout << "[9] B::ib = ";
    cout << *((int*)(&d)+9) << endl; //B::ib
    cout << "[10] B::cb = ";
    cout << (char)*((int*)(&d)+10) <<endl;//B::cb
}

执行结果如下：

下面给出运行后的结果（分VC++和GCC两部份）

GCC 3.4.4

VC++ 2003

[0] B1::_vptr ->     [0] : D::f()     [1] : D::f1()     [2] : B1::Bf1()     [3] : D::f2()     [4] : D::Df()     [5] : 1 [1] B1::ib1 : 11 [2] B1::cb1 : 1 [3] B2::_vptr ->     [0] : D::f()     [1] : D::f2()     [2] : B2::Bf2()     [3] : 0 [4] B2::ib2 : 12 [5] B2::cb2 : 2 [6] D::id : 100 [7] D::cd : D [8] B::_vptr ->     [0] : D::f()     [1] : B::Bf()     [2] : 0 [9] B::ib : 0 [10] B::cb : B [11] NULL : 0

[0] D::B1::_vptr->      [0] D::f1()      [1] B1::Bf1()      [2] D::Df()      [3] 00000000 [1] = 0x0013FDC4  ç 该地址取值后是-4 [2] B1::ib1 = 11 [3] B1::cb1 = 1 [4] D::B2::_vptr->      [0] D::f2()      [1] B2::Bf2()      [2] 00000000 [5] = 0x4539260   ç 该地址取值后是-4 [6] B2::ib2 = 12 [7] B2::cb2 = 2 [8] D::id = 100 [9] D::cd = D [10]  = 0x00000000 [11] D::B::_vptr->      [0] D::f()      [1] B::Bf()      [2] 00000000 [12] B::ib = 0 [13] B::cb = B

关于虚拟继承的运行结果我就不画图了（前面的作图已经让我产生了很严重的厌倦感，所以就偷个懒了，大家见谅了）

//                  __|B1:vfptr |--->|D::f1()  |
//  _________      |  |_________|    |_________|
// |   -4    |<----|--|B1:vbptr |    |B1::Bf1()|
// |_________|       |  |_________|    |_________|
// |   40    |<----|  |B1::ib1  |    |D::Df()  |
// |_________|       |  |_________|    |_________|
//                   |  |B1::cb1  |
//                   |  |_________|      _________
//                _|__|B2:vfptr |--->|D::f2()  |      
//  _________     | |  |_________|    |________ |
// |   -4    |<--|-|--|B2:vbptr |    |B2::Bf2()|    
// |_________|   | |  |_________|    |_________|
// |   24    |<--| |  |B2::ib2  |
// |_________|   | |  |_________|           
//                 | |  |B2::cb2  |
//                 | |  |_________|
//                 | |  |D::id    |
//                 | |  |_________|
//                 | |  |D::cd    | 
//                 | |  |_________|      _________
//                  |_|__|B:vfptr  |--->| D::f()  |             
//                      |_________|    |_________|
//                      |B::ib    |    | B::Bf() |    
//                      |_________|     |_________|
//                      |B::cb    |       

在上面的输出结果中，我用不同的颜色做了一些标明。我们可以看到如下的几点：

 

1）无论是GCC还是VC++，除了一些细节上的不同，其大体上的对象布局是一样的。也就是说，先是B1（黄色），然后是B2（绿色），接着是D（灰色），而B这个超类（青蓝色）的实例都放在最后的位置。

2）关于虚函数表，尤其是第一个虚表，GCC和VC++有很重大的不一样。但仔细看下来，还是VC++的虚表比较清晰和有逻辑性。

3）VC++和GCC都把B这个超类放到了最后，而VC++有一个NULL分隔符把B和B1和B2的布局分开。GCC则没有。

4）VC++中的内存布局有两个地址我有些不是很明白，在其中我用红色标出了。取其内容是-4。接道理来说，这个指针应该是指向B类实例的内存地址（这个做法就是为了保证重复的父类只有一个实例的技术）。但取值后却不是。这点我目前还并不太清楚，还向大家请教。

5）GCC的内存布局中在B1和B2中则没有指向B的指针。这点可以理解，编译器可以通过计算B1和B2的size而得出B的偏移量。

结束语

C++这门语言是一门比较复杂的语言，对于程序员来说，我们似乎永远摸不清楚这门语言背着我们在干了什么。需要熟悉这门语言，我们就必需要了解C++里面的那些东西，需要我们去了解他后面的内存对象。这样我们才能真正的了解C++，从而能够更好的使用C++这门最难的编程语言。