【Geekband】专题二：虚指针和内存分配

1. 传统继承类的设计

static void print_object(const char* name, void* this_, size_t size) {
      void** ugly = reinterpret_cast<void**>(this_);
      size_t i;
      printf("created %s at address %p of size %zu\n", name, this_, size);
      for(i = 0 ; i < size / sizeof(void*) ; i++) {
        printf("  pointer[%zu] == %p\n", i, ugly[i]);
      }
      cout << "-------------------------------"<<endl;
} 
class A {
public:
    long iA;
    long iAA;
    A ():iA (1), iAA(10) {print_object(__FUNCTION__, this, sizeof(*this));
}
    virtual void f() { cout << "A::f()" << endl; }
    virtual void g() { cout << "A::g()" << endl; }
    virtual void h() { cout << "A::h()" << endl; }
 
};
 
class B : public A {
public:
    long iB;
    B():iB(2) {
        print_object(__FUNCTION__, this, sizeof(*this));
 
}
    virtual void f() { cout << "B::f()" << endl; }
    virtual void g_B() { cout << "B::g_B()" << endl; }
    virtual void h_B() { cout << "B::h_B()" << endl; }
}; 
class C : public B{
public:
    long iC;
    C():iC(3) {
        print_object(__FUNCTION__, this, sizeof(*this));
}
    virtual void f() { cout << "C::f()" << endl; }
    virtual void g_B() { cout << "C::g_B()" << endl; }
    virtual void h_C() { cout << "C::h_C()" << endl; }
};

• A、B、C依次继承
• 并在构造过程中通过print_object输出构建信息

2. 构造Class C，并通过指针来一次访问内容或调用函数

int main()
{
    typedef void(*Fun)(void);
    long** pVtab = nullptr;
    Fun pFun = nullptr; 
    C c; 
    pVtab = (long**)&c;
    long *cAsLong = (long *)&c;        
    cout << "C::vptr memory address is " << &cAsLong[0] <<endl;
    cout << "[0] C::_vptr->" << endl;
    cout << "    No.\tMemory Address\t Value\t\tFunction" << endl;
    for (int i=0; (Fun)pVtab[0][i] != nullptr; i++){
                pFun = (Fun)pVtab[0][i];
                cout << "    ["<<i<<"]\t";
                cout << " " << &pVtab[0][i]<<"\t";;
                cout << " " << (void *)pVtab[0][i]<<"\t";;
                pFun();
    }
    cout << " ----------------------------------" << endl;
    cout << "[0] addr = " << &pVtab[0] << "\t";
    cout << "value = " << pVtab[0] << endl;
    cout << "[1] addr = "<< &pVtab[1] << "\t";
    cout << "value = " << pVtab[1] << "\t";    
    cout << "A.iA = " << *((long *) &pVtab[1] )<< endl;
    cout << c.iA << endl;   
    cout << "[2] addr = " << &pVtab[2] << "\t";
    cout << "value = " << pVtab[2] << "\t";    
    cout << "A.iAA = " << *((long *) &pVtab[2]) << endl;        
    cout << "[3] addr = " << &pVtab[3] << "\t";
    cout << "value = " << pVtab[3] << "\t";    
    cout << "B.iB = " << *((long *) &pVtab[3]) << endl;
    cout << "[4] addr = " << &pVtab[4] << "\t";
    cout << "value = " << pVtab[4] << "\t";    
    cout << "C.iC = " << *((long *) &pVtab[4]) << endl;
    return 0;

• pVtab用于显示虚表的位置，并用于后续的直接调用
• 通过循环，一次输出虚表中各个成员的地址和对应的存储内容
• 最后，通过直接调用相应的存储内容来调用成员或调用函数。

3. gdb的使用方法

• 编译过程

g++ -g -o test.cpp test

• gdb运行程序

(gdb) r

• gdb中设置断点

(gdb) break line_number

• gdb中查询虚表

(gdb) info vtbl c

• gdb中查询内存内容 - 选择输出的格式

(gdb) x /5xg 0xfffffffd70

(gdb) x /32xg 0x401440

4. 具体逻辑关系见下图

大图：http://7xlos6.com1.z0.glb.clouddn.com/%E6%9D%82.png