逆向分析一个完整的C++程序包含寄存器与参数传递详解
最近在分析C++ dump 文件的时候觉得有必要将一些必要的反汇编东西总结一下以备别人参考,自己有时间的时候也可以进行更多的改进。下面通过一个简单的C++代码转成汇编代码后的详细解释说明一下C++和汇编的对应关系,以及如何识别汇编代码中进行的一些操作的意义。代码的调用关系如下图所示:
完整C++代码下:
int InternalFunctionA(int nSizeA1, int nSizeA2)
{
int localnSizeA1 = nSizeA1;
int localnSizeA2 = nSizeA2;
int nFunctionA = localnSizeA1 + localnSizeA2;
return nFunctionA;
}
int InternalFunctionB(int nSizeB1, int nSizeB2)
{
int nFunctionA = InternalFunctionA(nSizeB1, nSizeB2);
return 0;
}
int _tmain(int argc, _TCHAR* argv[])
{
int nFunctionVal = InternalFunctionB(36, 64);
cout<<"Hello SolidMango!"<<endl;
return 0;
}
{
int localnSizeA1 = nSizeA1;
int localnSizeA2 = nSizeA2;
int nFunctionA = localnSizeA1 + localnSizeA2;
return nFunctionA;
}
int InternalFunctionB(int nSizeB1, int nSizeB2)
{
int nFunctionA = InternalFunctionA(nSizeB1, nSizeB2);
return 0;
}
int _tmain(int argc, _TCHAR* argv[])
{
int nFunctionVal = InternalFunctionB(36, 64);
cout<<"Hello SolidMango!"<<endl;
return 0;
}
那么这段简单的C++代码在转换成汇编代码之后是什么样子的呢?让我们拭目以待,首先让我们看看main函数转换后的代码(debug版), 如下,我们逐条来进行分析,
int _tmain(int argc, _TCHAR* argv[])
{
00411570 push ebp //栈底压栈
00411571 mov ebp,esp //栈底下移,更详细的请参考我关于ebp,esp的解释
00411573 sub esp,0CCh //局部变量预留空间
00411579 push ebx //保存ebx
0041157A push esi //保存esi
0041157B push edi //保存edi
0041157C lea edi,[ebp-0CCh] //下移edi到栈顶
00411582 mov ecx,33h //0CCh/4 = 33h
00411587 mov eax,0CCCCCCCCh //eax赋值
0041158C rep stos dword ptr es:[edi] //从edi开始做33h次赋值0CCCCCCCCh ,初始化栈内存
int nFunctionVal = InternalFunctionB(36, 64);
0041158E push 40h //参数64入栈,
00411590 push 24h //参数36入栈
00411592 call InternalFunctionB (41101Eh) );//到41101Eh处函数调用
00411597 add esp,8 //函数调用后将参数弹出,清理栈
0041159A mov dword ptr [nFunctionVal],eax
cout<<"Hello SolidMango!"<<endl;
0041159D mov esi,esp
0041159F mov eax,dword ptr [__imp_std::endl (41A338h)]
004115A4 push eax
004115A5 push offset string "Hello SolidMango!" (417800h)
004115AA mov ecx,dword ptr [__imp_std::cout (41A33Ch)]
004115B0 push ecx
004115B1 call std::operator<<<std::char_traits<char> > (411163h)
004115B6 add esp,8
004115B9 mov ecx,eax
004115BB call dword ptr [__imp_std::basic_ostream<char,std::char_traits<char> >::operator<< (41A320h)]
004115C1 cmp esi,esp
004115C3 call @ILT+430(__RTC_CheckEsp) (4111B3h)
return 0;
004115C8 xor eax,eax
}
004115CA pop edi //恢复edi
004115CB pop esi //恢复esi
004115CC pop ebx //恢复ebx
004115CD add esp,0CCh //栈顶上移
004115D3 cmp ebp,esp //检查栈平衡
004115D5 call @ILT+430(__RTC_CheckEsp) (4111B3h)
004115DA mov esp,ebp //恢复上一个栈帧的ebp,esp
004115DC pop ebp
004115DD ret//函数返回
{
00411570 push ebp //栈底压栈
00411571 mov ebp,esp //栈底下移,更详细的请参考我关于ebp,esp的解释
00411573 sub esp,0CCh //局部变量预留空间
00411579 push ebx //保存ebx
0041157A push esi //保存esi
0041157B push edi //保存edi
0041157C lea edi,[ebp-0CCh] //下移edi到栈顶
00411582 mov ecx,33h //0CCh/4 = 33h
00411587 mov eax,0CCCCCCCCh //eax赋值
0041158C rep stos dword ptr es:[edi] //从edi开始做33h次赋值0CCCCCCCCh ,初始化栈内存
int nFunctionVal = InternalFunctionB(36, 64);
0041158E push 40h //参数64入栈,
00411590 push 24h //参数36入栈
00411592 call InternalFunctionB (41101Eh) );//到41101Eh处函数调用
00411597 add esp,8 //函数调用后将参数弹出,清理栈
0041159A mov dword ptr [nFunctionVal],eax
cout<<"Hello SolidMango!"<<endl;
0041159D mov esi,esp
0041159F mov eax,dword ptr [__imp_std::endl (41A338h)]
004115A4 push eax
004115A5 push offset string "Hello SolidMango!" (417800h)
004115AA mov ecx,dword ptr [__imp_std::cout (41A33Ch)]
004115B0 push ecx
004115B1 call std::operator<<<std::char_traits<char> > (411163h)
004115B6 add esp,8
004115B9 mov ecx,eax
004115BB call dword ptr [__imp_std::basic_ostream<char,std::char_traits<char> >::operator<< (41A320h)]
004115C1 cmp esi,esp
004115C3 call @ILT+430(__RTC_CheckEsp) (4111B3h)
return 0;
004115C8 xor eax,eax
}
004115CA pop edi //恢复edi
004115CB pop esi //恢复esi
004115CC pop ebx //恢复ebx
004115CD add esp,0CCh //栈顶上移
004115D3 cmp ebp,esp //检查栈平衡
004115D5 call @ILT+430(__RTC_CheckEsp) (4111B3h)
004115DA mov esp,ebp //恢复上一个栈帧的ebp,esp
004115DC pop ebp
004115DD ret//函数返回
另外两层函数调用的汇编代码如下,感兴趣的读者可以对比一下,和main函数的过程相似,
int InternalFunctionA(int nSizeA1, int nSizeA2)
{
004114C0 push ebp
004114C1 mov ebp,esp
004114C3 sub esp,0E4h
004114C9 push ebx
004114CA push esi
004114CB push edi
004114CC lea edi,[ebp-0E4h]
004114D2 mov ecx,39h
004114D7 mov eax,0CCCCCCCCh
004114DC rep stos dword ptr es:[edi]
int localnSizeA1 = nSizeA1;
004114DE mov eax,dword ptr [nSizeA1]
004114E1 mov dword ptr [localnSizeA1],eax
int localnSizeA2 = nSizeA2;
004114E4 mov eax,dword ptr [nSizeA2]
004114E7 mov dword ptr [localnSizeA2],eax
int nFunctionA = localnSizeA1 + localnSizeA2;
004114EA mov eax,dword ptr [localnSizeA1]
004114ED add eax,dword ptr [localnSizeA2]
004114F0 mov dword ptr [nFunctionA],eax
return nFunctionA;
004114F3 mov eax,dword ptr [nFunctionA]
}
004114F6 pop edi
004114F7 pop esi
004114F8 pop ebx
004114F9 mov esp,ebp
004114FB pop ebp
004114FC ret
int InternalFunctionB(int nSizeB1, int nSizeB2)
{
00411510 push ebp
00411511 mov ebp,esp
00411513 sub esp,0CCh
00411519 push ebx
0041151A push esi
0041151B push edi
0041151C lea edi,[ebp-0CCh]
00411522 mov ecx,33h
00411527 mov eax,0CCCCCCCCh
0041152C rep stos dword ptr es:[edi]
int nFunctionA = InternalFunctionA(nSizeB1, nSizeB2);
0041152E mov eax,dword ptr [nSizeB2]
00411531 push eax
00411532 mov ecx,dword ptr [nSizeB1]
00411535 push ecx
00411536 call InternalFunctionA (411140h)
0041153B add esp,8
0041153E mov dword ptr [nFunctionA],eax
return 0;
00411541 xor eax,eax
}
00411543 pop edi
00411544 pop esi
00411545 pop ebx
00411546 add esp,0CCh
0041154C cmp ebp,esp
0041154E call @ILT+430(__RTC_CheckEsp) (4111B3h)
00411553 mov esp,ebp
00411555 pop ebp
00411556 ret
{
004114C0 push ebp
004114C1 mov ebp,esp
004114C3 sub esp,0E4h
004114C9 push ebx
004114CA push esi
004114CB push edi
004114CC lea edi,[ebp-0E4h]
004114D2 mov ecx,39h
004114D7 mov eax,0CCCCCCCCh
004114DC rep stos dword ptr es:[edi]
int localnSizeA1 = nSizeA1;
004114DE mov eax,dword ptr [nSizeA1]
004114E1 mov dword ptr [localnSizeA1],eax
int localnSizeA2 = nSizeA2;
004114E4 mov eax,dword ptr [nSizeA2]
004114E7 mov dword ptr [localnSizeA2],eax
int nFunctionA = localnSizeA1 + localnSizeA2;
004114EA mov eax,dword ptr [localnSizeA1]
004114ED add eax,dword ptr [localnSizeA2]
004114F0 mov dword ptr [nFunctionA],eax
return nFunctionA;
004114F3 mov eax,dword ptr [nFunctionA]
}
004114F6 pop edi
004114F7 pop esi
004114F8 pop ebx
004114F9 mov esp,ebp
004114FB pop ebp
004114FC ret
int InternalFunctionB(int nSizeB1, int nSizeB2)
{
00411510 push ebp
00411511 mov ebp,esp
00411513 sub esp,0CCh
00411519 push ebx
0041151A push esi
0041151B push edi
0041151C lea edi,[ebp-0CCh]
00411522 mov ecx,33h
00411527 mov eax,0CCCCCCCCh
0041152C rep stos dword ptr es:[edi]
int nFunctionA = InternalFunctionA(nSizeB1, nSizeB2);
0041152E mov eax,dword ptr [nSizeB2]
00411531 push eax
00411532 mov ecx,dword ptr [nSizeB1]
00411535 push ecx
00411536 call InternalFunctionA (411140h)
0041153B add esp,8
0041153E mov dword ptr [nFunctionA],eax
return 0;
00411541 xor eax,eax
}
00411543 pop edi
00411544 pop esi
00411545 pop ebx
00411546 add esp,0CCh
0041154C cmp ebp,esp
0041154E call @ILT+430(__RTC_CheckEsp) (4111B3h)
00411553 mov esp,ebp
00411555 pop ebp
00411556 ret
总结:通过这几篇文章的总结,相信大家已经可以看懂一些常规的C++反汇编代码,应该可以对付一般的应用,如果大家还有什么问题,或者建议欢迎讨论。