how to pop a "hello world!"messagebox?

[smallwolf]

hello:
i am a beginner in assembly,and i can't wait to test some code. i want to write a simple "hello world" program using inline assembly, how can i pass the string argument to the MessageBoxA()?
if i pass all the NULL value, it works,but not what i want.
i am using wxDevC++7.0, windows xp sp3. thanks, have a good day!

#include <stdio.h>
#include <stdlib.h>
#include <windows.h>

int main(int argc, char *argv[])
{
   LPCTSTR user32="user32.dll";
   HMODULE hd=LoadLibrary(user32);
   if(hd)
        printf("User32.dll is loaded!\n");
        
        //0x7e4507ea is the address of MessageBoxA
        
    __asm__(
     
         
         "xor %eax,%eax\n"
         "xor %ebx,%ebx\n"
         "mov $0x7e4507ea,%ebx\n"
         "push %eax\n"
         "push %eax\n"
         "push %eax\n"
         "push %eax\n"
         "call %ebx\n"
             
             
             );
  
  
  
  system("PAUSE");	
  return 0;
}

[Eri523]

Well, I don't know wxDevC++, but this is how I did it in MS VC++ (Unicode project):

 
#include "stdafx.h"
#include <Windows.h>
 
const _TCHAR achMessage[]=_T("Hello world!");
 
int _tmain(int argc, _TCHAR* argv[])
{
  __asm {
    xor  eax,eax
    push eax
    push eax
    push offset achMessage
    push eax
    call dword ptr [MessageBox]
  }
  return 0;
}

As to your code: Hardcoding the address of the MessageBox() function is pretty dangerous! How can you rely on it always being located at the same address, especially when importing it using LoadLibrary()!? And why do you zero out EBX just before loading it with an immediate value in the next instruction? This is as redundant as can be.

BTW: The assembly syntax used by your compiler is pretty unusual and definitely non-Intel! How come they use such a home-grown syntax? Could it be they would have needed a pretty expensive license from Intel to use the standard syntax?

HTH

[Bluefox815]

BTW: The assembly syntax used by your compiler is pretty unusual and definitely non-Intel! How come they use such a home-grown syntax? Could it be they would have needed a pretty expensive license from Intel to use the standard syntax?

HTH

The syntax he's using is the AT&T syntax, which is the only syntax I know of besides Intel. I do not believe any license is required to implement the Intel syntax, many compilers recognize it. But I could be wrong.

The AT&T syntax has been choice for GNU applications I've seen, such as their assembler (gas) and for inline assembly in their C compiler (gcc).

[Eri523]

The syntax he's using is the AT&T syntax, which is the only syntax I know of besides Intel. I do not believe any license is required to implement the Intel syntax, many compilers recognize it. But I could be wrong.

No, knowing that now, I don't think you're wrong. AT&T is a mighty company by itself, and they certainly don't have to bother paying some bucks to Intel for an assembly syntax license.

In fact I never heard that anyone had to pay any license fee for any assembler syntax at all, I was just wondering about that syntax I never saw yet.

If it's the AT&T syntax, I don't have to wonder any longer why so many C++ compilers use it: You certainly know that it's AT&T that stands behind Bjarne Stroustrup, and that Bjarne Stroustrup is the guy who stands behind the initial design of C++...