CODEWITHSUNDEEP
cassemblycompilationobject-files
Tryer
Tryer

Reputation: 4050

Disassembly of object file with offset from instruction pointer

The following example is provided in this book.

There are two .c files and one .h file.

main.c

#include "function.h"
extern int nCompletionStatus;
int main(int argc, char* argv[])
{
    float x = 1.0;
    float y = 5.0;
    float z;

    z = add_and_multiply(x,y);
    nCompletionStatus = 1;
    return 0;
}

function.h

#pragma once

#define FIRST_OPTION
#ifdef FIRST_OPTION
    #define MULTIPLIER (3.0)
#else
    #define MULTIPLIER (2.0)
#endif

float add_and_multiply(float x, float y);

function.c

#include "function.h"
int nCompletionStatus = 0;

float add(float x, float y)
{
    float z = x + y;
    return z;
}

float add_and_multiply(float x, float y)
{
    float z = add(x,y);
    z *= MULTIPLIER;
    return z;
}

To produce the .o files, the following command is provided:

gcc -c function.c main.c

Then, to view the contents of main.o, we have:

objdump -D -M intel main.o.

The author in his book lists the (snippets of) content of the objdump output thus focussing on the unresolved external reference in main.o to add_and_multiply(,) and the externed nCompletionStatus:

27: 89 04 24             mov    DWORD PTR [esp],eax
2a: e8 fc ff ff ff       call   2b <main + 0x2b>    ;//highlighted (a) by author
2f: d9 5c 24 1c          fstp   DWORD PTR [esp+0x1c]
33: c7 05 00 00 00 00 01 mov    DWORD PTR ds:0x0,0x1 ;//highlighted (b) by author

On my machine, which is slightly different from what the author has produced, I get the following output of objdump (relevant parts only):

3c: e8 00 00 00 00       call   41 <main+0x41>;//Equivalent to highlight (a) of author?
41: 66 0f 7e c0          movd   eax,xmm0
45: 89 45 fc             mov    DWORD PTR [rbp-0x4],eax
48: c7 05 00 00 00 00 01 mov    DWORD PTR [rip+0x0],0x1        # 52 <main+0x52>;//equivalent to highlight (b) of author?

My questions are indicated as comments above. That is, how is the output on my machine (running gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.10)) equivalent to what the author reports?

In particular, in the author's version, in highlight (b), ds: seems to be accessed, while in my version, some offset from the rip register seems to be in place. How are they equivalent?

Upvotes: 1

Views: 1383

Answers (1)

fuz
fuz

Reputation: 92986

The book you refer to is using 32 bit assembly while your compiler emits 64 bit assembly. Pass -m32 to gcc to compile your C code into 32 bit machine code instead. Also consider using -d instead of -D for objdump as the former only tries to disassemble sections that are known to contain machine code.

Upvotes: 2

Related Questions