flat assembler

Index > Main > ..

Goto page 1, 2 Next

Author

Thread

pool

Joined: 08 Jan 2007
Posts: 97

pool 21 Dec 2012, 16:04

Last edited by pool on 17 Mar 2013, 12:17; edited 1 time in total

21 Dec 2012, 16:04

JohnFound

Joined: 16 Jun 2003
Posts: 3499
Location: Bulgaria

JohnFound 21 Dec 2012, 16:20

It depends on "arg". In general, there is no "if" construction in assembly programming.

21 Dec 2012, 16:20

HaHaAnonymous

Joined: 02 Dec 2012
Posts: 1178
Location: Unknown

HaHaAnonymous 21 Dec 2012, 16:23

[ Post removed by author. ]

Last edited by HaHaAnonymous on 28 Feb 2015, 22:09; edited 1 time in total

21 Dec 2012, 16:23

cod3b453

Joined: 25 Aug 2004
Posts: 618

cod3b453 21 Dec 2012, 18:34

if statements usually translate to compare and branch-if-not-condition:

Code:

cmp [arg],0 ; if (arg)
je @f       ; {

@@:         ; }

cmp [arg],0 ; if (!arg)
jne @f      ; {

@@:         ; }

but this can (and usually will) be optimised in a number of ways.

Last edited by cod3b453 on 21 Dec 2012, 18:59; edited 2 times in total

21 Dec 2012, 18:34

pool

Joined: 08 Jan 2007
Posts: 97

pool 21 Dec 2012, 18:36

Last edited by pool on 17 Mar 2013, 12:17; edited 1 time in total

21 Dec 2012, 18:36

baldr

Joined: 19 Mar 2008
Posts: 1651

baldr 21 Dec 2012, 18:50

cod3b453,

You got it completely wrong: if (expression) statement executes statement if expression evaluates to non-zero; therefore proper conditional jump mnemonic is je, so it jumps over when expression evaluates to zero.

21 Dec 2012, 18:50

cod3b453

Joined: 25 Aug 2004
Posts: 618

cod3b453 21 Dec 2012, 18:58

Sorry pool/thanks baldr

That's what happens when I copy&paste code on an overtime Friday but as long as you do what I say and not what I do Very Happy

I'll fix it now

21 Dec 2012, 18:58

baldr

Joined: 19 Mar 2008
Posts: 1651

baldr 21 Dec 2012, 20:15

pool,

Many C compilers have option to output assembly listing (/Fafilename.asm for Visual C), you may examine it to see how code corresponding to high-level statements is generated. Beware of optimization though.

21 Dec 2012, 20:15

pool

Joined: 08 Jan 2007
Posts: 97

pool 21 Dec 2012, 22:21

Last edited by pool on 17 Mar 2013, 12:17; edited 1 time in total

21 Dec 2012, 22:21

Bargest

Joined: 09 Feb 2012
Posts: 79
Location: Russia

Bargest 22 Dec 2012, 19:27

Quote:

Many C compilers have option to output assembly listing

... or you can download any free disassembler and use it to get ASM-code of any construction you want. Smile

For examle, you can write like this:

Code:

_asm nop;
_asm nop;
_asm nop;
_asm nop;
if (arg)
{
    MessageBoxA(0,0,0,0);
}
if (!arg)
{
    MessageBoxW(0,0,0,0);
}
_asm nop;
_asm nop;
_asm nop;
_asm nop;

Then disassemlby it and find code between four NOPs.
Also, you should disable optimization, or good compiler can count the value of arg and if it is constant - it will compile code destroying all your IFs:

Code:

push 0
push 0
push 0
push 0
call [MessageBoxA]

22 Dec 2012, 19:27

pool

Joined: 08 Jan 2007
Posts: 97

pool 22 Dec 2012, 23:09

Last edited by pool on 17 Mar 2013, 12:18; edited 2 times in total

22 Dec 2012, 23:09

HaHaAnonymous

Joined: 02 Dec 2012
Posts: 1178
Location: Unknown

HaHaAnonymous 22 Dec 2012, 23:17

[ Post removed by author. ]

Last edited by HaHaAnonymous on 28 Feb 2015, 22:09; edited 3 times in total

22 Dec 2012, 23:17

baldr

Joined: 19 Mar 2008
Posts: 1651

baldr 23 Dec 2012, 07:21

pool,

You're right, those lea edi, [edi] are 7-bytes nops used to align _main to 16-bytes boundary.

----8<----
HaHaAnonymous,

Your comment about setne is partly wrong: instruction sets al to 1 (not 0 as you've said) if ZF==0 (i.e. [var1] is non-zero).

23 Dec 2012, 07:21

AsmGuru62

Joined: 28 Jan 2004
Posts: 1618
Location: Toronto, Canada

AsmGuru62 23 Dec 2012, 10:42

Weird, maybe it depends on compiler, but Visual Studio
evaluates IF() into zero and one.
I can see that 1 is a non-zero, but a non-zero is wider, it gives
an impression that it can be 3 or -5, but it is actually 1.

I could do this in C (Visual Studio):

Code:

HCURSOR      cursors [2];
...
cursor [0] = LoadCursor (...);
cursor [1] = LoadCursor (...);
...
SetCursor (cursor [(value1 > value2)]);

I wonder what other compilers are doing here?
Will that code work with GCC or Watcom or Intel?

23 Dec 2012, 10:42

baldr

Joined: 19 Mar 2008
Posts: 1651

baldr 23 Dec 2012, 12:20

AsmGuru62,

Compilers that adhere to standard (ISO/IEC 9899:2011, AKA C11) yield 0 or 1 as a result of relational/equality operators.

23 Dec 2012, 12:20

HaHaAnonymous

Joined: 02 Dec 2012
Posts: 1178
Location: Unknown

HaHaAnonymous 23 Dec 2012, 14:55

[ Post removed by author. ]

Last edited by HaHaAnonymous on 28 Feb 2015, 22:08; edited 1 time in total

23 Dec 2012, 14:55

typedef

Joined: 25 Jul 2010
Posts: 2909
Location: 0x77760000

typedef 23 Dec 2012, 16:13

How did you get Pelles-C to bring that debug window. Every time I put a debug breakpoint and compile with debug symbols it doesn't stop it just runs without breaking on them(break points).

It pisses me off. I just use Visual Studio instead.

23 Dec 2012, 16:13

typedef

Joined: 25 Jul 2010
Posts: 2909
Location: 0x77760000

typedef 23 Dec 2012, 16:32

pool wrote:

what is the equivalent of:

if(arg){

}

if(!arg){

}

to assembly?

thanks!!

^^^ This is amateurish code. This code will basically achieve the same thing in the end. The first if statement is enough. Your compiler might optimize it for you though.

Do this:

Code:

if(arg)
{
 ...
}else{
 ...
}

Which would translate to (in ASM)

Code:

; The main goal is to test for null.

cmp  [arg], 0  ; if (arg) or if (arg != null) or if(arg > 0) or if(arg != 0)
je      is_null   ; Jump if null or you could jump if not null

no_null:

jmp resume

is_null:


resume:

23 Dec 2012, 16:32

revolution
When all else fails, read the source

Joined: 24 Aug 2004
Posts: 20299
Location: In your JS exploiting you and your system

revolution 23 Dec 2012, 19:56

typedef wrote:

pool wrote:
what is the equivalent of:

if(arg){

}

if(!arg){

}

to assembly?

thanks!!
^^^ This is amateurish code. This code will basically achieve the same thing in the end. The first if statement is enough. Your compiler might optimize it for you though.

Do this:
Code:
if(arg)
{
 ...
}else{
 ...
}
    

What about the case where arg changes within the execution of the first if clause?

Or what if arg is a volatile external device status?

Or what if arg is a complex compound statement?

There can be many reasons why your replacement code would not be the same.

23 Dec 2012, 19:56

typedef

Joined: 25 Jul 2010
Posts: 2909
Location: 0x77760000

typedef 23 Dec 2012, 23:30

revolution wrote:

typedef wrote:
pool wrote:
what is the equivalent of:

if(arg){

}

if(!arg){

}

to assembly?

thanks!!
^^^ This is amateurish code. This code will basically achieve the same thing in the end. The first if statement is enough. Your compiler might optimize it for you though.

Do this:
Code:
if(arg)
{
 ...
}else{
 ...
}
    
What about the case where arg changes within the execution of the first if clause?

Or what if arg is a volatile external device status?

Or what if arg is a complex compound statement?

There can be many reasons why your replacement code would not be the same.

In this case we're given two expressions with the same goal. Either one of them will lead to the same result. That's what I'm saying. Plus it's a waste of time. Look at this:

Code:


input = getInput()

if(input)  // Test for non-zero value
{
  Debug.Print("Hello");
}


if(!input)
{
  Debug.Print("Good Bye");
}

As you can see, the last if statement will be assembled as a different branch of code. That's where ELSE comes in, so that way only the first if statement will be used to evaluate the expression. Wink

23 Dec 2012, 23:30

Goto page 1, 2 Next

< Last Thread | Next Thread >

Forum Rules:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum