Backward short jmp problem in long mode

Index > Main > Backward short jmp problem in long mode

Author

Thread

fonolite

Joined: 14 Dec 2005
Posts: 32

fonolite 04 Jan 2007, 04:19

Backward jmp to local label cannot be assembled in a new fasm 1.67.18 for DOS.
(The things are inside a procedure in use64 long mode)

like this,

@@:
jmp @B

or

.label23
loop .label23

=> error: address sizes do not agree.

jmp @F
@@:
=> OK in 1.67.18

But with an old version(1.67.14) , everything was ok.
what is the problem?

04 Jan 2007, 04:19

kohlrak

Joined: 21 Jul 2006
Posts: 1421
Location: Uncle Sam's Pad

kohlrak 04 Jan 2007, 04:24

Code:

use64
format PE gui

include 'FASMDIRECTORY\include\win32ax.inc'

entry start
section '.code' readable executable
start:
@@:
jmp @b
invoke ExitProcess, 0

section '.import' import readable writeable

library kernel, 'kernel32.dll'

import kernel, ExitProcess, 'ExitProcess'

Not dos, but since you posted this in the main forum i'm assuming that you're not restricting it to DOS. Well, the above code works, though i usually don't use @b. I think that should referance an anonymous lable that was declared before the last declared label. Perhaps you want to use jmp @@.

04 Jan 2007, 04:24

fonolite

Joined: 14 Dec 2005
Posts: 32

fonolite 04 Jan 2007, 08:11

I made a coff object with use64.

But really backward jmp has a problem with new 1.67.18 fasm.

04 Jan 2007, 08:11

vid
Verbosity in development

Joined: 05 Sep 2003
Posts: 7105
Location: Slovakia

vid 04 Jan 2007, 08:38

fonolite:
did you create 64bit coff?

Code:

format MS64 COFF

04 Jan 2007, 08:38

fonolite

Joined: 14 Dec 2005
Posts: 32

fonolite 04 Jan 2007, 23:43

Thanks vid,
MS64 COFF does work for backward jmp.

But I must use COFF, because my 16bit linker(dos programming) can't recognize MS64 COFF.
(Exactly coff2omf doesn't.)

( my program : COFF(use64) -> COFF2OMF -> MS 16bit linker 5.60 + 16bit omf objects)

What's wrong with backward jmp in use64 COFF?
It works fine with fasm 1.67.14.

04 Jan 2007, 23:43

LocoDelAssembly
Your code has a bug

Joined: 06 May 2005
Posts: 4624
Location: Argentina

LocoDelAssembly 05 Jan 2007, 01:24

Quote:

version 1.67.15 (Nov 20, 2006)

[-] Some final (?) fixes and cleanup in the relative-offsets calculations.

Maybe the reason of your problem is using a 32-bit address on an instruction requiring a 64-bit address.

Edit: though, a clarification of what exactly happens is welcomed. Tomasz Rolling Eyes

05 Jan 2007, 01:24

Tomasz Grysztar

Joined: 16 Jun 2003
Posts: 8359
Location: Kraków, Poland

Tomasz Grysztar 05 Jan 2007, 09:33

First of all: it doesn't depend on whether it is a backward jump or forward jump, you must have been misled by some additional factors.

As for why this kind of error happens, it is possible that I was a bit too paranoid with sanity checks. Wink

I will look into this later.

05 Jan 2007, 09:33

Tomasz Grysztar

Joined: 16 Jun 2003
Posts: 8359
Location: Kraków, Poland

Tomasz Grysztar 10 Feb 2007, 13:32

Corrected in 1.67.19. It appears that one of the check was not doing exactly what I intended because of a swap in values.

10 Feb 2007, 13:32

< 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