Re: Looking for disassembler, decompiler, discompiler or whatever.

andyjnsn@ma.ultranet.com
5 Sep 2001 22:10:18 -0400

          From comp.compilers

Related articles
Looking for disassembler, decompiler, discompiler or whatever. zuyihe@163.net (2001-09-03)
Re: Looking for disassembler, decompiler, discompiler or whatever. andyjnsn@ma.ultranet.com (2001-09-05)
Re: Looking for disassembler, decompiler, discompiler or whatever. vbdis@aol.com (2001-09-05)
Re: Looking for disassembler, decompiler, discompiler or whatever. mpointie@eden-studios.fr (MickaŽl Pointier) (2001-09-11)
Re: Looking for disassembler, decompiler, discompiler or whatever. joachim_d@gmx.de (Joachim Durchholz) (2001-09-11)
Re: Looking for disassembler, decompiler, discompiler or whatever. mpointie@eden-studios.fr (MickaŽl Pointier) (2001-09-11)
Re: Looking for disassembler, decompiler, discompiler or whatever. debray@CS.Arizona.EDU (2001-09-11)
Re: Looking for disassembler, decompiler, discompiler or whatever. eanders@argus.EECS.Berkeley.EDU (2001-09-11)
| List of all articles for this month |

From: andyjnsn@ma.ultranet.com
Newsgroups: comp.compilers
Date: 5 Sep 2001 22:10:18 -0400
Organization: Compilers Central
References: 01-09-011
Keywords: decompile
Posted-Date: 05 Sep 2001 22:10:18 EDT

On 3 Sep 2001 23:05:37 -0400, zuyihe@163.net (zyy) wrote:
> My question is: is there any tool that, given exe, will give an
>assembly which looks like the one generated by "gcc -S *.c"? Or, in
>the example, given a.out, will generate hello0.s?


In order to do this, you would need to have some symbolic information
from the program. Java decompilers have a much easier task because
all bindings to class fields and methods are done by name. Only local
method variables are accessed by "offset" values, but decompilers do a
reasonable job of making up names in these cases, based on the
synthesized types.


I have seen a comple of attempts at both disassembling and decompiling
x86 binary files, but none that have taken advantage of existing debug
information (either STABS or DWARF). The PPC disassemblers that I
have seen don't even show the "fixed-up" label definition points (i.e.
forward branch points), so you have to also turn on the expansion of
the relocation information to see the final branch destinations.


-AndyJ


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.