Re: Randomized Stack
From: Oldani Massimiliano (sgrakkyu_at_antifork.org)
Date: 11/25/05
- Previous message: Nish_at_securityfocus.com,: "Exploiting the Stack (Part I-IV)"
- In reply to: veider_at_ank-pki.ru: "Randomized Stack"
- Next in thread: Rik Bobbaers: "Re: Randomized Stack"
- Reply: Rik Bobbaers: "Re: Randomized Stack"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Date: Fri, 25 Nov 2005 17:47:42 +0100 To: vuln-dev@securityfocus.com
On Nov 23, 2005, at 10:21 PM, veider@ank-pki.ru wrote:
> Hello, All!
>
> I am trying to exploit a stack based overflow on a system with stack
> randomization. Knowing function addresses with "objdump -D" I am able
> to run thouse functions. I've read phrack article about bypassing
> PaX ASLR, but there the author is able to control function arguments
> that is
> impossible in my case. Any ideas on what i may try?
>
> Bye, All.
>
Stack random? only random stack? or with random mmap()/stack and
no-exec workaround ?
If you have only random stack and you can execute code in the stack,
you can check for interesting pointer in the stack and chain a
ret-into-ret until you get it
or find somewhere jmp *%esp instruction and jump on your payload.
Alternatively you can construct argument with ret-into-PLT strcpy()
chain in some RW place and then use them.
>
>
-----------
Oldani (sgrakkyu) Massimiliano
Antifork Research, Inc.
Metro Olografix
- Previous message: Nish_at_securityfocus.com,: "Exploiting the Stack (Part I-IV)"
- In reply to: veider_at_ank-pki.ru: "Randomized Stack"
- Next in thread: Rik Bobbaers: "Re: Randomized Stack"
- Reply: Rik Bobbaers: "Re: Randomized Stack"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]