OSEC

Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com
 
Subject: Re: No-Exec Stack Smashing 101
From: Michael H. Warfield (mhwWITTSEND.COM)
Date: Thu Apr 20 2000 - 20:07:04 CDT

On Thu, Apr 20, 2000 at 10:36:13AM -0700, Granquist, Lamont wrote:
> Well, its not an inherant feature of non-executable stacks that libc is
> mapped to an address with a zero in it. That's just a feature of Solar
> Designer's patch for Linux. My next question would have been, of course,
> how to get around that.

        Find a spot in the code which calls the library function and
return to that instead. Can't all of it have zeros in it. That may
be what they are referring to as a PLT.

> And it is still useful to discuss the return-into-libc approach just for
> educational benefit. And also because in addition to getting around
> no-exec stacks it also lets you bypass writing shellcode and means that
> your exploits will not have NOPs in them that IDSes could trigger off of
> (for remote exploits). And then there's exploits which don't just copy a
> null-terminated string and for those it doesn't matter if libc has a zero
> in it.

> Of course it sounds like what you're talking about below works as well and
> more generally.

> What is a PLT, if i might ask?

> Also, I'm still trying to figure out what you need to make the stack look
> like in order to do a return-into-libc exploit. I've got Solar Designer's
> lpr -C exploit, but I don't quite understand his comments (I mean I
> understand what he's making the stack look like, but I don't understand
> why).

> On Thu, 20 Apr 2000, M.C.Mar wrote:
> > > Thought I'd try to liven the discussion up here a bit...
> > >
> > > Okay, so I'm setting off to exploit the /usr/bin/man MANPATH exploit on
> > > RH6.1 (actually my system is RH6.2 i686 with man-1.5g-6 installed on it).
> > > And I'm looking for a little help here. What I've been playing with so
> > > far is things like the following trying to sort out the parameters of the
> > > buffer overflow:
> > >
> > > setenv MANPAGER `perl -e 'print "A" x 4057; print "\xe0\xca\x05\x40";
> > > /usr/bin/man man
> > > gdb /usr/bin/man core
> > >
> > > Now the goal is to make man do a system("/bin/echo i 0wn y3w") call via
> > > returning into system() with the string being a parameter to the call with
> > > no code being excuted on the stack.
> >
> > I'm affraid you made wrong aproach...
> > If you have nonexeck stack patch installed you cannot jump into libc, cos
> > libc is mmaped undex 0x00XXXXXX address!
> > The best idea is to jump into PLT. To find system() PLT entry do the
> > following:
> >
> > (gdb) p system
> > $2 = {<text variable, no debug info>} 0x8048d38 <system>
> >
> > 0x8048d38 is a PLT addr of system() call.
> >
> > As the system() call takes only one argument try to fing correct offset
> > in the bufer where is the value that overwrites the RET. Put there
> > 0x8048d38 and after that any value (fo saved %ebp, which should be there :)
> > and then addr where is your string to execute via system();
> >
> >
> > If anyone need I can write full working explot that uses system() or
> > strcpy() functions to execute anything you want! :)
> >
> > --
> > Mariusz Wołoszyn
> > Internet Security Specialist, IT -- Internet Partners
> > E-mail: Mariusz.Woloszynit.pl, woloszynit.pl
> > 

--
 Michael H. Warfield    |  (770) 985-6132   |  mhwWittsEnd.com
  (The Mad Wizard)      |  (770) 331-2437   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!