Re: OT? Are chroots immune to buffer overflows?

[SpaceWalker <spacewalker () altern org>]

Hi, your question is interresting, I've a good response for you
I'm speeking on the linux kernel, on a X86 box, but could be usable in most archs.
The chroot limitations breaks you only the accesses to the local filesystem. In most cases, you don't have an access to 
/proc ,/dev/*, nor to /bin/sh.
But If you are able to run code as root, a few syscalls are still available to you :
inserting modules and ptrace().
Both can be used to own the entire system, I coded two weeks ago a shellcode which uses ptrace to get out of the 
chroot, tracing his ppid (usualy inetd in the case of a chrooted ftp server), inserting a shellcode and leaving.

The case of inserting modules is more complex. You have to do a lkm who breaks the chroot only for your own process, 
and of course you must insmod it.
It used to have a wuftpd exploit who did compile staticaly insmod, compiled a such lkm and uploaded them on the remote 
ftp.
Of course, lkm support have to be enabled.

I think I responded to your question ..

On Wed, 22 May 2002 15:48:16 +1200
Jason Haar <Jason.Haar () trimble co nz> wrote:

> [note: my question is WRT non-root chrooted jails - we all know about
> chroot'ing root processes!]
>
> Most buffer overflows I've seen attempt to infiltrate the system enough to
> run /bin/sh. In chroot'ed environments, /bin/sh doesn't (shouldn't!) exist -
> so they fail.
>
> Is it as simple as that? As 99.999% of the system binaries aren't available
> in the jail, can a buffer overflow ever work?
>
> -- 
> Cheers
>
> Jason Haar
>
> Information Security Manager
> Trimble Navigation Ltd.
> Phone: +64 3 9635 377 Fax: +64 3 9635 417