Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 18 Dec 1998 22:41:45 -0800
From:      Don Lewis <Don.Lewis@tsc.tdk.com>
To:        Poul-Henning Kamp <phk@critter.freebsd.dk>, "Marco Molteni" <molter@tin.it>
Cc:        "Jordan K. Hubbard" <jkh@zippy.cdrom.com>, freebsd-security@FreeBSD.ORG
Subject:   Re: A better explanation (was: buffer overflows and chroot)
Message-ID:  <199812190641.WAA11564@salsa.gv.tsc.tdk.com>
In-Reply-To: Poul-Henning Kamp <phk@critter.freebsd.dk> "Re: A better explanation (was: buffer overflows and chroot)" (Dec 18,  9:00pm)

next in thread | previous in thread | raw e-mail | index | archive | help
On Dec 18,  9:00pm, Poul-Henning Kamp wrote:
} Subject: Re: A better explanation (was: buffer overflows and chroot)

} I have a set of patches which makes a chroot jail escape proof.  These 
} were developed under contract and will end up in FreeBSD sometime over
} the next year.  My client wants to get a head start, and that is only
} fair.

A year or so ago I implemented a more limited scheme to prevent access
to the filesystem outside the chroot area.  I'm in the process of cleaning
it up and hope to post my patches soon.

} The basic concept is that root is only root in a jail if the filesystem
} protects the rest of the system, otherwise he isn't.  For instance he
} can change the owner or modes on a file, but he cannot change IP# on
} an interface.  He can bind to a priviledged TCP port, but only on the
} IP# which belongs to the jail.  And so forth.  Works pretty well.

The IP restrictions would be very handy for some of the stuff that I do.

Can a process in jail kill() a process outside jail?  Can the compartments
nest?

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-security" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199812190641.WAA11564>