Home page logo
/

bugtraq logo Bugtraq mailing list archives

Security Advisory: A simple TCP spoofing attack
From: oliver () SECNET COM (Oliver Friedrichs)
Date: Sun, 9 Feb 1997 20:15:53 -0700


                        ######    ##   ##    ######
                        ##        ###  ##      ##
                        ######    ## # ##      ##
                            ##    ##  ###      ##
                        ###### .  ##   ## .  ######.

                            Secure Networks Inc.

                             Security Advisory
                             February 10, 1997

                        A simple TCP spoofing attack


Over the past few years TCP sequence number prediction attacks have become a
real threat against unprotected networks, taking advantage of the inherent
trust relationships present in many network installations.  TCP sequence
number prediction attacks have most commonly been implemented by opening a
series of connections to the target host, and attempting to predict the
sequence number which will be used next.  Many operating systems have
therefore attempted to solve this problem by implementing a method of
generating sequence numbers in unpredictable fashions.  This method does
not solve the problem.

This advisory introduces an alternative method of obtaining the initial
sequence number from some common trusted services.  The attack presented here
does not require the attacker to open multiple connections, or flood a port
on the trusted host to complete the attack.  The only requirement is that
source routed packets can be injected into the target network with fake
source addresses.

This advisory assumes that the reader already has an understanding of how
TCP sequence number prediction attacks are implemented.

The impact of this advisory is greatly diminished due to the large number of
organizations which block source routed packets and packets with addresses
inside of their networks.  Therefore we present the information as more of
a 'heads up' message for the technically inclined, and to re-iterate that
the randomization of TCP sequence numbers is not an effective solution
against this attack.


Technical Details
~~~~~~~~~~~~~~~~~

The problem occurs when particular network daemons accept connections
with source routing enabled, and proceed to disable any source routing
options on the connection.  The connection is allowed to continue, however
the reverse route is no longer used.  An example attack can launched against
the in.rshd daemon, which on most systems will retrieve the socket options
via getsockopt() and then turn off any dangerous options via setsockopt().

An example attack follows.

Host A is the trusted host
Host B is the target host
Host C is the attacker

Host C initiates a source routed connection to in.rshd on host B, pretending
to be host A.

Host C spoofing Host A         <SYN>    -->  Host B in.rshd

Host B receives the initial SYN packet, creates a new PCB (protocol
control block) and associates the route with the PCB.  Host B responds,
using the reverse route, sending back a SYN/ACK with the sequence number.

Host C spoofing Host A  <--  <SYN/ACK>       Host B in.rshd

Host C responds, still spoofing host A, acknowledging the sequence number.
Source routing options are not required on this packet.

Host C spoofing Host A         <ACK>    -->  Host B in.rshd

We now have an established connection, the accept() call completes, and
control is now passed to the in.rshd daemon.  The daemon now does IP
options checking and determines that we have initiated a source routed
connection.  The daemon now turns off this option, and any packets sent
thereafter will be sent to the real host A, no longer using the reverse
route which we have specified.  Normally this would be safe, however the
attacking host now knows what the next sequence number will be.  Knowing
this sequence number, we can now send a spoofed packet without the source
routing options enabled, pretending to originate from Host A, and our
command will be executed.

In some conditions the flooding of a port on the real host A is required
if larger ammounts of data are sent, to prevent the real host A from
responding with an RST.  This is not required in most cases when performing
this attack against in.rshd due to the small ammount of data transmitted.

It should be noted that the sequence number is obtained before accept()
has returned and that this cannot be prevented without turning off source
routing in the kernel.

As a side note, we're very lucky that TCP only associates a source route with
a PCB when the initial SYN is received.  If it accepted and changed the ip
options at any point during a connection, more exotic attacks may be possible.
These could include hijacking connections across the internet without playing
a man in the middle attack and being able to bypass IP options checking
imposed by daemons using getsockopt().  Luckily *BSD based TCP/IP stacks will
not do this, however it would be interesting to examine other implementations.

Impact
~~~~~~

The impact of this attack is similar to the more complex TCP sequence
number prediction attack, yet it involves fewer steps, and does not require
us to 'guess' the sequence number.  This allows an attacker to execute
arbitrary commands as root, depending on the configuration of the target
system.  It is required that trust is present here, as an example, the use
of .rhosts or hosts.equiv files.


Solutions
~~~~~~~~~

The ideal solution to this problem is to have any services which rely on
IP based authentication drop the connection completely when initially
detecting that source routed options are present.  Network administrators
and users can take precautions to prevent users outside of their network
from taking advantage of this problem.  The solutions are hopefully already
either implemented or being implemented.

1. Block any source routed connections into your networks
2. Block any packets with internal based address from entering your network.

Network administrators should be aware that these attacks can easily be
launched from behind filtering routers and firewalls.  Internet service
providers and corporations should ensure that internal users cannot launch
the described attacks.  The precautions suggested above should be implemented
to protect internal networks.

Example code to correctly process source routed packets is presented here
as an example.  Please let us know if there are any problems with it.
This code has been tested on BSD based operating systems.

        u_char optbuf[BUFSIZ/3];
        int optsize = sizeof(optbuf), ipproto, i;
        struct protoent *ip;

        if ((ip = getprotobyname("ip")) != NULL)
                ipproto = ip->p_proto;
        else
                ipproto = IPPROTO_IP;
        if (!getsockopt(0, ipproto, IP_OPTIONS, (char *)optbuf, &optsize) &&
            optsize != 0) {
                for (i = 0; i < optsize; ) {
                        u_char c = optbuf[i];
                        if (c == IPOPT_LSRR || c == IPOPT_SSRR)
                                exit(1);
                        if (c == IPOPT_EOL)
                                break;
                        i += (c == IPOPT_NOP) ? 1 : optbuf[i+1];
                }
        }


One critical concern is in the case where TCP wrappers are being used.  If
a user is relying on TCP wrappers, the above fix should be incorporated into
fix_options.c.  The problem being that TCP wrappers itself does not close
the connection, however removes the options via setsockopt().  In this case
when control is passed to in.rshd, it will never see any options present,
and the connection will remain open (even if in.rshd has the above patch
incorporated).  An option to completely drop source routed connections will
hopefully be provided in the next release of TCP wrappers.  The other option
is to undefine KILL_IP_OPTIONS, which appears to be undefined by default.
This passes through IP options and allows the called daemon to handle them
accordingly.


Disabling Source Routing
~~~~~~~~~~~~~~~~~~~~~~~~

We believe the following information to be accurate, however it is not
guaranteed.

--- Cisco

To have the router discard any datagram containing an IP source route option
issue the following command:

no ip source-route

This is a global configuration option.


--- NetBSD

Versions of NetBSD prior to 1.2 did not provide the capability for disabling
source routing.  Other versions ship with source routing ENABLED by default.
We do not know of a way to prevent NetBSD from accepting source routed packets.
NetBSD systems, however, can be configured to prevent the forwarding of packets
when acting as a gateway.

To determine whether forwarding of source routed packets is enabled,
issue the following command:

# sysctl net.inet.ip.forwarding
# sysctl net.inet.ip.forwsrcrt

The response will be either 0 or 1, 0 meaning off, and 1 meaning it is on.

Forwarding of source routed packets can be turned off via:

# sysctl -w net.inet.ip.forwsrcrt=0

Forwarding of all packets in general can turned off via:

# sysctl -w net.inet.ip.forwarding=0


--- BSD/OS

BSDI has made a patch availible for rshd, rlogind, tcpd and nfsd.  This
patch is availible at:

ftp://ftp.bsdi.com/bsdi/patches/patches-2.1

OR via their patches email server <patches () bsdi com>

The patch number is
U210-037 (normal version)
D210-037 (domestic version for sites running kerberized version)


BSD/OS 2.1 has source routing disabled by default

Previous versions ship with source routing ENABLED by default.  As far as
we know, BSD/OS cannot be configured to drop source routed packets destined
for itself, however can be configured to prevent the forwarding of such
packets when acting as a gateway.

To determine whether forwarding of source routed packets is enabled,
issue the following command:

# sysctl net.inet.ip.forwarding
# sysctl net.inet.ip.forwsrcrt

The response will be either 0 or 1, 0 meaning off, and 1 meaning it is on.

Forwarding of source routed packets can be turned off via:

# sysctl -w net.inet.ip.forwsrcrt=0

Forwarding of all packets in general can turned off via:

# sysctl -w net.inet.ip.forwarding=0


--- OpenBSD

Ships with source routing turned off by default.  To determine whether source
routing is enabled, the following command can be issued:

# sysctl net.inet.ip.sourceroute

The response will be either 0 or 1, 0 meaning that source routing is off,
and 1 meaning it is on.  If source routing has been turned on, turn off via:

# sysctl -w net.inet.ip.sourceroute=0

This will prevent OpenBSD from forwarding and accepting any source routed
packets.


--- FreeBSD

Ships with source routing turned off by default.  To determine whether source
routing is enabled, the following command can be issued:

# sysctl net.inet.ip.sourceroute

The response will be either 0 or 1, 0 meaning that source routing is off,
and 1 meaning it is on.  If source routing has been turned on, turn off via:

# sysctl -w net.inet.ip.sourceroute=0


--- Linux

Linux by default has source routing disabled in the kernel.


--- Solaris 2.x

Ships with source routing enabled by default.  Solaris 2.5.1 is one of the
few commercial operating systems that does have unpredictable sequence
numbers, which does not help in this attack.

We know of no method to prevent Solaris from accepting source routed
connections, however, Solaris systems acting as gateways can be prevented
from forwarding any source routed packets via the following commands:

# ndd -set /dev/ip ip_forward_src_routed 0

You can prevent forwarding of all packets via:

# ndd -set /dev/ip ip_forwarding 0

These commands can be added to /etc/rc2.d/S69inet to take effect at bootup.


--- SunOS 4.x

We know of no method to prevent SunOS from accepting source routed
connections, however a patch is availible to prevent SunOS systems from
forwarding source routed packets.

This patch is availible at:

ftp://ftp.secnet.com/pub/patches/source-routing-patch.tar.gz

To configure SunOS to prevent forwarding of all packets, the following
command can be issued:

# echo "ip_forwarding/w 0" | adb -k -w /vmunix /dev/mem
# echo "ip_forwarding?w 0" | adb -k -w /vmunix /dev/mem

The first command turns off packet forwarding in /dev/mem, the second in
/vmunix.


--- HP-UX

HP-UX does not appear to have options for configuring an HP-UX system to
prevent accepting or forwarding of source routed packets.  HP-UX has IP
forwarding turned on by default and should be turned off if acting as a
firewall.  To determine whether IP forwarding is currently on, the following
command can be issued:

# adb /hp-ux
ipforwarding?X      <- user input
ipforwarding:
ipforwarding: 1
#

A response of 1 indicates IP forwarding is ON, 0 indicates off.  HP-UX can
be configured to prevent the forwarding of any packets via the following
commands:

# adb -w /hp-ux /dev/kmem
ipforwarding/W 0
ipforwarding?W 0
^D
#

--- AIX

AIX cannot be configured to discard source routed packets destined for itself,
however can be configured to prevent the forwarding of source routed packets.
IP forwarding and forwarding of source routed packets specifically can be
turned off under AIX via the following commands:

To turn off forwarding of all packets:

# /usr/sbin/no -o ipforwarding=0

To turn off forwarding of source routed packets:

# /usr/sbin/no -o nonlocsrcroute=0

Note that these commands should be added to /etc/rc.net



If shutting off source routing is not possible and you are still using
services which rely on IP address authentication, they should be disabled
immediately (in.rshd, in.rlogind).  in.rlogind is safe if .rhosts and
/etc/hosts.equiv are not used.


Attributions
~~~~~~~~~~~~

Thanks to Niels Provos <provos () physnet uni-hamburg de> for providing
the information and details of this attack.  You can view his web
site at http://www.physnet.uni-hamburg.de/provos

Thanks to Theo de Raadt, the maintainer of OpenBSD for forwarding this
information to us.  More information on OpenBSD can be found at
http://www.openbsd.org

Thanks to Keith Bostic <bostic () bsdi com> for discussion and a quick
solution for BSD/OS.

Thanks to Brad Powell <brad.powell () west sun com> for providing information
for Solaris 2.x and SunOS 4.x operating systems.

Thanks go to CERT and AUSCERT for recommendations in this advisory.

You can contact the author of this advisory at oliver () secnet com

Type Bits/KeyID    Date       User ID
pub  1024/0E7BBA7D 1996/09/18 Oliver Friedrichs <oliver () secnet com>

-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: 2.6.3ia

mQCNAzJATn0AAAEEAJeGbZyoCw14fCoAMeBRKiZ3L6JMbd9f4BtwdtYTwD42/Uz1
A/4UiRJzRLGhARpt1J06NVQEKXQDbejxGIGzAGTcyqUCKH6yNAncqoep3+PKIQJd
Kd23buvbk7yUgyVlqQHDDsW0zMKdlSO7rYByT6zsW0Rv5JmHJh/bLKAOe7p9AAUR
tCVPbGl2ZXIgRnJpZWRyaWNocyA8b2xpdmVyQHNlY25ldC5jb20+iQCVAwUQMkBO
fR/bLKAOe7p9AQEBOAQAkTXiBzf4a31cYYDFmiLWgXq0amQ2lsamdrQohIMEDXe8
45SoGwBzXHVh+gnXCQF2zLxaucKLG3SXPIg+nJWhFczX2Fo97HqdtFmx0Y5IyMgU
qRgK/j8KyJRdVliM1IkX8rf3Bn+ha3xn0yrWlTZMF9nL7iVPBsmgyMOuXwZ7ZB8=
=xq4f
-----END PGP PUBLIC KEY BLOCK-----

Copyright Notice
~~~~~~~~~~~~~~~~
The contents of this advisory are Copyright (C) 1997 Secure Networks Inc,
and may be distributed freely provided that no fee is charged for
distribution, and that proper credit is given.

 You can find Secure Networks papers at ftp://ftp.secnet.com/pub/papers
 and advisories at ftp://ftp.secnet.com/advisories

 You can browse our web site at http://www.secnet.com

 You can subscribe to our security advisory mailing list by sending mail to
 majordomo () secnet com with the line "subscribe sni-advisories"



  By Date           By Thread  

Current thread:
[ Nmap | Sec Tools | Mailing Lists | Site News | About/Contact | Advertising | Privacy ]