AOH :: HP Unsorted D :: BT-21622.HTM

Dnsmasq Heap Overflow and Null-pointer Dereference on TFTP Server



CORE-2009-0820 - Dnsmasq Heap Overflow and Null-pointer Dereference on TFTP Server
CORE-2009-0820 - Dnsmasq Heap Overflow and Null-pointer Dereference on TFTP Server



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

        Core Security Technologies - CoreLabs Advisory
http://www.coresecurity.com/corelabs/ 

 Dnsmasq Heap Overflow and Null-pointer Dereference on TFTP Server


1. *Advisory Information*

Title: Dnsmasq Heap Overflow and Null-pointer Dereference on TFTP Server
Advisory ID: CORE-2009-0820
Advisory URL: http://www.coresecurity.com/content/dnsmasq-vulnerabilities 
Date published: 2009-08-31
Date of last update: 2009-08-31
Vendors contacted: Simon Kelley
Release mode: Coordinated release


2. *Vulnerability Information*

Class: Buffer overflow
Remotely Exploitable: Yes
Locally Exploitable: No
Bugtraq ID: 36120, 36121
CVE Name: CVE-2009-2957, CVE-2009-2958


3. *Vulnerability Description*

Dnsmasq is a lightweight DNS forwarder and DHCP server. A vulnerability
has been found that may allow an attacker to execute arbitrary code on
servers or home routers running dnsmasq[1] with the TFTP service[2][3]
enabled ('--enable-tfp'). This service is not enabled by default on most
distributions; in particular it is not enabled by default on OpenWRT or
DD-WRT. Chances of successful exploitation increase when a long
directory prefix is used for TFTP. Code will be executed with the
privileges of the user running dnsmasq, which is normally a
non-privileged one.

Additionally there is a potential DoS attack to the TFTP service by
exploiting a null-pointer dereference vulnerability.


4. *Vulnerable packages*

   . dnsmasq 2.40.
   . dnsmasq 2.41.
   . dnsmasq 2.42.
   . dnsmasq 2.43.
   . dnsmasq 2.44.
   . dnsmasq 2.45.
   . dnsmasq 2.46.
   . dnsmasq 2.47.
   . dnsmasq 2.48.
   . dnsmasq 2.49.
   . Older versions are probably affected too, but they were not checked.


5. *Non-vulnerable packages*

   . dnsmasq 2.50


6. *Vendor Information, Solutions and Workarounds*

If the TFTP service is enabled and patching is not available
immediately, a valid workaround is to filter TFTP for untrusted hosts in
the network (such as the Internet). This is the default configuration
when enabling TFTP on most home routers.

Patches are already available from the software author. Most
distributions should release updates for binary packages soon.


7. *Credits*

The heap-overflow vulnerability (CVE-2009-2957) was discovered during
Bugweek 2009 by Pablo Jorge and Alberto Solino from the team "Los
Herederos de Don Pablo" of Core Security Technologies.

The null-pointer dereference (CVE-2009-2958) was reported to the author
of dnsmasq independently by an uncredited code auditor. It was merged
with this advisory for user's convenience.


8. *Technical Description*

8.1. *Heap Overflow vulnerability (CVE-2009-2957, BID 36121)*

First let's focus on the overflow vulnerability. The 'tftp_request'
calls 'strncat' on 'daemon->namebuff', which has a predefined size of
'MAXDNAME' bytes (defaulting to 1025).

/-----------
else if (filename[0] == '/')
   daemon->namebuff[0] = 0;
strncat(daemon->namebuff, filename, MAXDNAME);
- -----------/

This may cause a heap overflow because 'daemon->namebuff' may already
contain data, namely the configured 'daemon->tftp_prefix' passed to the
daemon via a configuration file.

/-----------
if (daemon->tftp_prefix)
{
  if (daemon->tftp_prefix[0] == '/')
    daemon->namebuff[0] = 0;
    strncat(daemon->namebuff, daemon->tftp_prefix, MAXDNAME)
- -----------/

The default prefix is '/var/tftpd', but if a longer prefix is used,
arbitrary code execution may be possible.

Sending the string resulting from the execution of the following python
snippet to a vulnerable server, with a long enough directory prefix
configured, should crash the daemon.

/-----------
import sys
sys.stdout.write( '\x00\x01' + "A"*1535 + '\x00' + "netascii" + '\x00' )
- -----------/

8.2. *Null-pointer Dereference vulnerability (CVE-2009-2958, BID 36120)*

Now onto the null-pointer dereference. The user can crash the service by
handcrafting a packet, because of a problem on the guard of the first if
inside this code loop:

/-----------
while ((opt = next(&p, end)))
 {
   if (strcasecmp(opt, "blksize") == 0 &&
       (opt = next(&p, end)) &&
       !(daemon->options & OPT_TFTP_NOBLOCK))
     {
       transfer->blocksize = atoi(opt);
       if (transfer->blocksize < 1)
         transfer->blocksize = 1;
       if (transfer->blocksize > (unsigned)daemon->packet_buff_sz - 4)
          transfer->blocksize = (unsigned)daemon->packet_buff_sz - 4;
       transfer->opt_blocksize = 1;
       transfer->block = 0;
     }

  if (strcasecmp(opt, "tsize") == 0 && next(&p, end) &&
      !transfer->netascii)
     {
       transfer->opt_transize = 1;
       transfer->block = 0;
     }
 }
- -----------/

The problem exists because the guard of the first if includes the result
of 'opt = next(&p, end)' as part of the check. If this returns 'NULL',
the guard will fail and in the next if 'strcasecmp(opt, "tsize")' will
derrefence the null-pointer.


9. *Report Timeline*

. 2009-08-20:
Core Security Technologies notifies Simon Kelley of the vulnerability,
including technical details of the vulnerability in an advisory draft.

. 2009-08-21:
Simon Kelley acknowledges the vulnerability and confirms to be working
on a patch. He also informs that he is aware that most home router
distributions have tftp turned off by default, and firewalled, and
suggests this should be mentioned on the advisory. Simon also mentions
that a NULL-pointer dereference bug has also been discovered on that
code, and suggests merging both bugs in the same advisory. Monday 31/08
is accepted as a possible release date for this advisory, and help is
offered in contacting package maintainers of dnsmasq for most operating
systems.

. 2009-08-21:
Core changes the advisory draft to accommodate Simon's suggestions.
About the NULL-pointer dereference, Core mentions the terms it thinks
appropriate for the bug to be merged into this advisory, and details how
this would affect the following procedures, such as asking for a
CVE/Bugtraq ID.

. 2009-08-23:
Simon Kelley contacts Core back, saying that the terms for the
null-pointer derrefence bug to be included in the advisory are ok. He
also mentions that the finder of this bug prefers to remain uncredited
in this advisory. Details are sent by him about the new bug so that the
advisory draft can be updated to include it.

. 2009-08-23:
Core asks for proper CVE and Bugtraq ID numbers, specifying it believes
each vulnerability reported in this advisory should be assigned its own.

. 2009-08-23:
Vincent Danen, from Red Hat's Security Response Team contacts Core in
order to discuss both vulnerabilities by a secure communications
channel, and offers its help in obtaining proper CVE numbers, specifying
they also believe a separate number should be assigned to each
vulnerability.

. 2009-08-23:
Core replies to Vincent Danen by sending its gpg key. Core also mentions
separate CVE numbers have already been asked.

. 2009-08-23:
Core replies to Simon Kelley, including a new advisory draft with both
bugs merged.

. 2009-08-23:
Core receives proper CVE and Bugtraq ID numbers for both bugs, and sends
them to Red Hat and Simon Kelley.

. 2009-08-31:
The advisory CORE-2009-0820 is published.


10. *References*

[1] http://www.thekelleys.org.uk/dnsmasq/doc.html 
[2] http://www.isi.edu/in-notes/ien/ien133.txt 
[3] http://en.wikipedia.org/wiki/Trivial_File_Transfer_Protocol 


11. *About CoreLabs*

CoreLabs, the research center of Core Security Technologies, is charged
with anticipating the future needs and requirements for information
security technologies. We conduct our research in several important
areas of computer security including system vulnerabilities, cyber
attack planning and simulation, source code auditing, and cryptography.
Our results include problem formalization, identification of
vulnerabilities, novel solutions and prototypes for new technologies.
CoreLabs regularly publishes security advisories, technical papers,
project information and shared software tools for public use at:
http://www.coresecurity.com/corelabs. 


12. *About Core Security Technologies*

Core Security Technologies develops strategic solutions that help
security-conscious organizations worldwide develop and maintain a
proactive process for securing their networks. The company's flagship
product, CORE IMPACT, is the most comprehensive product for performing
enterprise security assurance testing. CORE IMPACT evaluates network,
endpoint and end-user vulnerabilities and identifies what resources are
exposed. It enables organizations to determine if current security
investments are detecting and preventing attacks. Core Security
Technologies augments its leading technology solution with world-class
security consulting services, including penetration testing and software
security auditing. Based in Boston, MA and Buenos Aires, Argentina, Core
Security Technologies can be reached at 617-399-6980 or on the Web at
http://www.coresecurity.com. 


13. *Disclaimer*

The contents of this advisory are copyright (c) 2009 Core Security
Technologies and (c) 2009 CoreLabs, and may be distributed freely
provided that no fee is charged for this distribution and proper credit
is given.


14. *PGP/GPG Keys*

This advisory has been signed with the GPG key of Core Security
Technologies advisories team, which is available for download at
http://www.coresecurity.com/files/attachments/core_security_advisories.asc. 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org 

iD8DBQFKp9rOyNibggitWa0RAjkbAJ0SLIFwI1CMF7IOHSDv+Fg0DwFNQwCfWsZm
wa3syAdyXlixVdQhdk5vcK0=tfqM
-----END PGP SIGNATURE-----

The entire AOH site is optimized to look best in Firefox® 3 on a widescreen monitor (1440x900 or better).
Site design & layout copyright © 1986-2014 AOH
We do not send spam. If you have received spam bearing an artofhacking.com email address, please forward it with full headers to abuse@artofhacking.com.