Full Disclosure mailing list archives

Re: Leveraging pam_env to steal DSA keys


From: Peter van Dijk <peter () 7bits nl>
Date: Tue, 31 May 2011 05:55:27 +0200


On May 31, 2011, at 12:48 AM, paul.szabo () sydney edu au wrote:

... http://7bits.nl/projects/pamenv-dsakeys/pamenv-dsakeys.html

Seems to me that CVE-2010-3435 may allow users to determine also:
 password in /etc/lilo.conf
 secret in /etc/bind/named.conf /etc/bind/rndc.conf /etc/bind/rndc.key
 bits of /etc/ssh/ssh_host_dsa_key /etc/ssh/ssh_host_rsa_key
which should all be protected.

- lilo.conf commonly has whitespace around '=', pam_env does not tolerate that
- bind configs don't even use '=' and are often indented, pam_env does not tolerate indents
- RSA appears to be uninteresting in that the amount of bits we can lift is not sufficient to make an attack feasible 
(this is in the article!)
- the DSA host key certainly is a target

If I understand correctly, Debian (and, I presume, Ubuntu) put this bug on low priority precisely because there were 
very few practical applications that they knew of.

Cheers, Peter
_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/


Current thread: