Home page logo

oss-sec logo oss-sec mailing list archives

Re: CVE request: cobbler lack of csrf protection, code execution
From: David Black <disclosure () d1b org>
Date: Thu, 12 Apr 2012 21:49:22 +1000

On Thu, Apr 12, 2012 at 9:46 PM, Jan Lieskovsky <jlieskov () redhat com> wrote:
Thank you for this post, David.

Just administrative note -- all of these security issues should
get CVE-2011-* CVE identifiers, as all of the Ubuntu bugs have
been reported in 2011 yet (2011-09-28 exactly).

On 04/12/2012 11:39 AM, David Black wrote:

Hi, I reported some bugs a while ago in cobbler which never received
CVE ID, could the follow bugs receive CVE ID ?
1. lack of csrf protection in the cobbler web interface (vulnerable to
csrf attacks)

Some further references / patches information I was able to found:
1) Ubuntu patch by Robie Basak:


2) Red Hat bugzilla entry:

2. code execution on the cobbler host through use of yaml.loads on
potentially untrusted user input

Though only yaml.load privilege escalation vector has been mentioned in this
post, from further look noticed two ways for privilege escalation:
1) (possibly remote) privilege escalation via yaml.load / by
  processing management parameters:

(Ubuntu bug)

  Ubuntu patch from Robie Basak:
  * Backport safe YAML load from upstream. (LP: #858883):


  https://bugzilla.redhat.com/show_bug.cgi?id=811920 (Red Hat bug)

2) local privilege escalation due to insecure use of PYTHON_EGG_CACHE

  https://bugs.launchpad.net/ubuntu/+source/cobbler/+bug/858875 (Ubuntu bug)
  https://fedorahosted.org/cobbler/ticket/688 (upstream ticket)

  (relevant upstream patch)
  https://bugzilla.redhat.com/show_bug.cgi?id=811926 (Red Hat bug)

Kurt, could you allocate three 2011 CVE ids for these issues?
i)   the first for CSRF issue,
ii)  the second for the yaml.load priv esc issue,
iii) the third for the PYTHON_EGG_CACHE local priv esc issue

David, would be great if you could confirm the three ids are necessary.

I left the third one (the PYTHON_EGG_CACHE  issue out) but it should
probably get a CVE ID as well.
Well what do you mean by 'necessary'? if you ask me I will say, "probably".


  By Date           By Thread  

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