Bugzilla Bug 154645

+++ This bug was initially created as a clone of Bug #154573 +++

The package mgv 3.1.5 also seem to contain the vulnerable code, although I
didn't have a closer look at it.

==========

GNU gv Stack Overflow Vulnerability


//----- Advisory


Program          : GNU gv
Homepage         : http://www.gnu.org/software/gv/
Tested version   : 3.6.2
Found by         : r.lifchitz at sysdream dot com
This advisory    : r.lifchitz at sysdream dot com
Discovery date   : 2006/11/06
Vendor notified  : 2006/11/09


//----- Application description


gv is a comfortable viewer of PostScript and PDF files for the X
Window System. It uses the ghostscript PostScript interpreter
and is based on the classic X front-end for gs, ghostview, which
it has replaced now.


//----- Description of vulnerability


The 'gv' viewer is prone to a remote stack overflow
vulnerability. This issue exists because the application fails
to perform proper boundary checks before copying user-supplied
data into process buffers. A remote attacker may execute arbitrary
code in the context of a user running the application. As a result,
the attacker can gain unauthorized access to the vulnerable computer.

This issue is present itself in the 'ps_gettext()' function residing
in the 'ps.c' file.

Long comments in some specific headers (such as '%%DocumentMedia:')
of PS files are unconditionally copied into 'text', a 257 character
buffer on the stack.

This issue is reported to affect gv 3.6.2, but earlier versions are
likely prone to this vulnerability as well. Applications using embedded
gv code may also be vulnerable.


//----- Proof Of Concept

[...]

/----- Solution


No known solution. You have to wait for a vendor upgrade and
be careful with unknown PS files.


//----- Impact


Successful exploitation leads to remote code execution.


//----- Credits


Renaud Lifchitz
r.lifchitz at sysdream dot com
http://www.sysdream.com/

This can be treecleaned. Upstream is dead, no release in 8 years.

Security let's start by masking it and let treecleaners do their job. Any
comments?

No objections, would somebody with the magick powers please do the trick?

masked. do we really need a maskglsa here?

The policy says "yes"... i would say "yes" too... (it's about an overflow so
it's rather severe)

I agree with both masking and GLSA'ing - if there's anyone still using it, they
need to know.

Though this bug is rather old. I've called for a maskglsa now with 2 YES votes.

Okay, our turn.

Treecleaners, please vote.

++

++

Upstream is dead, wasn't able to find another source for this package. Voting
yes for that.

Just FWI it was (masking) GLSA 200703-24

# Raúl Porcel <armin76@gentoo.org> (09 Apr 2007)
# Pending removal 09 Jun 2007, for treecleaners
# app-admin/cpu -> bug 173064
# app-admin/quickswitch -> bug 134335
# app-misc/jive -> bug 142838
# app-text/mgv -> bug 154645
# net-misc/dhcp-agent -> bug 168565
# x11-plugins/wmmail -> bug 73987
app-admin/cpu
app-admin/quickswitch
app-misc/jive
app-text/mgv
net-misc/dhcp-agent
x11-plugins/wmmail

Removed