RE: TippingPoint IPS Signature Evasion

["Paul Craig" <paul.craig () security-assessment com>]

This is exploitable (and tested) against IIS 5/5.1 (IIS6/7 are not
vulnerable)
However, potentially other web servers are also vulnerable if they are
capable of decoding alternate unicode characters.

I also agree with you, blaming an IPS for not detecting attack which is
impossible in the wild would be very pointless.
Although IIS 5 is old, it is still relatively common.

Any further questions, feel free to ask.


Cheers,



Paul Craig
Security Consultant
Security-Assessment.com


-----Original Message-----
From: 3APA3A [mailto:3APA3A () SECURITY NNOV RU] 
Sent: Thursday, 12 July 2007 2:30 a.m.
To: Paul Craig
Cc: bugtraq () securityfocus com; full-disclosure () lists grok org uk
Subject: Re: TippingPoint IPS Signature Evasion

Dear Paul Craig,

--Wednesday, July 11, 2007, 1:37:03 AM, you wrote to
bugtraq () securityfocus com:


PC> http://www.test.com/scripts%c0%afcmd.exe
PC> http://www.test.com/scripts%e0%80%afcmd.exe
PC> http://www.test.com/scripts%c1%9ccmd.exe

PC> Web servers located behind a Tippingpoint IPS device which are capable
PC> of decoding alternate Unicode characters can be accessed, and exploited
PC> without triggering the IPS device.

Can  you,  please, provide example of such server? Fatih Ozavci reported
similar   problem   with  Checkpoint  and  Halfwidth/Fullwidth  Unicode,
potential  attack  vector  was IIS with .Net framework, in this case IIS
seems not to be exploitable.

Blaming IPS it does not detect attack which is impossible in-the-wild is
nonsense. Blaming corporate-level IPS doesn't detect attack against SOHO
web server is acceptable nonsense :)

-- 
~/ZARAZA http://securityvulns.com/