Re: RE: FAX virus
('binary' encoding is not supported, stored as-is)
Scott,
The question was originally posed as “Can anyone send a fax that includes a file infected with the virus/ worm” (Wed, 07 Mar)
My concern was not with sanitisation as you are trying to suggest. It is with the idea that a buffer overflow is the attack vector. That for instance a virus / worm could be embedded. This is a suggestion that I remain in disbelief of.
What I suggested is an alternative. Rather then sending a virus/worm, send a XSS attack and rely on the users in the organisation to exploit this.
If this is sent in a PDF, it is going to display as the scripted entry. So a conversion to an attached PDF is still not going to work as what is displayed is what is on the page. It will need to be sent directly to a web enabled email or web server.
So it is not that I am suggesting an attack against the document processor, but rather extending this by adding user interaction. It is thus the user who extends this through reading email with the link or opening a page. In this case the site would still also have a simpler attack against the user in any instance.
I also believe that you suggested “If you allow the asterisk and parenthesis through, you run the risk of allowing SQL injection passed to your service.” The idea you stated other then a buffer overflow was a SQL injection. Neither of these are valid. You failed to consider XSS and having user involvement at the time. I did not think of this either. If you had suggested this I would have conceded that as an attack vector has I now have.
The suggestion that an embedded buffer overflow or binary attack against the fax server is still out of the question.
You for example stated:
“The communication is one-way as Craig so eloquently pointed out. But what if the command is to drop a database? In that case there was never any intention of receiving data back, it's a malicious vandalism of your database.”
Again, this is not a valid path or attack vector Scott. You are attempting to add too much complexity. So consider a XSS as a simplification of your idea. By over complicating the idea to send SQL commands to an unknown database or worse embed a buffer overflow (which I am still wondering how you could even propose as I see no way to fax a NOP sled) you take the thesis to a level where it may not be supported.
Regards,
Dr Craig Wright
------------------------------------------------------------------------
This list is sponsored by: Cenzic
Need to secure your web apps NOW?
Cenzic finds more, "real" vulnerabilities fast.
Click to try it, buy it or download a solution FREE today!
http://www.cenzic.com/downloads
------------------------------------------------------------------------
This archive was generated by hypermail 2.1.7
: Sat Apr 12 2008 - 10:58:13 EDT