Most Home Routers Vulnerable to Flash UPnP Attack
An Anonymous reader noted that some folks at GNU Citizen have been researching
UPNP Vulnerabilities in home routers, and have produced a flash swf file capable of opening open ports into your network simply by visiting an unfortunate URL. Looks like Firefox & Safari users are safe for now.
Yes. You may not be sure if a site is malicious or not, without visiting it.
And some sites may become malicious suddenly because of all those syndicated ads around.
Patents Drive Free Software as Hurricanes Drive Construction Industry
I thought the recommended steps for setting up a router were:
A. Unbox
B. Throw away the disk
C. Plug in your machine, Turn on the router and navigate to the webgui
D. Turn off UPNP
E. ??? (Change default name and password, set WPA, Turn off SSID etc....)
F. Profit...
The point is, I'd always been told to turn off UPNP 'cos sooner or later something is going to open ports that you don't know about.
Well yes. If you never visit a site with adverts. Or the Internet as it's otherwise known. Sure, you can block them (and I do) but sometimes sites switch to new providers and you are vulnerable for the time it takes to update the block file.
I'm not really surprised to be honest - I always thought UPnP looked fishy to me so I disabled it on my router. I don't like the idea that anyone coming to visit can plug in their malware-ridden Windows laptop and reconfigure my router. Sure, having it turned off means X-Box Live is less happy but that only decreases the number of people who can call me "fag" on a daily basis. I wonder if Microsoft will update the X-Box Live support page where they say that UPnP doesn't make your network insecure...
I also have Flash disabled by default because it is well known to be insecure and buggy and a delivery system for malware. Most proper web-browsers either let you enable flash on a per-site basis or will allow you to do so with a plug-in and this is really the way to go.
as usual opera is resistant.
[...] a flash swf file capable of opening open ports into your network [...]
:D
Hold on, now I'm confused: does this attack open open ports, or does it open ports open? Or even worse, does it open open open ports?
Hello! I'm a disaster waiting to happen!
Turn off UPnP! Why on Earth do you want it on anyway? That's the problem here - an XSS is one matter, although being able to send SOAP-style requests across your local network is a major concern. But having a router that automatically opens ports based on virtually zero authentication? A nightmare waiting to happen.
Never used it. Never wanted it. Never turned it on. Always turned it off on EVERYTHING. UPnP is the problem here - a simple (unauthenticated) HTTP-style page requested in a browser suddenly starts opening ports to your network. It should not happen. Even my DSL router/wireless router/Linux router has SSL only, passworded access to do anything even approaching opening ports. And if a webpage pops up with an authentication dialog with the header "Wireless Router" and you type in your password, then you're a fool, unless you specifically requested the router's configuration page.
There's rarely even a log of what UPnP has done - which ports it's opened in the past etc. for whom.
Just turn the damn thing off. It's too dangerous.
If the firmware has UPnP IGD enabled, then your machine is vulnerable to this attack.
The vulnerability is really Flash not restricting what untrusted scripts can do. The router's UPnP IGD profile is working as designed - an application on a machine within the firewall requests that an incoming port be forwarded, so the router does that. This is useful for VoIP, IM, P2P and other applications that need to be contactable from the outside world. Malicious programs that are running on your machine can always initiate outgoing connections, so generally the UPnP IGD is not allowing anything that cannot already be done. In the case of Flash, it is probably blocking most outgoing connections, so UPnP does expand the possibilities for a malicious Flash app to initiate connections with your machine. But unless Flash also allows you to open server sockets, the attacker would also need to find an exploitable service running on your machine.
All this should be detectable by a decent firewall program running on your local machine.
Yup, I have seen people computers infected from msn.com the banner ad's were at one time installing spyware from the default IE home page.
All it takes is to get your nastyness in a bunch of Ad rotations from doubleclick and other scumbag webad companies and you can hose a huge swath of the net.
Do not look at laser with remaining good eye.
Yes, but the social engineering requirement is more or less gone in this case. It takes substantially less work to convince someone to click a link than to download a file. (Granted, Bonzai Buddy got people by just being a purple ape.)
Why, look no further than the MyMiniCity/Goatse/2girls1cup links being posted here in every thread! At least one person clicks and ends up warning others. (Either by downmodding or posting.) Why, you just need someone who's curious enough to click.
On the other hand, it requires a bit of work to get someone familiar with malware to click on a 'you just won' banner and download the mystery prize. Don't even get me started on random email attachments following nonsense messages.
I just read Slashdot for the articles.
If a flash plugin can make outgoing XML requests it can persuade a upnp server to make your machine wide open, thus completely disabling your firewall. Making those kind of requests sounds like the kind of thing you want Flash to do, so I'd imagine all versions are vulnerable.
There are some ports.. 137,139,445,etc. that you really don't want on the open internet. If the plugin does something like a port forward of 0-65535 to your machine suddenly *every* service on there is wide open to any attack. It'll bypass protections from eg. the default XP firewall as the packets will appear to be coming from the local LAN (the router) rather than the original source.
It's not just flash (although a malicious advert on a page is the most obvious vector for this). Anything that runs on your machine can do it.. I reckon you could craft such an attack in javascript even (XMLHttpRequest with the right code).
Once the ports are open anything that manages to run on your machine can leave itself wide open without having to make telltale outgoing port connections (although it's often said that outgoing connections are the reason upnp is 'not worse' than existing protections, no working trojan would work in that manner, since the target of the outgoing connection would quickly be found and shut down.. OTOH leaving a trojan on your machine listening on your machine waiting for the command to send spam/infect others/distribute child porn/whatever is much more real a thread).
Challenging an anon coward...
Yeah, that'll work.
I use Linux with Seamonkey and..... uuhhh nothing else. :)
No infections either.
It looks like your doing everything except the simplest solution.
Oh and yes I use UPNP.