Slashdot Mirror
Critical Vulnerabilities In Call of Duty: Modern Warfare 3, CryEngine 3
hypnosec writes with news that two security consultants have found vulnerabilities in Call of Duty: Modern Warfare 3 and the CryEngine 3 graphics engine that could harm game makers and players alike. Presenting at the Power of Community (POC2012) security conference, the researchers demonstrated how a denial-of-service attack could affect Modern Warfare 3, and how a server-level attack on CryEngine 3 allowed them to "create a remote shell on a game-player's computer."
"'Once you get access to the server, which is basically the interface with the company, you can get access to all of the information on the players through the server,' Ferrante said. In general, game companies don't seem to be very focused on security but rather on performance of the game itself, Ferrante said. Adding security checks can slow down games, and if the companies don't deem the problem a very critical issue, it will usually be ignored. 'These are games that have a very large market,' Auriemma said."
Well of course they care only about performance Its all their user base really cares about.
If there only was a way to remedy this problem, a "patch" if you will.
Well of course they care only about performance Its all their user base really cares about.
To be fair...nobody is interested in security until things go wrong, they will and they do. Then its look for a scapegoat, and the solution is to remove rights and privacy of the individual for the illusion protection, throw in a few laws, that only affect the law abiding and decent. Then we live in fear.
I have to do triple double or level security passes, including a one time security token, to get into quite a few MMOs. They had to; many RMT organizations made a profit hacking and looting accounts by using keyloggers to obtain passwords.
Occasionally living proof of the Ballmer peak.
Wouldn't the rest of the series down to the original COD also be affected?
... by you know having LAN and private servers again so hacks don't take down the community. Security wouldn't be an issue for Diablo 3 if you could play the fucking game offline. But corporate greed and the dumb masses that feed the move to "online only" games this will become more frequent.
On Tuesday the patch for MW3 will be released. Some know it as Black Ops II but it will practically ensure that nobody is left playing MW3.
The importance of the remote shell is not that "if you can get arbitrary code execution, you can get a remote shell" (this is pretty much a tautology). The importance is that it demonstrates the possibility of arbitrary code execution at all. A lot of security vulnerabilities are difficult to actually exploit. In most cases, the best that an attacker will ever achieve is denial of service ( a crash, or forced disconnect, or using up all the RAM so the game runs too slowly, or soemthing like that).
Contrary to what the movies would have you believe, actual exploits are (especially in a modern environment full of vulnerability mitigations) very difficult to produce in most cases. Many security researchers don't even bother with that step; it's enough to find the vulnerability and flag it "probably exploitable".
There's no place I could be, since I've found Serenity...
They pretty much are. Some of these exploits have existed since the original id Tech 3 engine, from which Modern Warfare 3's engine is originally based. I've been using Luigi's proof of concept tools to do testing on old id Tech 3 engine games that I used to host servers on for years. With his advice I was able to work around certain problems, but not all of them.
I am not sure how bad the vulnerabilities have become, but back then it was generally buffer overflow exploits that allowed player clients to be crashed, servers to be crashed or even the master server to be crashed. There weren't any exploits that I would consider critical, but they were highly annoying.
That's an oversimplification. If the patches are signed and the update system verifies the signatures using well tested libraries, it's probably much harder to attack it that way instead of using any of the other "data entry points", even if that data isn't supposed to contain code.
Dilbert RSS feed
The common will always serve the main. Please continue to serve up your shiny hardware for use, as if you even had a clue to what it means to open up ports to arbitrary root level apps. Bitches.
Anyone know if the Dunia codebase forked from CryEngine before or after this vulnerability was introduced? I'd really like to enjoy some FarCry 3 during my year end holiday but I'd prefer not to get hacked.
HA! This is yet another piece of proof that consoles are better! NAH NAH :-)~ :) Go ahead, hack my console. Whatyagonnado? Jack up my Skyrim campaign. *Feigned Horror Scary Face* :O
Contrary to what the movies would have you believe, actual exploits are (especially in a modern environment full of vulnerability mitigations) very difficult to produce in most cases. Many security researchers don't even bother with that step; it's enough to find the vulnerability and flag it "probably exploitable".
On another hand, unpatched, unresolved, unfixed security issues will attract hackers until they find a way to exploit them. So, no need to find an easy exploitable scenario to flag them as probably exploitable. Why someone should sit and wait it becomes exploitable to fix it? It's a kind of security through obscurity you are talking about. I'm sorry, but this must be secure by design.
Achille Talon
Hop!
You will have to pay our software or you will get all of your computers cracked to the bones. We made sure that a hole was there for that matters. After all if a customer's computer gets penetrated, that's ... collateral damage. Besides you all accepted no guarantees when you purchased and you are the only ones who are going to suffer the consequences of our actions. So who cares :).
The MW3 Staff
Vulnerable to hacks indeed. WELL... if these Call of Duty Black Ops server thingies have become a problem, a way to hurt people I say maybe we should call it a day and just shut them down.
'Cause we don't want to hurt people now do we.
Do we??
<blink>down the rabbit hole</blink>
The attack stipulates the server-side is compromised. Updates come from the server-side. This is not a remote code execution, this is a compromised update server scenario, no need for any exploit at all.There is not much that can be done on client side to defend.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
I agree, but only if the signature keys are off-line and well protected. This rarely seems to be the case though.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
What do you mean, I'm talking about security through obscurity? That makes nothing resembling sense. I certainly didn't suggest that the vuln shouldn't be reported if the researcher doesn't develop an exploit, nor that the developer shouldn't fix it. Some devs won't take a vuln seriously without a PoC, it's true, but that's a failure on their part, not on the researcher's. Some developers don't take security seriously regardless.
I also don't understand why you seem to think that flagging a vuln as probably exploitable is something that requires it be easily exploitable, or where you think I even implied as such. Are you familiar with security testing? You can tell the probable exploitability of a vulnerability by reading a few lines of disassembly, in most cases. Attempt to read a NULL pointer that the attacker cannot control and there's no way for the attacker to have mapped page 0 in this process' memory space? Not exploitable (except for DoS). NX (Data Execute Prevention) violation? Exploitable. Read fail on the instruction pointer at or near 0? Probably exploitable (and can most likely be refined with a bit more research).
My claim was, quite simply, that most people would merely have done the wrok needed to say "yep, looks exploitable, probably arbitrary code execution", flagged the vuln as such, and moved on. The time and effort to develop working exploits, while fun, is rarely cost-effective to a white-hat.
There's no place I could be, since I've found Serenity...
Good point on the distinction between server and client side, and the fact that a meaningful cliam is actually being made here (two, really: first that it's possible to get arbitrary code execution on the server, second that it's possible to leverage that into arbitrary code execution on the client).
However, I don't quite buy your argument about updates. The update server is not usually the game server. Compromising a game server doesn't (in theory) let you send an update to the client, much less force them to install it. In practice, it sounds like the game servers have way more control over the connected clients than they should. Especially in the case of games where anybody who wants to can host a game (act as server), connecting to that server absolutely should not expose the client to arbitrary code execution. That's a huge security flaw.
There's no place I could be, since I've found Serenity...
Hello!! Fashion,low price,the good shopping places, Cheap wholesale and retail Gucci/Shoes $45, ( Discount UGG/Boots ) LV Shoes $46, DG Shoes $46, BURBERRY Shoes $46, LACOSTE Shoes $46, Women Boots $55, handbags(Coach lv fendi d&g/Gucci) $39, Sunglasses(Oakey,coach/Gucci,Armaini) $25, free shipping and quantity discount, Accept credit card and PAYPAL ==== http://www.cbssbase.com/ ==== ==== http://www.cbssbase.com/ ====