New IE Remote Code Execution Vulnerability Discovered
An anonymous reader writes "Microsoft is investigating a new remote code execution vulnerability in Internet Explorer and preparing a security update for all supported versions of its browser (IE6, IE7, IE8, IE9, IE10, and IE11). The company has issued a security advisory in the meantime because it has confirmed reports that the issue is being exploited in a 'limited number of targeted attacks' specifically directed at IE8 and IE9."
I'm not sure how this is news for nerds, or even anything that matters...
Always, always.
Which is way better than having an advisory and then having to wait weeks for a fix that requires a reboot,
Even Microsoft sent flowers to the mock funerals. And now they're digging out the grave to patch a corpse?
Get free satoshi (Bitcoin) and Dogecoins
Things like this happen, but I have to say that these days Microsoft has mostly taped Windows together quite well. We don't anymore see sensational headlines like "Blaster worm infects millions of computers". So for the 6.x core things are way better than in the past. However the EOL'ing of Windows XP will probably zombify heaps of machines.
"A limited Number of Targeted Attacks"
Must be that all 25 of their IE users got hit so now they are trying to patch it.
The bad guys could have kept this secret till after the end-of-life for XP and made a mint.
I thought IE 10 and after were sand-boxed? Or is it the nature of the buffer overrun that the injection gets CPU level access?
According to the advisory they only get current user-level access. How do they run a buffer overrun exploit that actual stays in the user-context and doesn't go all the way to the CPU?
It sounds like the destruction of objects is incomplete, so the attacker can still write to that area of memory. It's certainly possible that it's writeable BECAUSE it's still associated with the process, which mean it runs in the context of that process. Additionally, it's likely that while the attacker can write to the memory, they can't arbitrarily execute it directly. Rather, they have to cause IE to execute it, in which case it would run with the privileges IE has when IE runs it.
A security problem there is that since IE4, IE has been integrated with the system shell. Therefore, IE privileges are shell privileges - anything the user can do, the browser can do. For this reason, I much prefer a browser that is only a browser, not another view of the system shell. A browser that's just a browser can only screw up web pages, not the entire system.
Yes, I'm aware that on Windows 8 Microsoft has attempted to sandbox the browser. Like putting a lion in a cage, that works until the lion reaches through the bars. It doesn't compare to using a browser such as Firefox which does not have the potential harmful abilities baked in. No need to sandbox something that doesn't exist.
Botnet Command and Control map:
https://www.shadowserver.org/wiki/pmwiki.php/Stats/BotnetMaps#botnet
New IE Remote Code Execution Vulnerability Discovered... 3 years ago, reported to Microsoft, that reported it to the NSA, that took advantage of it all that time. Now a new, safer backdoor that only they should exploit is being deployed thru the fix for this vulnerability.
Is all those new slashdot redesigns, headlines can't hold all the relevant information anymore.
If I were to guess, NSA and MS must have coded the back door themselves.
NSA: Dear Microsoft, too many foreign parties are now using our vulnerability, time to replace it with a new one.
I think I remember using it once. But the alternative was Netscape 4.
But don't mention the critical Firefox flaws, because its against /. groupthink
http://www.theregister.co.uk/2013/09/18/firefox_24_update/