Privilege Escalation Vulnerabilities and the Value of Restricted Accounts

Found these links on another forum:

Windows Privilege Escalation Vulnerability...
Windows
This Windows-based exploit has been seen in the wild...
Windows
Not just Windows...
Linux
Even the very secure...
OpenBSD

A poster there claims the following:

Seeing that privilege escalation vulnerabilities are quite common among all operating systems (not just Windows), and that it looks to be easy, maybe very easy to get root on linux systems if an attacker is already in a user account, and that linux is supposedly more secure by design than XP or even Vista:

How valuable are restricted accounts against malicious attacks?

 

It´s just another layer, but apparently it´s very hard to write bugfree code, so no surprise that all these things can be bypassed. Same goes for HIPS, but then again, the chances that certain security restrictions will be bypassed is not that big, I think.

 

And one that doesn't slow down performance at all.

An engineering management proverb (pointing out that resources are finite) goes: good, fast, cheap - choose two.

Maybe home users want "cheap" and "fast", though we often say we want otherwise.

Is it because of obscurity that malware writers don't usually even try to bypass these (few at home use either HIPS or restricted accounts w/XP), or is it because of something else?

The HIPS comparison raises another question: between the two, is one any more secure than the other?

For instance, if the human element is considered, would a restricted account be safer? If the human element is not an issue, would a HIPS be safer, given that a (the?) major security weakness of HIPS is buffer overflow vulnerabilities, and that these can probably be avoided more easily with better programming practices than privilege escalation vulnerabilities can?

 

Hello,
Those nix vulnerabilities refer to very old versions of kernels. Nothing but a simple update so easily available to fix things out. Current RH is 5 and BSD stands at 4.2, if I'm not mistaken.

RH 4 was released in Feb 2005! That's more than 2 years ... almost 3. So I would not really sweat ... considering that the most up to date Windows XP kernel has on average 3-4 vulnerabilities every month.

Mrk

 

I see what you're saying. I keep in mind that those 3-4 vulnerabilities per month are for all types of vulnerabilities, if I'm not mistaken.