Using Kernel Exploits to Bypass Sandboxes for Fun and Profit

Discussion in 'other security issues & news' started by Thankful, Mar 20, 2013.

Thread Status:
Not open for further replies.
  1. Thankful

    Thankful Savings Monitor

    Joined:
    Feb 28, 2005
    Posts:
    3,741
    Location:
    New York City
  2. I find myself wondering if there's a better way to deal with kernel vulnerabilities. A kernel written mostly in managed code, maybe? That might be a good use of modern computers' processing power.

    BTW, I don't think font rendering on Linux uses kernel calls, probably just calls to X11. But X11 runs as root. So a vulnerability in e.g. FreeType could probably result in arbitrary code being run with root privileges.
     
  3. BoerenkoolMetWorst

    BoerenkoolMetWorst Registered Member

    Joined:
    Dec 22, 2009
    Posts:
    3,771
    Location:
    Outer space
    Interesting read, thanks :)
     
  4. Hungry Man

    Hungry Man Registered Member

    Joined:
    May 11, 2011
    Posts:
    9,148
    @GJ,

    If you want to deal with a secure kernel you want Grsecurity and Seccomp. The two best methods.

    When MRW labs released the info the night of pwn2own I posted pretty quickly about this. No matter how powerful your MAC is the kernel is too tempting a target. Until you secure your kernel your system is vulnerable.
     
  5. Yeah I know about GrSec and Seccomp. :) I'm trying to imagine more radical strategies; e.g. writing most drivers in a language with type safety and memory management, instead of portable assembly. This is what Singularity does. IOW stuff that isn't on the market yet, but might offer a longer term solution.

    Also I would point out that GrSec is (very unfortunately) not applicable to Windows. :mad:
     
Loading...
Thread Status:
Not open for further replies.