Re:Towards Linux Kernel Memory Safety
by StreamBright on Thursday November 16, 2017 @03:18AM
(#15710977)
Good to know. I think there is no silver bullet in security so anything making exploitation harder is good.
This already exists on x86_64 Linux in the form of SMEP and SMAP.
> With SMEP enabled, itβs no longer possible to map exploit payloads in userland, as the CPU will trigger a fault if it attempts to execute those user pages in kernel mode
http://vulnfactory.org/blog/2011/06/05/smep-what-is-it-and-h...
> SMEP can be used to prevent supervisor-mode code from unintentionally executing user-space code. SMAP extends this protection to reads and writes
https://en.wikipedia.org/wiki/Supervisor_Mode_Access_Prevent...
These mitigations aren't overly novel at this point and don't prevent exploits (just complicates them). Here's a recent CVE PoC containing an SMEP and KASLR bypass: https://github.com/xairy/kernel-exploits/blob/master/CVE-201...
Full memory safety takes this much further, and should be far harder to bypass.