1 thought on “Eek! /. on an exploit against processors …”

  1. While it is something to be patched, its not that scary a hole. To exploit it you already need ring 0 access (ie. typical kernel level access). If you have ring 0 then it provides a way to perhaps hide some nasty code in smram (but it’d have to be coded very much against your particular platform’s config). But its quite unlikely to let you get around a hypervisor as they’ll tend to virtualize mtrr access in any event. It isn’t really an exploit against the processor either – it simply uses well known features of the x86 processors since 386 days.

Comments are closed.