The AMDFLAWS folks are getting a lot of flak, much of it rightfully so: for overhyping, bad disclosure process,etc. But escalation from root to secure enclave is actually not nothing. Why is this so easily dismissed? Malware in VTL1 is also a threat that seems valid to me.
Frankly, I've just been skimming the "white paper". But code execution in VTL1 equals problems for MS HyperGuard and CredentialGuard, weakening their defenses. But yeah, that chipset problem too.
-
-
Yes, but it remains something done as admin via a signed driver. It is a design flaw but hardly dramatic for the vast majority of users. What it does show is a) dreadful QA on those critical drivers and b) dubious design choices for the PSP which appears to be over-powerful.
-
CredentialGuard is there to stop software with admin access from reading credentials, HyperGuard is there to stop software with admin access from hooking syscalls. Might not affect everybody, but still legit.
-
Agreed but is this really a QA issue with the driver or a bad design choice for the PSP? The latter might mean a sw update/redesign for the PSP, the former is plain silly.
-
Certainly nothing that I would have hired a PR agency for. Might have made for an entertaining talk at a con, max.
-
That is sort of besides the point now as the short on AMD clearly failed (stock is up!). I am wondering if the choice to put the PSP on-core has meant that AMD has been superficial in its risk assessment compared to, say, ME which is off-core.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.