Re:Technical feasibility of decrypting https by replacing th...
by acqq on Sunday July 14, 2013 @11:34AM
(#6041117)
Sorry I don't see the relevance of CPU having any hidden global state in the thought experiment I've addressed. If all I give you is a PGP file made on the Intel CPU you simply don't have access to the global state of my CPU and I can be sure that I'd get exactly the same file up to the last bit using non-Intel CPU starting from the same inputs (including same 'seeds'). So I can actually be absolutely sure in AES instructions for any scenario where I only give you the results of encryption.
[ Parent
| on HN ]
Re:Technical feasibility of decrypting https by replacing th...
by lucian1900 on Sunday July 14, 2013 @12:14PM
(#6041252)
You're open to a different attack: If someone has access to (part of) your CPU state before and after you make the PGP file, they may be able to decrypt the message if they manage to intercept it.