Message from C, C++ talks

June 2019

— The start key

— Yep, why not

— You then modify the key

— But the start key is enough information to do a side-channel attack


In October 2005, Dag Arne Osvik, Adi Shamir and Eran Tromer presented a paper demonstrating several cache-timing attacks against the implementations in AES found in OpenSSL and Linux's dm-crypt partition encryption function.[31] One attack was able to obtain an entire AES key after only 800 operations triggering encryptions, in a total of 65 milliseconds. This attack requires the attacker to be able to run programs on the same system or platform that is performing AES.

— Btw, if you want to read a bit more of the source 😉

— Sure, no implementation is perfect

— Also, this proves that publishing a paper is not illegal

— Any standard compliant implementation isn't 😉

— You don't get how this attack works, do you?

— This is not related with what you said about the key. This is a bug in the implementation, that btw you 'will do much better'

Message permanent page

— Do you? You are basically stating random facts, copying parts of wikipedia, changing your statements

— Not really

— The attack works because the program was able to read the source of randomness used in AES

— Meaning that it could effectively predict how the key changed