New cache designs for thwarting software cache-based side channel attacks

Abstract

Software cache-based side channel attacks are a serious new class of threats for computers. Unlike physical side channel attacks that mostly target embedded cryptographic devices, cache-based side channel attacks can also undermine general purpose systems. The attacks are easy to perform, effective on most platforms, and do not require special instruments or excessive computation power. In recently demonstrated attacks on software implementations of ciphers like AES and RSA, the full key can be recovered by an unprivileged user program performing simple timing measurements based on cache misses. We first analyze these attacks, identifying cache interference as the root cause of these attacks. We identify two basic mitigation approaches: the partition-based approach eliminates cache interference whereas the randomization-based approach randomizes cache interference so that zero information can be inferred. We present new security-aware cache designs, the Partition-Locked cache (PLcache) and Random Permutation cache (RPcache), analyze and prove their security, and evaluate their performance. Our results show that our new cache designs with built-in security can defend against cache-based side channel attacks in general-rather than only specific attacks on a given cryptographic algorithm-with very little performance degradation and hardware cost.

A Non-Inclusive Memory Permissions architecture for protection against cross-layer attacks

“…We do not protect against side-channels and covert channels; many previous works have addressed these issues [57,58]. We also do not protect against Denial-of-Service attacks initiated by higher privilege levels against lower levels, as higher levels can always deny service by preventing lower levels from executing.…”

Putting Trojans on the Horns of a Dilemma: Redundancy for Information Theft Detection

“…-Testing to guide system design: The basic framework for information flow measurement via repeated deterministic replays presented in this paper could be used to test existing systems for inference and side channels [7][8][9][10]. -Information flow security systems: Researchers have begun to explore the possibility of building systems that enforce information flow security by comparing different outputs for different inputs [11].…”

“…This inequality says that the entropy in the log file can only come from three sources: D, A, and N . The deterministic behavior that we did not need to log and replay from the log but did (I(L; D)) is chosen nondeterministically, meaning that before the transaction we do not know what determinism will be logged, so this term can be non-zero 10 . Any entropy from the confidential data A that was logged (I(L; A)) and then replayed will mask the entropy and can hide a covert channel from us.…”

“…9 Conditional entropy is also sometimes written as HN (B). 10 Recall that D is the system specification and is therefore only deterministic when the system S behaves as specified, so that I(L; D) can be non-zero for S ′ during replay.…”

A Cache Timing Analysis of HC-256

“…Subsequent research verified the correctness of the findings [11,10,9,15], improved the attack technically [14,3,8] or algorithmically [5], and devised and analysed countermeasures [6,4,16].…”