Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

New Cache Attacks and Defenses

Guo, Yanan (2020) New Cache Attacks and Defenses. Master's Thesis, University of Pittsburgh. (Unpublished)

Download (747kB) | Preview


The sharing of last-level cache (LLC) among different physical cores makes cache vulnerable to side channel attacks. An attacker can get private information about co-running applications (victims) by monitoring their accesses. Cache side channel attacks can be mitigated by partitioning cache between the victim and attacker. However, previous partition works make the incorrect assumption that only the victim's cache misses are visible to attackers.
In this work, we provide the key insight that both cache hits and cache misses from the victim are vulnerable. For a cache hit,
although it does not affect the existence state, it can still change the replacement state and coherence state, which can also leak information to attackers. Based on this, we propose Invisible-Victim cache (IVcache), a new cache design that can mitigate both traditional LLC attacks and the new variants. IVcache classifies all processes as protected and unprotected. For accesses from protected processes, IVcache handles state changes in a slightly different way to make those accesses absolutely invisible to any other processes. We evaluate the defense effectiveness and performance of IVcache in the gem5 simulator. We show that IVcache can defend against real-world attacks, and that it introduces negligible performance effect to both protected and unprotected processes.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Guo, Yananyag45@pitt.eduyag45
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairYang, Junjuy9@pitt.edujuy9
Committee MemberDickerson, Samueldickerson@pitt.edusjdst31
Committee MemberHu, Jingtongjthu@pitt.edujthu
Thesis AdvisorYang, Junjuy9@pitt.edujuy9
Date: 29 July 2020
Date Type: Publication
Defense Date: 11 February 2020
Approval Date: 29 July 2020
Submission Date: 11 March 2020
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 50
Institution: University of Pittsburgh
Schools and Programs: Swanson School of Engineering > Computer Engineering
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: Cache, Hardware Security
Date Deposited: 29 Jul 2020 17:21
Last Modified: 29 Jul 2020 17:21


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item