Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ UNSWorksarrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
UNSWorks
Doctoral thesis . 2021
License: CC BY NC ND
https://dx.doi.org/10.26190/un...
Doctoral thesis . 2021
License: CC BY NC ND
Data sources: Datacite
DBLP
Doctoral thesis . 2021
Data sources: DBLP
versions View all 2 versions
addClaim

Software-only Rowhammer Attacks and Countermeasures

Authors: Zhang, Zhi;

Software-only Rowhammer Attacks and Countermeasures

Abstract

Rowhammer is a hardware vulnerability in DRAM memory, where repeated access to memory (i.e., hammer rows) can induce bit flips in neighboring memory locations (i.e., victim rows). Being a hardware vulnerability, rowhammer bypasses all of the system memory protection, allowing adversaries to compromise the integrity and confidentiality of data. To defend against the rowhammer attacks, a class of recent software-only countermeasures suggests exploiting the spatial proximity between the accessed memory location and the location of the bit flip. Particularly, these countermeasures deny the attacker's permission to access exploitable hammer rows that are adjacent to rows of sensitive data. As an example, CATT is the first practical software-only defense. It enforces static physical kernel isolation to prevent attackers from accessing hammer rows next to rows of kernel data. In this thesis, we propose two new rowhammer attacks to defeat the above software-only defenses. We first present CATTmew that breaks CATT and gains privilege escalation. CATTmew is motivated by a key observation that modern OSes have kernel buffers that can be accessed by unprivileged users. The existence of such kernel buffers invalidates the physical kernel isolation enforced by CATT. CATTmew still requires access permission to exploitable hammer rows (i.e., the kernel buffers) and it can be mitigated by other defenses. To this end, we propose a new class of deputy-confused rowhammer attacks, termed Implicit Hammer that has no such requirement. To demonstrate the practicality of implicit hammer, we provide a concrete instance, called PThammer, that overcomes multiple software-only defenses and results in a system compromise. To mitigate the rowhammer attacks as well as the rowhammer effect, we present two countermeasures accordingly. One countermeasure is SoftTRR that defends against page table based privilege escalation attacks, as we observe that rowhammer kernel privilege escalation attacks including CATTmew and PThammer focus on corrupting Level-1 page tables. The other countermeasure is a rowhammer test tool, DRAMDig, that detects rowhammer vulnerability. As understanding a DRAM address mapping is a necessity to perform efficient and effective rowhammer tests, it is not publicly available in Intel-based microarchitectures. DRAMDig efficiently and deterministically uncovers the DRAM address mapping on any Intel-based machine, thus inducing a significant number of rowhammer bit flips.

Country
Australia
Related Organizations
Keywords

Privilege Escalation, 000, Physical Kernel Isolation, DRAM Address Mapping, Address Translation, Double-owned Buffer, Rowhammer, 004

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green