uXOM: Efficient eXecute-Only Memory on ARM Cortex-M

Authors: 

Donghyun Kwon, Jangseop Shin, and Giyeol Kim, Seoul National University; Byoungyoung Lee, Seoul National University, Purdue University; Yeongpil Cho, Soongsil University; Yunheung Paek, Seoul National University

Abstract: 

Code disclosure attacks are one of the major threats to a computer system, considering that code often contains security sensitive information, such as intellectual properties (e.g., secret algorithm), sensitive data (e.g., cryptographic keys) and the gadgets for launching code reuse attacks. To stymie this class of attacks, security researchers have devised a strong memory protection mechanism, called eXecute-Only-Memory (XOM), that defines special memory regions where instruction execution is permitted but data reads and writes are prohibited. Reflecting the value of XOM, many recent high-end processors have added support for XOM in their hardware. Unfortunately, however, low-end embedded processors have yet to provide hardware support for XOM.

In this paper, we propose a novel technique, named uXOM, that realizes XOM in a way that is secure and highly optimized to work on Cortex-M, which is a prominent processor series used in low-end embedded devices. uXOM achieves its security and efficiency by using special architectural features in Cortex-M: unprivileged memory instructions and an MPU. We present several challenges in making XOM non-bypassable under strong attackers and introduce our code analysis and instrumentation to solve these challenges. Our evaluation reveals that uXOM successfully realizes XOM in Cortex-M processor with much better efficiency in terms of execution time, code size and energy consumption compared to a software-only XOM implementation for Cortex-M.

USENIX Security '19 Open Access Videos Sponsored by
King Abdullah University of Science and Technology (KAUST)

Open Access Media

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.

BibTeX
@inproceedings {236270,
author = {Donghyun Kwon and Jangseop Shin and Giyeol Kim and Byoungyoung Lee and Yeongpil Cho and Yunheung Paek},
title = {{uXOM}: Efficient {eXecute-Only} Memory on {ARM} {Cortex-M}},
booktitle = {28th USENIX Security Symposium (USENIX Security 19)},
year = {2019},
isbn = {978-1-939133-06-9},
address = {Santa Clara, CA},
pages = {231--247},
url = {https://www.usenix.org/conference/usenixsecurity19/presentation/kwon},
publisher = {USENIX Association},
month = aug
}

Presentation Video