Exploring Novel File System Objects for Data-Only Attacks on Linux Systems

The paper titled "Beyond Control: Exploring Novel File System Objects for Data-Only Attacks on Linux Systems" by Jinmeng Zhou and colleagues delves into the realm of non-control data attacks on OS kernel exploits. The study focuses on identifying critical non-control objects within the file subsystem of Linux kernels and assessing their exploitability. By utilizing a custom analysis and testing framework, the researchers categorize these objects into types suitable for various exploit strategies, including a novel approach to bypass defenses. The identified objects offer the advantage of being exploitable without the need for Kernel Address Space Layout Randomization (KASLR), simplifying and enhancing the reliability of exploits. The research evaluates the exploitability of these file system objects using 18 real-world Common Vulnerabilities and Exposures (CVEs) and develops 10 end-to-end exploits against the kernel with state-of-the-art mitigations enabled. This work is submitted for presentation at the 31st ACM Conference on Computer and Communications Security (CCS) in 2024.