Software projects grow very large as they become popular, due to the support of a large number of new and legacy features. This results in complex systems that often expose a large attack surface for attackers to exploit. The Linux kernel exemplifies this problem, due to the large amount of features that are included by default in current Linux distributions.
In this project, we explore various techniques to reduce the attack surface of the Linux kernel, by identifying and disabling access to unnecessary features. We investigate run time and per-process attack surface reduction (e.g., by automatically deducing the set of kernel functions a process requires), as well as compilation time and system-wide attack surface reduction (e.g., by automatically producing small kernel configurations). We also create metrics to measure those attack surface, to better compare the effectiveness of each approach.
Our results show that such "economy of mechanism" approaches greatly improve overall system security, and also indicates that other large software projects can benefit from such approaches.
Titel | Art | Betreuer | Status |
---|---|---|---|
Linux Kernel Attack Surface Reduction Measurement | Masterarbeit | Prof. Dr. Rüdiger Kapitza | abgeschlossen |
Kernel as a Service - Custom tailored kernels for the cloud | Bachelorarbeit | Prof. Dr. Rüdiger Kapitza | abgeschlossen 2013 |
If you are interested in writing a thesis regarding this project, please feel free to contact us.