Security in innovative new operating systems

A principal criterion by which new operating systems are judged is the level of performance that they provide for applications. To this end, new operating systems have sought novel approaches to performance enhancement. A theme common to many of these initiatives is that of specialization. Instead of an operating system designed to serve all applications (either equally well or equally badly), the operating system is adapted to serve the needs of the application. The intent is not to provide a different static operating system for each application but to allow the operating system to be dynamically modified or specialized to best serve each application. The five operating system efforts presented are: the Exokernel Project, the Fluke Project, the Fox Project, the Scout Project, and the SPIN Project. The authors hope to give an overview of the innovative techniques being used to enhance performance in these systems and to discuss the effect of those enhancements on one's ability to reason about the security properties of systems.