LOCK trek: navigating uncharted space

Abstract

The design principles of the logical coprocessing kernel (LOCK) project are considered. LOCK is an advanced development of hardware-based computer security and cryptographic service modules. Much of the design and some of the implementation specifications are complete. The formal top level specification (FTLS) also is complete and the advanced noninterference proofs are beginning. This hardware-based approach has brought the LOCK project into many uncharted areas in the design, verification, and evaluation of an integrated information security system. System integration currently appears to be the single largest programatic problem. The authors examine two important design areas: design verification and porting Unix System V to a LOCK host. The verification tools seem able to verify design only and not implementation.<>