A Scalable Architecture for Proof-Carrying Code

Abstract

Proof-Carrying Code (PCC) is a general mechanism for verifying that a code fragment can be executed safely on a host system. The key technical detail that makes PCC simple yet very powerful is that the code fragment is required to be accompanied by a detailed and precise explanation of how it satisfies the safety policy. This leaves the code receiver with the simple task of verifying that the explanation is correct and that it matches the code in question.

Previous implementations of PCC used safety explanations in the form of explicit formal proofs of code safety, thus gaining leverage from a substantial amount of previous research in the area of proof representation and checking, but at the expense of poor scalability due to large proof sizes. In this paper we describe a series of changes that are necessary to achieve a truly scalable architecture for PCC. These include a new proof representation form along with a better integration of the various components of a PCC checker. We also present experimental results that show this architecture to be effective for checking the type safety of even very large programs expressed as machine code.

Acknowledgments

We would like to thank Shree Rahul for the help with the collection of the experimental data presented in this paper and to Peter Lee, Mark Plesko, Chris Colby, John Gregorski, Guy Bialostocki and Andrew McCreight from Cedilla Systems Corporation who have implemented the certifying compiler for Java used in these experiments.