Modular verification of hardware systems

Abstract

As hardware systems are becoming bigger and more complex, it is becoming increasingly harder to design and reason about these systems in a monolithic fashion. While hardware is often designed in a modular manner, its verification is rarely performed modularly. Moreover, any modular refinement to an existing system requires a full system verification to guarantee correctness, even if only a few components of the system have been refined. In this thesis, we present a new framework for modular verification of hardware designs written in the Bluespec language. That is, we formalize the idea of components in a hardware design, with well-defined input and output channels, and we show how to specify and verify components individually. For verifying a full system, we show how the proofs of its components can be composed, treating the components as black-boxes and not looking beyond their specifications. As a demonstration of this methodology, we verify a fairly realistic implementation of a multicore shared-memory system with two types of components: memory system and processor, with machine-checked proofs in the Coq proof assistant. Both components include nontrivial optimizations, with the memory system employing an arbitrary hierarchy of cache nodes that communicate with each other concurrently, and with the processor doing speculative execution of many concurrent read operations. Nonetheless, we prove that the combined system implements sequential consistency. To our knowledge, our memory-system proof is the first machine verification of a cache-coherence protocol parameterized over an arbitrary cache hierarchy, and our full-system proof is the first machine verification of sequential consistency for a multicore hardware design that includes caches and speculative processors. We also embed the Bluespec language in the Coq proof assistant and formalize its modular semantics, enabling a verification engineer to obtain machine-checked proofs for Bluespec designs using our framework.