Automatic Generation of Veri able Cache Coherence
نویسنده
چکیده
Performance modelling and veriication are vital steps in the development cycle of any cache coherency protocol. Two separate models are usually required to perform each analysis step and as protocols become increasingly complex each can become correspondingly unwieldy. We examine how stochastic process algebra can be used to describe cache coherency protocols in such a way as to allow both the correctness and performance of the protocol to be investigated together. We reintroduce a formalism called \Spade" and show how a simulation model can be generated from a Spade description of a memory system. We also show how the coherency protocol underlying the memory system can be veriied as correct by showing observational equivalence between the system and a speciication for the memory system based on a memory consistency model. We demonstrate the approach by applying it to a model of the write-through coherency protocol for bus-based shared-memory multiprocessors. We validate simulation results from the model by comparing against a hand coded simulation.
منابع مشابه
Parallelizing the Mur' Veriier
With the use of state and memory reduction techniques in veri cation by explicit state enumeration, runtime becomes a major limiting factor. We describe a parallel version of the explicit state enumeration veri er Mur' for distributed memory multiprocessors and networks of workstations that is based on the message passing paradigm. In experiments with three complex cache coherence protocols, pa...
متن کاملcient Veri cation of Symmetric Concurrent Systems
Previously, we proposed a reduction technique ID93] based on symmetries to alleviate the state explosion problem in automatic veriication of concurrent systems. This paper describes the results of testing the technique on a wide range of algorithms and protocols, including realistic multiprocessor synchronization algorithms and cache coherence protocols. Memory requirements were reduced by amou...
متن کاملStructured Parallel Programming and Cache Coherence in Multicore Architectures
It is clear that multicore processors have become the building blocks of today’s high-performance computing platforms. The advent of massively parallel singlechip microprocessors further emphasizes the gap that exists between parallel architectures and parallel programming maturity. Our research group, starting from the experiences on distributed and shared memory multiprocessor, was one of the...
متن کاملHardware/Software Coherence in Hybrid Memory Models
Current cache coherence protocols limit the scalability of chip multiprocessor (CMP) architectures. The expected increase of the number of cores in next generation CMPs call for an evolution of the memory subsystem. One solution is to introduce a local memory side to the cache hierarchy, forming a hybrid memory model. On the one hand, local memories are more power-efficient than caches and they...
متن کاملCompiler-Enforced Cache Coherence Using a Functional Language
The cost of hardware cache-coherence, both in terms of execution delay and operational cost, is substantial for scalable systems. Fortunately, compiler generated cache management can reduce program seri-alization due to cache-contention; increase execution performance; and reduce the cost of parallel systems by eliminating the need for more expensive hardware support. In this paper, we use the ...
متن کامل