System design of a CC-NUMA multiprocessor architecture using formal specification, model-checking, co-simulation, and test generation

Garavel, Hubert; Viho, César; Zendri, Massimo

doi:10.1007/s100090100044

Cited by 22 publications

(13 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We plan to continue our work by applying equivalence checking and model checking techniques to industrial GALS systems described in GRL. Hardware/software co-simulation is also possible using the EXEC/CAESAR framework [14] of CADP, which enables the C code generated from a GRL description to be integrated with a physical platform. We also plan to investigate the connection of GRL to verification frameworks based on the synchronous paradigm to analyse the behaviour of individual blocks corresponding to synchronous subsystems.…”

Section: Discussionmentioning

confidence: 99%

GRL: A Specification Language for Globally Asynchronous Locally Synchronous Systems

Jebali

Lang

Mateescu

2014

Formal Methods and Software Engineering

View full text Add to dashboard Cite

International audienceA GALS (Globally Asynchronous, Locally Synchronous) system consists of several synchronous subsystems that evolve concurrently and interact with each other asynchronously. Most formalisms and design tools support either the synchronous paradigm or the asynchronous paradigm but rarely combine both, which requires an intricate modeling of GALS systems. In this paper, we present a new language, called GRL (GALS Representation Language) designed to model GALS systems in an abstract and versatile manner for the purpose of formal verification. GRL has formal semantics combining the synchronous reactive model underlying dataflow languages and the asynchronous concurrent model underlying process algebras. We present the basic concepts and the main constructs of the language, together with an illustrative example

show abstract

Section: Discussionmentioning

confidence: 99%

GRL: A Specification Language for Globally Asynchronous Locally Synchronous Systems

Jebali

Lang

Mateescu

2014

Formal Methods and Software Engineering

View full text Add to dashboard Cite

show abstract

“…Finally, we will be able to develop a stand-alone compiler to generate a prototype distributed implementation of an LNT composition of tasks, as a family of remote task and protocol processes. The code for each task could be obtained by extending the EXEC/CAESAR framework [13] of CADP (which currently generates sequential code simulating a concurrent or sequential process) to make it fit the synchronization protocol interface.…”

Section: Discussionmentioning

confidence: 99%

Formal Verification of Distributed Branching Multiway Synchronization Protocols

Evrard

Lang

2013

Formal Techniques for Distributed Systems

View full text Add to dashboard Cite

Distributed systems are hard to design, and formal methods help to find bugs early. Yet, there may still remain a semantic gap between a formal model and the actual distributed implementation, which is generally handwritten. Automated generation of distributed implementations requires an elaborate multiway synchronization protocol. In this paper, we explore how to verify correctness of such protocols. We generate formal models, written in the LNT language, of synchronization scenarios for three protocols and we use the CADP toolbox for automated formal verifications. We expose a bug leading to a deadlock in one protocol, and we discuss protocol extensions. This work was partly funded by the French Fonds national pour la Société Numérique (FSN), Pôles Minalogic, Systematic and SCS (project OpenCloudware).

show abstract

“…Figure 3 gives an overview of how DLC proceeds to generate a distributed implementation. DLC relies on the EXEC/CAESAR [9] tool of CADP to obtain a sequential implementation (in C) of each task process. However, the implementation produced by EXEC/CAESAR is not complete: it can list the currently possible actions of a process, but does not decide which action shall be realized.…”

Section: Environment Interaction With Hook Functionsmentioning

confidence: 99%

“…Implementation Correctness. The sequential implementation of each task is obtained by the existing EXEC/CAESAR tool of CADP, which has already been employed in a formal context [9]. Interaction between tasks is achieved by a synchronization protocol [6] that we verified [4] using a formal approach that detected possible deadlocks in other protocols of the literature [5].…”

Section: Overview Of Compilation Internalsmentioning

confidence: 99%

DLC: Compiling a Concurrent System Formal Specification to a Distributed Implementation

Evrard

2016

Tools and Algorithms for the Construction and Analysis of Systems

View full text Add to dashboard Cite

International audienceFormal methods can verify the correctness of a concurrentsystem by analyzing its model. However, if the actual implementation iswritten by hand, subtle and hard to detect bugs may be unintentionallyintroduced, thus ruining the verification effort. In this paper, we presentDLC (Distributed LNT Compiler), a tool that automatically generatesdistributed implementation of concurrent systems modeled in the LNTlanguage, which can be formally verified using the CADP toolbox

show abstract

System design of a CC-NUMA multiprocessor architecture using formal specification, model-checking, co-simulation, and test generation

Cited by 22 publications

References 22 publications

GRL: A Specification Language for Globally Asynchronous Locally Synchronous Systems

GRL: A Specification Language for Globally Asynchronous Locally Synchronous Systems

Formal Verification of Distributed Branching Multiway Synchronization Protocols

DLC: Compiling a Concurrent System Formal Specification to a Distributed Implementation

Contact Info

Product

Resources

About