An Overview of the mCRL2 Toolset and Its Recent Advances

Abstract

Abstract. The analysis of complex distributed systems requires dedicated software tools. The mCRL2 language and toolset have been developed to support such analysis. We highlight changes and improvements made to the toolset in recent years. On the one hand, these affect the scope of application, which has been broadened with extended support for data structures like infinite sets and functions. On the other hand, considerable progress has been made regarding the performance of our tools for state space generation and model checking, due to improvements in symbolic reduction techniques and due to a shift towards parity gamebased solving. We also discuss the software architecture of the toolset, which was well suited to accommodate the above changes, and we address a number of case studies to illustrate the approach.

Combine and conquer: Relating BIP and Reo

“…Controlling and constraining multiparty synchronization is, however, more complex in ACP than it is in BIP and Reo (because additional operators, communication and block, need to be used beside parallel composition to specify admissible synchronizations). This is illustrated in work by Krause et al [30], who encoded Reo's semantics (i.e., Reo's composition operator and a number of primitives) in mCRL2 [31], a modern process specification language based on ACP.…”

Family-Based Model Checking with mCRL2

“…Our approach consists of two steps: (i) translation of the μL f -formula at hand; (ii) translation of the FTS describing the family behavior into an LTS with parametrized actions. Since μL FO is a fragment of the logic from [36,37], we can use off-the-shelf tools such as the mCRL2 toolset [21,22] to perform family-based model checking of properties expressed in μL f . We first review μL FO before we proceed to describe the above translations.…”

“…However, there have been earlier proposals for using the μ-calculus to analyze SPLs (cf., e.g., [2,10,12,19,20]). In [19], for instance, mCRL2 and its toolset [21,22] were used for product-based model checking. The flexibility of mCRL2's data language allowed to model and select valid product configurations and to model and check the behavior of individually generated products.…”

Using Domain Specific Languages to Improve the Development of a Power Control Unit

“…From the configuration DSL we generated models for the simulation tooling of POOSL [24] and the model checker mCRL2 [25]. The generators for mCRL2 and POOSL combine the behaviour of the high-level state machine and the low-level state behaviour described by the (generated) configuration file, into one state machine describing the complete system start-up and shutdown behaviour.…”