SYRF Project

Task 4.1: "Code distribution"

Abstract of deliverable 4.1

Back to the SYRF Home Page



Approach:

Modularity is advocated as the ultimate solution for the design of large systems, and this holds in particular for embedded systems, and their software and architecture. Modularity allows the designer to scale down design problems, and facilitate reuse of predefined modules. Task 4.1 of the project is devoted to address this issue for code generation. The mathematical translation of the concept of modularity is often that of compositionality. Paying attention to the composition of specifications is central to any system model involving concurrency or parallelism. More recently, significant effort has been devoted toward introducing compositionality in verification with aiming at deriving proofs of large programs from partial proofs involving (abstractions of) components. Compilation and code generation has been given less attention from the same point of view, however. This is unfortunate, as it is critical for the designer to scale down the design of large systems by

This paper is devoted the issues of compositionality aiming at modular code generation, for dataflow synchronous languages. As dataflow synchronous is rather a paradigm more than a few concrete languages or visual formalisms, it was desirable to abstract from such and such particular language. Thus we have chosen to work with a formalism proposed by Amir Pnueli, that of Symbolic Transition Systems (STS), which is at the same time very lightweight, and fully general to capture the essence of synchronous paradigm. Achieved Results: Publications:
  1. A. Benveniste, P. Le Guernic, and P. Aubry
    Compositionality in dataflow synchronous languages: specification and code generation
    Malente Workshop on Compositionality, W.P. de Roever, A. Pnueli Eds., September 1997