Section: New Results

Dataflow static analyses and optimisations

Participants : Laure Gonnord, Lionel Morel, Szabolcs-Marton Bagoly, Romain Fontaine.

Nowadays, parallel computers have become ubiquitous and current processors contain several execution cores, anywhere from a couple to hundreds. This multi-core tendency is due to constraints preventing the increase of clock frequencies, such as heat generation and power consumption. A variety of low-level tools exist to program these chips efficiently, but they are considered hard to program, to maintain, and to debug, because they may exhibit non-deterministic behaviors. We explore the potentiality of the data flow programming, which allows the programmer to specify only the operations to perform and their dependencies, without actually scheduling them. The work is published in two research reports: [48] and [49].

In [48], we explore the combination of a dataflow paradigm language, SigmaC, with the Polyhedral Model, which allows automatic parallelization and optimization of loop nests, in order to make the programming easier by delegating work to the compilers and static analyzers, in various case studies.

In [49], we explore the expressivity of the horn clause format for static analyses of Lustre programs with arrays. We propose a translation from a Lustre core language to horn clauses, with or without array variables.