Section: New Results

New Programming Languages for Embedded Systems

Participants : Alain Girault [contact person] , Pascal Fradet, Vagelis Bebelis, Bertrand Jeannet, Peter Schrammel.

Analysis and scheduling of parametric dataflow models

Recent data-flow programming environments support applications whose behavior is characterized by dynamic variations in resource requirements. The high expressive power of the underlying models (e.g., Kahn Process Networks, the CAL actor language) makes it challenging to ensure predictable behavior. In particular, checking liveness (i.e., no part of the system will deadlock) and boundedness (i.e., the system can be executed in finite memory) is known to be hard or even undecidable for such models. This situation is troublesome for the design of high-quality embedded systems.

We have introduced the schedulable parametric data-flow (SPDF) MoC for dynamic streaming applications [20] . SPDF extends the standard dataflow model by allowing rates to be parametric (e.g., of the form $2xy$, where $x$ and $y$ are parameters, the value of which can change at run-time). SPDF was designed to be statically analyzable while retaining sufficient expressive power. We formulated sufficient and general static criteria for boundedness and liveness. In SPDF, parameters can be changed dynamically even within iterations. The safety of dynamic parameter changes can be checked and their implementation made explicit in the graph. These different analyses are made possible using well-defined static operations on symbolic expressions. The same holds for quasi-static scheduling which is the first step towards code generation for multi-core systems.

We are now focusing on the parallel scheduling of parametric dataflow models. We have proposed a generic and flexible framework to generate parallel ASAP schedules targeted to the new Sthorm many-core platform designed by STMicroelectronics. The parametric dataflow graph is associated with generic or user-defined specific constraints aimed at minimizing, timing, buffer sizes, power consumption, or other criteria. The scheduling algorithm executes with minimal overhead and can be adapted to different scheduling policies just by changing some constraints. The safety of both the dataflow graph and constraints can be checked statically and all schedules are guaranteed to be bounded and deadlock free. Our case studies are video decoders for high definition video streaming such as VC-1 or HEVC.

This research is the central topic of Vagelis Bebelis' PhD thesis. It is conducted in collaboration with STMicroelectronics.