A compiler infrastructure for embedded heterogeneous MPSoCs

• We examine why programming heterogeneous MPSoCs is a grand challenge today.
• A C-language extension design (CPN) is proposed to model streaming applications.
• The retargetable source-to-source compiler (cpn-cc) automates MPSoC compilation process.
• MAPS is successfully used as compiler infrastructure for many real-world MPSoC architectures.
• Using MAPS improves programmer productivity in MPSoC software mapping exploration.

Programming heterogeneous MPSoCs (Multi-Processor Systems on Chip) is a grand challenge for embedded SoC providers and users today. In this paper, we argue the need for and significance of positioning the language and tool design from the perspective of practicality to address this challenge. We motivate, describe and justify such a practical design of a compilation framework for heterogeneous MPSoCs targeting the domain of streaming applications, named MAPS (MPSoC Application Programming Studio). MAPS defines a clean, light-weight C language extension to capture streaming programming models. A retargetable source-to-source compiler is developed to provide key capabilities to construct practical compilation frameworks for real-world, complex MPSoC platforms. Our results have shown that MAPS is a promising compiler infrastructure that enables programming of heterogeneous MPSoCs and increases productivity of MPSoC software developers.