Two Emerging Approaches: Model-based and Rule-driven

Abstract

This chapter describes the function of two practical compilers for embedded processors developed and used by the author and colleagues in industrial contexts. While the two operate on very different principles, they each have their strengths with regard to the goals of retargetability and ability to generate efficient code. Naturally, each approach also has a certain number of weaknesses. However, each of the two tools has made particular contributions which have added to the understanding of the compilation problem for embedded processor targets. Furthermore, each approach has shown to have its place as an effective approach to compilation for embedded processors.