Efficient implementation of DSP applications is critical for many embedded systems. Optimising C compilers for embedded processors largely focus on code generation and instruction scheduling which, with their growing maturity, are providing diminishing returns. This paper empirically evaluates another approach, namely source-level transformations and the probabilistic feedback-driven search for “good ” transformation sequences within a large optimisation space. This novel approach combines two selection methods: one based on exploring the optimisation space, the other focused on localised search of good areas. This technique was applied to the UTDSP benchmark suite on two digital signal and multimedia processors (Analog Devices TigerSHARC TS-101, Philips TriMedia TM-1100) and an embedded processor derived from a popular general-purpose processor architecture (Intel Celeron 400). On average, our approach gave a factor of 1.71 times improvement across all platforms equivalent to an average 41 % reduction in execution time, outperforming existing approaches. In certain cases a speedup of up to ≈ 7 was found for individual benchmarks.

Modern compilers have limited ability to exploit the performance improvement potential of complex transformation compositions. This is due to the ad-hoc nature of di#erent transformations. Various frameworks have been proposed to provide a unified representation of di#erent transformations, among them is Pugh's Unified Transformation Framework (UTF) (Kelly and Pugh (1993)). It presents a unified and systematic representation of iteration reordering transformations and their arbitrary combination, which results in a large and complex optimisation space for a compiler to explore. This paper presents a heuristic search algorithm capable of e#ciently locating good program optimisations within such a space. Preliminary experimental results on Java show that it can achieve an average speedup of 1.14 on Linux+Celeron and 1.10 on Windows+PentiumPro, and more than 75% of the maximum performance available can be obtained within 20 evaluations or less.