COLA: Cache Optimizations for LEON Analyses

COLA (Cache Optimizations for LEON Analyses) is a research project funded by the European Space Agency (ESA) under the basic Technology Research Programme (TRP). COLA is a follow-on project to PEAL2 (Prototype Execution-time Analyser for LEON). The purpose of COLA is to investigate how software running on a processor with cache can achieve maximum performance while remaining testable, predictable and analyzable. This work is being done with particular reference to the LEON, which is widely used in space applications. The use of cache raises difficulties when attempting to determine the worst-case application time (WCET) of software that runs on a processor equipped with a cache. The COLA project is approaching the problem from several perspectives, including:

• Making requirements and recommendations for software tools to support predictable cache usage
• Recommendations to structure application software using LEON caches
• Recommendations for future LEON cache architectures
• Specified, implemented and validated low level software for cache management for the LEON processor

Rapita Systems' contribution to COLA is in the area of tools to advise cache layout.