A Stack Cache for Real-Time Systems

Real-time systems need time-predictable computing platforms to allowfor static analysis of the worst-case execution time. Caches are important for good performance, but data caches arehard to analyze for the worst-case execution time. Stack allocated data has different properties related to locality, lifetime, and static analyzability of access addresses comparedto static or heap allocated data. Therefore, caching of stack allocateddata benefits from having its own cache.

In this paper we present a cache architecture optimized for stack allocateddata. This cache is additional to the normal data cache. As stack allocated datahas a high locality, even a small stack cache gives a high hit rate. A stack cache added to a write-through data cache considerablyimproves the performance, while a stack cache compared tothe harder to analyze write-back cache has about the sameaverage case performance.