Finding representative sets of optimizations for adaptive multiversioning applications

Iterative compilation is a widely adopted technique to opti- mize programs for difierent constraints such as performance, code size and power consumption in rapidly evolving hardware and software envi- ronments. However, in case of statically compiled programs, it is often re- stricted to optimizations for a speciflc dataset and may not be applicable to applications that exhibit difierent run-time behavior across program phases, multiple datasets or when executed in heterogeneous, reconflg- urable and virtual environments. Several frameworks have been recently introduced to tackle these problems and enable run-time optimization and adaptation for statically compiled programs based on static func- tion multiversioning and monitoring of online program behavior. In this article, we present a novel technique to select a minimal set of representa- tive optimization variants (function versions) for such frameworks while avoiding performance loss across available datasets and code-size explo- sion. We developed a novel mapping mechanism using popular decision tree or rule induction based machine learning techniques to rapidly select best code versions at run-time based on dataset features and minimize selection overhead. These techniques enable creation of self-tuning static binaries or libraries adaptable to changing behavior and environments at run-time using staged compilation that do not require complex recom- pilation frameworks while efiectively outperforming traditional single- version non-adaptable code.