Performance Driven Optimization Tuning in Vista
Prasad A. Kulkarni · 2003
This thesis describes the support provided in VISTA for finding effective sequences of optimization phases. VISTA is a software tool for a compiler which supports an interactive compilation paradigm. It has long been known that a single ordering of optimization phases will not produce the best code for every application. This phase ordering problem can be more severe when generating code for embedded systems due to the need to meet conflicting constraints on time, code size and power consumption. Given that many embedded application developers are willing to spend time tuning an application, we believe a viable approach is to allow the developer to steer the process of optimizing a function. With this in mind, we have enhanced VISTA with many new features and programming-language-like constructs. VISTA also provides the user with dynamic and static performance information that can be used during an interactive compilation session to gauge the progress of improving the code. In addition, VISTA provides support for automatically using performance information to select the best optimization sequence among several attempted. One such feature is the use of a genetic algorithm to search for the most efficient sequence based on specific fitness criteria. This thesis also includes a number of experimental results that evaluate the effectiveness of using a genetic algorithm in VISTA to find effective optimization phase sequences.%%%%A Thesis submitted to the Department of Computer Science in partial fulfillment of the requirements for the degree of Master of Science.%%%%Summer Semester, 2003.%%%%May 30, 2003.%%%%David Whalley, Professor Directing Dissertation; Xin Yuan, Committee Member; Kyle Gallivan, Committee Member.