Abstract | ||
---|---|---|
In this paper we propose a new technique and a framework to select inlining heuristic constraints - referred to as an inlining vector, for program optimization. The proposed technique uses machine learning to model the correspondence between inlining vectors and performance (completion time). The automatic selection of a machine learning algorithm to build such a model is part of our technique and we present a rigorous selection procedure. Subject to a given architecture, such a model evaluates the benefit of inlining combined with other global optimizations and selects an inlining vector that, in the limits of the model, minimizes the completion time of a program. We conducted our experiments using the GNU GCC compiler and optimized 22 combinations (program, input) from SPEC CINT2006 on the state-of-the-art Intel Xeon Westmere architecture. Compared with optimization level, i.e., -O3, our technique yields performance improvements ranging from 2% to 9%. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1007/978-3-642-37051-9_9 | CC |
Keywords | Field | DocType |
optimization level,proposed technique,technique yields performance improvement,inlining vector,completion time,program optimization,inlining heuristic constraint,new technique,automatic selection,rigorous selection procedure | Program optimization,Heuristic,Programming language,Computer science,Support vector machine,Parallel computing,Compiler,Ranging,Xeon,Spec#,Approximation error | Conference |
Volume | ISSN | Citations |
7791 | 0302-9743 | 3 |
PageRank | References | Authors |
0.41 | 33 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Rosario Cammarota | 1 | 111 | 12.05 |
Alexandru Nicolau | 2 | 2265 | 307.74 |
Alexander V. Veidenbaum | 3 | 757 | 78.24 |
Arun Kejariwal | 4 | 281 | 26.23 |
Debora Donato | 5 | 1665 | 83.29 |
Mukund Madhugiri | 6 | 3 | 0.41 |