In this paper we summarize our experiences in designing and running CSiBE, the new code size benchmark for GCC. Since its introduction in 2003, it has been widely used by GCC developers in their daily work to help them keep the size of the generated code as small as possible. We have been making continuous observations on the latest results and informing GCC developers of any problem when necessary. We overview some concrete “success stories” of where GCC benefited from the benchmark. This paper overviews the measurement methodology, providing some information about the test bed, the measuring method, and the hardware/software infrastructure. The new version of CSiBE, launched in May 2004, has been extended with new features such as code performance measurements and a test bed - four times larger - with even more versatile programs.
BibTeX:
@InProceedings{BFG04,
author = {Besz\'edes, {\'A}rp\'ad and Ferenc, Rudolf and Gergely, Tam\'as and Gyim\'othy, Tibor and L\'oki, G\'abor and Vid\'acs, L\'aszl\'o},
title = {{CSiBE} Benchmark: One Year Perspective and Plans},
booktitle = {Proceedings of the 2004 GCC Developers' Summit (GCC 2004)},
year = {2004},
pages = {7--15},
address = {Ottawa, Canada},
month = jun,
}