Messages in this thread | | | Date | 22 Mar 1999 15:46 PST | From | "Richard J. Farnsworth" <> | Subject | HINT benchmark and Linux |
| |
Just for laughs, I ran the HINT benchmark program (see http://www.scl.ameslab.gov/Projects/HINT/ for details) on Linux-2.2.1 and FreeBSD-3.0 using identical hardware and a 2.7.2 gcc.
To grossly simplify, this benchmark purports to measure performance a wide range of problem sizes. And sure enough, when run on FreeBSD the performance curve shows the classic ``3-step'' drop-off when the data size moves from cache to main memory to paging. The Linux graph is similar but with a much less clearly defined ``shape'' and the small-problem peak performance is about 15% less that FreeBSD. (See http://www.cs.washington.edu/lab/sw/HINT/hintgraphs.html.)
I don't plan on making a career of tracking down what is causing this behavior, but the most likely thing would seem to be ``cache trashing'' for whatever reason. So I thought I would ask the experts if any them knew off the top of their head what the likely culprits could be -- or actually _are_, for that matter.
Thanks for satisfying my curiousity...
--RJF
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |