> I'm still 99% convinced that you're not providing the right input to the > benchmarks. How many instructions get run for each benchmark when you > run it by itself? What command line do you use when you do that?
I agree it sounds like you're not specifying the correct benchmark command line. Good references can be found at: http://www.cag.lcs.mit.edu/~kbarr/cag/spec2000-commandlines.html and http://www.cag.lcs.mit.edu/~kbarr/cag/spec2000fp-commandlines.html I've written some Python classes which encapsulate each of the SPEC2000 benchmarks. They were designed for m5 v1 so might need a few changes for m5 v2. You can find them at: http://www.cl.cam.ac.uk/~jrs53/src/Benchmarks.py >From what I remember, it makes local copies of the input files from one read-only root. This is a little space inefficient but avoided having to have long paths in the command lines. James -- [EMAIL PROTECTED] http://www.cl.cam.ac.uk/~jrs53/ _______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
