On Fri, Jul 24, 2009 at 10:23 PM, Gabe Black<gbl...@eecs.umich.edu> wrote: > > We could also add a command line argument to the script. One thing to > note is that I think se.py is supposed to just be an example. I have the > impression it's often used exactly as is though, so that may not be a > valid way to look at it.
You're right... it's supposed to be an example but most people seem to use it as is (or with minimal hacking). I think what Korey's done is fine: the CPU model itself makes sure the workload length and the thread count are consistent, but the script makes life easier for people in the common case where you just want the thread count to scale with the workload. It did seem to break the regressions though, which is not good... Steve _______________________________________________ m5-dev mailing list m5-dev@m5sim.org http://m5sim.org/mailman/listinfo/m5-dev