This page describes the situation with Splash benchmarks, but it basically applies to running parallel codes in general:
http://www.m5sim.org/wiki/index.php/Splash_benchmarks If you want to run codes that you've developed yourself on MIPS, you'll have to do some work to implement fork() or something similar in SE mode. In general it's easiest to just use pthreads and run in FS mode, but unfortunately we don't have MIPS_FS up yet. Steve On Dec 21, 2007 2:58 PM, Panayiotis Petridis <[EMAIL PROTECTED]> wrote: > > Hi all, > > I am trying to run some simple benchmarks on M5 using CMP architecture. My > question is how can i run my benchmark parallel i.e. if i have 2 processors > how do i schedule the instructions for each one? > > Do i have to configure my system, for example on se.py file , and define > the workload to each processor or do i have to make the scheduling within > the benchmark, for example using pthreads? > > I want to simulate MIPS architecture on SE mode > > Thanks > Panayiotis > > ________________________________ > Express yourself instantly with MSN Messenger! MSN Messenger > > _______________________________________________ > m5-users mailing list > m5-users@m5sim.org > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > _______________________________________________ m5-users mailing list m5-users@m5sim.org http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
