I added the -s option to execution line, after include se.py and I see differences into the execution time, I tried it with the helloWorld test. I also tried gcc but after 3,5h the execution has not finished, is it normal? without the -s option it takes around 1,5h but I don't have any performance measure. Rodrigo
From: rodrigo.r...@hotmail.com To: gem5-users@gem5.org Date: Fri, 25 Jan 2013 13:15:05 +0000 Subject: Re: [gem5-users] About statistics I think I am using timing. I compiled the simulator with:scons build/X86/gem5.opt an run the benchmark using these command:./build/X86/gem5.opt configs/example/se.py --caches --l1d_size=32kB --l1i_size=32kB --l2cache --l2_size=256kB --cmd=/scratch/rodri/spec2006/benchspec/CPU2006/403.gcc/run/run_base_train_rrr-static-gcc45-64bit.0000/gcc_base.rrr-static-gcc45-64bit --input=/scratch/rodri/spec2006/benchspec/CPU2006/403.gcc/run/run_base_train_rrr-static-gcc45-64bit.0000/integrate.in --option=" -o m5out/bench.s" --output=m5out/bench.out --errout=m5out/bench.err For a real measure of performance do I need to use the -s option for se.py? > From: andreas.hans...@arm.com > To: gem5-users@gem5.org > Date: Fri, 25 Jan 2013 13:01:36 +0000 > Subject: Re: [gem5-users] About statistics > > Atomic or timing? > > From: Rodrigo Reynolds Ram�rez > <rodrigo.r...@hotmail.com<mailto:rodrigo.r...@hotmail.com>> > Reply-To: gem5 users mailing list > <gem5-users@gem5.org<mailto:gem5-users@gem5.org>> > Date: Friday, 25 January 2013 12:26 > To: gem5-users <gem5-users@gem5.org<mailto:gem5-users@gem5.org>> > Subject: Re: [gem5-users] About statistics > > I am changing the latency of the simple memory in se.py, I changed the line > in this way: > > first case: > physmem = SimpleMemory(range=AddrRange("512MB"),latency="1000ns") > > second case: > physmem = SimpleMemory(range=AddrRange("512MB"),latency="30ns") > > I checked config.ini and the change is done but the results in stats.txt are > unchanged. I am using gem5-stable-f75ee4849c40. > > Rodrigo > > > Date: Fri, 25 Jan 2013 05:59:10 -0600 > > Subject: Re: [gem5-users] About statistics > > From: ni...@cs.wisc.edu<mailto:ni...@cs.wisc.edu> > > To: gem5-users@gem5.org<mailto:gem5-users@gem5.org> > > CC: rodrigo.r...@hotmail.com<mailto:rodrigo.r...@hotmail.com> > > > > On Thu, January 24, 2013 6:44 pm, Rodrigo Reynolds Ram�rez wrote: > > > > > > I think my simulation finish satisfactorily, when it finishes I get this > > > message: > > > Exiting @ tick 3822080322000 because target called exit() > > > and in both cases I get the same result in stats.txtsim_seconds > > > 3.822080 > > > and the simulations takes a lot of timehost_seconds > > > 4091.93 > > > > > > I am running gcc from spec2006 with train inputs > > > > It is hard to believe that the run time would not change on changing the > > latency of the memory. Can you mention exactly the parameter you changed > > between the two simulations? > > > > -- > > Nilay > > > > -- IMPORTANT NOTICE: The contents of this email and any attachments are > confidential and may also be privileged. If you are not the intended > recipient, please notify the sender immediately and do not disclose the > contents to any other person, use it for any purpose, or store or copy the > information in any medium. Thank you. > > _______________________________________________ > gem5-users mailing list > gem5-users@gem5.org > http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users _______________________________________________ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
_______________________________________________ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
