Is there any evidence that whatever you're doing to select a different simpoint is having any effect? For example, does the simulator output differ at all? My guess is that the option you're using doesn't do what you think it does.
On Tue, May 12, 2009 at 11:00 PM, Pavan Kumar <[email protected]> wrote: > hi, > > I am running spec cpu 2006(alpha) benchmarks on M5-2.0b6 for different > simpoints. Means, every time i increasing the simpoint by 100 Million. > for example; > *for 1 simulation* > i put the simpoint =6*100E6 > i got the follwoing results. > >> system.cpu.dcache.ReadReq_accesses 22409799 >> system.cpu.dcache.ReadReq_misses 970709 >> system.cpu.dcache.WriteReq_accesses 9879012 >> system.cpu.dcache.WriteReq_misses 197596 >> system.cpu.dcache.overall_accesses 32288811 >> system.cpu.dcache.overall_misses 1168305 >> system.cpu.icache.ReadReq_accesses 100090404 >> system.cpu.icache.ReadReq_misses 1305749 >> system.cpu.icache.overall_accesses 100090404 >> system.cpu.icache.overall_misses 1305749 >> system.l2.ReadReq_accesses 2276458 >> system.l2.ReadReq_misses 115852 >> system.l2.overall_accesses 2397139 >> system.l2.overall_misses 236533 >> > > *for 2nd simulaltion* > i put the simpoint =7*100E6 > i got the same as above > >> system.cpu.dcache.ReadReq_accesses 22409799 >> system.cpu.dcache.ReadReq_misses 970709 >> system.cpu.dcache.WriteReq_accesses 9879012 >> system.cpu.dcache.WriteReq_misses 197596 >> system.cpu.dcache.overall_accesses 32288811 >> system.cpu.dcache.overall_misses 1168305 >> system.cpu.icache.ReadReq_accesses 100090404 >> system.cpu.icache.ReadReq_misses 1305749 >> system.cpu.icache.overall_accesses 100090404 >> system.cpu.icache.overall_misses 1305749 >> system.l2.ReadReq_accesses 2276458 >> system.l2.ReadReq_misses 115852 >> system.l2.overall_accesses 2397139 >> system.l2.overall_misses 236533 >> > > I ran it for 8 more times by increasing simpoint value by 100 Million, > still all the results are same. > > I couldn't understand what could be the reason for getting the same values > for different simpoints. > > > thanks, > pavan > > _______________________________________________ > m5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users >
_______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
