Hello Fernando, Thanks so much for your advice.
We also found out how to do this in SE mode: In src/cpu/simple/atomic.cc line 238: "schedule(tickEvent, clockEdge(0));" If we change the value there, we can control start time of cpu. Best, -Lunkai On Mon, Aug 8, 2016 at 10:55 AM, Fernando Endo <[email protected]> wrote: > Hello, > > What I'd do is have access to perf counters inside the application to > monitor how many insts have been executed. When it goes past 100 M, trigger > the second app. Perf counters work in AArch64 , but only in Full-system. > > You could also modify the pseudo-instruction 'rpns' to return the number > of committed instructions (it returns the number of nano-seconds elapsed). > > Regards, > > -- > Fernando A. Endo, Post-doc > > INRIA Rennes-Bretagne Atlantique > France > > > 2016-08-03 19:28 GMT+02:00 Lunkai Zhang <[email protected]>: > >> Hello, >> >> I am running identical workloads in a multi-processor version of GEM5 in >> SE mode. However, I need the workloads to be in different phases. >> >> For example, if I run two libquantum benchmarks from SPEC2006, I want one >> libquantum to be 100 million instructions before the other one. >> >> Could anyone tell me the easiest way to do that? >> (Note that, right now I fast-forward the applications by 2 billions >> cycles.) >> >> Thanks so much in advance! >> >> Best, >> >> -Lunkai >> >> _______________________________________________ >> gem5-users mailing list >> [email protected] >> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users >> > > > _______________________________________________ > gem5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users >
_______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
