From my experience, this error implies that the simulation has no event to schedule and rapidly proceeds to its simulation limit (maxtick) without anything to do. I have encountered this problem when the opcode latency of square root is set too high (24 cycles) on OOO cpu model. You might try reducing sqrt root latency to 12. If that doesn't work, I suggest comparing the debug output of Exec for a system using all atomic cpus and a system using all O3 (or timing depending on what you used for your simulation).
Best, -Rick Devraj Chapagain wrote: > > hi there, > > We are using 16 copies of SPEC CPU2000 benchmarks for one simulation > with 2GB physical memory in the configuration file. While simulating, > we face the following error, > > ======================================================= > warn: Entering event queue @ 0. Starting simulation... > Exiting @ cycle 9223372036854775807 because simulate() limit reached > ====================================================== > > Could you please inform me what type of error is this and what would > be the solution??? > > Thanking you, > devraj > > > ------------------------------------------------------------------------ > > _______________________________________________ > 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
