I also have met this problem before on v20.0.0.3，for any other test binaries，and have no idea about it. 发件人： VEDIKA JITENDRA KULKARNI via gem5-users<gem5-users@gem5.org<mailto:gem5-users@gem5.org>> 收件人： gem5-users<gem5-users@gem5.org<mailto:gem5-users@gem5.org>> 抄送： VEDIKA JITENDRA KULKARNI<ved...@iitg.ac.in<mailto:ved...@iitg.ac.in>> 主题： [gem5-users] Simulation aborts when number of instructions increases due to LSQUnit ERROR 时间： 2021-05-22 15:29:30
Hello, I want to run gem5 simulation for 40,00,000 insts. It works when benchmarks like namd and gromacs are used, however I want to use soplex, leslie3d etc (high MPKI). Any workload configuration works for 10,00,000 insts successfully. However, when I give 40,00,000 insts it shows the following error: gem5.opt: build/ALPHA/cpu/o3/lsq_unit.hh:621: Fault LSQUnit<Impl>::read(LSQUnit<Impl>::LSQRequest*, int) [with Impl = O3CPUImpl; Fault = std::shared_ptr<FaultBase>; LSQUnit<Impl>::LSQRequest = LSQ<O3CPUImpl>::LSQRequest]: Assertion `!load_inst->isExecuted()' failed. I came across a mail-archive discussing the same problem, but couldn't find a solution there. Has anyone resolved this issue? I am working with the following configuration Arch = ALPHA Number of CPUs = 16, CPU Type = DerivO3CPU System Clock = 2GHz, Topology = Mesh_XY L1-Icache = 16kB, 4-way associative L1-Dcache = 16kB, 4-way associative L2-Cache = 123kB, 8-way associative Memory Size = 8GB Memory System = Ruby, Coherence Protocol = MESI Two Level, Network = Garnet2.0 Thank you, Vedika.
_______________________________________________ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
