Hello, I am new to running multicore simulations in gem5.
Currently, I am running multiple single-threaded applications in a multicore environment in Full System simulations. I want to specify these in a rcS file so that I can preload my dynamically linked library before running the applications. So I have my rcS script as follows: /sbin/m5 dumpresetstats cd /path/to/benchmark/ LD_PRELOAD=my_library.so ./mcf_base.ia64-gcc42 inp.in LD_PRELOAD=my_library.so ./libquantum_base.ia64-gcc42 143 25 /sbin/m5 exit I am now running my gem5 simulation as follows: build/X86/gem5.fast -d output_folder configs/example/fs.py --cpu-type=detailed --cache --l2cache --mem-size=2GB --mem-channels=4 --disk-image=disks/disk-final.img --script=rcS/spec1.rcS --num-cpus=2 -F 1000000000 --maxinsts=100000000 --I see in my system.pc.com_1.terminal that only the first application has finished execution (only mcf and libquantum). I wanted to FF both mcf and libquantum to 1 billion and execute for 100 million instructions. What am I missing in my script? Is there a way in which I can run this simulation by fast-forwarding both the cpu by a different value? Say, in the above example I need to FF mcf application by 2billion and libquantum by 1 billion. Is it possible? Thanks a lot, Aditya Narayan
_______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
