Hello everyone, I'm recently running gem5 FS simulation with alpha processors using parsec-alpha v2.1 benchmarks. The running command is following:
./build/ALPHA/gem5.opt \ configs/custom/fs_elastic.py \ --kernel=$M5_PATH/binaries/vmlinux_2.6.27-gcc_4.3.4 \ --disk-image=$M5_PATH/disks/linux-parsec-2-1-m5-with-test-inputs.img \ --script=$M5_PATH/scripts/canneal/canneal_4C/canneal_4c_simdev_ckpts.rcS \ --caches --l1d_size=256B --l1d_assoc=2 \ --l2cache --l2_size=1kB --l2_assoc=4 \ --cacheline_size=64 \ --cpu-clock=1GHz \ --sys-clock=1GHz \ --mem-size=1GB \ --cpu-type=atomic \ -n 4 \ So the benchmark is canneal, number of CPUs is 4. After normally finishing simulation I look into the simulation statistics file stats.txt. Interestingly there are 4 statistics sheets one after another. Each of those sheets starts with ---------- Begin Simulation Statistics ---------- and ends with ---------- End Simulation Statistics ---------- I'm very curious about this partitioning of simulation into 4 parts. The question is: how to understand these 4 parts? which execution does each part correspond to? My guess: part 1: boot up linux; part 2: initialization of benchmark; part 3: Region of interest of benchmark (ROI) with feature of data parallelism; part 4: Shutdown phase of benchmark. Does anybody have official docs to explain the meaning of each part of this simulation stats? Thanks a lot in advance! Best regards, Mengyu Liang
_______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
