# Einen Prozess starten, der 1000 Prozesse startet, einen nach dem anderen
[arrowlake:~:4557] perf stat bash -c "for i in {0,1,2,3,4,5,6,7,8,9}{0,1,2,3,4,5,6,7,8,9}{0,1,2,3,4,5,6,7,8,9}; do /bin/true; done"
Performance counter stats for 'bash -c for i in {0,1,2,3,4,5,6,7,8,9}{0,1,2,3,4,5,6,7,8,9}{0,1,2,3,4,5,6,7,8,9}; do /bin/true; done':
287.69 msec task-clock # 0.949 CPUs utilized
1_025 context-switches # 3.563 K/sec
7 cpu-migrations # 24.331 /sec
105_610 page-faults # 367.093 K/sec
1_249_529_757 cpu_atom/cycles/ # 4.343 GHz
<not counted> cpu_core/cycles/ (0.00%)
0.303234645 seconds time elapsed
0.240837000 seconds user
0.078414000 seconds sys
# Und jetzt zwinge ich das System dazu, das auf einem P-Kern zu machen:
[arrowlake:~:4558] taskset -c 1 perf stat bash -c "for i in {0,1,2,3,4,5,6,7,8,9}{0,1,2,3,4,5,6,7,8,9}{0,1,2,3,4,5,6,7,8,9}; do /bin/true; done"
Performance counter stats for 'bash -c for i in {0,1,2,3,4,5,6,7,8,9}{0,1,2,3,4,5,6,7,8,9}{0,1,2,3,4,5,6,7,8,9}; do /bin/true; done':
230.48 msec task-clock # 0.922 CPUs utilized
1_002 context-switches # 4.347 K/sec
0 cpu-migrations # 0.000 /sec
104_564 page-faults # 453.683 K/sec
<not counted> cpu_atom/cycles/ (0.00%)
1_109_789_653 cpu_core/cycles/ # 4.815 GHz
0.250021005 seconds time elapsed
0.195993000 seconds user
0.053664000 seconds sys