调度的ｌｏｇ 1.5ms 12ms 4ms

36
  37
  38            loopM 24369 [001] 60789.192708: sched:sched_switch: prev_comm=loopM prev_pid=24369 prev_prio=120 prev_state=R ==> next_comm=loo     p next_pid=24361 next_prio=120
  39             loop 24361 [001] 60789.196705: sched:sched_stat_runtime: comm=loop pid=24361 runtime=3998313 [ns] vruntime=20031728161993 [ns]
  40             loop 24361 [001] 60789.200705: sched:sched_stat_runtime: comm=loop pid=24361 runtime=3999604 [ns] vruntime=20031732161597 [ns]
  41             loop 24361 [001] 60789.200706: sched:sched_stat_runtime: comm=loop pid=24361 runtime=1047 [ns] vruntime=20031732162644 [ns]
  42             loop 24361 [001] 60789.200707: sched:sched_stat_wait: comm=loopM pid=24369 delay=7998964 [ns]
  43             loop 24361 [001] 60789.200707: sched:sched_switch: prev_comm=loop prev_pid=24361 prev_prio=120 prev_state=R ==> next_comm=loop     M next_pid=24369 next_prio=120
  t
为啥是每4ms进行一次 1.5ms, small_granularity=1.5ms, sched_latency_ns=12ms

为啥

时钟中断从/proc/接口读出来不固定，但是从内核do_timer处打点后发现是固定，应该是以内核为标准的