Linux虛擬運(yùn)行時(shí)間的計(jì)算

虛擬運(yùn)行時(shí)間的計(jì)算

關(guān)于tick:

tick是周期性的時(shí)鐘中斷，時(shí)鐘中斷驅(qū)動(dòng)調(diào)度器runing，其周期間隔根據(jù)硬件頻率的設(shè)定相關(guān)(T=1/f)，如下：

T=1/250 = 4ms，也就說時(shí)鐘中斷以4ms為周期【tips:每發(fā)生一次時(shí)鐘中斷，jiffies的數(shù)值就加上1】,每一個(gè)進(jìn)程的虛擬時(shí)間在時(shí)鐘中斷里面被維護(hù)，每次時(shí)鐘中斷都要更新當(dāng)前進(jìn)程的虛擬時(shí)間。更新調(diào)用的主要函數(shù)如下：

static void update_curr(struct cfs_rq *cfs_rq)
{
 struct sched_entity *curr = cfs_rq- >curr;
 u64 now = rq_clock_task(rq_of(cfs_rq));
 u64 delta_exec;

 if (unlikely(!curr))
  return;

 delta_exec = now - curr- >exec_start;
 if (unlikely((s64)delta_exec <= 0))
  return;

 curr- >exec_start = now;

 schedstat_set(curr- >statistics.exec_max,
        max(delta_exec, curr- >statistics.exec_max));

 curr- >sum_exec_runtime += delta_exec;
 schedstat_add(cfs_rq- >exec_clock, delta_exec);

 curr- >vruntime += calc_delta_fair(delta_exec, curr);
 update_min_vruntime(cfs_rq);

 if (entity_is_task(curr)) {
  struct task_struct *curtask = task_of(curr);

  trace_sched_stat_runtime(curtask, delta_exec, curr- >vruntime);
  cgroup_account_cputime(curtask, delta_exec);
  account_group_exec_runtime(curtask, delta_exec);
 }

 account_cfs_rq_runtime(cfs_rq, delta_exec);
}