Anatomizing Linux scheduler: CFS 3

superfluous

As mentioned above
       delta_time = now - prev_stored_now
       se.wait_runtime = delta_time - delta_exec_time
According to these definitions, CFS scheduler can work smoothly. But there are some other factors to affect the fairness. We should take these factors into account. Then CFS scheduler
    1. tunes se.fair_key:
       se.fair_key = cfs_rq.fair_clock - se.wait_runtime, if se.load_weight == NICE_0_LOAD;
       se.fair_key = cfs_rq.fair_clock - se.wait_runtime * NICE_0_LOAD / se.load_weight,
                         if se.load_weight != NICE_0_LOAD && se.wait_runtime = 0
    2. tunes se.wait_runtime
       a. a task is in run queue, but not run on the cpu
          se.wait_runtime += (now - prev_stored_now) * se.load_weight / NICE_0_LOAD
       b. a task gets added back to the runnable tasks and gets a small credit for the cpu time it missed out on.
          se.wait_runtime += delta * cfs_rq.nr_running
       c. a task wakes-up from sleeping or blockint state
          se.wait_runtime += delta_sleep_fair * se.load_weight / NICE_0_LOAD
               
               
               

本文来自ChinaUnix博客，如果查看原文请点：http://blog.chinaunix.net/u3/103132/showart_2040027.html