Re: [ANNOUNCE/RFC] Really Simple Really Fair Scheduler

[Ingo Molnar <mingo@elte.hu>]

* Roman Zippel <zippel@linux-m68k.org> wrote:

> On Mon, 3 Sep 2007, Ingo Molnar wrote:
> 
> > If this basic model is correct, we can look further.
> 
> The basic model is correct insofar I use an absolute time instead of a 
> relative time, but it's not the essence of my math, so I don't quite 
> understand the point of this exercise.

thanks. (and i did not claim nor do i want to claim this to be the 
essence of your efforts - it is very clear from your mails where your 
focus is.)

My next question then is about this code of yours in the wakeup path:

 +static void
 +enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se)
 +{
 +       kclock_t min_time;
 +
 +       verify_queue(cfs_rq, cfs_rq->curr != se, se);
 +       min_time = get_time_avg(cfs_rq) - se->req_weight_inv;
 +       if ((kclock_t)(se->time_norm - min_time) < 0)
 +               se->time_norm = min_time;

why do you only use the "min_time" if the pre-sleep time_norm is smaller 
than the min_time? Here 'min_time' is close to the current average. 
Shouldnt here the woken up task be set to the average time, like i did 
it in the crude prototype:

+               se->exec_runtime = avg_exec_runtime(cfs_rq);

(and lets again only consider the special case of only having nice-0 
tasks.)

Or is it set in a similar way as my prototype does, and i missed some 
detail why that branch is there?

	Ingo