Re: [RFC patch 1/2] sched: dynamically adapt granularity with nr_running

[Mathieu Desnoyers <mathieu.desnoyers@efficios.com>]

* Steven Rostedt (rostedt@goodmis.org) wrote:
> On Mon, 2010-09-13 at 16:16 +0200, Peter Zijlstra wrote:
> 
> > > > -     if (delta_exec < sysctl_sched_min_granularity)
> > > > -             return;
> > > 
> > > Well, the reason why this test is here seems to be that we don't want to trigger
> > > "resched_task" more often than needed, and here it's defined by the granularity.
> > 
> > Right, but its wrong for the weighted case. Letting a light task run
> > that long will make its latency suck.
> > 
> > > I don't quite see with what you are replacing this, other than "let's set the
> > > resched flag all the time to save a 32-bit division". I figure out it's more
> > > expensive the call the scheduler than to do a 32-bit div.
> > 
> > The more divs we put it, the more expensive it all becomes.
> > 
> 
> What about:
> 
> 	if (delta_exec < sysctl_sched_min_granularity ||
> 	    delta_exec < __sched_gran(cfs_rq->nr_running))
> 
> This way we avoid the div when delta_exec is less than the
> min_granularity, and then we can do the div to perhaps avoid a needless
> resched?

Please note that my patch does not even take the 32-bit div when there are less
than 4 threads running on the system.

So I'm not at all sure we're trying to optimize an important case here.

Mathieu

> 
> -- Steve
> 
> 

-- 
Mathieu Desnoyers
Operating System Efficiency R&D Consultant
EfficiOS Inc.
http://www.efficios.com