lkml.org 
[lkml]   [2003]   [Aug]   [18]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: scheduler interactivity: timeslice calculation seem wrong


Eric St-Laurent wrote:

>currently, nicer tasks (nice value toward -20) get larger timeslices,
>and less nice tasks (nice value toward 19) get small timeslices.
>

Funny, isn't it!

>
>this is contrary to all process scheduling theory i've read, and also
>contrary to my intuition.
>

Yep.

>
>maybe it was done this way for fairness reasons, but that's another
>story...
>
>high priority (interactive) tasks should get small timeslices for best
>interactive feeling, and low priority (cpu hog) tasks should get large
>timeslices for best efficiency, anyway they can be preempted by higher
>priority tasks if needed.
>

Its done this way because this is really how the priorities are
enforced. With some complicated exceptions, every task will be
allowed to complete 1 timeslice before any task completes 2
(assuming they don't block).

So higher priority tasks need bigger timeslices.

>
>also, i think dynamic priority should be used for timeslice calculation
>instead of static priority. the reason is, if a low priority task get a
>priority boost (to prevent starvation, for example) it should use the
>small timeslice corresponding to it's new priority level, instead of
>using it's original large timeslice that can ruin the interactive feel.
>

Among other things, yes, I think this is a good idea too. I'll be
addressing both these issues in my scheduler fork.

I do have dynamic timeslices, but currently high priority tasks
still get big timeslices.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:47    [W:0.299 / U:0.220 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site