lkml.org 
[lkml]   [1998]   [Mar]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
Subjectsched.c change.




Subject: [uPATCH] sched.c change; snappy_response++

Well increasing the goodness will no doubt make the current process
snappier.
But it is still a little far from making an effectivly round robin
scheduler into a multi-level feedback scheduler with multiple queues.

I think the goal should be a multi-queue system that aims
at using nice as just one factor in determining priority, vs the
QNX-like
one Adam has made. Where nice alone can determine feast or famine.

Adam, you should add a type of decay to your reset of your default run
queues. Instead of just moving task back to default run queues for
blocking/starving, you should ajust the factor by the work load.
So we do not have a v3 type scheduler that tends to raise process
prioritys, lower their run_q, when the system is under extreme loads.

That way under heavy loads, it will not try to renice everything, that
starves due to the load.

Just an idea.

On Tue, 10 Mar 1998, Rik van Riel wrote:

> Hi there,
>
> I've made a (very small) change to goodness() in
> linux/kernel/sched.c to make the current scheduler
> more resemblent of the QNX scheduler, but without
> the problem of deadlocks.
>
> I've done some testing, and it seems to make Linux's
> interactive response just a little bit snappier.
>
> But it might have some side-effects in other (exotic?)
> situations, so I'd like to see it tested before sending
> it to Linus.
>
> Rik.
> +-----------------------------+------------------------------+
> | For Linux mm-patches, go to | "I'm busy managing memory.." |
> | my homepage (via LinuxHQ). | H.H.vanRiel@fys.ruu.nl |
> | ...submissions welcome... | http://www.fys.ruu.nl/~riel/ |
> +-----------------------------+------------------------------+
>
> --- linux/kernel/sched.c.orig Mon Mar 9 20:58:54 1998
> +++ linux/kernel/sched.c Mon Mar 9 20:59:22 1998
> @@ -231,6 +231,7 @@
> /* .. and a slight advantage to the current process */
> if (p == prev)
> weight += 1;
> + weight += p->priority;
> }

_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

\
 
 \ /
  Last update: 2005-03-22 13:41    [W:0.035 / U:0.616 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site