lkml.org 
[lkml]   [2001]   [Nov]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [patch] scheduler cache affinity improvement for 2.4 kernels
On Thu, 8 Nov 2001, Ingo Molnar wrote:

>
> Davide,
>
> On Thu, 8 Nov 2001, Davide Libenzi wrote:
>
> > Maybe you missed this :
> >
> > http://www.xmailserver.org/linux-patches/mss.html
> >
> > where the patch that does this is here :
> >
> > http://www.xmailserver.org/linux-patches/lnxsched.html#CPUHist
>
> i'm not sure what the patch is trying to achieve, but this part of
> mcsched-2.4.13-0.4.diff looks incorrect:
>
> + prev->cpu_jtime += (jiffies - prev->sched_jtime) + jiffies;
>
> (this is "2*jiffies - prev->sched_jtime" which doesnt appear to make much
> sense - does it?)

The optimization is not good ( i left it in that way to make it more clear
what that operation is meant ) but the mean of the code is ok.
It sets the time ( in jiffies ) at which the process won't have any more
scheduling advantage.


> and your patch adds a scheduling advantage to processes with more cache
> footprint, which is the completely opposite of what we want.

It is exactly what we want indeed :
<quote>
it's a fix for a UP and SMP scheduler problem Alan described to me
recently, the 'CPU intensive process scheduling' problem. The essence of
the problem: if there are multiple, CPU-intensive processes running,
intermixed with other scheduling activities such as interactive work or
network-intensive applications, then the Linux scheduler does a poor job
of affinizing processes to processor caches. Such scheduler workload is
common for a large percentage of important application workloads: database
server workloads, webserver workloads and math-intensive clustered jobs,
and other applications.
</quote>

and if you take a look at the LatSched sampling it is achived very well.


> but in any case, changing the goodness() function was not a goal of my
> patch, i change the granularity of how processes lose their 'effective
> priority'.

I'll test the patch asap with the LatSched sampler and i'll let you know.




- Davide



-
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 13:18    [W:0.072 / U:0.008 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site