>
> On Sun, 26 Mar 2000, Kanoj Sarcar wrote:
> > > On Sun, 26 Mar 2000, Russell King wrote:
> > >
> > > > I think I've solved (very dirtily) my kswapd problem
> > >
> > > Your patch is the correct one. I've added an extra reschedule
> > > point and cleaned up the code a little bit. I wonder who sent
> > > the brown-paper-bag patch with the superfluous while loop to
> > > Linus ... (please raise your hand and/or buy rmk a beer)
> >
> > That would be me ...
> >
> > What is the problem that your patch is fixing?
>
> Removing the superfluous while loop.
>
Maybe I am being stupid, but where is the superfluous loop? Remember,
as I just pointed out, 2.3.43 had the same loop, are you claiming it
was buggy even then?
> Without my patch kswapd uses between 50 and 70% CPU time
> in a particular workload. Now it uses between 3 and 5%.
Can you explain how this is happening? I can see that in your patch,
kswapd does not go thru the loop if need_resched is set, but with
a single node, 3 zones, I would find it hard to explain such a
difference.
> Oh, and the latency problem probably has been fixed too...
What latency problem? I still believe that the pre3 code is doing
the right thing, assuming 2.3.43 was doing the right thing.
Kanoj
>
> cheers,
>
> Rik
> --
> The Internet is not a network of computers. It is a network
> of people. That is its real strength.
>
> Wanna talk about the kernel? irc.openprojects.net / #kernelnewbies
> http://www.conectiva.com/ http://www.surriel.com/
>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Mar 31 2000 - 21:00:18 EST