Re: [PATCH] HZ change for ix86
Horvath Karoly (rh@hold.sch.bme.hu)
Sun, 10 Jan 1999 01:25:50 +0100
On Sat, Jan 09, 1999 at 07:05:16PM -0500, Richard B. Johnson wrote:
> > > There seems to be a general misinformation about what the HZ value is.
> > > I will "simplicate and add lightness".
> >
> > still darkness here....
> >
> > > However, if your code is doing:
> > >
> > > UncompressFonts(...........);
> > > ReadAudioFromDsp(...........);
> > > ConvolveImageData(..........);
> > >
> > > you don't want the CPU taken away until you are done.
> >
> > We can't prevent the kernel from taking away the CPU - and a higher HZ value
> > helps us get back the CPU as soon as possible.
>
> If we don't do blocking system calls, we don't lose the CPU until our
> time-slice is up.
>
> Once you lose the CPU, you may lose it for N time slices because there
> are N computable tasks that will get it before you get it back. This is a
> multitasking operating system with many more tasks than just yours. You
> get it CPU back when it's your turn, not the next timer-tick.
>
> If N tasks are CPU bound, you sure want a long time-slice before the
> CPU gets stolen from you.
Why? Each of the N processes gets the 1/N of the CPU - wheter there are
long or short timeslices. If you have the CPU for long time, you'll miss
it for long time...
> > > doesn't produce visual effects. The higher the HZ value, the more
> > > often the CPU gets stolen from the interactive user.
> >
> > Why is that a problem? Here is an example:
> > You earn 1000$ a month. Your boss decides to pay you not monthly but weekly,
> > and you begin to complain that you earn only 250$.
>
> Not accurate.
I think so.
Anybody, who agrees with Johnson?
--
[:)]---------------------------
[:)] Horvath Karoly Zoltan / Technical University of Budapest
[:)] rh@the.rulez.org / an active member of AC Studio & Live
[:)]-------------------------
-
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/