Re: [PATCH] HZ change for ix86

Peter T. Breuer (ptb@it.uc3m.es)
Mon, 11 Jan 1999 03:40:56 +0100 (MET)


"A month of sundays ago Richard B. Johnson wrote:"
> 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.

Nope. Still the wrong argument. You work with N people and you all get
paid 1K per month. One person gets paid every 31/N days by an extremely
slowly moving aged accountant who pays him the amount due since his last
visit. You all get paid once per month. Now you give the accountant
some caffein and he doubles his speed. He takes half as long to move
from desk to desk. His Hz has doubled. You get paid twice every month
(he's not crazy however, so he only pays you half as much as before for
your work since the last visit since he visited you last half as long
ago as he used to).

> If N tasks are CPU bound, you sure want a long time-slice before the
> CPU gets stolen from you.

Why? Only "overhead" springs to mind as a reason.

> > 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 only earn 250$.
>
> Not accurate.

On the contrary, an excellent and very logical argument. Would you prefer
that he put it in terms of x and y :-) ? You do yourself a disservice by
dismissing it instead of trying to point out what you think is wrong.

> Cheers,
> Dick Johnson

Peter

-
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/