Re: Report: Big Improvement in -test3

From: Linus Torvalds (torvalds@transmeta.com)
Date: Thu Jul 13 2000 - 00:09:01 EST


In article <396D156D.80D097B7@toyota.com>, Joel Sloan <jjs@toyota.com> wrote:
>
>When I returned to my desk a couple of minutes later, I noticed
>the "gears" 3D screensaver running at pretty much normal speed.
>
>My first thought was "The compile must have died", since in my
>previous experience, the screensaver would always slow to a crawl
>during a kernel compile. I logged in and found to my suprise that
>the kernel compile was still chugging away, the load average was
>meandering around between 1.5 and 2, and it was 16M into swap.

Heh.

This was actually almost certainly due to a _really_ simple improvement.

As of test4-pre4, the default time-slice for a normal process is just
50ms, while it used to be 200ms.

200ms is way too long a timeslice when working with interactive things:
it's easily noticeable. 50ms should be much better.

I'd be interested to hear if the "Quake-III with direct rendering"
choppy mouse problem on UP is also better due to this. (It was always
fine on a SMP kernel, because the second CPU really helps interactive
feel, but I'm told it was not nearly as smooth on UP..)

                Linus

-
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 : Sat Jul 15 2000 - 21:00:16 EST