Subject: Re: more on performance of "make build" in -current vs. 1.5
To: Chuck Silvers <>
From: Frank van der Linden <>
List: tech-kern
Date: 02/19/2001 09:59:35
On Sun, Feb 18, 2001 at 03:09:58PM -0800, Chuck Silvers wrote:
> 1.5 32MB
> 3126.891u 596.892s 2:00:53.25 51.3%     0+0k 151929+207357io 25155pf+0w
> -current+balancing 32MB
> 3253.558u 723.490s 1:39:03.70 66.9%     0+0k 10687+30709io 14532pf+0w
> 1.5 128MB
> 3092.888u 560.706s 1:36:37.29 63.0%     0+0k 51347+197915io 11051pf+0w
> -current+balancing 128MB
> 3252.077u 746.587s 1:34:11.46 70.7%     0+0k 7003+30014io 4295pf+0w

First, I'm running the balance diffs, and they seem to do the job
of keeping the system fast for both I/O and interactive work. I did
some comparisons by running bonnie with file sizes that would fit
into memory, whilst editing several files and switching virtual
desktops, and compared them to a system without the change.

Anyway.. it's a bit strange that having 128M instead of 32 doesn't
buy you that much more speed, isn't it? Of course, it could just be
that the working set of files that can be usefully cached for
a make build is such that the optimimum is reached when main
memory is not too far above 32M.

- Frank

Frank van der Linden                 
Quality NetBSD CDs, Support & Service.