I haven't see this problem in a 1.5_ALPHA kernel from about a week ago.
what date was your kernel from where you saw the problem?

-Chuck


On Wed, Jul 05, 2000 at 10:05:55PM -0400, Nathan J. Williams wrote:
> <lennart@augstsson.net> (Lennart Augustsson) writes:
> 
> > Do we have a swap leak in NetBSD?
> > I've been compiling packages on one machine for about 24 hours now
> > and the swap usage was about 80M.  I see no reason for this since the
> > machine as 128M and it wasn't really doing anything but compiling
> > packages.  (It never finished, btw.  I did a `top' and it died.)
> > 
> > I think a swap leak is a really serious bug, since it will prevent us from
> > really long up times.
> 
> I found today that my desktop x86 box had nearly used up its 512M of
> swap. Killing the X server recovered about 400M of that. I haven't yet
> figured out exactly why it was chewing up that much swap without
> seeing it accounted for somewhere....
> 
>         - Nathan