Subject: Re: CVS commit: src/sys
To: enami tsugutomo <enami@but-b.or.jp>
From: Daniel Carosone <dan@geek.com.au>
List: source-changes
Date: 02/11/2004 08:59:15
--7gGkHNMELEOhSGF6
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Wed, Feb 11, 2004 at 06:29:20AM +0900, enami tsugutomo wrote:
> Was that better than calculating the targets first and then do the
> both scans?

yamt pointed out yet another units confusion in this calcualtion
yesterday. We have (had) different parts of the system counting in
bytes, pages, and buffers.

The mathematical fix is clear and simple, but it will require some
additional testing and revisiting some of these other tuning questions
before being applied, because it will change the sizing behaviour
again.

What we have now seems to work fairly well, even if its not quite for
the right reasons.

--
Dan.

--7gGkHNMELEOhSGF6
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (NetBSD)

iD8DBQFAKVQzEAVxvV4N66cRAh7NAKCon+krb+ItX52YB5v4XN8KEqbd3gCeNXQk
mZVlzHtKySuocCD9Lhp54s8=
=03Il
-----END PGP SIGNATURE-----

--7gGkHNMELEOhSGF6--