From owner-freebsd-current Wed Sep 2 01:37:27 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id BAA15277 for freebsd-current-outgoing; Wed, 2 Sep 1998 01:37:27 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.15.68.22]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id BAA15186 for ; Wed, 2 Sep 1998 01:36:07 -0700 (PDT) (envelope-from bde@godzilla.zeta.org.au) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.7/8.8.7) id SAA18187; Wed, 2 Sep 1998 18:30:53 +1000 Date: Wed, 2 Sep 1998 18:30:53 +1000 From: Bruce Evans Message-Id: <199809020830.SAA18187@godzilla.zeta.org.au> To: bde@zeta.org.au, tlambert@primenet.com Subject: Re: ELF binaries size Cc: current@FreeBSD.ORG, jdp@polstra.com, reilly@zeta.org.au Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >> >I think Bruce is mistaking the dual mapping for a single mapping, >> >on the theory that in a unified VM and buffer cache, there can be >> >only one instance of a page hung off the VP. >> >> Nah. The dual mapping is precisely what costs a full page of real memory. > >The page is wasted through padding in the image, or it is wasted >through padding in the execution class loader. Either way, it is >wasted in the in-memory image of the application. No, padding in (relatively cheap) disk space saves an average of half a page in (relatively expensive) memory. >The savings are disk-space only. Perhaps not even there, because file systems allocate in blocks, or at least fragments. Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message