From owner-freebsd-stable@FreeBSD.ORG Mon Oct 10 02:07:31 2005 Return-Path: X-Original-To: stable@freebsd.org Delivered-To: freebsd-stable@FreeBSD.ORG Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 465AF16A41F for ; Mon, 10 Oct 2005 02:07:31 +0000 (GMT) (envelope-from david@catwhisker.org) Received: from bunrab.catwhisker.org (adsl-63-193-123-122.dsl.snfc21.pacbell.net [63.193.123.122]) by mx1.FreeBSD.org (Postfix) with ESMTP id C6E4043D45 for ; Mon, 10 Oct 2005 02:07:30 +0000 (GMT) (envelope-from david@catwhisker.org) Received: from bunrab.catwhisker.org (localhost [127.0.0.1]) by bunrab.catwhisker.org (8.13.3/8.13.3) with ESMTP id j9A27UYm056430 for ; Sun, 9 Oct 2005 19:07:30 -0700 (PDT) (envelope-from david@bunrab.catwhisker.org) Received: (from david@localhost) by bunrab.catwhisker.org (8.13.3/8.13.1/Submit) id j9A27U4a056429 for stable@freebsd.org; Sun, 9 Oct 2005 19:07:30 -0700 (PDT) (envelope-from david) Date: Sun, 9 Oct 2005 19:07:29 -0700 From: David Wolfskill To: stable@freebsd.org Message-ID: <20051010020729.GA56351@bunrab.catwhisker.org> Mail-Followup-To: David Wolfskill , stable@freebsd.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.4.2.1i Cc: Subject: 5.x: how do I get a *swap*-backed /tmp via rc.conf? X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: stable@freebsd.org List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 10 Oct 2005 02:07:31 -0000 I made the somewhat unexpected discovery that in FreeeBSD 5-STABLE, if I use the "tmp*" variables in /etc/rc.conf to have an MFS /tmp created, it is apparentyly not swap-backed -- as I expected from the part of the mdmfs man page that reads: By default, mdmfs creates a swap-based (MD_SWAP) disk with soft-updates enabled and mounts it on mount-point. and a review of the way mdmfs is invoked by /etc/rc; rather, it appears to be malloc-backed. Here's what's in /etc/{defaults/,}rc.conf about it on one such machine: g1-59(5.4-S)[42] grep tmp /etc/{default*/,}rc.conf /etc/defaults/rc.conf:tmpmfs="AUTO" # Set to YES to always create an mfs /tmp, NO to never /etc/defaults/rc.conf:tmpsize="20m" # Size of mfs /tmp if created /etc/defaults/rc.conf:tmpmfs_flags="-S" # Extra mdmfs options for the mfs /tmp /etc/defaults/rc.conf:isdn_traceflags="-f /var/tmp/isdntrace0" # Flags for isdntrace /etc/defaults/rc.conf:clear_tmp_enable="NO" # Clear /tmp at startup. /etc/rc.conf:tmpmfs="YES" /etc/rc.conf:tmpsize="512m" /etc/rc.conf:tmpmfs_flags="-i4096" g1-59(5.4-S)[43] And: g1-59(5.4-S)[43] sudo mdconfig -l -u md0 md0 malloc 524288 KBytes g1-59(5.4-S)[44] Oh, for some sense of what we're working with: g1-59(5.4-S)[44] uname -a FreeBSD g1-59.catwhisker.org. 5.4-STABLE FreeBSD 5.4-STABLE #6: Sun Oct 9 06:24:17 PDT 2005 root@g1-59.catwhisker.org.:/usr/obj/usr/src/sys/LAPTOP i386 g1-59(5.4-S)[45] So, back to my original question: how do I get a *swap*-backed /tmp? [The way I was alerted to the possibliity that my /tmp might not be swap-backed is that I was using a machine running 5-STABLE as a CVS server, in order to update /usr/ports on another machine. That's why I have the -i4096 argument in there, BTW: to double the number of inodes for when the CVS server starts consuming them with wild abandon as it builds an isomorphic hierarchy to /usr/ports in /tmp. I think it only took me 3 panics before I poked around in the mailing list archives and noted scottl's comment about using a swap-backed /tmp instead, which got me wondering what backing store my /tmp was using. I did look at /etc/rc.d/tmp, as well a /etc/rc.subr, but I'm still failing to see why I'm getting a malloc-backed /tmp. FWIW, although the /tmp in question is an MFS in 4.x, I do this (use the box as a CVS server for /usr/ports) without problem on 4.x.] (I don't need separate copies of any replies; I read -stable, so I set Reply-To as an appropriate hint.) Thanks. Peace, david -- David H. Wolfskill david@catwhisker.org Prediction is difficult, especially if it involves the future. -- Niels Bohr See http://www.catwhisker.org/~david/publickey.gpg for public key.