From owner-freebsd-current Wed Nov 12 13:36:25 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id NAA11042 for current-outgoing; Wed, 12 Nov 1997 13:36:25 -0800 (PST) (envelope-from owner-freebsd-current) Received: from hobbes.saturn-tech.com (drussell@drussell.internode.net [198.161.228.154]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id NAA11031 for ; Wed, 12 Nov 1997 13:36:20 -0800 (PST) (envelope-from drussell@saturn-tech.com) Received: from localhost (drussell@localhost) by hobbes.saturn-tech.com (8.8.4/8.8.2) with SMTP id OAA15928; Wed, 12 Nov 1997 14:35:35 -0700 (MST) Date: Wed, 12 Nov 1997 14:35:35 -0700 (MST) From: Doug Russell To: John-Mark Gurney cc: Burton Sampley , current@FreeBSD.ORG Subject: Re: make world time???/ In-Reply-To: <19971111152250.16665@hydrogen.nike.efn.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Tue, 11 Nov 1997, John-Mark Gurney wrote: > Burton Sampley scribbled this message on Nov 11: > > I'm really curious what people are doing to complete make world in such > > disgustingly low amount of time. I just recently upgraded from IDE to > > SCSI (IDE drives are still installed, but are presently not being used) > > and installed a 233-MMX (overclocked to 266 using 75MHz bus speed) and > > the best time I can get for make world is 3:00 hours. I started with > > source code from 9/11 with /usr/obj/ empty. I'm only using the > > plain-vanilla 'make world' from /usr/src. Any suggestion? > > well.. sounds like you need to enable -pipe on the CFLAGS (see > /etc/make.conf)... and you also need to pass something like -j4 on the > make commandline to enble parallel building... this dropped my time > bye 25%... note though, that some of those times are with some parts of > the build disabled... Even without doing parallel builds, his worldstone still seems low for such a powerful system. I routinely do make world (*** on 2.2-STABLE ***) in about 1h30min, including cleaning, etc on a K6-187.5 (166 chip at 75x2.5) with 32 megs SDRAM and an IDE disk. (Usually some recent Western Digital, right now it's a 3.1 gig). His system should be (I would think) significantly faster than my build box. The 1h30 varies a little by the exact configuration at the time (sometimes I have older HDs in there or whatever, but lately I've been burning CD-Rs with that machine, so I have a big 3.1 gig disk in it), but I always have it well under 3 hours, even when I'm building TO AND FROM NFS! I even had it down to just a hair over an hour on a K6-250 (233 chip at 83Mhz x 3) with dual WD IDE disks on seperate controllers mounting seperate /usr/src and /usr/obj, both async, noatime. It seems that at that kind of processing speed, you need better disk than the IDEs, because it should go even faster. Although, I haven't tried a fast K6 using parallel building yet. Later......