From owner-freebsd-stable Sun Sep 16 17: 9:41 2001 Delivered-To: freebsd-stable@freebsd.org Received: from smtp.noos.fr (camus.noos.net [212.198.2.70]) by hub.freebsd.org (Postfix) with ESMTP id D342D37B408 for ; Sun, 16 Sep 2001 17:09:36 -0700 (PDT) Received: (qmail 62769961 invoked by uid 0); 17 Sep 2001 00:09:35 -0000 Received: from unknown (HELO gits.dyndns.org) ([212.198.231.187]) (envelope-sender ) by 212.198.2.70 (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 17 Sep 2001 00:09:35 -0000 Received: (from root@localhost) by gits.dyndns.org (8.11.6/8.11.6) id f8H09Y921584; Mon, 17 Sep 2001 02:09:34 +0200 (CEST) (envelope-from root) Message-Id: <200109170009.f8H09Y921584@gits.dyndns.org> Subject: Re: make -j4 vs -j8... 4 works, but 8 does not In-Reply-To: <20010916003359.A48953@student.uu.se> To: Erik Trulsson Date: Mon, 17 Sep 2001 02:09:34 +0200 (CEST) Cc: stable@FreeBSD.ORG Reply-To: clefevre@citeweb.net From: Cyrille Lefevre Organization: ACME X-Face: V|+c;4!|B?E%BE^{E6);aI.[<97Zd*>^#%Y5Cxv;%Y[PT-LW3;A:fRrJ8+^k"e7@+30g0YD0*^^3jgyShN7o?a]C la*Zv'5NA,=963bM%J^o]C X-Mailer: ELM [version 2.4ME+ PL94b (25)] MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Erik Trulsson wrote: > On Sat, Sep 15, 2001 at 06:06:10PM -0400, Garance A Drosihn wrote: > > > [aside on those multiple builds that I did: It was interesting that > > even though it was on a dual-processor system, there was not much of > > a speed improvement (on 4.3-stable) when going from -j4 to -j10. Big > > improvement going from -j1 to -j4, but after that it didn't help much] > > That sounds exactly like what I would expect. More than 2 (or maybe 3) > jobs per CPU is normally not going to make things go faster but might > actually slow it down. I'm not right w/ you. let's try : # setup a reusable environment cd /sys/compile/MYKERNEL make depend make # run the jobs for i in 1 2 4 8; do (date;touch opt_global.h;time make -j$i all;date) >| .j$i 2>&1 done results are (on my old P200 processor) : j1 12:36 j2 12:46 j4 11:36 j8 11:21 so, for a 12 min job, you could win 1 min using -j4, which may gives you 1 hour for a 12 hours job (such as buildworld :). ok, it's not so many, but that may help a lot. for instace, I currently use -j4 since -j8 is too CPU and memory intensive (the machine slow down significally). PS : I'm using SCSI disks. using IDE disks probably gives you different results. I/O are very differents between the two. Cyrille. -- Cyrille Lefevre mailto:clefevre@citeweb.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message