From owner-freebsd-current Thu Aug 22 17:28:49 2002 Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 264CD37B400 for ; Thu, 22 Aug 2002 17:28:47 -0700 (PDT) Received: from canning.wemm.org (canning.wemm.org [192.203.228.65]) by mx1.FreeBSD.org (Postfix) with ESMTP id DA82943E6A for ; Thu, 22 Aug 2002 17:28:46 -0700 (PDT) (envelope-from peter@wemm.org) Received: from wemm.org (localhost [127.0.0.1]) by canning.wemm.org (Postfix) with ESMTP id BBF082A7D6 for ; Thu, 22 Aug 2002 17:28:46 -0700 (PDT) (envelope-from peter@wemm.org) X-Mailer: exmh version 2.5 07/13/2001 with nmh-1.0.4 To: current@FreeBSD.ORG Subject: Re: Memory corruption in -CURRENT [was Re: Plea to committers to only commit to HEAD if you run -current {from developers@FreeBSD.org}] In-Reply-To: <20020822233846.GJ90596@procyon.firepipe.net> Date: Thu, 22 Aug 2002 17:28:46 -0700 From: Peter Wemm Message-Id: <20020823002846.BBF082A7D6@canning.wemm.org> Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Will Andrews wrote: > On Thu, Aug 22, 2002 at 12:12:02PM -0700, David O'Brien wrote: > > What -j value did you use? Anyway, "so what"? `make buildworld' isn't > > the most stressful thing. To better stress things, do > > `make -j<10*number CPU> buildworld', and a few other processes that > > allocate/deallocate memory, span processes, etc.. Ie, add some > > dissimilarity to the mix. > > Try 6 to 12 simultaneous portbuild jobs, that will beat make -j20 > buildworld on any metric (mostly when you are building something > as large as KDE). I've done that on P4 1.7GHz / 768MB / 40GB ATA > boxes and when a few large (1000+ file) packages are being untarred > simultaneously, there frequently becomes a tar(1) or two exiting > with a signal 4, or a gcc exiting with a signal 11 before it's > done. That's on a July 31 -CURRENT kernel FWIW, and before ~July > 1, this never happened. Normally I just cron another run to > finish the ones that spontaneously SIG4/SIG11'd, and that > finishes the remaining jobs. > > Currently I'm doing another test build with DISABLE_PSE and > DISABLE_PG_G in the kernel to see how that handles. As of yet, > there have been exactly *ZERO* SIG4/SIG11's this build. That > has never happened since KSE MIII, and I've done at least ten or > fifteen full builds since then. :-) Is this on a uniprocessor system? Note that the old SMP code never had PG_G active for SMP before the last round of pmap changes. DISABLE_PG_G almost goes back to the old way for SMP systems. DISABLE_PSE should be irrelevant since we've been using it all along. I've become aware of some nasty races in the pmap code for SMP boxes about a week ago while working on PAE stuff. DISABLE_PG_G *might* minimize the effect of them but they are still there. Cheers, -Peter -- Peter Wemm - peter@wemm.org; peter@FreeBSD.org; peter@yahoo-inc.com "All of this is for nothing if we don't go to the stars" - JMS/B5 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message