From owner-freebsd-hackers@FreeBSD.ORG Sun Feb 29 12:40:54 2004 Return-Path: Delivered-To: freebsd-hackers@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6C8CB16A4CE for ; Sun, 29 Feb 2004 12:40:54 -0800 (PST) Received: from mail.sandvine.com (sandvine.com [199.243.201.138]) by mx1.FreeBSD.org (Postfix) with ESMTP id A1CF243D39 for ; Sun, 29 Feb 2004 12:40:53 -0800 (PST) (envelope-from don@sandvine.com) Received: by mail.sandvine.com with Internet Mail Service (5.5.2657.72) id <15P458HZ>; Sun, 29 Feb 2004 15:40:52 -0500 Message-ID: From: Don Bowman To: 'Deepak Jain' , freebsd-hackers@freebsd.org Date: Sun, 29 Feb 2004 15:40:44 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2657.72) Content-Type: text/plain; charset="iso-8859-1" Subject: RE: 4.9 boot problem on em0 platform. X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 29 Feb 2004 20:40:54 -0000 > From: Deepak Jain [mailto:deepak@ai.net] > As a part of tracking down a performance issue, I tried building a > custom kernel (with just IPFW, DUMMYNET added, NMBCLUSTERS, > commenting > out MATH_EMULATE, INET6, I386, I486). The system is currently > running a > kernel from a similar machine with the same settings. The > machine does > run on this kernel: 4.9-RELEASE FreeBSD 4.9-RELEASE #8, with > the above > options, but I have not been able to compile a 4.9-RELEASE #2 > (which is > the source tree on the machine) kernel that has an identical > config file. > > So, when it builds itself from -RELEASE sources, it hangs at: > > "pmap_mapdev: Couldn't alloc kernel virtual memory" I couldn't find a > reference to anything recent. Nothing non-default (from a GENERIC > kernel) with respect to ACPI has been touched. I see a reference to > -CURRENT from 9/03, but that's it. Just a guess, but i think you've bumped nmbclusters or nmbufs up too much (or perhaps maxsockets, maxfds, ...) and have run out of KVA. You can tune clusters & mbufs in loader.conf without recompiling kernel. You will want to see what vm.zone_kmem_pages, vm.zone_kmem_kvaspace are showing you, vmstat -z, vmstat -m, etc. You may want to alter VM_KMEM_SIZE_SCALE to e.g. '2' if you are trying to put more into the kernel mem space.