From owner-freebsd-hackers@FreeBSD.ORG Tue Mar 9 20:40:30 2010 Return-Path: Delivered-To: freebsd-hackers@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id D68FB1065670 for ; Tue, 9 Mar 2010 20:40:30 +0000 (UTC) (envelope-from toasty@dragondata.com) Received: from mail-yx0-f184.google.com (mail-yx0-f184.google.com [209.85.210.184]) by mx1.freebsd.org (Postfix) with ESMTP id 930B48FC12 for ; Tue, 9 Mar 2010 20:40:30 +0000 (UTC) Received: by yxe14 with SMTP id 14so1194465yxe.7 for ; Tue, 09 Mar 2010 12:40:29 -0800 (PST) Received: by 10.101.11.14 with SMTP id o14mr564081ani.196.1268167229570; Tue, 09 Mar 2010 12:40:29 -0800 (PST) Received: from vpn177.ord02.your.org (vpn177.ord02.your.org [204.9.55.177]) by mx.google.com with ESMTPS id 21sm5894390iwn.15.2010.03.09.12.40.27 (version=TLSv1/SSLv3 cipher=RC4-MD5); Tue, 09 Mar 2010 12:40:28 -0800 (PST) From: Kevin Day Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: quoted-printable Date: Tue, 9 Mar 2010 14:40:26 -0600 Message-Id: <2C7A849F-2571-48E7-AA75-B6F87C2352C1@dragondata.com> To: FreeBSD Hackers Mime-Version: 1.0 (Apple Message framework v1077) X-Mailer: Apple Mail (2.1077) Subject: Extremely slow boot on VMWare with Opteron 2352 (acpi?) X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 09 Mar 2010 20:40:30 -0000 I'm troubleshooting a pretty weird problem with running FreeBSD 8.0 = (amd64) inside VMware ESX/ESXi servers. We've got a wide range of = physical servers running identical copies of VMware and identical = FreeBSD virtual machines. Everything works fine on all of our servers = for Windows and Linux VMs, but FreeBSD running on Opteron 2352 physical = servers takes an average of about 20 minutes to boot. Normally I would = chalk this up to being a VMware bug, but the situation that's actually = occurring is somewhat interesting. If I boot up on an Opteron 2218 system, it boots normally. If I boot the = exact same VM moved to a 2352, I get: acpi0: on motherboard PCIe: Memory Mapped configuration base @ 0xe0000000 (very long pause) ioapic0: routing intpin 9 (ISA IRQ 9) to lapic 0 vector 48 acpi0: [MPSAFE] acpi0: [ITHREAD] then booting normally. The pause is between 1 and 60 minutes (somewhat variable, and i'm not = sure on what). After it eventually boots, everything seems fine. Doing a = diff between the 2218(-) and 2352(+) servers' verbose boots, I see: -ACPI timer: 0/90 0/23 1/1 1/1 1/1 1/1 0/12 1/1 1/1 1/1 -> 7 +ACPI timer: 0/162 0/1546 0/1119 0/150 0/165 0/778 0/123 0/203 0/83 0/93 = -> 0 This looks more interesting. If I'm reading acpi_timer_test() correctly, = the ACPI timer isn't particularly good on the 2218, but is flat out = unusable on the 2352. I don't know if this is a symptom or the actual = cause of the problem, though. Disabling ACPI results in an instant boot up, but the SCSI PCI device = isn't found which makes it kinda pointless. Before I delve too deeply into this, does this ring any bells to anyone?