From owner-freebsd-current@FreeBSD.ORG Tue Sep 21 03:42:41 2004 Return-Path: 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 3DA3916A4CE for ; Tue, 21 Sep 2004 03:42:41 +0000 (GMT) Received: from spxau01.smeglobalnet.net (spxau01.smeglobalnet.net [203.57.65.8]) by mx1.FreeBSD.org (Postfix) with ESMTP id 635DC43D39 for ; Tue, 21 Sep 2004 03:42:40 +0000 (GMT) (envelope-from andy@bradfieldprichard.com.au) Received: from bpgate.speednet.com.au ([203.41.15.9]) by spxau01.smeglobalnet.net with Microsoft SMTPSVC(5.0.2195.6713); Tue, 21 Sep 2004 13:15:07 +1000 Received: from bpgate.speednet.com.au (localhost [127.0.0.1]) i8L3F6W6061029 for ; Tue, 21 Sep 2004 13:15:06 +1000 (EST) (envelope-from andy@bradfieldprichard.com.au) Received: from localhost (andy@localhost)i8L3F5qA061026 for ; Tue, 21 Sep 2004 13:15:05 +1000 (EST) (envelope-from andy@bradfieldprichard.com.au) X-Authentication-Warning: bpgate.speednet.com.au: andy owned process doing -bs Date: Tue, 21 Sep 2004 13:15:05 +1000 (EST) From: Andy Farkas X-X-Sender: andy@bpgate.speednet.com.au To: freebsd-current@freebsd.org Message-ID: <20040921125501.L60864@bpgate.speednet.com.au> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed X-OriginalArrivalTime: 21 Sep 2004 03:15:07.0798 (UTC) FILETIME=[32643360:01C49F89] Subject: panic: APIC: Previous IPI is stuck X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 21 Sep 2004 03:42:41 -0000 Why are my IPI's stuck? They never used to be.... This box has been tracking RELENG_5 and was running fine until about the end of August. Now I get constant panics, even during a fsck when booted into single user mode! The situation seems to have become worse :( Its a Dell PowerEdge 6100/200 quad Pentium-Pro 200MHz with 512MB RAM. I have tried both ULE and 4BSD schedulers. GENERIC has been panicing for the last 4 weeks. It was suggested I try infinite waiting in the ipi wait loop, but that didn't work. I tried s/panic/printf/ and this seemed to work... for a while. Once, it printed 3 "Previous IPI is stuck" messages at various times it was doing a buildworld, and eventually finished! But the box will eventually crash, in different ways, with this kludge. This box is not ready to be recycled yet, but is unable to run the latest BETA5. Can someone explain why an IPI might get stuck? Thanks, -andyf ps. I have posted twice previously with dmesg and other output. If you want me to repeat, please ask.