From owner-freebsd-current Tue Dec 24 11: 6:51 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 7596237B401 for ; Tue, 24 Dec 2002 11:06:50 -0800 (PST) Received: from kayak.xcllnt.net (209-128-86-226.bayarea.net [209.128.86.226]) by mx1.FreeBSD.org (Postfix) with ESMTP id 749B643ED1 for ; Tue, 24 Dec 2002 11:06:49 -0800 (PST) (envelope-from marcel@xcllnt.net) Received: from dhcp01.pn.xcllnt.net (dhcp01.pn.xcllnt.net [192.168.4.201]) by kayak.xcllnt.net (8.12.6/8.12.6) with ESMTP id gBOJ6irT003895; Tue, 24 Dec 2002 11:06:44 -0800 (PST) (envelope-from marcel@kayak.pn.xcllnt.net) Received: from dhcp01.pn.xcllnt.net (localhost [127.0.0.1]) by dhcp01.pn.xcllnt.net (8.12.6/8.12.6) with ESMTP id gBOJ7e73000816; Tue, 24 Dec 2002 11:07:40 -0800 (PST) (envelope-from marcel@dhcp01.pn.xcllnt.net) Received: (from marcel@localhost) by dhcp01.pn.xcllnt.net (8.12.6/8.12.6/Submit) id gBOJ7doK000815; Tue, 24 Dec 2002 11:07:39 -0800 (PST) Date: Tue, 24 Dec 2002 11:07:39 -0800 From: Marcel Moolenaar To: Yu-Shun Wang Cc: current@FreeBSD.org Subject: Re: mpt(4): berserk interrupt frequency Message-ID: <20021224190739.GA782@dhcp01.pn.xcllnt.net> References: <20021224004748.GA1134@athlon.pn.xcllnt.net> <3E08A798.6080400@isi.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3E08A798.6080400@isi.edu> User-Agent: Mutt/1.5.1i 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 On Tue, Dec 24, 2002 at 10:29:44AM -0800, Yu-Shun Wang wrote: > Hi, > > Can't comment on the reasons for your case, but that number was > definitely strange. Here's ours: I figured out what was the problem. On the ia64 box this was happening we had sio2 interrupting at irq 82, while there was only room to keep counters for 64 interrupt sources. We over- flowed the interrupt name table and wrote into the interrupt counter table and happened to globber the counters for mpt0, mpt1 and bge0. I increased the table size for 128 interrupt sources and all looks normal again. If it wasn't for vmstat I'd nuke the whole table and have a more dynamic approach... Ah well.. -- Marcel Moolenaar USPA: A-39004 marcel@xcllnt.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message