From owner-freebsd-current@FreeBSD.ORG Fri Jan 16 08:18:14 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 9D89716A4CE for ; Fri, 16 Jan 2004 08:18:14 -0800 (PST) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6DC3D43D49 for ; Fri, 16 Jan 2004 08:18:13 -0800 (PST) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.12.10/8.12.10) with ESMTP id i0GGIC5P025132 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 16 Jan 2004 11:18:12 -0500 (EST) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.12.9p2/8.12.9/Submit) id i0GGI7dS089035; Fri, 16 Jan 2004 11:18:07 -0500 (EST) (envelope-from gallatin) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <16392.3775.357598.168310@grasshopper.cs.duke.edu> Date: Fri, 16 Jan 2004 11:18:07 -0500 (EST) To: Peter Losher In-Reply-To: <200401160250.54593.Peter_Losher@isc.org> References: <200401160250.54593.Peter_Losher@isc.org> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid cc: current@freebsd.org Subject: Re: Fatal trap under heavy load on 5.1-p10... 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: Fri, 16 Jan 2004 16:18:14 -0000 Peter Losher writes: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > After dealing w/ the traffic loads for 5.2-REL for three days, the system > popped into the debugger and halted. Here are the relevant details - If this > is a known issue and fixed in 5.2, then say so... :) > The panic is in the zero-copy sockets code. There have been many VM system changes which may have fixed it. Specifically, Alan Cox improved the sf_buf system quite a bit between 5.1 and 5.2. Can you map socow_setup+0xc9 to a source file line number in your system please? Do gdb kernel.debug <...> (gdb) list *socow_setup+0xc9 BTW, A certain way to prevent this from ever happening is to disable that code by removing options ZERO_COPY_SOCKETS from your kernel config file. Thanks, Drew