From owner-freebsd-current@FreeBSD.ORG Wed Oct 26 07:11:18 2005 Return-Path: X-Original-To: current@freebsd.org 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 C7E0116A41F for ; Wed, 26 Oct 2005 07:11:18 +0000 (GMT) (envelope-from bkoenig@cs.tu-berlin.de) Received: from efacilitas.de (smtp.efacilitas.de [85.10.196.108]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5EAF643D46 for ; Wed, 26 Oct 2005 07:11:17 +0000 (GMT) (envelope-from bkoenig@cs.tu-berlin.de) Received: from eurystheus.local (port-212-202-37-245.dynamic.qsc.de [212.202.37.245]) by efacilitas.de (Postfix) with ESMTP id 150274B25B for ; Wed, 26 Oct 2005 09:18:39 +0200 (CEST) Received: from [192.168.1.4] (syn.local [192.168.1.4]) by eurystheus.local (Postfix) with ESMTP id 272B2332F48 for ; Wed, 26 Oct 2005 09:10:11 +0200 (CEST) Message-ID: <435F2BCE.1040704@cs.tu-berlin.de> Date: Wed, 26 Oct 2005 09:10:06 +0200 From: Bjoern Koenig User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.7.10) Gecko/20050824 X-Accept-Language: en-us, en MIME-Version: 1.0 To: current@freebsd.org Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit Cc: Subject: deadlocks with SMP and Pentium 3 X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 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: Wed, 26 Oct 2005 07:11:18 -0000 Hello, I have an Asus CUVX4-D mainboard with an VIA Apollo Pro 133A chipset and two Intel Pentium III 733 MHz processors. It runs fine with FreeBSD 4.11 and SMP enabled, but something happened before 5.0-RELEASE that it don't work anymore now; neither with 5.4-RELEASE, 6.0-RC1 nor 7-CURRENT. In case I have many concurrent processes the system just locks up after a few minutes. I disabled ACPI, compiled a kernel with WITNESS, set all debug mpsafe tunables to zero and many other things, but nothing helped to find a reason for this behaviour. Right now I noticed that top(1) shows a huge bunch of processes with the state *Giant shortly before the machine freezed. Therefore I think it's related to the giant lock. ;-) Currently I prepare to do a binary search in HEAD between march 2000 and januar 2003, but I would be glad if somebody still has a hint for me to prevent me from doing this. Is there still a way to gather debug information? Regards Björn