From owner-freebsd-current@FreeBSD.ORG Wed Sep 5 18:32:43 2007 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 8130E16A417 for ; Wed, 5 Sep 2007 18:32:43 +0000 (UTC) (envelope-from mfp49_freebsd@plass-family.net) Received: from plass-family.net (adsl-68-127-22-237.dsl.pltn13.pacbell.net [68.127.22.237]) by mx1.freebsd.org (Postfix) with ESMTP id 5CF4D13C467 for ; Wed, 5 Sep 2007 18:32:43 +0000 (UTC) (envelope-from mfp49_freebsd@plass-family.net) Received: from nat.plass-family.net (nat.plass-family.net [68.127.22.235]) by plass-family.net (Postfix) with ESMTP id 746C02284B for ; Wed, 5 Sep 2007 11:22:26 -0700 (PDT) Received: from [13.2.117.100] (presto.parc.xerox.com [13.2.117.100]) by garage.plass-family.net (Postfix) with ESMTP id E8D785C22 for ; Wed, 5 Sep 2007 11:22:25 -0700 (PDT) Mime-Version: 1.0 (Apple Message framework v752.2) In-Reply-To: <20070905120016.2470416A4F8@hub.freebsd.org> References: <20070905120016.2470416A4F8@hub.freebsd.org> Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Message-Id: <2D3D02AF-9586-4908-997D-E889A70E5A90@plass-family.net> Content-Transfer-Encoding: 7bit From: Michael Plass Date: Wed, 5 Sep 2007 11:22:19 -0700 To: freebsd-current@freebsd.org X-Mailer: Apple Mail (2.752.2) Subject: Re: Unkillable and runaway processes 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, 05 Sep 2007 18:32:43 -0000 On Wed, 05 Sep 2007, Benjamin Close wrote: > I'd love to debug here this but can't as the box uses a USB > mouse/keyboard so every time I drop to a debugger I lose keyboard > support :( I have had some success using DDB with a USB keyboard by removing atk and kbdmux from the kernel config (nodevice lines for atkbdc, atkbd, psm, and kbdmux). A downside is that when I do this, dumps stop working - they hang on the call to cncheckc() when checking for user abort (the cncheckc() never returns). Replacing this call with (-1) allows the dump to proceed. For me, the relevant file is /usr/src/sys/amd64/amd64/ minidump_machdep.c. - Michael