From owner-freebsd-current Sun May 18 04:07:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA24680 for current-outgoing; Sun, 18 May 1997 04:07:21 -0700 (PDT) Received: from Campino.Informatik.RWTH-Aachen.DE (campino.Informatik.RWTH-Aachen.DE [137.226.116.240]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA24660; Sun, 18 May 1997 04:07:15 -0700 (PDT) Received: from gil.physik.rwth-aachen.de (gilberto.physik.rwth-aachen.de [137.226.31.2]) by Campino.Informatik.RWTH-Aachen.DE (RBI-Z-5/8.6.12) with ESMTP id NAA29338; Sun, 18 May 1997 13:08:27 +0200 (MET DST) Received: (from kuku@localhost) by gil.physik.rwth-aachen.de (8.8.5/8.6.9) id NAA07993; Sun, 18 May 1997 13:08:39 +0200 (MEST) From: Christoph Kukulies Message-Id: <199705181108.NAA07993@gil.physik.rwth-aachen.de> Subject: Re: can anyone confirm my NFS/VM problem? In-Reply-To: from Doug Rabson at "May 17, 97 04:52:46 pm" To: dfr@nlsystems.com (Doug Rabson) Date: Sun, 18 May 1997 13:08:39 +0200 (MEST) Cc: kuku@gilberto.physik.RWTH-Aachen.DE, dyson@freebsd.org, current@freebsd.org Reply-To: Christoph Kukulies X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > On Fri, 16 May 1997, Christoph Kukulies wrote: > > > > > I have created a test case to repeat that weird NFS/VM problem I > > have here with a physics production program - the exposed bug is > > really a showstopper for our production and I'm not sure if it is > > not also in 2.2.1R. > > (Don't know if Doug is already going to deal with this group > > of NFS/VM/MMAP problems). Anyway, to assure it is really a > > repeatable problem independent of some local peculiarity here > > at my site, I'm seeking for someone with the following > > configuration: > > Well I think I have a fix. It was bloody painful too. There were some > extremely ugly problems associated with NFS' use of b_validoff and > b_validend. I attempted to fix them but it still feels pretty fragile > although it does pass Ghristoph's test case. I tested it here and can confirm the latter - at least it cures my test case. > > John, could you have a look at these changes and tell me whether they make > sense? > [ patch of kern/vfs_bio.c vm/vm_fault.c vm/vnode_pager.c ommitted ] > -- > Doug Rabson Mail: dfr@nlsystems.com > Nonlinear Systems Ltd. Phone: +44 181 951 1891 > -- Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de