From owner-freebsd-hardware Sun May 31 22:11:12 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id WAA01716 for freebsd-hardware-outgoing; Sun, 31 May 1998 22:11:12 -0700 (PDT) (envelope-from owner-freebsd-hardware@FreeBSD.ORG) Received: from implode.root.com (implode.root.com [198.145.90.17]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id WAA01711 for ; Sun, 31 May 1998 22:11:05 -0700 (PDT) (envelope-from root@implode.root.com) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.5/8.8.5) with ESMTP id WAA09519; Sun, 31 May 1998 22:10:37 -0700 (PDT) Message-Id: <199806010510.WAA09519@implode.root.com> To: Scott Drassinower cc: freebsd-hardware@FreeBSD.ORG Subject: Re: Intel EtherExpress 100+ problems In-reply-to: Your message of "Sun, 31 May 1998 21:06:20 EDT." From: David Greenman Reply-To: dg@root.com Date: Sun, 31 May 1998 22:10:37 -0700 Sender: owner-freebsd-hardware@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >I've been having some problems with an Intel EtherExpress 100+ card (which >2.2.6-RELEASE is seeing as a 100B card, fwiw) running in 100 megabit mode, >full duplex, to a Cisco Catalyst 2900XL. Running tail or grep on a large >(>10mb) file via NFS from another 2.2.6 machine (with an Intel 100+ too) >will simply freeze the tail or grep process, and it won't die. ps shows >the process in disk wait, even though other operations on the mount will >be fine. > >When I replace the Cisco with a 10 megabit hub, the cards drop to 10 >megabit and half duplex, and there are no nfs problems. > >Cisco said they had some internal docs talking about problems with the >Intel 100B cards running 100 megabit, full duplex, that were cleared up >with the 100+ cards. Intel was completely useless. > >I'm wondering if there is perhaps a problem with the fxp driver in 2.2.6, >or some weird issue with the Cisco. I don't have another 100 megabit >switch laying around, and I really wouldn't want to have to switch from >the Intel cards not knowing what the problem is. All that Cisco could >suggest was different cards or a sniffer to look for more clues. > >Any ideas? This seems pretty weird. There are no known bugs in the fxp driver, but there are plenty of NFS bugs, some of which show up as link speed sensitive race conditions. -DG David Greenman Co-founder/Principal Architect, The FreeBSD Project To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message