From owner-freebsd-net Tue May 26 23:32:20 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA11174 for freebsd-net-outgoing; Tue, 26 May 1998 23:32:20 -0700 (PDT) (envelope-from owner-freebsd-net@FreeBSD.ORG) Received: from ifi.uio.no (0@ifi.uio.no [129.240.64.2]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA11107 for ; Tue, 26 May 1998 23:31:59 -0700 (PDT) (envelope-from dag-erli@ifi.uio.no) Received: from nastrond.ifi.uio.no (2602@nastrond.ifi.uio.no [129.240.64.65]) by ifi.uio.no (8.8.8/8.8.7/ifi0.2) with ESMTP id IAA25383 for ; Wed, 27 May 1998 08:31:56 +0200 (MET DST) Received: (from dag-erli@localhost) by nastrond.ifi.uio.no ; Wed, 27 May 1998 08:31:55 +0200 (MET DST) Mime-Version: 1.0 To: net@FreeBSD.ORG Subject: ep0 problems: shed some light? Organization: University of Oslo, Department of Informatics X-url: http://www.stud.ifi.uio.no/~dag-erli/ X-Stop-Spam: http://www.cauce.org From: dag-erli@ifi.uio.no (Dag-Erling Coidan =?iso-8859-1?Q?Sm=F8rgrav?= ) Date: 27 May 1998 08:31:54 +0200 Message-ID: Lines: 31 X-Mailer: Gnus v5.5/Emacs 19.34 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org OK, I have begun to have problems with my ep0 getting stuck again, despite setting NMBCLUSTERS to 8192. I believe I have gotten a little bit closer to tracking down the bug. First, it is not mbuf-related. Although augmenting the number of mbufs seems to help, it can't be much more than an interesting side-effect. mbuf usage on my box never even reached 250 from my last reboot a couple of days ago till the interface got stuck last night. Second, it seems that 'ifconfig ep0 up' alone is enough to unstick the interface (which is good news, because 'ifconfig ep0 down' zaps my NFS connections). Third, 'ifconfig -au' before 'ifconfig ep0 up' reveals that ep0 is stuck with the OACTIVE bit on. Although 'ifconfig ep0 up' should theoretically be a no-op since ep0 is already up, it apparently clears the OACTIVE flag, which unsticks the interface. Fourth, the bug always manifests itself *between* two large NFS transfers. The only other machine on that physical network is a scratch box which does nothing but play MP3 files over NFS. The point at which the interface gets stuck (on the server, never on the client) is at the beginning of an MP3 file (16 frames into the file, to be precise). Does anyone feel like diving into the ep0 code to try to find out why the OACTIVE flag gets stuck? I would, but I have an exam in half an hour :) -- Noone else has a .sig like this one. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message