From owner-freebsd-bugs Fri Mar 8 11:13:39 2002 Delivered-To: freebsd-bugs@freebsd.org Received: from patrocles.silby.com (d59.as26.nwbl0.wi.voyager.net [169.207.65.187]) by hub.freebsd.org (Postfix) with ESMTP id 3DBB237B41B for ; Fri, 8 Mar 2002 11:13:34 -0800 (PST) Received: from patrocles.silby.com (localhost [127.0.0.1]) by patrocles.silby.com (8.12.2/8.12.2) with ESMTP id g28DI0Nu008341; Fri, 8 Mar 2002 13:18:00 GMT (envelope-from silby@silby.com) Received: from localhost (silby@localhost) by patrocles.silby.com (8.12.2/8.12.2/Submit) with ESMTP id g28DHvHJ008338; Fri, 8 Mar 2002 13:17:57 GMT X-Authentication-Warning: patrocles.silby.com: silby owned process doing -bs Date: Fri, 8 Mar 2002 13:17:57 +0000 (GMT) From: Mike Silbersack To: The Anarcat Cc: freebsd-bugs@FreeBSD.org, "Daniel O'Callaghan" Subject: Re: kern/35640: heavy collision rate hangs vr network interface In-Reply-To: <20020308182518.GA261@lenny.anarcat.dyndns.org> Message-ID: <20020308130845.K8029-100000@patrocles.silby.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-bugs@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, 8 Mar 2002, The Anarcat wrote: > That was it. Linking the 2 boxes with a crosslink and putting > both at autoselect makes them switch to 100baseTX. > So this means the patch on the rl works, since this (media autoselect) > wasn't working before. > > But now I feel kind of bad, since there's got to be something I clearly > misunderstand about the protocol. > > Both interfaces are capable of doing 10baseT/UTP. Why are they doing > 100baseTX? Can't I run those in a 10baseT network??? Does this mean I > must buy a 100baseTX hub? There are four possible configurations: 10baseT - half duplex 10baseT - full duplex 100baseTX - half duplex 100baseTX - full duplex My best guess is that the vr card isn't autoselecting 10baseT - half duplex correctly. Unfortunately, I don't have any vr cards here, so I wouldn't be able to help you much in working on that. Perhaps it's worth sending a message to -questions or -stable asking if anyone has the same network card you do and if they have used it on a 10baseT network. Buying a 100baseTX switch sounds like it would solve your problem, but I'm not sure that's the most efficient way to solve the problem. If you know that the vr card is to blame, perhaps you could try replacing it with a different NIC to see if that solves the problem temporarily. > I still believe there is something wrong with the vr since it is > something external to the server that made it freeze, so it's a kind of > DOS. It shouldn't happen, IMHO. If the duplex is being negotiated incorrectly, I wouldn't be surprised that the network card is acting very strangely. That being said, I see a few problem reports already filed about the vr cards... there's probably a driver bug in there. So, ask around on -questions and -stable, maybe someone else has a fix. Mike "Silby" Silbersack To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message