From owner-freebsd-questions Fri May 18 7:48:42 2001 Delivered-To: freebsd-questions@freebsd.org Received: from courier.netrail.net (courier.netrail.net [205.215.10.53]) by hub.freebsd.org (Postfix) with ESMTP id 31C3637B422 for ; Fri, 18 May 2001 07:48:38 -0700 (PDT) (envelope-from cschreiber@netrail.net) Received: from cschriaber (localhost.netrail.net [127.0.0.1]) by courier.netrail.net (Postfix) with SMTP id B5EF3DD for ; Fri, 18 May 2001 10:48:30 -0400 (EDT) Reply-To: From: "Christian S." To: Subject: RE: TX underrun, increasing TX threshold Date: Fri, 18 May 2001 10:47:54 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0) In-Reply-To: <003801c0df21$4874c1a0$2813933f@cat> X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Importance: Normal Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I've had the same problem with my xl0 driver (3C905).. I get hideously slow performance transferring between 2 100Mb NIC's, but they are connected via a 100Mb hub.. I haven't tried a x-connect cable 'twixt the two, so I can't blame the NIC's yet.. :/ Christian > -----Original Message----- > From: owner-freebsd-questions@FreeBSD.ORG > [mailto:owner-freebsd-questions@FreeBSD.ORG]On Behalf Of pan > Sent: Thursday, May 17, 2001 6:33 PM > To: freebsd-questions@FreeBSD.ORG > Subject: Re: TX underrun, increasing TX threshold > > > > From: "Peter" > Subject: TX underrun, increasing TX threshold > > > > What exactly does this error mean and how can I avoid it? > > [It appears while doing heavy/big file transfers between my LAN > on my dc0 > [linksys] ] > > Newer Linksys seem to have (ADMtek AN985) chips instead of the > (82c169 PNIC) > ones. > (check your boot message). > The dc0 driver never produces the underrun event with the PNIC > while with the ADMtek > it does. > What it is, bascially, is that your system could not produce data > fast enough for your > nic. There's more to it - buffer size etc, but the message says > that your system has > decided to increase the number of bytes to fill the buffer before > the nic can send. > This can occur more than once after a boot until the system and > the nic are > happy with > each other - once a reset is found the messages stop and everything > is ok. if you reboot > you will see the messages again. the event occurs only within a LAN > where you have > a linespeed greater than 10Mbs - if you were only trafficking > through your gateway to > the net (via a dsl bridge/router for example) at the max > linespeed of 10Mbs > you wouldn't > see the message. I bet you only see it when you reach > n x > 1024kbps linespeed within your > LAN. > > I think the dc0 driver intializes the nic with too small a number > of bytes - > it certainly > isn't true on the physical layer that the system hardware can't > keep up with > the nic - > it's a handshaking thing. > > Corrections and additions welcome. > > Pan > > > > > > > > _________________________________________________________ > > Do You Yahoo!? > > No. > > > Get your free @yahoo.com address at http://mail.yahoo.com > > No thanks. > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-questions" in the body of the message -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 6.5.8 for non-commercial use iQA/AwUBOwU1uCkK9qTvGvteEQIg2QCeIA+HSrW+7O4zz+mUqmMQG6HzZtMAnipj Oi3yfM742YWKoFUwcCrgRN0E =vod8 -----END PGP SIGNATURE----- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message