From owner-freebsd-mobile Sat Oct 14 16: 0: 5 2000 Delivered-To: freebsd-mobile@freebsd.org Received: from mx-out.daemonmail.net (mx-out.daemonmail.net [209.75.5.3]) by hub.freebsd.org (Postfix) with ESMTP id 647CA37B66C for ; Sat, 14 Oct 2000 16:00:00 -0700 (PDT) Received: from mx0.emailqueue.net (localhost.daemonmail.net [127.0.0.1]) by mx-out.daemonmail.net (8.9.3/8.9.3) with SMTP id QAA41702; Sat, 14 Oct 2000 16:00:09 -0700 (PDT) (envelope-from kkonstan@webdaemon.net) Received: from webdaemon.net (webdaemon.net [213.5.85.63]) by mail.webdaemon.net with ESMTP id KqA0quG2 Sat, 14 Oct 2000 16:00:03 -0700 (PDT) Message-ID: <39E8C70A.4E5F8160@webdaemon.net> Date: Sat, 14 Oct 2000 23:50:18 +0300 From: Konstantinos Konstantinidis Reply-To: kkonstan@duth.gr Organization: I've heard of it. X-Mailer: Mozilla 4.75 [en] (X11; U; FreeBSD 4.1.1-RELEASE i386) X-Accept-Language: el, en MIME-Version: 1.0 To: freebsd-mobile@freebsd.org Cc: sklauder@ibd-web.de Subject: Re: "device timeout" with DFE-650 Content-Type: text/plain; charset=iso-8859-7 Content-Transfer-Encoding: 7bit Sender: owner-freebsd-mobile@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org (quoted from the web archives) > BTW, since we're on this topic; IMHO this card is just junk. > > I've got mine working a couple of weeks ago, but it's in a rather > useless state: cvsup/ftp/scp etc are not working. > > Any connection that transfers more than some 40K at a time stalls and > finally times out. Oddly enough, http,nfs,ssh,telnet are just fine. > With ftp I get a speed of ~7K/sec :( > > I'm running 4.1-STABLE and it makes no difference whether I use polling > mode or not. I've just tried a update to 4.1.1-STABLE (via NFS mount), > but it seems that sendmail-8.11.1 (MFC'ed yesterday) is broken... I have exactly the same symptoms with my xircom. it was working perfectly for several months with 4.0-STABLE, but when I upgraded to 4.1.1 it started behaving this way. Needless to say, I'm pissed :) I am relieved to find out that it doesn't only affect the xircoms, since nothing relevant chagned on the xircom driver since 4.0 and I was really lost trying to figure out where the breakage was - I am now convinced that something somewhere else is really, really broken. Consider giving 4.0 a try - see if that works - it's not the card that's junk, I'm convinced something broke in FreeBSD sometime in july, august or september. Unfortunately I have a libretto, and 4.0 does not support the xircom properly (that was MFCd a bit after the release). Back then I had a windows partition which I used to setup freebsd from a dos partition. I nuked this before I setup 4.1.1 (since there was no need for it anyway). Needless to say, setting up windows again will be a major PITA (20+ floppies, and the floppy drive is unbeleivably slow). I made a post regarding this on freebsd-mail last week but it seems to have been ignored - since then I've been spending all my free time debugging and can't honestly figure out why for some reason when large flows build up (so to say) the kernel stops sending packets to the card. Hope someone smarter than me figures this out before I go nuts. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-mobile" in the body of the message