From owner-freebsd-questions Thu Oct 26 15:51: 6 2000 Delivered-To: freebsd-questions@freebsd.org Received: from hotmail.com (f50.law6.hotmail.com [216.32.241.50]) by hub.freebsd.org (Postfix) with ESMTP id 25B8737B479; Thu, 26 Oct 2000 15:51:01 -0700 (PDT) Received: from mail pickup service by hotmail.com with Microsoft SMTPSVC; Thu, 26 Oct 2000 15:51:01 -0700 Received: from 203.11.225.5 by lw6fd.law6.hotmail.msn.com with HTTP; Thu, 26 Oct 2000 22:51:00 GMT X-Originating-IP: [203.11.225.5] From: "Aaron Hill" To: julian@elischer.org Cc: freebsd-questions@FreeBSD.ORG, freebsd-net@FreeBSD.ORG Subject: Re: More on PPPoE & ADSL (Telstra Bigpond) Date: Thu, 26 Oct 2000 22:51:00 GMT Mime-Version: 1.0 Content-Type: text/plain; format=flowed Message-ID: X-OriginalArrivalTime: 26 Oct 2000 22:51:01.0015 (UTC) FILETIME=[36534270:01C03F9F] Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >The provider treats these differently >we probably need to reverse the order of the Host_uniq and Service >name fields.. It does look like that hence my searching through the source code to check if I could make a simple hack to test it. I ran out of time though. I've been reading the RFC (2516) lately which says (I quote)... The PADI packet MUST contain exactly one TAG of TAG_TYPE Service-Name, indicating the service the Host is requesting, and any number of other TAG types. ... so from the order of that statement it seems putting the Service-Name tag first is the correct thing to do. The RFC doesn't explicitly mention what order the tags should be in. It's entirely plausible that the ISPs equipment has a requirement (bug?) that the service name comes last. >(I don't know why it comes up with "UTF8" though, I think that's >TCPDUMP misreading something, (I forget what I put there)) I agree, I think it's tcpdump trying, incorrectly, to interpret the tag contents. In case you'd like to see what the Host-Uniq tags actually contained here's some hex of the Windows PADI frame... 0:10:5a:0:d3:de Broadcast 8863 60: PPPoE PADI [Host-Uniq UTF8] [Service-Name "bigpond"] 0x0000 1109 0000 0015 0103 0006 0010 5a00 d3de ............Z... 0x0010 0101 0007 6269 6770 6f6e 6400 0000 0000 ....bigpond..... 0x0020 0000 0000 0000 0000 0000 0000 0000 .............. ... and the FreeBSD PADI frame... 0:e0:29:73:81:dd Broadcast 8863 60: PPPoE PADI [Service-Name "bigpond"] [Host-Uniq UTF8] 0x0000 1109 0000 0013 0101 0007 6269 6770 6f6e ..........bigpon 0x0010 6401 0300 0480 7067 c300 0000 0000 0000 d.....pg........ 0x0020 0000 0000 0000 0000 0000 0000 0000 .............. ... so Windows' Host-Uniq is "0010 5a00 d3de" and FreeBSD's is "0480 7067 c3". The RFC states this value can be anything the Host chooses and it is not interpreted by the Access Concentrator. Interestingly the RFC also states that the AC MUST (!) include the Host-Uniq value in any PADO/PADS replies, which isn't happening in my capture. So I think the AC's PADO continues in another frame which tcpdump isn't showing me. This would explain why we don't see the AC-Name tag being sent but FreeBSD knows the AC-Name in it's PADR. This is a side issue perhaps, it still doesn't explain why the AC doesn't understand FreeBSD's inital PADI. Sorry if this is information overload or heading down the wrong track. I'm just trying to understand the problem. Aaron _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. Share information about yourself, create your own public profile at http://profiles.msn.com. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message