From owner-freebsd-hackers Tue Jul 21 05:50:22 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id FAA13423 for freebsd-hackers-outgoing; Tue, 21 Jul 1998 05:50:22 -0700 (PDT) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from skynet.ctr.columbia.edu (skynet.ctr.columbia.edu [128.59.64.70]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id FAA13417 for ; Tue, 21 Jul 1998 05:50:15 -0700 (PDT) (envelope-from wpaul@skynet.ctr.columbia.edu) Received: (from wpaul@localhost) by skynet.ctr.columbia.edu (8.6.12/8.6.9) id IAA13651; Tue, 21 Jul 1998 08:53:00 -0400 From: Bill Paul Message-Id: <199807211253.IAA13651@skynet.ctr.columbia.edu> Subject: Re: Obtaining 3Com programming documentation To: hamada@astec.co.jp (HAMADA Naoki) Date: Tue, 21 Jul 1998 08:52:58 -0400 (EDT) Cc: hackers@FreeBSD.ORG In-Reply-To: <199807210916.SAA10730@stone.astec.co.jp> from "HAMADA Naoki" at Jul 21, 98 06:16:25 pm X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Of all the gin joints in all the towns in all the world, HAMADA Naoki had to walk into mine and say: > Bill Paul wrote: > >Anybody know the right place/person from which to obtain programming > >documentation for 3Com ethernet chips? > > See http://infodeli.3com.com/infodeli/services/online2.htm#facts and > try their interactive fax delivery service. Request an information > request form for developers. After you get it, fill out the form and > send it back via fax. This procedure is quite irritating, but their > response is not so bad! > > - nao I ended up calling tech support and from there was told about the faxback service. For the record, the document to request from the faxback menu is #9070. However, be advised that the version of the document there is out of date: it lists choices for ordering the 3c50x and 3c59x tech documents, but it says nothing about the 3c90x. The 3c90x documentation is available though: if it's not listed on the form, you can write in that you want it at the bottom of the sheet and fax it back. I filled out and faxed back the forms this past Friday, and lo and behond, they arrived on Monday via Airborne Express. Included with the bundle was an updated form which did indeed list the 3c90x documents in the tech briefs section. (It also had larger boxes for writing in your name and address, which is another big improvement: the original form had tiny little boxes that were much too small. The new form also has an entry at the bottom form the 3Com 3c985 gigabit ethernet adapter, but it says documentation isn't available yet ("no ETA"). :( Looking over the 3c90x book, I'm a little dismayed that they don't have a complete register summary anywhere: it looks like they describe all the important registers, but the descriptions are spead out all through the manual. There is a summary of those registers not described anywhere else, but I would have prefered an entire listing: the entire last third of the ThunderLAN manual contains a complete summary of all the available registers with short descriptions of each and what their various bits do, which is very handy when you're cranking code and just need to look something up quickly. I know now that the reason the Vortex driver doesn't work with the 3c905B chip is that they got rid of the PIO interface for transfering packets to/from the chip's on-board RAM: now you have to use the bus-master DMA mechanism. The Boomerang chip in the 3c905 supported bus-master DMA, but retained the PIO interface for compatibility. Now the PIO interface is totally gone. The DMA mechanism appears to work very much like that of the ThunderLAN (it has upload/download lists with fragment pointers, a forward pointer and a status word), which I suppose is not that surprising. The 3c905B's improvements over the 3c905 include a 64-bit multicast hash filter (previously you could only program the chip to receive all multicast packets) and hardware TCP/IP checksumming. This means you can get the chip to generate IP, TCP and UDP checksums on the way out and verify IP, TCP and UDP checksums on the way in. I'm not quite sure how to take advantage of this in the BSD TCP/IP structure, so I'm probably going to leave it alone for now. The manual states that it only works for IPv4 packets at this point. -Bill -- ============================================================================= -Bill Paul (212) 854-6020 | System Manager, Master of Unix-Fu Work: wpaul@ctr.columbia.edu | Center for Telecommunications Research Home: wpaul@skynet.ctr.columbia.edu | Columbia University, New York City ============================================================================= "It is not I who am crazy; it is I who am mad!" - Ren Hoek, "Space Madness" ============================================================================= To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message