From owner-freebsd-questions Tue Apr 9 14:23:34 1996 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id OAA15238 for questions-outgoing; Tue, 9 Apr 1996 14:23:34 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id OAA15233 for ; Tue, 9 Apr 1996 14:23:31 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id OAA06015; Tue, 9 Apr 1996 14:15:43 -0700 From: Terry Lambert Message-Id: <199604092115.OAA06015@phaeton.artisoft.com> Subject: Re: Artisoft AE-3 Ethernet card: How to use large buffer? To: Brett_Glass@ccgate.infoworld.com (Brett Glass) Date: Tue, 9 Apr 1996 14:15:42 -0700 (MST) Cc: questions@FreeBSD.org In-Reply-To: <9603098290.AA829070379@ccgate.infoworld.com> from "Brett Glass" at Apr 9, 96 09:11:24 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-questions@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk > I'm running FreeBSD on a machine with an Artisoft AE-3 card -- an NE2000 > clone with 64K (rather than the usual 16K) of buffer space. How can I get > FreeBSD to recognize, and use, this additional space for transmit (and, > especially, receive) buffering? The kernel source, which is almost > uncommented, suggests that the OS will assume an NE2000-compatible > card has 16K of buffering. There's also some conditionaly compiled code > having to do with a Gateway Ethernet card. It's not normally compiled in, > but it looks as if it *may* test the buffer size. > > Will the Gateway code enable the use of the full 64K of buffer space? If > so, is the code tested and reliable? Finally, if both of the above are > true, what is the best way to cause it to be compiled in? Will an "options > GWETHER" in the kernel configuration file work? You could always try it. I don't think it will work, since it's unlikely to have compatable gate array settings. You need to set mem_end to mem_start + 64k. This should be done in whichever probe routine is coming true. Hopefully you can provide a way to uniquely and non-destructively identify the card, and just submit a probe code patch. This assumes that, unlike some cards, there is not only an 8k window to get at the memeory (in which case you want to leave it alone). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.