From owner-freebsd-net Fri Oct 19 18:20: 5 2001 Delivered-To: freebsd-net@freebsd.org Received: from cage.simianscience.com (cage.simianscience.com [64.7.134.1]) by hub.freebsd.org (Postfix) with ESMTP id 83BB037B401 for ; Fri, 19 Oct 2001 18:20:01 -0700 (PDT) Received: (from root@localhost) by cage.simianscience.com (8.11.6/8.11.6) id f9K1K0K96595; Fri, 19 Oct 2001 21:20:00 -0400 (EDT) (envelope-from mike@sentex.net) Received: from chimp.sentex.net (fcage [192.168.0.2]) by cage.simianscience.com (8.11.6/8.11.6av) with ESMTP id f9K1Ju696587; Fri, 19 Oct 2001 21:19:56 -0400 (EDT) (envelope-from mike@sentex.net) Message-Id: <5.1.0.14.0.20011019211817.05a05890@192.168.0.12> X-Sender: mdtancsa@192.168.0.12 X-Mailer: QUALCOMM Windows Eudora Version 5.1 Date: Fri, 19 Oct 2001 21:19:54 -0400 To: Marko Zec From: Mike Tancsa Subject: Re: fxp driver - receive interrupt bundling Cc: freebsd-net@freebsd.org In-Reply-To: <3BD0D063.C084B21D@tel.fer.hr> References: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed X-Virus-Scanned: by AMaViS perl-10 Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org At 03:16 AM 10/20/2001 +0200, Marko Zec wrote: >It doesn't matter how many fxp cards you have installed - if your box acts >as a >server than most probably you can leave INT_DELAY at default value (Intel >proposes 0x600, but I think 0x400 would be more appropriate). > >If you use your multi-fxp-card BSD box as a router, than the microcode will >impose additional delay *twice* (once in each direction), so in that case the >default value of 0x600 might be too high for achieving full 100 megE >throughput, because of TCP windowing scheme having to wait for ACK frames, >which will be held in fxp receive buffers too long. On the other hand, setting >INT_DELAY too low minimizes the benefits of bundling interrupts, as fewer >received frames get "bundled" on a single interrupt. > >To summarize: if you are doing any routing (or bridging as I do), find the >best >value for INT_DELAY for your specific environment experimentally, it should be >definitely smaller than or equal to 0x400. If you don't do packet forwarding >between fxp interfaces, use the defaults. Thanks! This is for a router pushing upwards of 20Mb/s with about 110K routes on 4 interfaces. I will try and experiment a bit on my test setup first to see what works best. ---Mike -------------------------------------------------------------------- Mike Tancsa, tel +1 519 651 3400 Sentex Communications, mike@sentex.net Providing Internet since 1994 www.sentex.net Cambridge, Ontario Canada www.sentex.net/mike To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message