From owner-freebsd-current@FreeBSD.ORG Sat May 26 14:01:52 2012 Return-Path: Delivered-To: current@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 74A0C106566B; Sat, 26 May 2012 14:01:52 +0000 (UTC) (envelope-from bz@FreeBSD.org) Received: from mx1.sbone.de (bird.sbone.de [46.4.1.90]) by mx1.freebsd.org (Postfix) with ESMTP id 10B4E8FC1A; Sat, 26 May 2012 14:01:52 +0000 (UTC) Received: from mail.sbone.de (mail.sbone.de [IPv6:fde9:577b:c1a9:31::2013:587]) (using TLSv1 with cipher ADH-CAMELLIA256-SHA (256/256 bits)) (No client certificate requested) by mx1.sbone.de (Postfix) with ESMTPS id EFBB525D389C; Sat, 26 May 2012 14:01:50 +0000 (UTC) Received: from content-filter.sbone.de (content-filter.sbone.de [IPv6:fde9:577b:c1a9:31::2013:2742]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mail.sbone.de (Postfix) with ESMTPS id 31F30BE7C68; Sat, 26 May 2012 14:01:50 +0000 (UTC) X-Virus-Scanned: amavisd-new at sbone.de Received: from mail.sbone.de ([IPv6:fde9:577b:c1a9:31::2013:587]) by content-filter.sbone.de (content-filter.sbone.de [fde9:577b:c1a9:31::2013:2742]) (amavisd-new, port 10024) with ESMTP id bNlmr7QW++0C; Sat, 26 May 2012 14:01:48 +0000 (UTC) Received: from orange-en1.sbone.de (orange-en1.sbone.de [IPv6:fde9:577b:c1a9:31:cabc:c8ff:fecf:e8e3]) (using TLSv1 with cipher AES128-SHA (128/128 bits)) (No client certificate requested) by mail.sbone.de (Postfix) with ESMTPSA id B4A96BE7C67; Sat, 26 May 2012 14:01:48 +0000 (UTC) Mime-Version: 1.0 (Apple Message framework v1084) Content-Type: text/plain; charset=us-ascii From: "Bjoern A. Zeeb" In-Reply-To: Date: Sat, 26 May 2012 14:01:47 +0000 Content-Transfer-Encoding: quoted-printable Message-Id: <9CBC841B-6339-41BC-A91B-83D47B84B9B4@FreeBSD.org> References: To: net@freebsd.org X-Mailer: Apple Mail (2.1084) Cc: current@FreeBSD.org Subject: Re: WARNING - DO NOT test: IPv6 offload support in HEAD + patch for stable/9 X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 26 May 2012 14:01:52 -0000 On 25. May 2012, at 16:55 , Bjoern A. Zeeb wrote: > Hey, >=20 > last night I pushed in the essential offloading support changes for = IPv6 > along with quite a bit of other "noise" into HEAD. There is more = locking > improvements etc. to come once I have looped things back to my working = tree > and Michael Tuexen will improve SCTP/v6 on loopback as well soonish. >=20 > This is a call for testing. The in-tree cxgb(4) and ixgbe(4) drivers WARNING - please refrain from testing IPv6 or updating your HEAD if you = do not have any of the above two NICs and rely on IPv6, or if you have = updated and are experiencing problems. Disabling -txcsum -tso for the moment should = be an often unhelpful workaround. It seems I was just lucky with my choice of = other NICs I tested but I cannot say which once are affected in the tree and = which once aren't. Andrew Gallatin has pointed out that I missed an essential IPv4 header = parsing thing beyond TSO in some (most?) NIC drivers and it went unnoticed in = review. I'll post an update in a few hours once I know how many drivers are = affected, or have the proper fix as it might also be a question in which = (old/cheap) silicon might not do what is expected. > have been updated to make use of the new features (TSO6/LRO6), and = more > drivers will follow (I already have cxgbe done, talking about mxge, = ..) > but others should also see improvements for at least upper layer = protocol > checksum calculations and I'd love people to test with as many drivers = as > possible, as I plan to merge it for the upcoming 9.1-RELEASE cycle and > wouldn't want to ship broken IPv6 in a few months;-) >=20 > Here's the patch that should just apply to stable/9 matching what I = put into > HEAD (+ an earlier cxgb change) (untested): >=20 > http://people.freebsd.org/~bz/20120525-01-ipv6-offload-mfc9.diff >=20 > If you need a patch for a specific release please drop me a private = email > and I'll try to publish one (8.2 and up only though most likely). >=20 >=20 > Please test and report to me or net@. >=20 >=20 > Thanks > /bz >=20 >=20 > PS: we now also disallow LRO automatically if forwarding is turned on, = just > in case you wonder; a change that should have been done years ago;-) --=20 Bjoern A. Zeeb You have to have visions! It does not matter how good you are. It matters what good you do!