From owner-freebsd-stable@FreeBSD.ORG Fri May 1 13:20:11 2009 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 96EFF106566C; Fri, 1 May 2009 13:20:11 +0000 (UTC) (envelope-from petefrench@ticketswitch.com) Received: from constantine.ticketswitch.com (constantine.ticketswitch.com [IPv6:2002:57e0:1d4e:1::3]) by mx1.freebsd.org (Postfix) with ESMTP id 59F0D8FC1A; Fri, 1 May 2009 13:20:11 +0000 (UTC) (envelope-from petefrench@ticketswitch.com) Received: from dilbert.rattatosk ([10.64.50.6] helo=dilbert.ticketswitch.com) by constantine.ticketswitch.com with esmtps (TLSv1:AES256-SHA:256) (Exim 4.69 (FreeBSD)) (envelope-from ) id 1Lzsev-000Cif-Ad; Fri, 01 May 2009 14:20:09 +0100 Received: from petefrench by dilbert.ticketswitch.com with local (Exim 4.69 (FreeBSD)) (envelope-from ) id 1Lzsev-000Ofm-92; Fri, 01 May 2009 14:20:09 +0100 To: bug-followup@freebsd.org Message-Id: From: Pete French Date: Fri, 01 May 2009 14:20:09 +0100 Cc: freebsd-stable@freebsd.org Subject: Re: kern/133756: [bce] bce commit r190582 breaks lagg in 7.2-PRERELASE X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 01 May 2009 13:20:11 -0000 This is just a quick update about some further investigations on this. I tested out the patch that Niki Denev kindly sent me which apparently fixes a length issue when zero copy sockets are not in use. This did not, however, solve the problem, but as part of this I ran tcpdump on the bce0 and bce1 interfaces for the working and non working kernels. On the kernel which does not work I never see any packets being received, though LACP packets are being transmitted. The switch to which the devices are connected is (I believe) configured in a mode where it will only send LACP packets back when it has received some. So the lack of incomming pa ckets in tcpdump does not necessarily mean that the recive side is failing, as it may be that the transmitted packets are not making it out of the interface. On identical hardware with no LACP in place the bce interfaces work fine. What I have yet to try is runnign a simple lagg bundle on this machine, to see if the problem is specific to LACP. I also rolled the system back to 7.1 and tested a patch from Doug Ambrisko which adds the support for the 5709 but contains none of the other changes. This did work, so that has excluded those changes from the cause of the problems. -pete.