From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752082AbbELAIJ (ORCPT ); Mon, 11 May 2015 20:08:09 -0400 Received: from sabe.cs.wisc.edu ([128.105.6.20]:38684 "EHLO sabe.cs.wisc.edu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751735AbbELAIG (ORCPT ); Mon, 11 May 2015 20:08:06 -0400 Message-ID: <5551445A.2050103@cs.wisc.edu> Date: Mon, 11 May 2015 19:07:54 -0500 From: Mike Christie User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.2.0 MIME-Version: 1.0 To: Chris Leech , linux-kernel@vger.kernel.org, Konrad Rzeszutek Wilk CC: open-iscsi@googlegroups.com, linux-scsi@vger.kernel.org Subject: Re: [PATCH] iscsi_ibft: filter null v4-mapped v6 addresses References: <1431062071-7074-1-git-send-email-cleech@redhat.com> In-Reply-To: <1431062071-7074-1-git-send-email-cleech@redhat.com> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 05/08/2015 12:14 AM, Chris Leech wrote: > I've had reports of UEFI platforms failing iSCSI boot in various > configurations, that ended up being caused by network initialization > scripts getting tripped up by unexpected null addresses (0.0.0.0) being > reported for gateways, dhcp servers, and dns servers. > > The tianocore EDK2 iSCSI driver generates an iBFT table that always uses > IPv4-mapped IPv6 addresses for the NIC structure fields. This results > in values that are "not present or not specified" being reported as > ::ffff:0.0.0.0 rather than all zeros as specified. > > The iscsi_ibft module filters unspecified fields from the iBFT from > sysfs, preventing userspace from using invalid values and making it easy > to check for the presence of a value. This currently fails in regard to > these mapped null addresses. > > In order to remain consistent with how the iBFT information is exposed, > we should accommodate the behavior of the tianocore iSCSI driver as it's > already in the wild in a large number of servers. > > Tested under qemu using an OVMF build of tianocore EDK2. > > Signed-off-by: Chris Leech Looks ok to me. Reviewed-by: Mike Christie