From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vlad Yasevich Subject: Re: [PATCH] IPv6: Implement RFC 4429 Optimistic Duplicate Address Detection Date: Fri, 02 Feb 2007 17:05:14 -0500 Message-ID: <45C3B59A.6010802@hp.com> References: <20070129213013.GA26841@hmsreliant.homelinux.net> <20070130.072536.129403235.yoshfuji@linux-ipv6.org> <20070130130208.GA3723@hmsreliant.homelinux.net> <20070131.011629.84005028.yoshfuji@linux-ipv6.org> <20070131205443.GA12237@hmsreliant.homelinux.net> <20070202190634.GA29001@hmsreliant.homelinux.net> <45C3ACE1.8010800@hp.com> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Cc: Neil Horman , YOSHIFUJI Hideaki / ???? , sri@us.ibm.com, davem@davemloft.net, kuznet@ms2.inr.ac.ru, pekkas@netcore.fi, jmorris@namei.org, kaber@coreworks.de, netdev@vger.kernel.org To: Brian Haley Return-path: Received: from atlrel9.hp.com ([156.153.255.214]:53382 "EHLO atlrel9.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1946113AbXBBWFR (ORCPT ); Fri, 2 Feb 2007 17:05:17 -0500 In-Reply-To: <45C3ACE1.8010800@hp.com> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Brian Haley wrote: > Hi Neil, > >> @@ -830,7 +836,8 @@ retry: >> ift = !max_addresses || >> ipv6_count_addresses(idev) < max_addresses ? >> ipv6_add_addr(idev, &addr, tmp_plen, >> - ipv6_addr_type(&addr)&IPV6_ADDR_SCOPE_MASK, >> IFA_F_TEMPORARY) : NULL; >> + ipv6_addr_type(&addr)&IPV6_ADDR_SCOPE_MASK, >> + IFA_F_TEMPORARY|IFA_F_OPTIMISTIC) : NULL; > Hi Brian > So why are you always adding these as optimistic now? Shouldn't this be > triggering off idev->cnf.optimistic_dad? I know you're clearing it in > ipv6_add_addr(), but I liked Vlad's suggestion of not setting it > initially since this way seems backwards. The troubling case seems to manually configured addresses (inet6_addr_add()). If we can clearly and easily distinguish between this case of address and all the other ones, then we can simply set the flag in ipv6_add_addr, like we set the tentative flag. So, we can introduce another parameter to ipv6_add_addr() or another flag that can distinguish manual config. Otherwise, we can keep the code as is, passing the optimistic flag from needed callers, and clearing it inside ipv6_add_addr(). My thought was to clear it from the 'flags' parameter before ifa->flags was set, but that doesn't really matter. -vlad