Received: with ECARTIS (v1.0.0; list netdev); Tue, 09 Nov 2004 06:57:23 -0800 (PST) Received: from yue.st-paulia.net (yue.linux-ipv6.org [203.178.140.15]) by oss.sgi.com (8.13.0/8.13.0) with ESMTP id iA9EvEuw028387 for ; Tue, 9 Nov 2004 06:57:15 -0800 Received: from localhost (localhost [127.0.0.1]) by yue.st-paulia.net (Postfix) with ESMTP id C8DFA33CE5; Tue, 9 Nov 2004 23:57:59 +0900 (JST) Date: Tue, 09 Nov 2004 09:57:59 -0500 (EST) Message-Id: <20041109.095759.109272072.yoshfuji@linux-ipv6.org> To: herbert@gondor.apana.org.au Cc: davem@davemloft.net, acme@conectiva.com.br, netdev@oss.sgi.com, jgarzik@pobox.com, buytenh@wantstofly.org, yoshfuji@linux-ipv6.org Subject: Re: Bug in ipv6_ifa_notify? From: YOSHIFUJI Hideaki / =?iso-2022-jp?B?GyRCNUhGIzFRTEAbKEI=?= In-Reply-To: <20041109114839.GA1942@gondor.apana.org.au> References: <20041108121040.GA3618@gondor.apana.org.au> <20041108203741.GA993@gondor.apana.org.au> <20041109114839.GA1942@gondor.apana.org.au> Organization: USAGI Project X-URL: http://www.yoshifuji.org/%7Ehideaki/ X-Fingerprint: 9022 65EB 1ECF 3AD1 0BDF 80D8 4807 F894 E062 0EEA X-PGP-Key-URL: http://www.yoshifuji.org/%7Ehideaki/hideaki@yoshifuji.org.asc X-Face: "5$Al-.M>NJ%a'@hhZdQm:."qn~PA^gq4o*>iCFToq*bAi#4FRtx}enhuQKz7fNqQz\BYU] $~O_5m-9'}MIs`XGwIEscw;e5b>n"B_?j/AkL~i/MEaZBLP X-Mailer: Mew version 2.2 on Emacs 20.7 / Mule 4.1 (AOI) Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-archive-position: 11632 X-ecartis-version: Ecartis v1.0.0 Sender: netdev-bounce@oss.sgi.com Errors-to: netdev-bounce@oss.sgi.com X-original-sender: yoshfuji@linux-ipv6.org Precedence: bulk X-list: netdev Content-Length: 1002 Lines: 23 In article <20041109114839.GA1942@gondor.apana.org.au> (at Tue, 9 Nov 2004 22:48:39 +1100), Herbert Xu says: > On Tue, Nov 09, 2004 at 07:37:41AM +1100, herbert wrote: > > > > Well I've found a way for this to occur, but I must say that it is > > fairly unlikely. When you bring the interface down, addrconf_ifdown > > will try to to delete the addrconf timer and then notify DELADDR. > > Since it doesn't wait for the timer to complete, it might still be > > executing. Therefore it is possible to have a NEWADDR after a DELADDR > > event. > > The attractive thing about this theory is that it is an SMP-only > race. This could explain why only Jeff/Lennert are seeing it. > > Jeff/Lennert, could you please go back to 2.6.9, and then apply > this patch? If this warning triggers, then that would confirm > this theory. I think this only happens when you shut down interface while DAD; for example, "ifup eth0 &; ifdown eth0" or something like that. --yoshfuji