From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: Problems with /proc/net/tcp6 - possible bug - ipv6 Date: Mon, 24 Jan 2011 14:31:41 -0800 (PST) Message-ID: <20110124.143141.104045281.davem@davemloft.net> References: <115440.16148.qm@web63902.mail.re1.yahoo.com> <1295731235.2651.66.camel@edumazet-laptop> <1295732444.2651.68.camel@edumazet-laptop> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: runningdoglackey@yahoo.com, linux-kernel@vger.kernel.org, netdev@vger.kernel.org To: eric.dumazet@gmail.com Return-path: In-Reply-To: <1295732444.2651.68.camel@edumazet-laptop> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org =46rom: Eric Dumazet Date: Sat, 22 Jan 2011 22:40:44 +0100 > Le samedi 22 janvier 2011 =E0 22:20 +0100, Eric Dumazet a =E9crit : >> Le samedi 22 janvier 2011 =E0 11:42 -0800, PK a =E9crit : >> > Eric Dumazet wrote: >> > >=20 >> > > I had some incidents, after hours of testing... >> > > >> > > After following patch, I could not reproduce it. >> >=20 >> >=20 >> > Looks like that patch solved the /proc/net/tcp6 problem. The caus= al commit was=20 >> > the one you identified... confirmed with bisect. >> >=20 >> > These warnings show up when I run the script (or I presume any tcp= 6 connection=20 >> > flooder) with /proc/sys/net/tcp/ipv4/tcp_tw_recycle enabled. Ther= e's textual=20 >> > corruption of the traces a lot of the time. Here's a sample trace= that doesn't=20 >> > appear to be corrupt. All the warnings I've seen are from route.c= :209, and I=20 >> > don't see how that would cause memory corruption. >>=20 >> Thats a different isse, already reported, under investigation. >>=20 >> David did some changes recently >>=20 >> http://comments.gmane.org/gmane.linux.network/179874 >>=20 >>=20 >>=20 >=20 > In my testings, I even have crashes in cleanup_once() if I > enable /proc/sys/net/ipv4/tcp_tw_recycle=20 I'm looking into this, thanks guys.