Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933258Ab1EYBQT (ORCPT ); Tue, 24 May 2011 21:16:19 -0400 Received: from c-67-188-111-120.hsd1.ca.comcast.net ([67.188.111.120]:34510 "EHLO saboo.goop.org" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752177Ab1EYBQS (ORCPT ); Tue, 24 May 2011 21:16:18 -0400 X-Greylist: delayed 3601 seconds by postgrey-1.27 at vger.kernel.org; Tue, 24 May 2011 21:16:18 EDT Message-ID: <4DDC2C91.6060408@goop.org> Date: Tue, 24 May 2011 15:09:21 -0700 From: Jeremy Fitzhardinge User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.17) Gecko/20110428 Fedora/3.1.10-1.fc15 Lightning/1.0b3pre Thunderbird/3.1.10 MIME-Version: 1.0 To: NetDev CC: Linux Kernel Mailing List Subject: 2.6.39: ipv6 tcp connections hanging on transmit X-Enigmail-Version: 1.1.2 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1607 Lines: 39 When I use ssh over ipv6 to connect to my machines, it works fine for command-line interaction until there's a large output (ps alxw, ps -l, etc), whereupon the connection hangs. It appears to be about when it would send an MTU-sized packet. Logging in with ssh over ipv4 works fine. The outputting processes terminate OK, but they leave queued data in the socket: # netstat Proto Recv-Q Send-Q Local Address Foreign Address State [...] tcp 0 5776 2001:470:a816:0:230:48f:ssh 2001:470:a816:0:e45e::47865 ESTABLISHED tcp 0 26688 2001:470:a816:0:230:48f:ssh 2001:470:a816:0:e45e::55428 ESTABLISHED tcp 0 0 2001:470:a816:0:230:48f:ssh 2001:470:a816:0:e45e::55426 ESTABLISHED tcp 0 5856 2001:470:a816:0:230:48f:ssh 2001:470:a816:0:e45e::55434 ESTABLISHED and the corresponding ssh sessions are hung forever. The path to the host is over openvpn bridged to the host's lan. The path MTU is 1500, verified with "traceroute6 --mtu" and "ping -s 1500" works. The network device is a bridge containing a single ethernet device: an Intel 82574L using the e1000e driver. I think this regression appeared in .38/.39, but unfortunately I haven't had a chance to go back and systematically test older kernels yet. Any clues, suggestions, further information I can provide, etc? Thanks, J -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/