From mboxrd@z Thu Jan 1 00:00:00 1970 From: Evgeniy Polyakov Subject: Re: Network hangs with 2.6.30.5 Date: Fri, 20 Nov 2009 15:04:52 +0300 Message-ID: <20091120120452.GA11462@ioremap.net> References: <20090907072143.GA5966@ff.dom.local> <20091001.154913.88345178.davem@davemloft.net> <20091119.154038.194307613.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: ilpo.jarvinen@helsinki.fi, jarkao2@gmail.com, holger.hoffstaette@googlemail.com, netdev@vger.kernel.org, eric.dumazet@gmail.com To: David Miller Return-path: Received: from cet.com.ru ([195.178.208.66]:50557 "EHLO tservice.net.ru" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1751887AbZKTMEs (ORCPT ); Fri, 20 Nov 2009 07:04:48 -0500 Content-Disposition: inline In-Reply-To: <20091119.154038.194307613.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: Hi. On Thu, Nov 19, 2009 at 03:40:38PM -0800, David Miller (davem@davemloft.net) wrote: > >> I also looked through all the TCP commits in 2.6.29 to 2.6.30 > >> and I could not find anything else that might cause stalls with > >> time-wait recycled connections. > > > > What about the more than 64k connections change a9d8f9110d7e953c2f2 (or > > its fixes), it might be another possibility? ...It certainly does > > something related to reuse and happens to be in the correct time frame... > > (I've added Evgeniy). > > So I've been studying this one quite a bit. > > This change could only cause problems in timewait recycling if it: > > 1) Would accept using a bind bucket that previously it would not > > 2) Caused socket corruption due to bad locking > > And I can find neither problem with the Evgeniy's bind changes. > > So we've back to square one I think. Wasn't solution found for this problem, I recall related discussion recently, or should I dig deeper into this one? -- Evgeniy Polyakov