From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [NET]: Please revert disallowing zero listen queues Date: Tue, 06 Mar 2007 10:37:06 -0800 (PST) Message-ID: <20070306.103706.68156515.davem@davemloft.net> References: <200703061332.09677@strip-the-willow> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, weid@np.css.fujitsu.com To: gerrit@erg.abdn.ac.uk Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:35062 "EHLO sunset.davemloft.net" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S965991AbXCFShJ (ORCPT ); Tue, 6 Mar 2007 13:37:09 -0500 In-Reply-To: <200703061332.09677@strip-the-willow> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org From: Gerrit Renker Date: Tue, 6 Mar 2007 13:32:09 +0000 > Please can you reconsider the patch regarding the accept_queue > > http://git.kernel.org/?p=linux/kernel/git/davem/net-2.6.22.git;a=commit;h=8488df894d05d6fa41c2bd298c335f944bb0e401 > > It disallows to set a `backlog' argument to listen(2) of zero. Using > a zero backlog is often done (e.g. ttcp), and disallowing a zero > backlog will break many applications. I had to recode several applications > which rely on this convention. > > The problem further spreads from TCP to DCCP (same behaviour). > > Below is a patch to revert this change. Everything I've ever seen clearly states that a backlog of zero means that zero connections are allowed. So we're not "disallowing" a backlog argument of zero to listen(). We'll accept that just fine, the only thing that happens is that you'll get what you ask for, that being no connections :-)