Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Tue, 5 Nov 2002 13:13:24 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Tue, 5 Nov 2002 13:12:10 -0500 Received: from to-velocet.redhat.com ([216.138.202.10]:34546 "EHLO touchme.toronto.redhat.com") by vger.kernel.org with ESMTP id ; Tue, 5 Nov 2002 13:12:03 -0500 Date: Tue, 5 Nov 2002 13:18:39 -0500 From: Benjamin LaHaise To: John Gardiner Myers Cc: linux-aio@kvack.org, Linux Kernel Mailing List , lse-tech@lists.sourceforge.net Subject: Re: pipe POLLOUT oddity Message-ID: <20021105131839.B3934@redhat.com> References: <20021031230215.GA29671@bjl1.asuk.net> <20021101020119.GC30865@bjl1.asuk.net> <3DC30DED.6040207@netscape.com> <20021102045535.GA21356@mark.mielke.cc> <20021031230215.GA29671@bjl1.asuk.net> <3DC80AC5.6080705@netscape.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i In-Reply-To: <3DC80AC5.6080705@netscape.com>; from jgmyers@netscape.com on Tue, Nov 05, 2002 at 10:15:33AM -0800 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 810 Lines: 18 On Tue, Nov 05, 2002 at 10:15:33AM -0800, John Gardiner Myers wrote: > I could see this going either way, depending on the application. > Holding off the POLLOUT readiness could improve performance by making > sure that whenever a process is scheduled to write to a pipe the pipe > has enough buffer to take all of the data. Aio write to pipes has a distinct advantage here as the pipe code can provide the write atomicity guarantees while preserving the non-blocking aspect of the io submission interface. -ben -- "Do you seek knowledge in time travel?" - 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/