From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756630AbXF0BjU (ORCPT ); Tue, 26 Jun 2007 21:39:20 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753122AbXF0BjM (ORCPT ); Tue, 26 Jun 2007 21:39:12 -0400 Received: from ns2.suse.de ([195.135.220.15]:33187 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752520AbXF0BjL (ORCPT ); Tue, 26 Jun 2007 21:39:11 -0400 Date: Wed, 27 Jun 2007 03:39:03 +0200 From: Nick Piggin To: Randy Dunlap Cc: Andrew Morton , Dave Kleikamp , Adrian Bunk , linux-kernel@vger.kernel.org Subject: Re: [-mm patch] remove nobh_{prepare,commit}_write() Message-ID: <20070627013903.GA21268@wotan.suse.de> References: <20070606220313.8f7c1fab.akpm@linux-foundation.org> <20070614221555.GY3588@stusta.de> <20070626133203.d1aa5383.akpm@linux-foundation.org> <1182890938.11569.27.camel@kleikamp.austin.ibm.com> <20070626142320.9336df5b.akpm@linux-foundation.org> <20070626143335.684887f9.randy.dunlap@oracle.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20070626143335.684887f9.randy.dunlap@oracle.com> User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Jun 26, 2007 at 02:33:35PM -0700, Randy Dunlap wrote: > On Tue, 26 Jun 2007 14:23:20 -0700 Andrew Morton wrote: > > > On Tue, 26 Jun 2007 15:48:58 -0500 > > Dave Kleikamp wrote: > > > > > On Tue, 2007-06-26 at 13:32 -0700, Andrew Morton wrote: > > > > On Fri, 15 Jun 2007 00:15:55 +0200 > > > > Adrian Bunk wrote: > > > > > > > > > nobh_{prepare,commit}_write() are no longer used. > > > > > > > > wth? What happened to ext2 and ext3 nobh mode? They seem to > > > > have magically and unchangeloggedly disappeared? > > > > > > They were removed with Nick's new aops patches. > > ^secretly > > > > > > > That much I worked out for myself. It's kinda staggering that a fairly > > major feature in two fairly major filesystems got removed without even a > > mention in the changelog. I don't recall having seen it discussed in email > > but I obviously missed that bit. > > > > Look, I'm one micron from just dropping the whole lot. These changes > > simply have not received the amount of energy, effort, care, attention and > > testing which a change of this magnitude requires. > > so be sure to discuss that (not the patches themselves so much, > but the process(es)) at the kernel summit etc.... I did of course mention that nobh wasn't converted when sending the patches. I asked for comments about how much it is used in real world. Badari was the only one who replied about that but we didn't reach a conclusion. I don't know about energy, but I have seen lots of other patches cause a lot more problems...