From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Sat, 4 Aug 2001 20:28:49 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Sat, 4 Aug 2001 20:28:39 -0400 Received: from marine.sonic.net ([208.201.224.37]:58998 "HELO marine.sonic.net") by vger.kernel.org with SMTP id ; Sat, 4 Aug 2001 20:28:33 -0400 X-envelope-info: Date: Sat, 4 Aug 2001 17:28:40 -0700 From: Mike Castle To: linux-kernel@vger.kernel.org Cc: Alexander Viro Subject: Re: fdatasync(2) is also there (was: intermediate summary of ext3-2.4-0.9.4 thread) Message-ID: <20010804172840.L437@thune.mrc-home.com> Reply-To: Mike Castle Mail-Followup-To: Mike Castle , linux-kernel@vger.kernel.org, Alexander Viro Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20010804054043.F16516@emma1.emma.line.org> User-Agent: Mutt/1.3.18i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Sat, Aug 04, 2001 at 05:40:43AM +0200, Matthias Andree wrote: > How portable is fsync()ing the directory? It is as portable as assuming that the directory entries are synced during fsync(). > > How USEFUL is it to the application if all other boxen fsync() the > directory entries along with the file? Except, you can NOT guarantee that this is the case. The spec does NOT require it. > You want to sync the files, but the directory only after you created all > of the files? Use fdatasync(2) for the files - it doesn't flush meta > data, then sync the directory. It's POSIX, it's SUS v2. Is fsync(dir) what you mean here? mrc -- Mike Castle dalgoda@ix.netcom.com www.netcom.com/~dalgoda/ We are all of us living in the shadow of Manhattan. -- Watchmen fatal ("You are in a maze of twisty compiler features, all different"); -- gcc