From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751191AbWGaV7q (ORCPT ); Mon, 31 Jul 2006 17:59:46 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751384AbWGaV7q (ORCPT ); Mon, 31 Jul 2006 17:59:46 -0400 Received: from gprs189-60.eurotel.cz ([160.218.189.60]:40881 "EHLO amd.ucw.cz") by vger.kernel.org with ESMTP id S1751191AbWGaV7q (ORCPT ); Mon, 31 Jul 2006 17:59:46 -0400 Date: Mon, 31 Jul 2006 23:59:33 +0200 From: Pavel Machek To: nathans@sgi.com, kernel list , "Rafael J. Wysocki" Subject: XFS vs. swsusp Message-ID: <20060731215933.GB3612@elf.ucw.cz> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline X-Warning: Reading this can be dangerous to your mental health. User-Agent: Mutt/1.5.11+cvs20060126 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi! Rafael has patches to add bdev freezing to swsusp. I'd like to know if they are neccessary (and why). 1) Is sync() enough to guarantee that all the data written before sync actually reach the platters? (Or is it that data only reach the journal? OTOH that would be okay, too). 2) If we stop all the user proceses and all the kernel threads, is that enough to prevent XFS from writing to disk? Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html