From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932103AbWBRSAl (ORCPT ); Sat, 18 Feb 2006 13:00:41 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932104AbWBRSAl (ORCPT ); Sat, 18 Feb 2006 13:00:41 -0500 Received: from gprs189-60.eurotel.cz ([160.218.189.60]:13442 "EHLO amd.ucw.cz") by vger.kernel.org with ESMTP id S932103AbWBRSAl (ORCPT ); Sat, 18 Feb 2006 13:00:41 -0500 Date: Sat, 18 Feb 2006 19:00:29 +0100 From: Pavel Machek To: Richard Mittendorfer Cc: linux-kernel@vger.kernel.org Subject: Re: [swsusp] not enough memory Message-ID: <20060218180029.GH1776@elf.ucw.cz> References: <20060218005814.6548092d.delist@gmx.net> <20060218151115.GC5658@openzaurus.ucw.cz> <20060218185759.248208f9.delist@gmx.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20060218185759.248208f9.delist@gmx.net> X-Warning: Reading this can be dangerous to your mental health. User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On So 18-02-06 18:57:59, Richard Mittendorfer wrote: > Also sprach Pavel Machek (Sat, 18 Feb 2006 16:11:16 > +0100): > > On Sat 18-02-06 00:58:14, Richard Mittendorfer wrote: > > > If I end all apps but the XServer it works. I've allready added some > > > more swapspace, but that didn't help. So, how much memory will I > > > need for a successful suspend or better (since i can't stuff any > > > more into it) is there a way to minimize the amount needed? > > > > 128MB should be enough for you. > > Swap? 128M was my first attempt -- OOM when the box was heavily loaded. > IIRC no effect on swsusp - there are 64M Ram + 2M Video + Kernel + X?. > The reason I did it were some google results talking about free pages. > There were some solutions talking about adding swap. > > (Ram? Would be the way to go, but it's fully loaded.) RAM. No ammount of swap will help you. > > Or try modifying try_to_free memory > > routine to retry shrink_all_mem few more times, with schedule() in > > between... > > Uhh. I don't think this will be a good idea /me ha[ck|rm]ing good ol' > linux source. ;-) It should be easy. It is quite hard to loose much data this way ;-). Pavel -- Web maintainer for suspend.sf.net (www.sf.net/projects/suspend) wanted...