From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753337AbaGVKzu (ORCPT ); Tue, 22 Jul 2014 06:55:50 -0400 Received: from atrey.karlin.mff.cuni.cz ([195.113.26.193]:41673 "EHLO atrey.karlin.mff.cuni.cz" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751823AbaGVKzs (ORCPT ); Tue, 22 Jul 2014 06:55:48 -0400 Date: Tue, 22 Jul 2014 12:55:46 +0200 From: Pavel Machek To: Joerg Roedel Cc: "Rafael J. Wysocki" , Len Brown , linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH 0/6 v2] PM / Hibernate: Memory bitmap scalability improvements Message-ID: <20140722105546.GA9814@amd.pavel.ucw.cz> References: <1405938422-21900-1-git-send-email-joro@8bytes.org> <20140721160346.GP30979@8bytes.org> <20140721230500.GA7019@amd.pavel.ucw.cz> <6793803.1FXN0a3O0V@vostro.rjw.lan> <20140722103443.GV30979@8bytes.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20140722103443.GV30979@8bytes.org> User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue 2014-07-22 12:34:44, Joerg Roedel wrote: > On Tue, Jul 22, 2014 at 02:41:29AM +0200, Rafael J. Wysocki wrote: > > It looks like some specific need motivated the Joerg's work, however, > > so let's just not dismiss the use case lightly without knowing it. > > The motivation was to optimize the data structures for machines with > large amounts of RAM without penalizing average machines. On a 12TB > machine you are close to 100000 pages just for one bitmap. Scanning > through that linearly to find a given bit just doesnt scale anymore in > this case. Writing out every single page on 12TB machine to disk does not scale, either :-). > I also see how the problem could be solved differently, but what I > didn't get from the discussion yet is: What is actually *wrong* with > *this* approach? It throws complex / tricky to review code at a problem... that is not a problem in any reasonable configuration. Now... should I spend half an hour reviewing your changes, or are we maybe better without them? > So we save around 50ms (or 62% of time) already on this 16GB machine. Or about 5 seeks or about 0.000% of total hibernation cycle. Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html