From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Rafael J. Wysocki" Subject: Re: [PATCH 0/6 v2] PM / Hibernate: Memory bitmap scalability improvements Date: Tue, 29 Jul 2014 23:22:39 +0200 Message-ID: <3290133.D59AnCbavX@vostro.rjw.lan> References: <1405938422-21900-1-git-send-email-joro@8bytes.org> <20140722103443.GV30979@8bytes.org> <20140728135933.GA5918@pd.tnic> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7Bit Return-path: Received: from v094114.home.net.pl ([79.96.170.134]:60380 "HELO v094114.home.net.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1752030AbaG2VEH (ORCPT ); Tue, 29 Jul 2014 17:04:07 -0400 In-Reply-To: <20140728135933.GA5918@pd.tnic> Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: Borislav Petkov Cc: Joerg Roedel , Pavel Machek , Len Brown , linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org On Monday, July 28, 2014 03:59:33 PM Borislav Petkov wrote: > On Tue, Jul 22, 2014 at 12:34:44PM +0200, Joerg Roedel wrote: > > So my test was to resume from a swap partition that contained no image. > > Here is the result from the 16GB machine. First with a v3.16-rc6 kernel > > without my changes: > > > > kv:~/base # time perf record /usr/sbin/resume /dev/sda1 > > resume: libgcrypt version: 1.5.3 > > [ perf record: Woken up 1 times to write data ] > > [ perf record: Captured and wrote 0.019 MB perf.data (~823 samples) ] > > > > real 0m0.084s > > user 0m0.012s > > sys 0m0.064s > > > > Here is the result with my patches on top: > > > > kv:~/hibernate # time perf record /usr/sbin/resume /dev/sda1 > > resume: libgcrypt version: 1.5.3 > > [ perf record: Woken up 1 times to write data ] > > [ perf record: Captured and wrote 0.014 MB perf.data (~602 samples) ] > > > > real 0m0.032s > > user 0m0.003s > > sys 0m0.027s > > > > So we save around 50ms (or 62% of time) already on this 16GB machine. > > So, let's see, with Joerg's patches we > > - solve the issue on huge boxes. And yes, we most definitely want to be > able to suspend them too. RAS is one very prominent use case here. > > *and* > > - we see improvement on smaller boxes, above numbers look good to me. > > and all that for an additional 8K for a S/R cycle?! And for some > additional complexity of a radix tree which is self-contained, well > tested and understood? > > This looks like certainly like net win to me. Yes, it does. I've queued up the Joergs patches for 3.17, thanks! Rafael