From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755387Ab0IBPgj (ORCPT ); Thu, 2 Sep 2010 11:36:39 -0400 Received: from mail-gx0-f174.google.com ([209.85.161.174]:40144 "EHLO mail-gx0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751279Ab0IBPgi (ORCPT ); Thu, 2 Sep 2010 11:36:38 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:cc:content-type; b=Jmfx/gG7xs57eSe2lLrGgHIqp39ylYu3AvP3u1qCevsfbawpTneRvq1pzovGulVyri Cc5aXGebRtrqX0Hexh9FZa9MTs/Wef8nPWe45cYIB/ZodgrYEDD4A6AM2qz/Sg441bxZ Y2OLvlmNLqPTlZxrt4jH0BWjNH8vitjbpbvkg= MIME-Version: 1.0 Date: Thu, 2 Sep 2010 17:36:35 +0200 Message-ID: Subject: Re: [PATCH 00/20] mm: Preemptibility -v4 From: Mat To: peterz@infradead.org Cc: Linux Kernel , piotr@hosowicz.com Content-Type: text/plain; charset=ISO-8859-1 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Peter, I'm currently also testing your preemptibility patchset and have to say that on top of 2.6.36-rc3 it really seems to make a positive difference for me :) since I'm only a regular (power)user utilizing linux-distributions for my everyday work I can report only about that: - the desktop stays very responsive even during copying, backing up and transferring of large amounts of data (e.g. backing up 1 TB of data via rsync between different filesystems); this includes the GUI staying responsive, the mouse not lagging and apps opening up almost instantaneously - there is very little to no noticable lags or repetitions when hearing webradio or audios during longer periods of time - and much more considering that 2.6.36 received several other improvements in this area - http://lkml.org/lkml/2010/8/1/40, the patches from Wu Fengguang and KOSAKI Motohiro - concurrency managed workqueues - and other performance improvements, especially VFS may also add up to this anyways: please consider adding this to next/mainline during the next merge cycle my linux-desktop didn't feel that reponsive for a long time ! ^^ if I recall correctly if felt comparably responsive when I used Con's RSDL or SD scheduler - CFS and CFQ have made great progress in terms of desktop usage, interactivity and delays Regards and thanks to everyone involved in the continued improvements in this area Mat