From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Luis Claudio R. Goncalves" Subject: Re: Cyclictest expected results Date: Mon, 8 Oct 2007 11:55:52 -0300 Message-ID: <20071008145552.GC27098@unix.sh> References: <1985e0f60709300837l1b16deedgb15575bc73a729ad@mail.gmail.com> <20070930183253.GW23573@pengutronix.de> <1985e0f60709302030q62d19dd3q398a5d0685353786@mail.gmail.com> <20071002065451.GM23573@pengutronix.de> <1985e0f60710070550w47b1ccd7na897920cbd0aeb40@mail.gmail.com> <20071008145508.GA29081@thunk.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Jaswinder Singh , Robert Schwebel , Thomas Gleixner , linux-rt-users@vger.kernel.org To: Theodore Tso Return-path: Received: from tucuriba.hst.terra.com.br ([200.176.10.214]:40634 "EHLO tucuriba.hst.terra.com.br" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751914AbXJHPEy (ORCPT ); Mon, 8 Oct 2007 11:04:54 -0400 Received: from sr06-09.mta.terra.com.br (sr06-09.mta.terra.com.br [200.154.152.78]) by tucuriba.hst.terra.com.br (Postfix) with ESMTP id 77DC0275EF9 for ; Mon, 8 Oct 2007 12:04:52 -0300 (BRT) Content-Disposition: inline In-Reply-To: <20071008145508.GA29081@thunk.org> Sender: linux-rt-users-owner@vger.kernel.org List-Id: linux-rt-users.vger.kernel.org On Mon, Oct 08, 2007 at 10:55:09AM -0400, Theodore Tso wrote: | On Sun, Oct 07, 2007 at 06:20:29PM +0530, Jaswinder Singh wrote: | > | > But I am not able to see performance difference in my GNOME | > Environment. It is still very very slow, how can I boost X-Windows by | > -rt patch. | | There are many reasons why GNOME performance might be slow. Stupid | application design going to helped by some magic kernel patch. Keep | in mind the GNOME doesn't run with real-time priority levels; neither | does the X server. | | The -rt patch isn't magic; no RTOS is magic. A large amount of | application design and reworking in order to properly utilize the | POSIX real-time interfaces. Also keep in mind that by its very nature, -rt is more sensible to bad coding and bad application design. Some applications that perform well in mainstream linux may present bugs in -rt (those bugs exist in the code but are more prone to show uo in -rt). People here have been working most of time focused on determinism. Eventually it comes with a performance enhancement bonus. Sometimes not. Luis -- [ Luis Claudio R. Goncalves Bass - Gospel - RT ] [ Fingerprint: 4FDD B8C4 3C59 34BD 8BE9 2696 7203 D980 A448 C8F8 ]