From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTP id 9B4DCDDDF4 for ; Wed, 5 Dec 2007 23:11:49 +1100 (EST) Received: from root by ciao.gmane.org with local (Exim 4.43) id 1IzslO-0005mc-Ml for linuxppc-dev@ozlabs.org; Wed, 05 Dec 2007 11:50:02 +0000 Received: from artipc9.vub.ac.be ([134.184.26.59]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 05 Dec 2007 11:50:02 +0000 Received: from fhimpe by artipc9.vub.ac.be with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for ; Wed, 05 Dec 2007 11:50:02 +0000 To: linuxppc-dev@ozlabs.org From: Frederik Himpe Subject: Re: [PATCH v2] Fix hardware IRQ time accounting problem. Date: Wed, 5 Dec 2007 11:39:25 +0000 (UTC) Message-ID: References: <20071204055144.GR24243@bakeyournoodle.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Sender: news Cc: linux-kernel@vger.kernel.org, linux-usb-devel@lists.sourceforge.net List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Tue, 04 Dec 2007 16:51:44 +1100, Tony Breeds wrote: > The commit fa13a5a1f25f671d084d8884be96fc48d9b68275 (sched: restore > deterministic CPU accounting on powerpc), unconditionally calls > update_process_tick() in system context. In the deterministic > accounting case this is the correct thing to do. However, in the > non-deterministic accounting case we need to not do this, and results in > the time accounted as hardware irq time being artificially elevated. > > Also this patch collapses 2 consecutive '#ifdef > CONFIG_VIRT_CPU_ACCOUNTING' checks in time.h into one for neatness. I can confirm this patch fixed the problem for me. Thank you! -- Frederik Himpe From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752298AbXLELjt (ORCPT ); Wed, 5 Dec 2007 06:39:49 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751442AbXLELjl (ORCPT ); Wed, 5 Dec 2007 06:39:41 -0500 Received: from main.gmane.org ([80.91.229.2]:52176 "EHLO ciao.gmane.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751432AbXLELjk (ORCPT ); Wed, 5 Dec 2007 06:39:40 -0500 X-Injected-Via-Gmane: http://gmane.org/ To: linux-kernel@vger.kernel.org From: Frederik Himpe Subject: Re: [PATCH v2] Fix hardware IRQ time accounting problem. Date: Wed, 5 Dec 2007 11:39:25 +0000 (UTC) Message-ID: References: <20071204055144.GR24243@bakeyournoodle.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-Complaints-To: usenet@ger.gmane.org X-Gmane-NNTP-Posting-Host: artipc9.vub.ac.be User-Agent: Pan/0.132 (Waxed in Black) Cc: linuxppc-dev@ozlabs.org, linux-usb-devel@lists.sourceforge.net Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, 04 Dec 2007 16:51:44 +1100, Tony Breeds wrote: > The commit fa13a5a1f25f671d084d8884be96fc48d9b68275 (sched: restore > deterministic CPU accounting on powerpc), unconditionally calls > update_process_tick() in system context. In the deterministic > accounting case this is the correct thing to do. However, in the > non-deterministic accounting case we need to not do this, and results in > the time accounted as hardware irq time being artificially elevated. > > Also this patch collapses 2 consecutive '#ifdef > CONFIG_VIRT_CPU_ACCOUNTING' checks in time.h into one for neatness. I can confirm this patch fixed the problem for me. Thank you! -- Frederik Himpe