Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265466AbTFRRvU (ORCPT ); Wed, 18 Jun 2003 13:51:20 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265468AbTFRRvU (ORCPT ); Wed, 18 Jun 2003 13:51:20 -0400 Received: from gateway-1237.mvista.com ([12.44.186.158]:39924 "EHLO hermes.mvista.com") by vger.kernel.org with ESMTP id S265466AbTFRRvP (ORCPT ); Wed, 18 Jun 2003 13:51:15 -0400 Subject: Re: [PATCH] 2.5.72 O(1) interactivity bugfix From: Robert Love To: Con Kolivas Cc: linux kernel mailing list , Ingo Molnar In-Reply-To: <200306190043.14291.kernel@kolivas.org> References: <200306190043.14291.kernel@kolivas.org> Content-Type: text/plain Message-Id: <1055959503.7069.1864.camel@localhost> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.0 (1.4.0-2) Date: 18 Jun 2003 11:05:05 -0700 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 876 Lines: 23 On Wed, 2003-06-18 at 07:43, Con Kolivas wrote: > While messing with the interactivity code I found what appears to be an > uninitialised variable (p->sleep_avg), which is responsible for all the > boost/penalty in the scheduler. The variable isn't uninitialized, it is inherited from the parent. Recall that the task_struct of the new child starts off as a duplicate of the parent's. If the variables are not explicitly set, they remain the same as the parent's. So this patch ends up resetting the sleep_avg to zero, instead of being the same as the parent, which may or may not result in better performance. Robert Love - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/