From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751633Ab1GYMh5 (ORCPT ); Mon, 25 Jul 2011 08:37:57 -0400 Received: from ozlabs.org ([203.10.76.45]:46738 "EHLO ozlabs.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751431Ab1GYMh1 (ORCPT ); Mon, 25 Jul 2011 08:37:27 -0400 X-Mailbox-Line: From anton@samba.org Mon Jul 25 12:34:26 2011 Message-Id: <20110725023426.307093692@samba.org> User-Agent: quilt/0.48-1 Date: Mon, 25 Jul 2011 12:33:15 +1000 From: Anton Blanchard To: mingo@elte.hu, peterz@infradead.org, benh@kernel.crashing.org Cc: linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org Subject: [PATCH 4/5] powerpc/numa: Disable NEWIDLE balancing at node level References: <20110725023311.175792493@samba.org> Content-Disposition: inline; filename=nodes_no_newidle Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On big POWER7 boxes we see large amounts of CPU time in system processes like workqueue and watchdog kernel threads. We currently rebalance the entire machine each time a task goes idle and this is very expensive on large machines. Disable newidle balancing at the node level and rely on the scheduler tick to rebalance across nodes. Signed-off-by: Anton Blanchard --- Index: linux-2.6-work/arch/powerpc/include/asm/topology.h =================================================================== --- linux-2.6-work.orig/arch/powerpc/include/asm/topology.h 2011-07-25 12:14:25.448671947 +1000 +++ linux-2.6-work/arch/powerpc/include/asm/topology.h 2011-07-25 12:14:26.568692651 +1000 @@ -75,7 +75,7 @@ static inline int pcibus_to_node(struct .forkexec_idx = 0, \ \ .flags = 1*SD_LOAD_BALANCE \ - | 1*SD_BALANCE_NEWIDLE \ + | 0*SD_BALANCE_NEWIDLE \ | 1*SD_BALANCE_EXEC \ | 1*SD_BALANCE_FORK \ | 0*SD_BALANCE_WAKE \