From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759580Ab1EMQhC (ORCPT ); Fri, 13 May 2011 12:37:02 -0400 Received: from mga01.intel.com ([192.55.52.88]:18167 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753337Ab1EMQhA (ORCPT ); Fri, 13 May 2011 12:37:00 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.64,365,1301900400"; d="scan'208";a="1876256" From: Andi Kleen To: James Bottomley Cc: x86@kernel.org, linux-mm , linux-kernel , Mel Gorman Subject: Re: Possible sandybridge livelock issue References: <1305303156.2611.51.camel@mulgrave.site> Date: Fri, 13 May 2011 09:36:21 -0700 In-Reply-To: <1305303156.2611.51.camel@mulgrave.site> (James Bottomley's message of "Fri, 13 May 2011 11:12:36 -0500") Message-ID: User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/23.2 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org James Bottomley writes: > > When the hang occurred, kswapd basically pegged one core in 100% system > time. This looks like there's something specific to sandybridge that > causes this type of bad interaction. I was wondering if it could be > something to to with a scheduling problem in turbo mode? Once kswapd > goes flat out, the core its on will kick into turbo mode, which causes > it to get preferentially scheduled there, leading to the live lock. Sounds unlikely to me. Turbo mode does not affect the scheduler and the cores are (reasonably) independent. > The only evidence I have to support this theory is that when I reproduce > the problem with PREEMPT, the core pegs at 100% system time and stays > there even if I turn off the load. However, if I can execute work that > causes kswapd to be kicked off the core it's running on, it will calm > back down and go to sleep. Turbo mode just makes the CPU faster, but it should not change the scheduler decisions. -Andi -- ak@linux.intel.com -- Speaking for myself only From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail190.messagelabs.com (mail190.messagelabs.com [216.82.249.51]) by kanga.kvack.org (Postfix) with SMTP id 301C36B0023 for ; Fri, 13 May 2011 12:36:49 -0400 (EDT) From: Andi Kleen Subject: Re: Possible sandybridge livelock issue References: <1305303156.2611.51.camel@mulgrave.site> Date: Fri, 13 May 2011 09:36:21 -0700 In-Reply-To: <1305303156.2611.51.camel@mulgrave.site> (James Bottomley's message of "Fri, 13 May 2011 11:12:36 -0500") Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-linux-mm@kvack.org List-ID: To: James Bottomley Cc: x86@kernel.org, linux-mm , linux-kernel , Mel Gorman James Bottomley writes: > > When the hang occurred, kswapd basically pegged one core in 100% system > time. This looks like there's something specific to sandybridge that > causes this type of bad interaction. I was wondering if it could be > something to to with a scheduling problem in turbo mode? Once kswapd > goes flat out, the core its on will kick into turbo mode, which causes > it to get preferentially scheduled there, leading to the live lock. Sounds unlikely to me. Turbo mode does not affect the scheduler and the cores are (reasonably) independent. > The only evidence I have to support this theory is that when I reproduce > the problem with PREEMPT, the core pegs at 100% system time and stays > there even if I turn off the load. However, if I can execute work that > causes kswapd to be kicked off the core it's running on, it will calm > back down and go to sleep. Turbo mode just makes the CPU faster, but it should not change the scheduler decisions. -Andi -- ak@linux.intel.com -- Speaking for myself only -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: email@kvack.org