From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751568AbbBULTN (ORCPT ); Sat, 21 Feb 2015 06:19:13 -0500 Received: from casper.infradead.org ([85.118.1.10]:57700 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751181AbbBULTK (ORCPT ); Sat, 21 Feb 2015 06:19:10 -0500 Date: Sat, 21 Feb 2015 12:19:04 +0100 From: Peter Zijlstra To: Nicolas Pitre Cc: linux-kernel@vger.kernel.org, mingo@kernel.org, rjw@rjwysocki.net, tglx@linutronix.de Subject: Re: [PATCH 27/35] sched/idle: Use explicit broadcast oneshot control function Message-ID: <20150221111904.GF23367@worktop.ger.corp.intel.com> References: <20150216121435.203983131@infradead.org> <20150216122413.591402191@infradead.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.22.1 (2013-10-16) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Feb 20, 2015 at 07:56:17PM -0500, Nicolas Pitre wrote: > On Mon, 16 Feb 2015, Peter Zijlstra wrote: > > > From: Thomas Gleixner > > > > Replace the clockevents_notify() call with an explicit function call. > > > > Signed-off-by: Thomas Gleixner > > This patch makes my test system hang solid after letting it sit idle for > 5 to 15 minutes. Reverting it and this issue goes away. > > The explicit function call is not an equivalent replacement. In > clockevents_notify() the clockevents_lock is held across the call to > tick_broadcast_enter() or tick_broadcast_exit(). This patch drops the > locking. Yeah, the Changelog introducing the explicit call states as much. I'll have to go figure out why that matters here.