From: Mike Galbraith Subject: Re: [patch 1/3] ipc/sem: fix -rt livelock Date: Fri, 04 Oct 2013 13:16:50 +0200 Message-ID: <1380885410.5432.5.camel@marge.simpson.net> References: <1379051751.5455.112.camel@marge.simpson.net> <1379052760.5455.127.camel@marge.simpson.net> <5234D94D.8010608@colorfullife.com> <1379220340.5545.93.camel@marge.simpson.net> <20131004104450.GB19953@linutronix.de> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: Manfred Spraul , linux-rt-users , Steven Rostedt , Thomas Gleixner , Peter Zijlstra To: Sebastian Andrzej Siewior Return-path: Received: from moutng.kundenserver.de ([212.227.17.9]:53463 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753104Ab3JDLQ5 (ORCPT ); Fri, 4 Oct 2013 07:16:57 -0400 In-Reply-To: <20131004104450.GB19953@linutronix.de> Sender: linux-rt-users-owner@vger.kernel.org List-ID: On Fri, 2013-10-04 at 12:44 +0200, Sebastian Andrzej Siewior wrote: > * Mike Galbraith | 2013-09-15 06:45:40 [+0200]: > > Hi, > > >On Sat, 2013-09-14 at 23:46 +0200, Manfred Spraul wrote: > >> Could you check that? > > > >I think you're right. > > > >> Or alternatively: Is my proposed sem_lock() function -rt friendly? > > > >Some way of making spin_unlock_wait() _go away_ along with the livelock > >would be better, but patches look good to me. I'll apply both and stare > >at the sum, and ask boxen what they think.. they're better at spotting > >locking troubles ;-) > > Do you post a new series of those three patches with the live lock > Manfred mentioned fixed or should I look at this serie? Manfred's race fix also kills loop, and thereby the -rt livelock, so the only question is does -rt want to use his completion wakeup scheme, and does it want to do something about spin_unlock_wait() unconditionally grabbing/releasing every lock in the array. -Mike