From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932333Ab1JDOua (ORCPT ); Tue, 4 Oct 2011 10:50:30 -0400 Received: from smtp105.prem.mail.ac4.yahoo.com ([76.13.13.44]:25052 "HELO smtp105.prem.mail.ac4.yahoo.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S932141Ab1JDOu3 (ORCPT ); Tue, 4 Oct 2011 10:50:29 -0400 X-Yahoo-Newman-Property: ymail-3 X-YMail-OSG: Um4QIsIVM1l0sQtNL_Y1NdGrwYRY1vyINSSQzHoL4NCEH9B 7wyN9Dm2HoRGif.0k3H5XyfZlBvAIkcL5F5QlRYYmrqA_X1BkM5NbQ57PijW o8ZngrI8w63xaNwlMp5adv9TNTpotLx9OoVemW6h2ha.aidYq_fOZaJ8.e6C HikNr6.juwaLy0dqFLesVOmtjn3_b2pRQg_5gN8ijOFFQpKS81xYRFUluQoW S0TAYOlE5t58LMr1cbqYC4JazP_70JoxC7geB_agUMt6oEGXW57LTy3SxDKu n16wpiKAiS7CpVcPKGBhQ_2Sf5jbzr9Tkap28w6cj5Ct_zwbOc3AzX1cySXG 0IbXz3gYLjIklsHNWJe2ya.iK9nzIEmQnsSCY7oAjlVz4x_Pt1sFckdiBTk9 nPxvmcjS_.OIFCyfjN4raKk4- X-Yahoo-SMTP: _Dag8S.swBC1p4FJKLCXbs8NQzyse1SYSgnAbY0- Date: Tue, 4 Oct 2011 09:50:25 -0500 (CDT) From: Christoph Lameter X-X-Sender: cl@router.home To: Peter Zijlstra cc: "Paul E. McKenney" , Sitsofe Wheeler , Ingo Molnar , linux-kernel@vger.kernel.org, penberg@kernel.org, mpm@selenic.com, linux-mm@kvack.org, David Rientjes Subject: Re: lockdep recursive locking detected (rcu_kthread / __cache_free) In-Reply-To: <1317739225.32543.9.camel@twins> Message-ID: References: <20111003175322.GA26122@sucs.org> <20111003203139.GH2403@linux.vnet.ibm.com> <20111003214739.GK2403@linux.vnet.ibm.com> <1317739225.32543.9.camel@twins> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) 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 On Tue, 4 Oct 2011, Peter Zijlstra wrote: > It could of course be I got confused and broke stuff instead, could > someone who knows slab (I guess that's either Pekka, Christoph or David) > stare at those patches? Why is the loop in init_lock_keys only running over kmalloc caches and not over all slab caches? It seems that this has to be especially applied to regular slab caches because those are the ones that mostly have off slab structures. So modify init_lock_keys to run over all slab caches?