Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753868AbZKTOsw (ORCPT ); Fri, 20 Nov 2009 09:48:52 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753834AbZKTOsv (ORCPT ); Fri, 20 Nov 2009 09:48:51 -0500 Received: from e8.ny.us.ibm.com ([32.97.182.138]:38569 "EHLO e8.ny.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753823AbZKTOsv (ORCPT ); Fri, 20 Nov 2009 09:48:51 -0500 Date: Fri, 20 Nov 2009 06:48:55 -0800 From: "Paul E. McKenney" To: Pekka Enberg Cc: Peter Zijlstra , linux-mm@kvack.org, cl@linux-foundation.org, mpm@selenic.com, LKML , Nick Piggin Subject: Re: lockdep complaints in slab allocator Message-ID: <20091120144855.GB22527@linux.vnet.ibm.com> Reply-To: paulmck@linux.vnet.ibm.com References: <20091118181202.GA12180@linux.vnet.ibm.com> <84144f020911192249l6c7fa495t1a05294c8f5b6ac8@mail.gmail.com> <1258709153.11284.429.camel@laptop> <84144f020911200238w3d3ecb38k92ca595beee31de5@mail.gmail.com> <1258714328.11284.522.camel@laptop> <4B067816.6070304@cs.helsinki.fi> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <4B067816.6070304@cs.helsinki.fi> User-Agent: Mutt/1.5.15+20070412 (2007-04-11) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1781 Lines: 36 On Fri, Nov 20, 2009 at 01:05:58PM +0200, Pekka Enberg wrote: > Peter Zijlstra kirjoitti: >> On Fri, 2009-11-20 at 12:38 +0200, Pekka Enberg wrote: >>> >>> On Fri, Nov 20, 2009 at 11:25 AM, Peter Zijlstra >>> wrote: >>>> 2) propagate the nesting information and user spin_lock_nested(), given >>>> that slab is already a rat's nest, this won't make it any less obvious. >>> spin_lock_nested() doesn't really help us here because there's a >>> _real_ possibility of a recursive spin lock here, right? >> Well, I was working under the assumption that your analysis of it being >> a false positive was right ;-) >> I briefly tried to verify that, but got lost and gave up, at which point >> I started looking for ways to annotate. > > Uh, ok, so apparently I was right after all. There's a comment in > free_block() above the slab_destroy() call that refers to the comment above > alloc_slabmgmt() function definition which explains it all. > > Long story short: ->slab_cachep never points to the same kmalloc cache > we're allocating or freeing from. Where do we need to put the > spin_lock_nested() annotation? Would it be enough to just use it in > cache_free_alien() for alien->lock or do we need it in cache_flusharray() > as well? Hmmm... If the nc->lock spinlocks are always from different slabs (as alloc_slabmgmt()'s block comment claims), why not just give each array_cache structure's lock its own struct lock_class_key? They are zero size unless you have lockdep enabled. Thanx, Paul -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/