Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758159AbZFKTRE (ORCPT ); Thu, 11 Jun 2009 15:17:04 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752311AbZFKTQy (ORCPT ); Thu, 11 Jun 2009 15:16:54 -0400 Received: from gir.skynet.ie ([193.1.99.77]:56662 "EHLO gir.skynet.ie" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751093AbZFKTQy (ORCPT ); Thu, 11 Jun 2009 15:16:54 -0400 Date: Thu, 11 Jun 2009 20:16:54 +0100 From: Mel Gorman To: Christoph Lameter Cc: Pekka Enberg , linux-kernel@vger.kernel.org, Larry.Finger@lwfinger.net Subject: Re: [PATCH 2/2] SLUB: Disable debugging if it increases the minimum page order Message-ID: <20090611191653.GK7302@csn.ul.ie> References: <1244728824.17483.51.camel@penberg-laptop> <1244733119.17483.56.camel@penberg-laptop> <1244734086.17483.59.camel@penberg-laptop> <20090611180422.GJ7302@csn.ul.ie> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-15 Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.17+20080114 (2008-01-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1559 Lines: 36 On Thu, Jun 11, 2009 at 02:26:02PM -0400, Christoph Lameter wrote: > On Thu, 11 Jun 2009, Mel Gorman wrote: > > > "With SLUB_DEBUG, enable debug on all caches unless the increased meta-data > > would force the minimum order up on order due to object sizes being near or > > equal the PAGE_SIZE. If SLUB_DEBUG must be enabled, specify slub_debug=A for > > 'All caches enable debug regardless'" > > Hmmm... No SLUB_DEBUG just compiles the code in for debugging. Does not > enable anything. > > But we could make SLUB_DEBUG_ON enable debugging except on those slabs > where the order increases (for which it does a printk saying that the > debugging was disabled) and then have a slub_debug=A option that > overrides the order check. Sorry, this is what I meant and I failed to explain myself properly. > Generally if you know that you have memory > corruption you would want to boot with slub_debug=A. If you just want > continuous checks then SLUB_DEBUG_ON would do the trick. > That's the general idea. It would catch a number of bugs that SLUB_DEBUG_ON catches without creating a different class of bug report such as this atomic-order-1 problem. -- Mel Gorman Part-time Phd Student Linux Technology Center University of Limerick IBM Dublin Software Lab -- 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/