Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932134AbVLLScO (ORCPT ); Mon, 12 Dec 2005 13:32:14 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932142AbVLLScO (ORCPT ); Mon, 12 Dec 2005 13:32:14 -0500 Received: from api.pobox.com ([208.210.124.75]:24015 "EHLO thorn.pobox.com") by vger.kernel.org with ESMTP id S932138AbVLLScN (ORCPT ); Mon, 12 Dec 2005 13:32:13 -0500 Date: Mon, 12 Dec 2005 13:32:01 -0500 From: Nathan Lynch To: Linus Torvalds Cc: Benjamin Herrenschmidt , Andrew Morton , Linux Kernel list , Paul Mackerras , Jens Axboe , Brian King Subject: Re: Memory corruption & SCSI in 2.6.15 Message-ID: <20051212183201.GA19599@localhost.localdomain> References: <1134371606.6989.95.camel@gaston> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4.2.1i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1415 Lines: 31 Linus Torvalds wrote: > > On Mon, 12 Dec 2005, Benjamin Herrenschmidt wrote: > > > > Current -git as of today (that is 2.6.15-rc5 + the batch of fixes Linus > > pulled after his return) was dying in weird ways for me on POWER5. I had > > the good idea to activate slab debugging, and I now see it detecting > > slab corruption as soon as the IPR driver initializes. > > > > Since I remember seeing a discussion somewhere on a list between Brian > > King and Jens Axboe about use-after-free problems in SCSI and possible > > other niceties of that sort, I though it might be related... > > > > Anything I can do to help track this down ? > > If it's easily repeatable, doing a "git bisect" to see when it starts > happening is the obvious big sledge-hammer thing to try. Even if you don't > bisect all the way, just narrowing it down a bit more might help. I manually narrowed this down to between 2.6.14-git14 (good) and 2.6.15-rc1 (bad). Will try git bisection later today. > There's a raid1 use-after-free bugfix that I just merged and pushed out, > but I doubt that one is relevant. But you might try to update. FWIW, I'm hitting this in a non-raid setup. - 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/