Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751272AbVLLVBD (ORCPT ); Mon, 12 Dec 2005 16:01:03 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751277AbVLLVBD (ORCPT ); Mon, 12 Dec 2005 16:01:03 -0500 Received: from gate.crashing.org ([63.228.1.57]:60895 "EHLO gate.crashing.org") by vger.kernel.org with ESMTP id S1751272AbVLLVBC (ORCPT ); Mon, 12 Dec 2005 16:01:02 -0500 Subject: Re: Memory corruption & SCSI in 2.6.15 From: Benjamin Herrenschmidt To: Linus Torvalds Cc: Andrew Morton , Linux Kernel list , Paul Mackerras , Jens Axboe , Brian King In-Reply-To: References: <1134371606.6989.95.camel@gaston> <1134419609.6989.116.camel@gaston> Content-Type: text/plain Date: Tue, 13 Dec 2005 07:57:59 +1100 Message-Id: <1134421080.6989.121.camel@gaston> Mime-Version: 1.0 X-Mailer: Evolution 2.2.3 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1042 Lines: 31 On Mon, 2005-12-12 at 12:51 -0800, Linus Torvalds wrote: > [ I applied Brian's patch, so hopefully it was the same issue, and current > git doesn't show this problem any more ] I'll be able to test in about 1h. > On Tue, 13 Dec 2005, Benjamin Herrenschmidt wrote: > > > > > Also, enabling DEBUG_PAGEALLOC might help, but that's not available on > > > powerpc. > > > > Remind me what is needed to get that working ? Unmapping of linear > > mapping pages ? (I suppose I could do that if I also disable using large > > pages for it). > > Yes, basically you'd have to allow the kernel mapping being unmapped one > page at a time. > > And yes, it's inefficient. Don't use it for performance measurements ;) Ok, I'll look into it. It's doable though it will definitely not be fast :) Ben. - 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/