From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752642Ab2LTNlS (ORCPT ); Thu, 20 Dec 2012 08:41:18 -0500 Received: from 173-166-109-252-newengland.hfc.comcastbusiness.net ([173.166.109.252]:37869 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751744Ab2LTNlJ (ORCPT ); Thu, 20 Dec 2012 08:41:09 -0500 Message-ID: <50D31546.7020402@kernel.dk> Date: Thu, 20 Dec 2012 14:40:22 +0100 From: Jens Axboe MIME-Version: 1.0 To: Arthur Marsh CC: shli@kernel.org, neilb@suse.de, linux-kernel@vger.kernel.org Subject: Re: git-bisect points to block: discard granularity might not be power of 2 References: <5806f82593938f8c5ec753ef9f1f27cbc5508ddc@webmail.internode.on.net> In-Reply-To: <5806f82593938f8c5ec753ef9f1f27cbc5508ddc@webmail.internode.on.net> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 2012-12-20 14:33, Arthur Marsh wrote: > Hi, > one of my machines is locking up fairly early in the boot process (about 1.4 seconds in) with recent kernels and a git-bisect lead to: > commit 8dd2cb7e880d2f77fba53b523c99133ad5054cfd > > Author: Shaohua Li > Date: Fri Dec 14 11:15:36 2012 +0800 > > block: discard granularity might not be power of 2 > > In MD raid case, discard granularity might not be power of 2, for example, a > 4-disk raid5 has 3*chunk_size discard granularity. Correct the calculation for > such cases. > > Reported-by: Neil Brown > Signed-off-by: Shaohua Li > Signed-off-by: Jens Axboe > > The machine is a dual core Pentium IV with 2 IDE and one SCSI disk. I've tried pulling the SCSI card out and the problem still occurs. > > There are no software raid arrays assembled in the boot process. > > Short of taking a photo of the monitor with the dump on it I'm a bit stuck as to what to do to further identify the problem. Should be fixed in Linus' tree as-of yesterday. So please try that and report back if it doesn't work. -- Jens Axboe