From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: [Bugme-new] [Bug 14214] New: BUG at drivers/scsi/scsi_lib.c:1108! Date: Wed, 30 Sep 2009 21:19:24 +0000 Message-ID: <1254345564.3925.533.camel@mulgrave.site> References: <20090930125607.752808d6.akpm@linux-foundation.org> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20090930125607.752808d6.akpm@linux-foundation.org> Sender: linux-raid-owner@vger.kernel.org To: Andrew Morton Cc: "Rafael J. Wysocki" , Jens Axboe , bugzilla-daemon@bugzilla.kernel.org, linux-raid@vger.kernel.org List-Id: linux-raid.ids On Wed, 2009-09-30 at 12:56 -0700, Andrew Morton wrote: > On Wed, 23 Sep 2009 11:13:26 GMT > bugzilla-daemon@bugzilla.kernel.org wrote: > > > http://bugzilla.kernel.org/show_bug.cgi?id=14214 > > Post-2.6.31 regression. This, in scsi_setup_fs_cmnd(): > > /* > * Filesystem requests must transfer data. > */ > BUG_ON(!req->nr_phys_segments); Well, the BUG_ON is correct: there can't be a FS type command with no data. This can't be a SCSI issue because the problem was caused before SCSI was entered. Looking at the backtrace (all functions listed as unreliable, sigh) this is either a deadline scheduler problem (less likely) or an md_raid one (more likely). Let's start with what was the mdraid configuration and condition of the filesystem being mounted? I've added linux-raid to the cc list so they can chime in for more details. James