From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756102Ab3FEOyH (ORCPT ); Wed, 5 Jun 2013 10:54:07 -0400 Received: from merlin.infradead.org ([205.233.59.134]:54196 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755974Ab3FEOyD (ORCPT ); Wed, 5 Jun 2013 10:54:03 -0400 Date: Wed, 5 Jun 2013 16:53:13 +0200 From: Jens Axboe To: hari krishnan Cc: linux-kernel@vger.kernel.org Subject: Re: Bug found in block/scsi_ioctl.c Message-ID: <20130605145313.GC32665@kernel.dk> References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Jun 05 2013, hari krishnan wrote: > I got a kernel crash, and dump is pointing to "IP: [] > sg_scsi_ioctl+0x166/0x3b0". > > Test case: *I was just deleting few NPIV port to which some of LUN's where > exported, after sometime I was running parted on some devices"* > > The crash is in sg_scsi_ioctl of block/scsi_ioctl.c. > From code and dump it seems sg_scsi_ioctl is allocating request by calling > blk_get_request and assuming it will never fail, but it will fail if device > is dead, here in this case some of SCSI devices are dead. It's a known bug, Joe Lawrence has posted a patch to fix it up that will be muscled into 3.10. Can you give it a go? http://www.spinics.net/lists/linux-scsi/msg66294.html -- Jens Axboe