From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bernhard Kohl Subject: Re: [Qemu-devel] lsi_scsi assertion Date: Fri, 17 Sep 2010 15:16:53 +0200 Message-ID: <4C936A45.2050203@nsn.com> References: <4C922233.6030602@dlh.net> <4C9352BA.3090808@nsn.com> <1494275E-AE25-47E7-B800-739B40C336F1@dlh.net> <4C935F81.7040103@siemens.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: Peter Lieven , ext Kevin Wolf , qemu-devel@nongnu.org, kvm@vger.kernel.org To: ext Jan Kiszka Return-path: Received: from demumfd002.nsn-inter.net ([93.183.12.31]:12791 "EHLO demumfd002.nsn-inter.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752738Ab0IQNRZ (ORCPT ); Fri, 17 Sep 2010 09:17:25 -0400 In-Reply-To: <4C935F81.7040103@siemens.com> Sender: kvm-owner@vger.kernel.org List-ID: Am 17.09.2010 14:30, schrieb ext Jan Kiszka: > Am 17.09.2010 14:26, Peter Lieven wrote: > >> Am 17.09.2010 um 13:36 schrieb Bernhard Kohl: >> >> >>> Am 16.09.2010 15:57, schrieb ext Peter Lieven: >>> >>>> Hi, >>>> >>>> I found the following assertion in my log files after a system reset was executed >>>> >>>> kvm: lsi_scsi: error: ORDERED queue not implemented >>>> last message repeated 5 times >>>> kvm: lsi_scsi: error: Unimplemented message 0x0d >>>> kvm: lsi_scsi: error: Unimplemented message 0x0c >>>> kvm: lsi_scsi: error: Unimplemented message 0x0d >>>> kvm: qemu-kvm-0.12.4.1: /usr/src/qemu-kvm-0.12.4/hw/lsi53c895a.c:596: lsi_reselect: Assertion `s->current == ((void *)0)' failed. >>>> kvm: Aborted >>>> kvm errno=134 >>>> >>>> The guest is a 64-bit ubuntu 10.04 server. Qemu-KVM is 0.12.4. >>>> >>>> Sounds this familiar to anyone? >>>> >>>> BR, >>>> Peter >>>> >>> For messages 0x0c and 0x0d I recently submitted a patch: >>> http://thread.gmane.org/gmane.comp.emulators.qemu/79660 >>> >> thanks. >> >> >>> ORDERED queue is still missing. >>> The assertion needs to be investigated. What guest OS? >>> >> ubuntu 10.04.1 lts 64-bit >> >> > Already checked if even the abort persists with 0.13-rc1? Just to > exclude that this is only a missing back-port. > > Jan > Maybe the assertion is a consequence of the improper handling of messages 0x0c and 0x0d which are SCSI ABORT messages.