From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262443AbTJXSJy (ORCPT ); Fri, 24 Oct 2003 14:09:54 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262441AbTJXSJx (ORCPT ); Fri, 24 Oct 2003 14:09:53 -0400 Received: from camus.xss.co.at ([194.152.162.19]:48399 "EHLO camus.xss.co.at") by vger.kernel.org with ESMTP id S262439AbTJXSJs (ORCPT ); Fri, 24 Oct 2003 14:09:48 -0400 Message-ID: <3F996AE7.5020409@xss.co.at> Date: Fri, 24 Oct 2003 20:09:43 +0200 From: Andreas Haumer Organization: xS+S User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3) Gecko/20030312 X-Accept-Language: en-us, en MIME-Version: 1.0 To: "Moore, Eric Dean" CC: linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org, mpt_linux_developer@lsil.com Subject: Re: [PATCH] 2.4.23-pre8 driver udpate for MPT Fusion (2.05.10) References: <0E3FA95632D6D047BA649F95DAB60E57035A944F@exa-atlanta.se.lsil.com> In-Reply-To: <0E3FA95632D6D047BA649F95DAB60E57035A944F@exa-atlanta.se.lsil.com> X-Enigmail-Version: 0.74.0.0 X-Enigmail-Supports: pgp-inline, pgp-mime Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi! Moore, Eric Dean wrote: > Here's a patch for 2.4.23-pre8 kernel for MPT Fusion driver, coming from LSI > Logic. > [...] > > (8) In a mixed device configuration (Ultra 320 and Non-Ultra > 320 devices are present on an HBA channel), code was fixed to prevent > negotiating for QAS on all Ultra 320 devices on that channel. > This keeps the Ultra 320 devices from dominating the SCSI bus, > preventing Non-Ultra 320 IO's from executing. > [...] Uh Oh... Is this a fix for the problems I reported back in June to LKML and to Pam Delaney? On June, 12th, I sent a mail to Pam reporting system freezes I had on an ASUS AP 1700-S5 server with onboard 53C1030 controller. In this mail I speculated about the possibility of a mixed device configuration beeing the source of the problem: [...] Another idea: is it possible that the SCSI controller has problems with U160 and U320 SCSI disks mixed together on the same bus? As far as I know there shouldn't be a problem with this combination, but who knows? I always tested with a mix of the following SCSI disks: IBM DDYS-T18350M (18GB U160 SCA) IBM IC35L036UCDY10-0 (36GB U320 SCA) I'm currently running 2.4.20 with 4 U320 disks and no U160 disk in the how swap cage. The system is now up for about 6 hours and no SCSI timeout occured so far (this doesn't say anything. Sometimes it takes 24 hours before a SCSI timeout or even a freeze occurs) [...] I haven't got an answer to this report, but with a all-U320 configuration I never had system freezes since. Does the new driver now work in a U160/U320 mixed configuration, too? - - andreas - -- Andreas Haumer | mailto:andreas@xss.co.at *x Software + Systeme | http://www.xss.co.at/ Karmarschgasse 51/2/20 | Tel: +43-1-6060114-0 A-1100 Vienna, Austria | Fax: +43-1-6060114-71 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.1 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQE/mWrjxJmyeGcXPhERAowMAJ9mTHr/qLObJ/OvUodLsrILx/MnkACfTj+U 7QVE6mUMmxAEYmIfvHeDItU= =DcTz -----END PGP SIGNATURE-----