From mboxrd@z Thu Jan 1 00:00:00 1970 From: Takuya Yoshikawa Subject: Re: PROBLEM: in 2.6.25.3 -- reply to your suggestions Date: Wed, 14 May 2008 17:13:29 +0900 Message-ID: <482A9F29.3040805@oss.ntt.co.jp> References: <4828E995.3080509@oss.ntt.co.jp> <2EB68762-7B3A-4E0E-A1AB-4581C6011E10@adaptec.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from serv2.oss.ntt.co.jp ([222.151.198.100]:35571 "EHLO serv2.oss.ntt.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1761635AbYENINZ (ORCPT ); Wed, 14 May 2008 04:13:25 -0400 In-Reply-To: <2EB68762-7B3A-4E0E-A1AB-4581C6011E10@adaptec.com> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Mark Salyzyn Cc: "linux-scsi@vger.kernel.org" , AACRAID , "linux-kernel@vger.kernel.org" , =?ISO-8859-1?Q?Fernando_Luis_V=E1zquez_Cao?= Reply to your suggestions, 1. with the kernel parameters, **aacraid.expose_physicals=0, aacraid.dacmode=0**: I tried both parameters, but failed. 2. about "[PATCH] Stop using num_physpages in aacraid": I am not using this patch: I saw drivers/scsi/aacraid/aachba.c (kernel-2.6.25.3) 3. about other kernels Our machine normally runs with kernel-2.6.24 and this failure occurs only with kernel-2.6.25 and later, including 2.6.26-rc1 Sincerely, ------------------------------------------------------------- Takuya Yoshikawa @ NTT Open Source Software Center e-mail: yoshikawa.takuya@oss.ntt.co.jp ------------------------------------------------------------- > Not a panic, merely a problem probing the physical devices. It may > seem 'forever' but that is what happens when it must issue commands to > every target and lun on two physical busses. You may be able to > mitigate this with the kernel parameter aacraid.expose_physicals=0, or > aacraid.dacmode=0 > > Status 8193 (0x2003) is not a known valid response code from the > Firmware, it is a result of the Firmware not processing the command in > the first place. > > Could be a result of 'Subject: Re: [PATCH] Stop using num_physpages in > aacraid '? Could you revert that one out to see? Could you instrument > the return value to the new referenced function > dma_get_required_mask(&dev->pdev->dev) that is incorporated in that > patch? > > Sincerely -- Mark Salyzyn