From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: Problem with pata_amd/sata_nv on ASRock Penryn1600SLI-110dB board Date: Thu, 06 Mar 2008 17:32:50 +0900 Message-ID: <47CFAC32.6000103@gmail.com> References: <20080224225322.061dc654@alkaid.a.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from wa-out-1112.google.com ([209.85.146.182]:28997 "EHLO wa-out-1112.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750740AbYCFIc6 (ORCPT ); Thu, 6 Mar 2008 03:32:58 -0500 Received: by wa-out-1112.google.com with SMTP id v27so2129858wah.23 for ; Thu, 06 Mar 2008 00:32:57 -0800 (PST) In-Reply-To: <20080224225322.061dc654@alkaid.a.lan> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Andreas Bierfert Cc: linux-ide@vger.kernel.org Andreas Bierfert wrote: > Hi there, > > I got the above mentioned board [1] two weeks ago with a Q6600 cpu. I am running > Fedora 8 x86_64. While using 2.6.23, the attached cdrom/dvd drives went all > crazy together with a nice error message of the hdd from time to time. > > After switching to 2.6.24 the cd drives were useful again but the hdd error > still occurs, effectively in the worst case making the system fs unusable and > requiring a hard reset. The error seems to appear randomly and I cannot > reproduce it by causing e.g. high disk usage... > > See http://lowlatency.de/kbug/{dmesg,error,lspci} for the 2.6.23 and > http://lowlatency.de/kbug/{dmesg,error,lspci}-2.6.24 for the corresponding logs. > > I first thought the bug was related to the pata drives so I switched some drives > around to only have sata hdds and only the cd/dvd drives on the pata controller > (see http://lowlatency.de/kbug/dmesg-newest). Still the same error (just other > ata numbers of course) for the hdd. Just to note => it also is a different > hdd so I would rule out defective hardware. Current setup is with these > harddisks http://lowlatency.de/kbug/hdparamI and I am still seeing the error. > > Also tried with noapic parameter set at boot time. > > Would be cool if anyone has some clues. If more information or tests are needed > let me know. > > Best Regards, > Andreas Bierfert > > [1] - http://www.asrock.com/mb/overview.asp?Model=Penryn1600SLI-110dB Does running "hal-disable-polling /dev/sr0" and "hal-disable-polling /dev/sr1" after boot remove the problems and keep the drives usable? -- tejun