From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: 2.6.22.6 sata_sil device errors & timeouts Date: Mon, 17 Sep 2007 11:18:49 +0200 Message-ID: <46EE4679.7030600@gmail.com> References: <76366b180709170135v6c46f3d6qd3586ea79f9abc2@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from rv-out-0910.google.com ([209.85.198.190]:10939 "EHLO rv-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753310AbXIQJUo (ORCPT ); Mon, 17 Sep 2007 05:20:44 -0400 Received: by rv-out-0910.google.com with SMTP id k20so1254850rvb for ; Mon, 17 Sep 2007 02:20:44 -0700 (PDT) In-Reply-To: <76366b180709170135v6c46f3d6qd3586ea79f9abc2@mail.gmail.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Andrew Paprocki Cc: linux-ide@vger.kernel.org Andrew Paprocki wrote: > 1) When built with modules disabled, and libata handling the sata + > pata (AMD CS5536) connections, the pata drives come _after_ the sata > drives (i.e. w/ 2 sata drives, the first IDE drive is sdc). This makes > boot configuration more complicated if booting off the pata drive. Is > there any way to control which order the drives are assigned when not > building w/ modules? Please use mount-by-LABEL or UUID. > 2) The drives display that they support udma6 in hdparm -I, but only > udma5 is being used. And hdparm -i only shows up to udma2.. ? For SATA, UDMA mode doesn't matter at all. As long as you're in DMA mode, the only thing that matters is PHY link speed and whether NCQ is enabled. > ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x2400000 action 0x2 frozen > ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x280000 action 0x0 In both cases, SError is indicating transmission problem. Handshake error and Unrecognized FIS type in the first case, 10b to 8b decode error and CRC error on the second case. I can't tell why but signals flying through those redish cables are getting corrupted. There have been quite a few cases of bad PSU causing transmission failures on SATA or you might have a bad controller and/or cables. The best way to debug this kind of problem is by elimination - by swapping hardware piece by piece you can find out which one is causing the problem. -- tejun