From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Garzik Subject: Re: Marvell 6121 AHCI: hotplug not reliable? Date: Thu, 13 Aug 2009 09:19:13 -0400 Message-ID: <4A8412D1.2070506@garzik.org> References: <4A8406EE.1040203@aquazul.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from srv5.dvmed.net ([207.36.208.214]:43103 "EHLO mail.dvmed.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754391AbZHMNTP (ORCPT ); Thu, 13 Aug 2009 09:19:15 -0400 In-Reply-To: <4A8406EE.1040203@aquazul.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Mourad De Clerck Cc: linux-ide@vger.kernel.org Mourad De Clerck wrote: > Hello, > > I've got a box with an Asus M2V motherboard and a Marvell 6121 SATA > controller. Kernel version 2.6.30. > > It works in general, but from time to time it stops detecting the disks > when I hotplug them. So usually hotplug works, but sometimes it just > stops working. > > If I unload the ahci module, and reload it (with the marvell_enable=1 > parm), it detects the disks again and works without any further issues - > until the next failed hotplug. > > Is there anything that can be done about this? > > In attachment, the kernel log with at the end (at [128508.539280]) a > failed hotplug, after which I unload/reload the ahci module and things > work again (starting at [147111.353555]). > [128508.539280] ata1: exception Emask 0x10 SAct 0x0 SErr 0x0 action 0xe frozen > [128508.539306] ata1: irq_stat 0x02400000, PHY RDY changed > [128508.539333] ata1: hard resetting link > [128514.300013] ata1: link is slow to respond, please be patient (ready=0) > [128518.556015] ata1: COMRESET failed (errno=-16) > [128518.556040] ata1: hard resetting link > [128524.320015] ata1: link is slow to respond, please be patient (ready=0) > [128528.576014] ata1: COMRESET failed (errno=-16) > [128528.576039] ata1: hard resetting link > [128534.340014] ata1: link is slow to respond, please be patient (ready=0) > [128563.628015] ata1: COMRESET failed (errno=-16) > [128563.628041] ata1: limiting SATA link speed to 1.5 Gbps > [128563.628062] ata1: hard resetting link > [128568.664014] ata1: COMRESET failed (errno=-16) > [128568.664038] ata1: reset failed, giving up > [128568.664059] ata1: EH complete hmmmm. Since PHY_RDY-leads-to-failure is a very rare report for this chip, my initial guess would be indicative of either a power, SATA port, or SATA cable problem. PHY_RDY is essentially an indication of whether or not the host-device SATA connection is alive and transmitting SATA packets (FIS's). If PHY_RDY changes, it means the hardware -thinks- that the SATA connection changed. Jeff