From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Lord Subject: Re: sata_mv on Orion: CONFIG_SATA_PMP=y needed even when no multiplier Date: Mon, 12 May 2008 11:49:06 -0400 Message-ID: <482866F2.5040901@rtr.ca> References: <20080512132840.GA15711@deprecation.cyrius.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from rtr.ca ([76.10.145.34]:2523 "EHLO mail.rtr.ca" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751309AbYELPtI (ORCPT ); Mon, 12 May 2008 11:49:08 -0400 In-Reply-To: <20080512132840.GA15711@deprecation.cyrius.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Martin Michlmayr Cc: linux-ide@vger.kernel.org, nico@cam.org, buytenh@wantstofly.org, Tejun Heo Martin Michlmayr wrote: > With 2.6.26-rc1 I have to enable CONFIG_SATA_PMP=y otherwise sata_mv > doesn't find a disk on my Orion based device (QNAP TS-209). Without > CONFIG_SATA_PMP I get: > > [42949381.790000] sata_mv sata_mv.0: version 1.20 > [42949381.790000] sata_mv sata_mv.0: slots 32 ports 2 > [42949381.800000] scsi0 : sata_mv > [42949381.800000] scsi1 : sata_mv > [42949381.810000] ata1: SATA max UDMA/133 irq 29 > [42949381.810000] ata2: SATA max UDMA/133 irq 29 > [42949382.170000] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300) > [42949382.520000] ata2: SATA link down (SStatus 0 SControl 300) > [and nothing else] > > So obviously it sees that something is connected to ata1, but it > doesn't actually print anything about the disk or create the device > node. > > This works correctly with 2.6.25 or when I set CONFIG_SATA_PMP=y. > However, my device doesn't have a multiplier. .. Okay, so the "SStatus 123" means that sata_mv::mv_hardreset must have succeeded in finding good link status. Great. After that, libata-core and pals do the actual device IDENTIFY/probing. With CONFIG_SATA_PMP=y, there's an added soft-reset done first, and without it, there is no soft-reset. Looks like your device either needs the extra soft-reset regardless, or maybe just needs a little more time in the probing sequence. Tejun? I suppose we could just always ask for the extra soft-reset here (?)