From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Garzik Subject: Re: [PATCH] libata test-unit-ready for ATAPI devices Date: Wed, 16 Nov 2005 07:46:28 -0500 Message-ID: <437B2A24.9050508@pobox.com> References: <20051116044841.GA11722@havoc.gtf.org> <20051116100117.GA22225@htj.dyndns.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20051116100117.GA22225@htj.dyndns.org> Sender: linux-scsi-owner@vger.kernel.org To: Tejun Heo Cc: linux-ide@vger.kernel.org, linux-scsi@vger.kernel.org List-Id: linux-ide@vger.kernel.org Tejun Heo wrote: > On Tue, Nov 15, 2005 at 11:48:41PM -0500, Jeff Garzik wrote: > >>The following patch clears the "I was just reset" condition from an >>ATAPI device, and waits for it to come online, before continuing with >>the probe. >> >>Not checking this into any upstream-bound branch, as I'm not yet >>convinced of its value. >> > > > How about the following one? It doesn't make libata wait for the > device to become ready before trying to configure ATA stuff. It just > makes sr wait for the device to become ready before performing SCSI > configuration. Maybe both are needed? [thinking out loud] My rationale for doing TUR in libata was to clear the UNIT ATTENTION that libata caused through software/hardware reset. Once the UNIT ATTENTION was cleared, my rationale for waiting for the NOT READY(BECOMING READY) condition was simply symmetry with other libata probe tasks, which similarly wait for device readiness. However, given that sr is already coded to clear "the initial startup UNIT_ATTENTION", I tend to think your patch is more appropriate, and my patch should not be applied at all. Jeff