From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [PATCH] libata: Handle bay devices in dock stations Date: Mon, 09 Jun 2008 10:44:01 +0900 Message-ID: <484C8AE1.3010908@gmail.com> References: <20080528143857.GB5585@homac.suse.de> <4845886E.2010505@garzik.org> <20080603181346.GA5013@srcf.ucam.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20080603181346.GA5013@srcf.ucam.org> Sender: linux-acpi-owner@vger.kernel.org To: Matthew Garrett Cc: Jeff Garzik , linux-kernel@vger.kernel.org, linux-ide@vger.kernel.org, linux-acpi@vger.kernel.org, akpm@linux-foundation.org List-Id: linux-ide@vger.kernel.org Matthew Garrett wrote: > On Tue, Jun 03, 2008 at 02:07:42PM -0400, Jeff Garzik wrote: > >> It would be nice if you and Holger could work together to produce a >> single patch[set]... right now I have patches from both of you, and I >> was sorta waiting on the thread to die down to see if competing patches >> might merge into a single set > > I think we were waiting for feedback from Tejun as to why removing the > port freeze call fixed the hang I was seeing? Beyond that, Holger's > latest patch looked good to me. Sorry, was off for the last week. The difference between freezing and scheduling EH is that the former immediately aborts all in-flight commands and resets the port while the latter waits till the in-flight commands finish or time out (EH scheduling kicks fast-drain and the timeout is reduced to three seconds). TF-based ATA controllers are very sensitive to how the registers are accessed and sometimes lock up the whole machine when they are not happy by indefinitely holding the PCI bus. This could have been the case if IOs were in flight when the dock event occurred. Were they? Thanks. -- tejun