From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751310AbWFOHh2 (ORCPT ); Thu, 15 Jun 2006 03:37:28 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751313AbWFOHh2 (ORCPT ); Thu, 15 Jun 2006 03:37:28 -0400 Received: from gw.goop.org ([64.81.55.164]:30903 "EHLO mail.goop.org") by vger.kernel.org with ESMTP id S1751310AbWFOHh1 (ORCPT ); Thu, 15 Jun 2006 03:37:27 -0400 Message-ID: <44910E2A.5090205@goop.org> Date: Thu, 15 Jun 2006 00:37:14 -0700 From: Jeremy Fitzhardinge User-Agent: Thunderbird 1.5.0.2 (X11/20060501) MIME-Version: 1.0 To: George Nychis CC: lkml Subject: Re: cdrom support with thinkpad x6 ultrabay References: <4490E776.7080000@cmu.edu> <4490F4BC.1040300@goop.org> <44910B54.8000408@cmu.edu> In-Reply-To: <44910B54.8000408@cmu.edu> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org George Nychis wrote: > it successfully is applied, and i notice that CONFIG_ACPI_DOCK needs to > be set, so I did a "make oldconfig" after applying the patch, expecting > it to ask me whether or not i wanted to support it... it didn't. So > then I manually added "CONFIG_ACPI_DOCK=y" to the .config and built the > kernel, but dock.o is never built... what else do i need to do? > Make sure you disable the (obsolete?) ACPI_IBM_DOCK stuff. > If i can't get hot swappable support yet, I might as well get what is > supported for now so I can atleast use it sometimes :) > > Maybe this cry for help will spark someone to finish off the work on hot > swapping the optical drive. > Yeah, I'm hoping all the work on power management in libata will make things "just work" soon, but I think there's more to it. When you press the dock eject button, it really needs to go out to acpid, activate a script to unmount any filesystems mounted off the device, and then poke the ata layer to remove the device, before OKing the dock eject so the hardware's "don't do that" light goes out. But in the meantime I'm having enough trouble getting plain old suspend/resume reliable. J