From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bartlomiej Zolnierkiewicz Subject: Re: [PATCH 2/2] ide: add proper PCI PM support Date: Sun, 3 Aug 2008 16:15:38 +0200 Message-ID: <200808031615.38477.bzolnier@gmail.com> References: <200808021917.04847.bzolnier@gmail.com> <200808022249.38969.rjw@sisk.pl> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from fk-out-0910.google.com ([209.85.128.184]:17363 "EHLO fk-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755421AbYHCOSj (ORCPT ); Sun, 3 Aug 2008 10:18:39 -0400 Received: by fk-out-0910.google.com with SMTP id 18so1485809fkq.5 for ; Sun, 03 Aug 2008 07:18:37 -0700 (PDT) In-Reply-To: <200808022249.38969.rjw@sisk.pl> Content-Disposition: inline Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: "Rafael J. Wysocki" Cc: linux-ide@vger.kernel.org, linux-kernel@vger.kernel.org, Jesse Barnes On Saturday 02 August 2008, Rafael J. Wysocki wrote: > On Saturday, 2 of August 2008, Bartlomiej Zolnierkiewicz wrote: > > * Keep pointer to ->init_chipset method also in > > struct ide_host and set it in ide_host_alloc_all(). > > > > * Add ide_pci_suspend() and ide_pci_resume() helpers > > (default ->suspend and ->resume implementations). > > > > * ->init_chipset can no longer be marked __devinit. > > > > * Add proper PCI PM support to IDE PCI host drivers > > (rz1000.c and tc86c001.c are skipped for now since > > they need to be converted from using ->init_hwif > > to use ->init_chipset instead). > > Well, first, the work on improving the PCI PM support by IDE drivers is much > appreciated. > > Still, since you're adding new routines etc., it may be a good time to > implement them using the new PM framework, as defined in > include/linux/pm.h (please also have a look at drivers/pci/pci-driver.c for the > PCI bus type's PM callbacks implementation details). Unfortunately, it's not > been well documented yet, so if you have any questions, please ask. > > In fact, we need some example implementations of the new PM callbacks and > since you know the drivers in question very well, we could use your > implementations as examples, if you don't mind. [Certainly, they would be > better than PM callbacks written by me for any driver. ;-) ] Thanks for reviewing the patch. I'll take a look into converting it to use new PM infrastructure (it may take a while so please be patient).