From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hannes Reinecke Subject: Re: [PATCH 6/6] Implement the AHCI suspend/resume Date: Fri, 02 Jun 2006 10:21:34 +0200 Message-ID: <447FF50E.4060609@suse.de> References: <1149234379.13451.59.camel@forrest26.sh.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from ns1.suse.de ([195.135.220.2]:57316 "EHLO mx1.suse.de") by vger.kernel.org with ESMTP id S1751034AbWFBIVg (ORCPT ); Fri, 2 Jun 2006 04:21:36 -0400 In-Reply-To: <1149234379.13451.59.camel@forrest26.sh.intel.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: "zhao, forrest" Cc: jeff@garzik.org, axboe@suse.de, htejun@gmail.com, jeremy@goop.org, lkml@rtr.ca, linux-ide@vger.kernel.org zhao, forrest wrote: > This patch implements the AHCI suspend/resume. > It puts the port suspend/resume operations in=20 > ahci_pci_device_suspend/resume(), which is in conformance with > Jeff's idea of host<->bus<->device suspend/resume sequence. >=20 >=20 > Signed-off-by: Forrest Zhao >=20 There is one quirky bit in here: we have to enable interrupts for the _adapter_ after all _devices_ have been re-initialized. This breaks the host<->bus<->device initialization sequence. Is there any chance libata can add some hooks for that sort of thing? The current approach (using a device map and run the host initialisatio= n after the last device is done) is a bit hackish. I'd rather like to see the upper layers being able to deal with such devices. But until then: Signed-off-by: Hannes Reinecke Cheers, Hannes --=20 Dr. Hannes Reinecke hare@suse.de SuSE Linux Products GmbH S390 & zSeries Maxfeldstra=DFe 5 +49 911 74053 688 90409 N=FCrnberg http://www.suse.de