From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Rafael J. Wysocki" Subject: Re: 900af0d breaks some embedded suspend/resume Date: Sun, 19 Apr 2009 00:44:57 +0200 Message-ID: <200904190044.58444.rjw@sisk.pl> References: <20090417231009.GB6900@flint.arm.linux.org.uk> <20090418135323.GA7148@flint.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from ogre.sisk.pl ([217.79.144.158]:45785 "EHLO ogre.sisk.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752733AbZDRWpa (ORCPT ); Sat, 18 Apr 2009 18:45:30 -0400 In-Reply-To: Content-Disposition: inline Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: Linus Torvalds Cc: Russell King , Len Brown , Linux Kernel List , pm list , ACPI Devel Maling List On Saturday 18 April 2009, Linus Torvalds wrote: > > On Sat, 18 Apr 2009, Russell King wrote: > > > > What I need to be able to do is to suspend most devices on the host side > > which may involve talking to a separate microcontroller via I2C to shut > > down power to peripherals. > > I suspect that for cases like this, the simplest thing to do is to just > add a marker for "don't mess with my power management, I'm doing > everything through sysdev" for the specified devices. In this particular case, if sysdev was used, there would be no problem, but the platform uses a platform device to suspend-resume the i2c controller. In principle we could convert it to use a sysdev, but that would be more difficult than the last patch I sent IMO (at least to me). Also, apparently it is not the only platform doing it. > Then those i2c controllers (and perhaps some PCI bridges etc) can just set > that bit, and the device would basically turn invisible as far as the PM > layer is concerned. > > Not that different from the IRQF_TIMER bit for timer interrupts. I generally agree and a patch to implemet such a flag has been submitted recently. Thanks, Rafael