From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH 2/3] misc: bmp085: Add DT bindings for EOC gpio line and direct irq. Date: Fri, 15 Nov 2013 19:47:50 +0100 Message-ID: <201311151947.51196.arnd@arndb.de> References: <1384465609-26485-1-git-send-email-marek@goldelico.com> <1384465609-26485-3-git-send-email-marek@goldelico.com> <20131115153015.GD24831@e106331-lin.cambridge.arm.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20131115153015.GD24831@e106331-lin.cambridge.arm.com> Sender: linux-doc-owner@vger.kernel.org To: Mark Rutland Cc: Marek Belisko , "gregkh@linuxfoundation.org" , "neilb@suse.de" , "hns@goldelico.com" , "rob.herring@calxeda.com" , Pawel Moll , "swarren@wwwdotorg.org" , "ijc+devicetree@hellion.org.uk" , "rob@landley.net" , "devicetree@vger.kernel.org" , "linux-doc@vger.kernel.org" , "linux-kernel@vger.kernel.org" List-Id: devicetree@vger.kernel.org On Friday 15 November 2013, Mark Rutland wrote: > There's some contention over the description of gpio-based IRQs in DT. > From the point of view of the device there is a logical IRQ output; the > fact that this happens to be wired up to a GPIO pin that can happen to > generate interrupts isn't anything to do with the device itself. There > are plenty of device we have now whose interrupt lines could be wired to > GPIOs. I see no reason to extend their bindings to support explicit > GPIOs for IRQs, and I see no reason the driver should have to handle > this. > > It would be far nicer for the device binding to just have the interrupts > property, and for the gpio controller to act as an interrupt-controller, > with the appropriate pin management. Yes, agreed. I missed this point in my review: the GPIO is used only as an interrupt pin here, so there is no reason to know the GPIO number. Arnd