From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jonathan Cameron Subject: Re: [RFC 0/4] iio: trigger: Add OF support and GPIO based trigger Date: Sat, 18 Feb 2017 19:07:12 +0000 Message-ID: <3a2d79be-d39d-10d8-6ff4-d8aea9560263@kernel.org> References: <1487347429-31761-1-git-send-email-fabrice.gasnier@st.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=m.gmane.org@lists.infradead.org To: Daniel Baluta , Fabrice Gasnier Cc: Mark Rutland , Devicetree List , benjamin.gaignard@linaro.org, Lars-Peter Clausen , alexandre.torgue@st.com, "linux-iio@vger.kernel.org" , Peter Meerwald , linux@armlinux.org.uk, Linux Kernel Mailing List , Rob Herring , mcoquelin.stm32@gmail.com, Hartmut Knaack , Linus Walleij , linux-arm-kernel@lists.infradead.org, benjamin.gaignard@st.com List-Id: devicetree@vger.kernel.org On 17/02/17 16:23, Daniel Baluta wrote: > On Fri, Feb 17, 2017 at 6:03 PM, Fabrice Gasnier wrote: >> Hi, >> >> I send this as an RFC, as I'm digging to try to get as generic >> as possible GPIO based trigger support, as discussed earlier in >> other patchset "Add EXTI GPIO trigger support to STM32 ADC": >> https://www.spinics.net/lists/arm-kernel/msg559091.html >> >> Following various comments on this, I sent this RFC to address two >> topics, that I see as possible prerequisite: >> - Add trigger OF support >> - GPIO based trigger > > I wonder if we could use configfs to create and configure > a GPIO based trigger. > > You should just create another trigger type (look into > drivers/iio/trigger/iio-trig-hrtimer.c) > the userspace API will look like this: > > # cd /config/iio/triggers > # insert your iio-trig-gpio > # setup the global parameters for the GPIO (the local params can be > set via dts/acpi) > # cd /config/iio/trigers/gpios/ > > #mkdir t1 > > And this will create an io-gpio-trigger named t1. > > Not sure this fits your usecase but would be nice to consider. > It might potentially make sense to support both DT or similar instantiation and instantiation from userspace. Primary route seems to me to always be DT based though as we are dealing with wiring... So I'm unconvinced on the configfs option, but might be talked round. > thanks, > Daniel. >