On Mon, Feb 26, 2018 at 7:24 AM, Baolin Wang <baolin.wang@xxxxxxxxxx> wrote: > On 21 February 2018 at 19:35, Baolin Wang <baolin.wang@xxxxxxxxxx> wrote: >> On 20 February 2018 at 02:11, Rob Herring <robh@xxxxxxxxxx> wrote: >>>> diff --git a/Documentation/devicetree/bindings/input/gpio-keys.txt b/Documentation/devicetree/bindings/input/gpio-keys.txt >>>> index a949404..e3104bd 100644 >>>> --- a/Documentation/devicetree/bindings/input/gpio-keys.txt >>>> +++ b/Documentation/devicetree/bindings/input/gpio-keys.txt >>>> @@ -29,6 +29,8 @@ Optional subnode-properties: >>>> - linux,can-disable: Boolean, indicates that button is connected >>>> to dedicated (not shared) interrupt which can be disabled to >>>> suppress events from the button. >>>> + - gpio-key,level-trigger: Boolean, indicates that button's interrupt >>>> + type is level trigger. Otherwise it is edge trigger as default. >>> >>> No. Just use 'interrupts' instead of 'gpios' and specify the trigger >>> type. Or put both if you need to read the state. >> >> Okay, so something as below to get the level type from the >> 'interrupts' property. >> if (fwnode_property_read_u32(child, "interrupts", &button->level_type)) >> button->level_type = IRQ_TYPE_NONE; > > After more thinking, if we use 'interrupts' to indicate the irq type > for this case, we cannot specify the irq number due to the irq number > should be get by gpiod_to_irq(). So the device nodes look weird, since > we should define the index of the interrupt controller instead of the > irq type if the #interrupt_cells is set to 1 according to the > interrupt controller documentation. What do you think about this? I think what you're ultimately seeing is a bad fit between this GPIO/irq-controller and the Linux gpio keys driver, it doesn't have very much to do with the device tree bindings. What I think is appropriate is to try to create a new input driver in Linux that just takes an interrupt, not a GPIO, and let the GPIO chip only act as an irqchip for this. This avoid the complicated step of converting a GPIO to an interrupt in order to use it, when all you really want to do is use an interrupt, you don't really care about the GPIO here, correct? So we would create drivers/input/keyboard/interrupt_keys.c however I suspect a bunch of code would need to be shared with gpio_keys.c so maybe it is necessary to break out the parts of gpio_keys.c into its own helper file. Or maybe even have the pure interrupt handling as part of gpio_keys.c, i.e. if the driver can't find any associated GPIO, it goes on to check if there is an interrupt assigned to the device node and use that directly instead. Either way, Dmitry must be involved. Yours, Linus Walleij -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
