SDK8.1 and SDK 9.0 app_gpiote and nrf_drv_gpiote conflict

[Susheel Nuguru]

In SDK8.1 and SDK9.0 we have new driver nrf_drv_gpiote which has GPIOTE interrfupt handler in conflict with app_gpiote library. How to fix this

[Susheel Nuguru]

@HSA: I have updated the merge file. try to copy it now. You are also trying to copy files from SDK 8.1 to SDK 8.0. So what ever is missing, you need to copy those declarations also from SDK 8.1

[HSA]

@Aryan, sorry for the delay, I will try and let you know. Thanks

[HSA]

@Aryan: I tried to use the updated file. when I port the button_debounce (which I had discussed you earlier here devzone.nordicsemi.com/.../) example in [--ble-app-lbs-master--], Its not encounter the 'nrf_drv_gpiote.x' file. It compile in normal way as it was before.

Ans when I tried to add nrf_driver/nrf_library software packs, it shows error.

Please suggest. Thanks.

[wim]

@ any Nordic employee: what is the reasoning behind the introduction of a gpiote driver? is it because the app_gpiote was not really fully implementing the whole functionality? why was this introduced so late, as it was part of the chip for a long time now? I just would like to know the philosophy around it, such that I can anticipate on possible changes in the future. For example, should I abandon using the app_gpiote altogether, and start working with the driver instead? Is that the idea, that it will disappear from the SDK, because of it being a somewhat limited implementation of the gpiote functionality?

[Krzysztof Chruscinski]

@wim, yes nrf_drv_gpiote was added to cover all functionality of GPIOTE module (task & events) and not just PORT. And why so late? it's hard to answer. In general, we are trying now to provide a driver for every peripheral. Driver should cover all or most of the peripheral functionality - since it usually implements interrupt handler there cannot be many drivers for one peripheral. In next SDK we will add app_gpiote implementation that is based on nrf_drv_gpiote.