Latch disabled GPIOTE interrupt?

[TomDavies]

Is it possible to disable a GPIOTE interrupt, while performing some processing, and latch any incoming interrupt while that processing is occurring, so it fires once the interrupt is reenabled?

My scenario, external SPI device receives some data, signals the nRF by an interrupt line, I want to disable the interrupt while I retrieve that data, if an interrupt occurs while I'm retrieving that data, I want to finish processing, reenable interrupts and then process go ahead and process it.

I tried nrfx_gpiote_in_event_disable but the interrupt is 'lost' as the interrupt is not latched while disabled.

Is this possible?

 

[Einar Thorsrud]

Hi,

You can read the latch, but as it seems lost, I wonder if you are seeing the same as in this thread (that the driver reads and clears it)?

[TomDavies]

Hi Einar,

I've not explicitly enabled any latching - I was expecting the interrupt to be latched through the section of code while disabled.

I have attached a project I put together to test the idea.

Button 1 is set up with a gpiote handler to toggle LED1

LED2 is toggled by a timer

When LED2 is on, button 1 interrupt is disabled. when LED2 turns off, the interrupt is reenabled.

What I want to happen:
LED2 turns on, disables interrupt. I press Button 1, nothing happens, when LED2 turns off and reenables the interrupt, I then want the interrupt to fire, turning on LED 1.

What happens is that LED 1 only turns on is Button 1 is pressed while LED2 is off.bsp_demo.zip

[Einar Thorsrud]

Hi,

I am sorry for the late reply. I have not had a chance to look more closely at this yet, have you made any progress?

I will be OoO next week, so if you urgently need feedback on this, please open a new case. If not, I will get back to you after Easter.

[TomDavies]

Hi Einar,

I've not had a chance to take a look at it any further. I've read some more posts but either I'm missing something or not finding the right posts to get me any further along.

[Einar Thorsrud]

I don't see any indication that you are trying to read the latched interrupt in the code you uploaded? There you use the driver, and when you call nrfx_gpiote_in_event_enable(), that calls nrfx_gpiote_trigger_enable(), which explicitly clears the event before enabling the interrupt. So you need to check if it is set before, or avoid using the driver do disable and re-enable the interrupt (you can use nrf_gpiote_int_enable() directly instead).

[TomDavies]

Thanks Einar, that sounds like the bit I was overlooking. I've had to put this on the back burner the last week and won't be able to pick it up again until next week but will look into this and report back. Thanks.

[TomDavies]

Hi Einar,

I managed to get a bit of time this afternoon to test this out. Thanks for the pointer, the clearing of the event is was the bit I was missing. Changing the code to gpiote int enable/disable did the trick and now, if the pin change occurs while the interrupt is disabled, I get the interrupt firing once it's reenabled.

Thank you!