nRF52, SDK12.1, Eclipse, gcc 5.4.1, FreeRTOS, no soft-device.
I am having a problem with GPIOTE. I have six interrupts programmed on six different pins. As long as the signals do not transition at the same time, the firmware receives interrupts from all pins independently; everything works fine. However, if any two signals change simultaneously, the GPIOTE seems to stop responding altogether: no further interrupts are generated.
Five of the interrupts are configured like so:
nrf_drv_gpiote_in_config_t config = GPIOTE_CONFIG_IN_SENSE_LOTOHI(false);
config.pull = NRF_GPIO_PIN_PULLDOWN;
ret_val = nrf_drv_gpiote_in_init(adxl_isr_pin[i], &config, adxl_event_handler);
The sixth interrupt is configured using NRF_GPIOTE_POLARITY_TOGGLE as it must interrupt on both edges. There are three different event handlers.
In this post is the following quote:
If power-consumption is not an issue and the signals may transition at the same time, then I would initialize the pins with nrf_drv_gpiote_in_init with a nrf_drv_gpiote_in_config_t struct where hi_accuracy is set to true and register a separate event handler for each of the pins.
This post implies that signals that transition simultaneously must use high accuracy. Is this true? If a GPIOTE event is configured for "low accuracy", can multiple simultaneous events be detected? Assuming that simultaneous events cannot be detected, is this restriction removed / eliminated if "high accuracy" is used? Why does this happen?
