Inference Blocking on LL_ATON_OSAL_WFE() with ThreadX

[Foahh]

Dear ST Development Teams,

I am working on implementing an object detection application for the STM32N6570-DK, starting from the reference projects:

• STMicroelectronics/x-cube-n6-ai-people-detection-tracking
• STMicroelectronics/STM32N6-GettingStarted-ObjectDetection

I'm encountering a blocking issue when implementing the inference pipeline:

During debugging, I found that the inference execution blocks indefinitely at LL_ATON_OSAL_WFE() on line 114 of `app_nn.c`. The code is stuck in the following loop:

static void NN_RunInference(void) {
  LL_ATON_RT_RetValues_t ret;

  do {
    ret = LL_ATON_RT_RunEpochBlock(&NN_Instance_od_yolo_x_person);

    if (ret == LL_ATON_RT_WFE) {
      LL_ATON_OSAL_WFE();  // <-- Blocks here
    }
  } while (ret != LL_ATON_RT_DONE);

  LL_ATON_RT_Reset_Network(&NN_Instance_od_yolo_x_person);
}

I've reviewed similar posts in the community

Inference on STM32N6 using FreeRTOS - STMicroelectronics Community

, but it didn't resolve my issue.

I have enable the clocks during sleep using 

__HAL_RCC_***_CLK_SLEEP_ENABLE()

Questions:

What could be the issue that blocks inference indefinitely at `LL_ATON_OSAL_WFE()`, and how to fix it?

Source Code:

My current implementation is available at: https://github.com/Foahh/stm32n6-ai-camera.git

I am using STEdgeAI 3.0.0.

(Note: Post-processing is not yet completed)

I’m new to STM32 and am actively learning it; any additional guidance or suggestions would be greatly appreciated!

Thank you for your support.