X-CUBE-AI issue with other peripherals hanging

[Jf.1]

Hi everyone,

I'm currently working on a project that involves implementing Edge AI using the X-CUBE-AI software package within STM32CubeIDE. After adding the AI model, which requires 44 bytes of RAM and generating the necessary code, We encountered a problem: one of the peripherals, specifically I2C, starts hanging after a while when interfacing with a sensor to obtain readings.

Here’s a summary of the steps we've taken so far:

1. Isolated the Issue: To determine if the hanging issue was related to the AI initialization, we commented out the AI functions from the main file. However, the I2C communication still hangs when waiting for a response from the sensor after some time.

2. Sensor Exemption: We tried excluding the specific sensor to see if other sensors would work correctly. Unfortunately, the other sensors are also exhibiting the same hanging behavior.

3. Peripherals in Use: Our project involves multiple active peripherals, including I2C1-3 and UART4-5.

Additional Information

• Workspace Version: 1.7.0
• Microcontroller: L4A6VGT6P (custom PCB)
• X-CUBE-AI Version: 6.0.0 (device application)

Questions

We would appreciate any guidance or suggestions on the following:

• Are there specific configurations or settings in X-CUBE-AI that might impact I2C performance?
• Should we be concerned about the priority settings of the NVIC for I2C and other peripherals when using X-CUBE-AI?
• Are there known issues with I2C performance in this context that we should be aware of?

Thank you for your assistance!

Kind regards,

J

[fauvarque.daniel]

The inference implementation on the target is pure software and doesn't use any interrupt.

On this version, it is required to activate the CRC. That's the only adherence with the hardware that we have.

FYI we are currently at version 9.1 so you may want to upgrade to the latest version although it won't change your issue (the 9.1 version doesn't require the activation of the CRC)

Regards

Daniel

[Jf.1]

Thanks for the response. Would this be more of a resource or memory management issue?

Kind regards,

J