It would help if you could share a bit more detail (error logs, board type, RT-Thread version, etc.). That makes it much easier for others to pinpoint the issue.
In general, many RT-Thread problems are related to toolchain setup or configuration, so double-check your build environment and dependencies as well.
https://www.rt-thread.io/document/site/drift boss
Hi,
It looks like a mismatch between TFLM and the CMSIS package being pulled in by RT-Thread. The missing arm_helium_utils.h typically means the build is trying to use Helium (MVE) kernels, but the CMSIS package doesn’t include them.
You could try checking whether your board/toolchain actually enables MVE (Helium) flags (e.g. -march=armv8.1-m.main+mve.fp slope). If not, those files shouldn’t be compiled.
Hope that helps narrow it down
I see you're running a TFLite model on the Vision Board! Have you tried different optimization levels within TFLite? I found that adjusting the optimization flags significantly impacted build times. In my experience, reducing optimization led to faster compiles, although performance suffered slightly. Any experience with snow rider
That's quite a techy situation you're wrestling with! Building with TFLM can sometimes feel like herding cats. I can see you're diving deep into embedded ML which is cool. Ever think about optimizing kernels manually for performance? I remember a time when I was trying to get a neural network to classify different types of fruit on a Raspberry Pi, resource management became critical, and everything slowed to a crawl and debugging felt endless. After all that the thing just crashed and became a Slither io useless.
