Closed mocleiri closed 2 years ago
I think I moved to fast on filing this pull request.
I did a clean build and got the HAL errors again which shouldn't happen if the STM32H7 update had fixed my issue.
Upon deeper inspection I can see that there are some HAL API differences between the STM32F4xx HAL files used in machine-i2s.c for stm32 and the HAL API provided by the STM32H7xx HAL files.
The H7 HAL could do with updating, but it seems that's a separate issue to getting I2S working. So I'll hold off on this.
I've confirmed that there are just differences between the STM32F4 and STM32H7 HAL's as they relate to i2s and this proposed update doesn't effect that in any way.
I should be able to can get machine.I2S working for STM32H7 using the existing HAL without needing this update.
I'm going to close this request. I don't expect to be able to look at the issue for a few months so will reopen or refile if it is needed for I2S support on STM32H7.
I tried to enable I2S on my Nucleo H743ZI2 but it failed due to missing HAL functions.
https://github.com/mocleiri/tensorflow-micropython-examples/issues/54
I was able to run the fetch_from_upstream.sh script to update to the files contained here and the build worked properly with I2S enabled.