I think I was writing this down, decided to write down the part about vbussense before actually testing it and when I got back this morning I saw that I hadn't posted this yet and.... I seem to have forgotten that I got it to work when testing the vbussense....
The first time it didn't work it ran into the board configuration specifying that LSE would run. Not implemented on my board... I double checked the HSE clock and found the multiplier of 432 "quite a lot, that probably won't work" but ran into trouble when 216 cannot be divided by 4.5 to get the 48MHz USB clock.... Checked the datasheet and: 432 is just fine (and precisely "max").
The chips share the reference manual. Including the "memory map" that in older models would've been in the datasheet and not in the reference manual.
Oh.. You mean the USB module? Yeah, that seems to have been stuck in reset. I thought you were talking about the whole chip.
Anyway:
Code: Select all
> info
Kernel: 4.0.0
Compiler: GCC 4.9.3 20150529 (prerelease)
Architecture: ARMv7E-M
Core Variant: Cortex-M7
Port Info: Advanced kernel mode
Platform: STM32F746 Very High Performance with DSP and FPU
Board: STMicroelectronics STM32F746G-Discovery
Build time: May 26 2020 - 22:20:00
It still has a few "strings" from the 746, but otherwise it is now working fine. Next step is some serial stuff (probably not through the chibios drivers because I need some stuff the hardware can do but the driver doesn't (and I have that code for a different STM already)).
I still have to "pretend" that my CPU is a '746. There are no CMSIS files for the '750. I checked current git and... no files there either. I would expect those to be published by ST somewhere but I couldn't find them. As they share the same reference manual I think I won't run into trouble due to this, but still it would be nice to know where to find the official files....