Ah that's a good spot. I'd always assumed it was running on core1 as I didn't see any code reference other than init, and it sort of made sense given USB can be quite demanding but going over the code again you're right, it selects the core it was called from and sets an an interrupt which presumably is doing the work behind the scenes rather than on a separate core. Perhaps the same is happening for the cyw43<-->rp2040 wifi interface handling too (all interrupt driven rather than separate core); I'll dig in more.In debugger it looks like the usb stdio runs on the core where and when is called, which by default is core0, the other core being stopped all the time. By the way it uses spinlocks, I think it can be used also by the second core, if it's called from there.
I got a bit off topic here but if I can get a RNDIS example together that at least means I have core1 available if that helps.
Statistics: Posted by mattpi — Fri Oct 18, 2024 9:39 am