Following up on this old post. I'm seeing the occasional dropped byte on the non-XMOS receiver end using uart_tx_buffered_if (using one GPIO of a 4-bit port, as that's all I have available right now). Reducing use of set_core_fast_mode_on and adding some core affinities appears to have improved things (so far).
Does anyone have any comments on the comparative reliability of multi_uart_tx? I'd be happy to put a 1.8432MHz crystal on board, but I need to make that decision sooner rather than later.
UART problem / solid solution?
-
lukehatpadl
- Member++
- Posts: 28
- Joined: Sat Jul 08, 2023 5:15 am