Hello, I’m using TX2i and I see a weird phenomena.
If there is a error print to the UART-debug, I see bad interrupt handling in another serial interface (UART 3, with FC).
This is easily reproduce for many dev_err prints drivers.
For example, if I shutdown a core using
echo -n 0 > /sys/devices/system/cpu/cpu3/online
there are some prints to the UART debug (/dev/ttyS0) regarding the shutdown of the core, followed by:
serial-tegra 3130000.serial: RxData PIO to tty layer failed (which is weird, again, using FC and no PIO)
and then bpmp: mrq 22 took 2744000 us (I guess this is some bug/delay caused by shutting down the core).
the serial-tegra message is reoccurring through all uart debug error messages that cause this problem.
Is this connected?
https://github.com/torvalds/linux/commit/f1681a9
I would like to get a reply, this harms my application badly
hello linus.s.gates,
may I also know which JetPack release you’re working with?
besides CPU3: shutdown, I don’t see such failure from uart logs from my side.
please share more details/steps for reproduce this issue,
thanks
I’m using the latest meta-tegra (Yocto).
The CPU3: shutdown is an example for a UART error print I can create (I specifically wrote that FOR EXAMPLE I can generate one by doing echo -n 0 > /sys/devices/system/cpu/cpu3/online which shuts down core 3 if it is on before)
As I mentioned, every error print that is thrown to the UART debug causes this behavior
hello linus.s.gates,
may I also know what’s the baudrate configuration it is?
There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks
hello linus.s.gates,
is there an error report by tty_insert_flip_char() or something from tty?
please also setup serial console to gather the details logs for reference,
thanks