[PATCH V3 1/3] serial: stm32: remove infinite loop possibility in putchar function

[Valentin CARON]

On 4/19/22 10:59, Geert Uytterhoeven wrote:
> Hi Valentin,
>
> On Tue, Apr 19, 2022 at 10:54 AM Valentin Caron
> <valentin.caron at foss.st.com> wrote:
>> Rework stm32_usart_console_putchar() function in order to anticipate
>> the case where the character can never be sent.
>>
>> Signed-off-by: Valentin Caron <valentin.caron at foss.st.com>
> Thanks for your patch!
>
>> --- a/drivers/tty/serial/stm32-usart.c
>> +++ b/drivers/tty/serial/stm32-usart.c
>> @@ -1640,10 +1640,16 @@ static void stm32_usart_console_putchar(struct uart_port *port, unsigned char ch
>>   {
>>          struct stm32_port *stm32_port = to_stm32_port(port);
>>          const struct stm32_usart_offsets *ofs = &stm32_port->info->ofs;
>> +       u32 isr;
>> +       int ret;
>>
>> -       while (!(readl_relaxed(port->membase + ofs->isr) & USART_SR_TXE))
>> -               cpu_relax();
>> -
>> +       ret = readl_relaxed_poll_timeout_atomic(port->membase + ofs->isr, isr,
>> +                                               (isr & USART_SR_TXE), 100,
>> +                                               STM32_USART_TIMEOUT_USEC);
>> +       if (ret != 0) {
>> +               dev_err(port->dev, "Error while sending data in UART TX : %d\n", ret);
> Does it make sense to print this message, i.e. will the user ever see it?
> Or is the failure above temporary?
> I assume that you have seen this trigger?
>
>> +               return;
>> +       }
>>          writel_relaxed(ch, port->membase + ofs->tdr);
>>   }
> Gr{oetje,eeting}s,
>
>                          Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                  -- Linus Torvalds
Hi Geert,

The failure is temporary. It can appears when the uart is too slow to 
send data.

I never tested the case, but I prefer to show a message to know if the 
console loses a character.

Thanks,
Valentin