Re: [PATCH v6] tty/serial: at91: fix hardware handshake on Atmel platforms

[Richard Genoud]

2016-10-28 11:51 GMT+02:00 Uwe Kleine-KÃnig <u.kleine-koenig@xxxxxxxxxxxxxx>:
> On Fri, Oct 28, 2016 at 01:13:31AM +0200, Alexandre Belloni wrote:
>> On 27/10/2016 at 20:02:29 +0200, Uwe Kleine-KÃnig wrote :
>> > Hello Richard,
>> >
>> > On Thu, Oct 27, 2016 at 06:04:06PM +0200, Richard Genoud wrote:
>> > > diff --git a/drivers/tty/serial/atmel_serial.c b/drivers/tty/serial/atmel_serial.c
>> > > index fd8aa1f4ba78..168b10cad47b 100644
>> > > --- a/drivers/tty/serial/atmel_serial.c
>> > > +++ b/drivers/tty/serial/atmel_serial.c
>> > > @@ -2132,11 +2132,29 @@ static void atmel_set_termios(struct uart_port *port, struct ktermios *termios,
>> > >           mode |= ATMEL_US_USMODE_RS485;
>> > >   } else if (termios->c_cflag & CRTSCTS) {
>> > >           /* RS232 with hardware handshake (RTS/CTS) */
>> > > -         if (atmel_use_dma_rx(port) && !atmel_use_fifo(port)) {
>> > > -                 dev_info(port->dev, "not enabling hardware flow control because DMA is used");
>> > > -                 termios->c_cflag &= ~CRTSCTS;
>> > > -         } else {
>> > > +         if (atmel_use_fifo(port) &&
>> > > +             !mctrl_gpio_to_gpiod(atmel_port->gpios, UART_GPIO_CTS)) {
>> > > +                 /*
>> > > +                  * with ATMEL_US_USMODE_HWHS set, the controller will
>> > > +                  * be able to drive the RTS pin high/low when the RX
>> > > +                  * FIFO is above RXFTHRES/below RXFTHRES2.
>> > > +                  * It will also disable the transmitter when the CTS
>> > > +                  * pin is high.
>> > > +                  * This mode is not activated if CTS pin is a GPIO
>> > > +                  * because in this case, the transmitter is always
>> > > +                  * disabled (there must be an internal pull-up
>> > > +                  * responsible for this behaviour).
>> > > +                  * If the RTS pin is a GPIO, the controller won't be
>> > > +                  * able to drive it according to the FIFO thresholds,
>> > > +                  * but it will be handled by the driver.
>> > > +                  */
>> > >                   mode |= ATMEL_US_USMODE_HWHS;
>> >
>> > You use
>> >
>> >     !mctrl_gpio_to_gpiod(atmel_port->gpios, UART_GPIO_CTS)
>> >
>> > as indicator that the cts mode of the respective pin is used. Is this
>> > reliable? (It's not if there are machines that don't use CTS, neither as
>> > gpio nor using the hardware function.) Maybe this needs a dt property to
>> > indicate that there is no (hw)handshaking available?
>> >
>>
>> We had a call today were we agreed that this should be added in a future
>> patch. Let's fix the regression for now.
>
> A machine without CTS (neither gpio nor hw function) used to work fine
> before the breaking commit, right? So this case is part of the
> regression and needs a fix?
Actually, a machine with a FIFO and without CTS didn't even exist at the
time of the breaking commit (v4.0), the FIFO handling was introduced later,
so it's not even a regression !

> Anyhow, this probably shouldn't stop the commit entering mainline
> because there are probably very few such machines (if any).
>
> So:
> Acked-by: Uwe Kleine-KÃnig <u.kleine-koenig@xxxxxxxxxxxxxx>
>
> Best regards
> Uwe


Thanks !

Greg, could you take this in your tree ?

regards,
Richard.