Re: [PATCH v1 2/3] serial: liteuart: separate RX loop from poll timer

["Gabriel L. Somlo" <gsomlo@gmail.com>]

On Thu, Nov 10, 2022 at 01:09:59AM +0000, Joel Stanley wrote:
> On Mon, 7 Nov 2022 at 17:15, Gabriel Somlo <gsomlo@gmail.com> wrote:
> >
> > Move the character-receive (RX) loop to its own dedicated function,
> > and (for now) call that from the poll timer, liteuart_timer().
> >
> > This is in preparation for adding IRQ support to the receive path.
> >
> > Signed-off-by: Gabriel Somlo <gsomlo@gmail.com>
> > ---
> >  drivers/tty/serial/liteuart.c | 25 +++++++++++++++----------
> >  1 file changed, 15 insertions(+), 10 deletions(-)
> >
> > diff --git a/drivers/tty/serial/liteuart.c b/drivers/tty/serial/liteuart.c
> > index 4b9cca249828..90a29ed79bff 100644
> > --- a/drivers/tty/serial/liteuart.c
> > +++ b/drivers/tty/serial/liteuart.c
> > @@ -69,29 +69,34 @@ static struct uart_driver liteuart_driver = {
> >  #endif
> >  };
> >
> > -static void liteuart_timer(struct timer_list *t)
> > +static void liteuart_rx_chars(struct uart_port *port)
> >  {
> > -       struct liteuart_port *uart = from_timer(uart, t, timer);
> > -       struct uart_port *port = &uart->port;
> >         unsigned char __iomem *membase = port->membase;
> > -       unsigned int flg = TTY_NORMAL;
> > -       int ch;
> > -       unsigned long status;
> > +       unsigned int status;
> > +       unsigned char ch;
> 
> u32, u8, void __iomem * would be better kernel types to use here.
> 
> You've also changed ch from a signed 32 to an unsigned 8.

uart_insert_char() expects both `status` and `ch` to be of type
`unsigned int`. Switching `ch` to 8-bit was a typo, thanks for
catching it!

I'm going to use `unsigned int status, ch` in v3, to match the
signature of `uart_insert_char()` -- hope that's OK. This will be
a separate commit preceding the "move rx loop out of poll timer"
change.
 
> >
> >         while ((status = !litex_read8(membase + OFF_RXEMPTY)) == 1) {
> >                 ch = litex_read8(membase + OFF_RXTX);
> >                 port->icount.rx++;
> >
> >                 /* necessary for RXEMPTY to refresh its value */
> > -               litex_write8(membase + OFF_EV_PENDING, EV_TX | EV_RX);
> > +               litex_write8(membase + OFF_EV_PENDING, EV_RX);
> 
> You're no longer clearing EV_TX, but don't mention why (I understand
> why with the context of the other changes, so perhaps add something to
> this commit message).

I'm adding a separate commit to document this (in v3) as well.

Thanks,
--Gabriel
 
> 
> >
> >                 /* no overflow bits in status */
> >                 if (!(uart_handle_sysrq_char(port, ch)))
> > -                       uart_insert_char(port, status, 0, ch, flg);
> > -
> > -               tty_flip_buffer_push(&port->state->port);
> > +                       uart_insert_char(port, status, 0, ch, TTY_NORMAL);
> >         }
> >
> > +       tty_flip_buffer_push(&port->state->port);
> > +}
> > +
> > +static void liteuart_timer(struct timer_list *t)
> > +{
> > +       struct liteuart_port *uart = from_timer(uart, t, timer);
> > +       struct uart_port *port = &uart->port;
> > +
> > +       liteuart_rx_chars(port);
> > +
> >         mod_timer(&uart->timer, jiffies + uart_poll_timeout(port));
> >  }
> >
> > --
> > 2.37.3
> >