serial: 8250: Optimize irq enable after console write

Commit 7f9803072ff6 ("serial: 8250: Support console on software emulated
rs485 ports") amended serial8250_console_write() with rs485 support, but
positioned the invocation of ->rs485_stop_tx() after re-enablement of
interrupts.  The irq handler and ->console_write() are serialized with
the port spinlock, so no problem there, but due to the rs485 delay, the
irq handler may unnecessarily spin for a while.  Avoid that by moving
->rs485_stop_tx() before re-enablement of interrupts, which also mirrors
the order at the beginning of serial8250_console_write().

Signed-off-by: Lukas Wunner <lukas@wunner.de>
Link: https://lore.kernel.org/r/019839cb1f61b01210b6ff9ac9f9079ca77f8411.1585319447.git.lukas@wunner.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c
index a1d3aef..f77bf82 100644 (file)
--- a/drivers/tty/serial/8250/8250_port.c
+++ b/drivers/tty/serial/8250/8250_port.c
@@ -3268,7 +3268,6 @@ void serial8250_console_write(struct uart_8250_port *up, const char *s,
         *      and restore the IER
         */
        wait_for_xmitr(up, BOTH_EMPTY);
-       serial_port_out(port, UART_IER, ier);
 
        if (em485) {
                mdelay(port->rs485.delay_rts_after_send);
@@ -3276,6 +3275,8 @@ void serial8250_console_write(struct uart_8250_port *up, const char *s,
                        up->rs485_stop_tx(up);
        }
 
+       serial_port_out(port, UART_IER, ier);
+
        /*
         *      The receive handling will happen properly because the
         *      receive ready bit will still be set; it is not cleared