On Mon, 2018-03-12 at 09:43 +1300, Joshua Scott wrote:
> The previous implementation has had a detrimental effect on devices
> using
> high bitrates (bluetooth), as the fifo being non-empty for a single
> check
> would result in a 10 µs delay.
>
> This patch updates the code such that it will allow 1000 retries to
> fail
> before delaying by 1 µs for the remaining retry limit.
>
> In addition, the maximum number of retries has been increased, both to
> cover the decreased delay length, and to cover a new worst-case seen
> on
> the Armada 385 SoC. "dmesg ; resize", filling the buffer with text to
> output before reconfiguring, required 13 ms to empty the buffer and
> avoid
> losing characters.
I gave a thought about this patch and initial approach. I think we need
a kind of quirk depending on which the workaround would be enabled or
disabled.
>
> Signed-off-by: Joshua Scott <joshua.scott@xxxxxxxxxxxxxxxxxxx>
> ---
> drivers/tty/serial/8250/8250_dw.c | 11 +++++++++--
> 1 file changed, 9 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/tty/serial/8250/8250_dw.c
> b/drivers/tty/serial/8250/8250_dw.c
> index 823c1dc..843c2a2 100644
> --- a/drivers/tty/serial/8250/8250_dw.c
> +++ b/drivers/tty/serial/8250/8250_dw.c
> @@ -127,13 +127,20 @@ static void dw8250_check_lcr(struct uart_port
> *p, int value)
> /* Returns once the transmitter is empty or we run out of retries */
> static void dw8250_tx_wait_empty(struct uart_port *p, int tries)
> {
> + unsigned int delay_threshold = tries - 1000;
> unsigned int lsr;
>
> while (tries--) {
> lsr = readb (p->membase + (UART_LSR << p->regshift));
> if (lsr & UART_LSR_TEMT)
> break;
> - udelay (10);
> +
> + /* The device is first given a chance to empty
> without delay,
> + * to avoid slowdowns at high bitrates. If after 1000
> tries
> + * the buffer has still not emptied, allow more time
> for low-
> + * speed links. */
> + if (tries < delay_threshold)
> + udelay (1);
> }
> }
>
> @@ -143,7 +150,7 @@ static void dw8250_serial_out(struct uart_port *p,
> int offset, int value)
>
> /* Allow the TX to drain before we reconfigure */
> if (offset == UART_LCR)
> - dw8250_tx_wait_empty(p, 1000);
> + dw8250_tx_wait_empty(p, 20000);
>
> writeb(value, p->membase + (offset << p->regshift));
>
--
Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
Intel Finland Oy
--
To unsubscribe from this list: send the line "unsubscribe linux-serial" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
