[PATCH v2] serial: 8250_dw: Improve unwritable LCR workaround

[ezequiel.garcia@free-electrons.com (Ezequiel Garcia)]

Hello,

On Tue, Oct 01, 2013 at 10:18:08AM -0700, Tim Kryger wrote:
> When configured with UART_16550_COMPATIBLE=NO or in versions prior to
> the introduction of this option, the Designware UART will ignore writes
> to the LCR if the UART is busy.  The current workaround saves a copy of
> the last written LCR and re-writes it in the ISR for a special interrupt
> that is raised when a write was ignored.
> 
> Unfortunately, interrupts are typically disabled prior to performing a
> sequence of register writes that include the LCR so the point at which
> the retry occurs is too late.  An example is serial8250_do_set_termios()
> where an ignored LCR write results in the baud divisor not being set and
> instead a garbage character is sent out the transmitter.
> 
> Furthermore, since serial_port_out() offers no way to indicate failure,
> a serious effort must be made to ensure that the LCR is actually updated
> before returning back to the caller.  This is difficult, however, as a
> UART that was busy during the first attempt is likely to still be busy
> when a subsequent attempt is made unless some extra action is taken.
> 
> This updated workaround reads back the LCR after each write to confirm
> that the new value was accepted by the hardware.  Should the hardware
> ignore a write, the TX/RX FIFOs are cleared and the receive buffer read
> before attempting to rewrite the LCR out of the hope that doing so will
> force the UART into an idle state.  While this may seem unnecessarily
> aggressive, writes to the LCR are used to change the baud rate, parity,
> stop bit, or data length so the data that may be lost is likely not
> important.  Admittedly, this is far from ideal but it seems to be the
> best that can be done given the hardware limitations.
> 
> Lastly, the revised workaround doesn't touch the LCR in the ISR, so it
> avoids the possibility of a "serial8250: too much work for irq" lock up.
> This problem is rare in real situations but can be reproduced easily by
> wiring up two UARTs and running the following commands.
> 
>   # stty -F /dev/ttyS1 echo
>   # stty -F /dev/ttyS2 echo
>   # cat /dev/ttyS1 &
>   [1] 375
>   # echo asdf > /dev/ttyS1
>   asdf
> 
>   [   27.700000] serial8250: too much work for irq96
>   [   27.700000] serial8250: too much work for irq96
>   [   27.710000] serial8250: too much work for irq96
>   [   27.710000] serial8250: too much work for irq96
>   [   27.720000] serial8250: too much work for irq96
>   [   27.720000] serial8250: too much work for irq96
>   [   27.730000] serial8250: too much work for irq96
>   [   27.730000] serial8250: too much work for irq96
>   [   27.740000] serial8250: too much work for irq96
> 
> Signed-off-by: Tim Kryger <tim.kryger@linaro.org>
> Reviewed-by: Matt Porter <matt.porter@linaro.org>
> Reviewed-by: Markus Mayer <markus.mayer@linaro.org>
> ---
> 
> Changes in v2:
>   - Rebased on tty-next
>   - Updated commit messsage to mention UART_16550_COMPATIBLE
>   - Removed potentially unnecessary read of LSR and MSR
>   - Only attempt workaround when LCR write is ignored
> 
>  drivers/tty/serial/8250/8250_dw.c | 41 ++++++++++++++++++++++++++++++---------
>  1 file changed, 32 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/tty/serial/8250/8250_dw.c b/drivers/tty/serial/8250/8250_dw.c
> index d04a037..4658e3e 100644
> --- a/drivers/tty/serial/8250/8250_dw.c
> +++ b/drivers/tty/serial/8250/8250_dw.c
> @@ -57,7 +57,6 @@
>  
>  struct dw8250_data {
>  	u8			usr_reg;
> -	int			last_lcr;
>  	int			last_mcr;
>  	int			line;
>  	struct clk		*clk;
> @@ -77,17 +76,33 @@ static inline int dw8250_modify_msr(struct uart_port *p, int offset, int value)
>  	return value;
>  }
>  
> +static void dw8250_force_idle(struct uart_port *p)
> +{
> +	serial8250_clear_and_reinit_fifos(container_of
> +					  (p, struct uart_8250_port, port));
> +	(void)p->serial_in(p, UART_RX);
> +}
> +
>  static void dw8250_serial_out(struct uart_port *p, int offset, int value)
>  {
>  	struct dw8250_data *d = p->private_data;
>  
> -	if (offset == UART_LCR)
> -		d->last_lcr = value;
> -
>  	if (offset == UART_MCR)
>  		d->last_mcr = value;
>  
>  	writeb(value, p->membase + (offset << p->regshift));
> +
> +	/* Make sure LCR write wasn't ignored */
> +	if (offset == UART_LCR) {
> +		int tries = 1000;
> +		while (tries--) {
> +			if (value == p->serial_in(p, UART_LCR))
> +				return;
> +			dw8250_force_idle(p);
> +			writeb(value, p->membase + (UART_LCR << p->regshift));
> +		}
> +		dev_err(p->dev, "Couldn't set LCR to %d\n", value);
> +	}
>  }
>  
>  static unsigned int dw8250_serial_in(struct uart_port *p, int offset)
> @@ -108,13 +123,22 @@ static void dw8250_serial_out32(struct uart_port *p, int offset, int value)
>  {
>  	struct dw8250_data *d = p->private_data;
>  
> -	if (offset == UART_LCR)
> -		d->last_lcr = value;
> -
>  	if (offset == UART_MCR)
>  		d->last_mcr = value;
>  
>  	writel(value, p->membase + (offset << p->regshift));
> +
> +	/* Make sure LCR write wasn't ignored */
> +	if (offset == UART_LCR) {
> +		int tries = 1000;
> +		while (tries--) {
> +			if (value == p->serial_in(p, UART_LCR))
> +				return;
> +			dw8250_force_idle(p);
> +			writel(value, p->membase + (UART_LCR << p->regshift));
> +		}
> +		dev_err(p->dev, "Couldn't set LCR to %d\n", value);
> +	}
>  }
>  
>  static unsigned int dw8250_serial_in32(struct uart_port *p, int offset)
> @@ -132,9 +156,8 @@ static int dw8250_handle_irq(struct uart_port *p)
>  	if (serial8250_handle_irq(p, iir)) {
>  		return 1;
>  	} else if ((iir & UART_IIR_BUSY) == UART_IIR_BUSY) {
> -		/* Clear the USR and write the LCR again. */
> +		/* Clear the USR */
>  		(void)p->serial_in(p, d->usr_reg);
> -		p->serial_out(p, UART_LCR, d->last_lcr);
>  
>  		return 1;
>  	}

Since v3.13-rc1, this commit seems to have introduced some oddities on
some of our boards. See this log snippet:

Serial: 8250/16550 driver, 4 ports, IRQ sharing disabled
????R?console [ttyS0] enabled
console [ttyS0] enabled
bootconsole [earlycon0] disabled
bootconsole [earlycon0] disabled
dw-apb-uart d0012100.serial: Couldn't set LCR to 191
dw-apb-uart d0012100.serial: Couldn't set LCR to 191
dw-apb-uart d0012100.serial: Couldn't set LCR to 224
dw-apb-uart d0012100.serial: Couldn't set LCR to 224
d0012100.serial: ttyS1 at MMIO 0xd0012100 (irq = 18, base_baud = 15625000) is a 16550A

This behavior appear in at least Armada 370 and Armada XP boxes.

I confirm reverting this commit fixes the issue and things get back to normal.
Here's the complete kernel log: sprunge.us/gMdL

Ideas?
-- 
Ezequiel Garc?a, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com