Re: [PATCH] 8250 UART backup timer

[Alex Williamson]

   Ping.  Any other comments on this?  Thanks,

	Alex

On Tue, 2006-06-27 at 13:04 -0600, Alex Williamson wrote:
>    The patch below works around a minor bug found in the UART of the
> remote management card used in many HP ia64 and parisc servers (aka the
> Diva UARTs).  The problem is that the UART does not reassert the THRE
> interrupt if it has been previously cleared and the IIR THRI bit is
> re-enabled.  This can produce a very annoying failure mode when used as
> a serial console, allowing a boot/reboot to hang indefinitely until an
> RX interrupt kicks it into working again (ie. an unattended reboot could
> stall).
> 
>    To solve this problem, a backup timer is introduced that runs
> alongside the standard interrupt driven mechanism.  This timer wakes up
> periodically, checks for a hang condition and gets characters moving
> again.  This backup mechanism is only enabled if the UART is detected as
> having this problem, so systems without these UARTs will have no
> additional overhead.
> 
>    This version of the patch incorporates previous comments from Pavel
> and removes races in the bug detection code.  The test is now done
> before the irq linking to prevent races with interrupt handler clearing
> the THRE interrupt.  Short delays and syncs are also added to ensure the
> device is able to update register state before the result is tested.
> Comments?  Thanks,
> 
> 	Alex
> 
> Signed-off-by: Alex Williamson <alex.williamson@xxxxxx>
> ---
> 
> diff -r e3554576c29a drivers/serial/8250.c
> --- a/drivers/serial/8250.c	Sun Jun 18 02:00:07 2006 +0000
> +++ b/drivers/serial/8250.c	Tue Jun 27 12:27:03 2006 -0600
> @@ -353,6 +353,23 @@ serial_out(struct uart_8250_port *up, in
>  	}
>  }
>  
> +static void
> +serial_out_sync(struct uart_8250_port *up, int offset, int value)
> +{
> +	switch (up->port.iotype) {
> +	case UPIO_MEM:
> +	case UPIO_MEM32:
> +#ifdef CONFIG_SERIAL_8250_AU1X00
> +	case UPIO_AU:
> +#endif
> +		serial_out(up, offset, value);
> +		(void)serial_in(up, UART_LCR); /* safe, no side-effects */
> +		break;
> +	default:
> +		serial_out(up, offset, value);
> +	}
> +}
> +
>  /*
>   * We used to support using pause I/O for certain machines.  We
>   * haven't supported this for a while, but just in case it's badly
> @@ -1436,6 +1453,13 @@ static void serial_unlink_irq_chain(stru
>  	serial_do_unlink(i, up);
>  }
>  
> +/* Base timer interval for polling */
> +static inline int
> +poll_timeout(int timeout)
> +{
> +	return timeout > 6 ? (timeout / 2 - 2) : 1;
> +}
> +
>  /*
>   * This function is used to handle ports that do not have an
>   * interrupt.  This doesn't work very well for 16450's, but gives
> @@ -1445,16 +1469,52 @@ static void serial8250_timeout(unsigned 
>  static void serial8250_timeout(unsigned long data)
>  {
>  	struct uart_8250_port *up = (struct uart_8250_port *)data;
> -	unsigned int timeout;
>  	unsigned int iir;
>  
>  	iir = serial_in(up, UART_IIR);
>  	if (!(iir & UART_IIR_NO_INT))
>  		serial8250_handle_port(up, NULL);
>  
> -	timeout = up->port.timeout;
> -	timeout = timeout > 6 ? (timeout / 2 - 2) : 1;
> -	mod_timer(&up->timer, jiffies + timeout);
> +	mod_timer(&up->timer, jiffies + poll_timeout(up->port.timeout));
> +}
> +
> +static void serial8250_backup_timeout(unsigned long data)
> +{
> +	struct uart_8250_port *up = (struct uart_8250_port *)data;
> +	unsigned int iir, ier = 0;
> +
> +	/*
> +	 * Must disable interrupts or else we risk racing with the interrupt
> +	 * based handler.
> +	 */
> +	if (is_real_interrupt(up->port.irq)) {
> +		ier = serial_in(up, UART_IER);
> +		serial_out(up, UART_IER, 0);
> +	}
> +
> +	iir = serial_in(up, UART_IIR);
> +
> +	/*
> +	 * This should be a safe test for anyone who doesn't trust the
> +	 * IIR bits on their UART, but it's specifically designed for
> +	 * the "Diva" UART used on the management processor on many HP
> +	 * ia64 and parisc boxes.
> +	 */
> +	if ((iir & UART_IIR_NO_INT) && (up->ier & UART_IER_THRI) &&
> +	    (!uart_circ_empty(&up->port.info->xmit) || up->port.x_char) &&
> +	    (serial_in(up, UART_LSR) & UART_LSR_THRE)) {
> +		iir &= ~(UART_IIR_ID | UART_IIR_NO_INT);
> +		iir |= UART_IIR_THRI;
> +	}
> +
> +	if (!(iir & UART_IIR_NO_INT))
> +		serial8250_handle_port(up, NULL);
> +
> +	if (is_real_interrupt(up->port.irq))
> +		serial_out(up, UART_IER, ier);
> +
> +	/* Standard timer interval plus 0.2s to keep the port running */
> +	mod_timer(&up->timer, jiffies + poll_timeout(up->port.timeout) + HZ/5);
>  }
>  
>  static unsigned int serial8250_tx_empty(struct uart_port *port)
> @@ -1523,6 +1583,36 @@ static void serial8250_break_ctl(struct 
>  		up->lcr &= ~UART_LCR_SBC;
>  	serial_out(up, UART_LCR, up->lcr);
>  	spin_unlock_irqrestore(&up->port.lock, flags);
> +}
> +
> +#define BOTH_EMPTY (UART_LSR_TEMT | UART_LSR_THRE)
> +
> +/*
> + *	Wait for transmitter & holding register to empty
> + */
> +static inline void wait_for_xmitr(struct uart_8250_port *up, int bits)
> +{
> +	unsigned int status, timeout = 10000;
> +
> +	/* Wait up to 10ms for the character(s) to be sent. */
> +	do {
> +		status = serial_in(up, UART_LSR);
> +
> +		if (status & UART_LSR_BI)
> +			up->lsr_break_flag = UART_LSR_BI;
> +
> +		if (--timeout == 0)
> +			break;
> +		udelay(1);
> +	} while ((status & bits) != bits);
> +
> +	/* Wait up to 1s for flow control if necessary */
> +	if (up->port.flags & UPF_CONS_FLOW) {
> +		timeout = 1000000;
> +		while (--timeout &&
> +		       ((serial_in(up, UART_MSR) & UART_MSR_CTS) == 0))
> +			udelay(1);
> +	}
>  }
>  
>  static int serial8250_startup(struct uart_port *port)
> @@ -1598,18 +1688,50 @@ static int serial8250_startup(struct uar
>  		serial_outp(up, UART_LCR, 0);
>  	}
>  
> +	if (is_real_interrupt(up->port.irq)) {
> +		/*
> +		 * Test for UARTs that do not reassert THRE when the
> +		 * transmitter is idle and the interrupt has already
> +		 * been cleared.  Real 16550s should always reassert
> +		 * this interrupt whenever the transmitter is idle and
> +		 * the interrupt is enabled.  Delays are necessary to
> +		 * allow register changes to become visible.
> +		 */
> +		spin_lock_irqsave(&up->port.lock, flags);
> +
> +		wait_for_xmitr(up, UART_LSR_THRE);
> +		serial_out_sync(up, UART_IER, UART_IER_THRI);
> +		udelay(1); /* allow THRE to set */
> +		(void)serial_in(up, UART_IIR);
> +		serial_out(up, UART_IER, 0);
> +		serial_out_sync(up, UART_IER, UART_IER_THRI);
> +		udelay(1); /* allow a working UART time to re-assert THRE */
> +		iir = serial_in(up, UART_IIR);
> +		serial_out(up, UART_IER, 0);
> +
> +		spin_unlock_irqrestore(&up->port.lock, flags);
> +
> +		/*
> +		 * If the interrupt is not reasserted, setup a timer to
> +		 * kick the UART on a regular basis.
> +		 */
> +		if (iir & UART_IIR_NO_INT) {
> +			pr_debug("ttyS%d - using backup timer\n", port->line);
> +			up->timer.function = serial8250_backup_timeout;
> +			up->timer.data = (unsigned long)up;
> +			mod_timer(&up->timer, jiffies +
> +			          poll_timeout(up->port.timeout) + HZ/5);
> +		}
> +	}
> +
>  	/*
>  	 * If the "interrupt" for this port doesn't correspond with any
>  	 * hardware interrupt, we use a timer-based system.  The original
>  	 * driver used to do this with IRQ0.
>  	 */
>  	if (!is_real_interrupt(up->port.irq)) {
> -		unsigned int timeout = up->port.timeout;
> -
> -		timeout = timeout > 6 ? (timeout / 2 - 2) : 1;
> -
>  		up->timer.data = (unsigned long)up;
> -		mod_timer(&up->timer, jiffies + timeout);
> +		mod_timer(&up->timer, jiffies + poll_timeout(up->port.timeout));
>  	} else {
>  		retval = serial_link_irq_chain(up);
>  		if (retval)
> @@ -1725,9 +1847,9 @@ static void serial8250_shutdown(struct u
>  	 */
>  	(void) serial_in(up, UART_RX);
>  
> -	if (!is_real_interrupt(up->port.irq))
> -		del_timer_sync(&up->timer);
> -	else
> +	del_timer_sync(&up->timer);
> +	up->timer.function = serial8250_timeout;
> +	if (is_real_interrupt(up->port.irq))
>  		serial_unlink_irq_chain(up);
>  }
>  
> @@ -2187,36 +2309,6 @@ serial8250_register_ports(struct uart_dr
>  
>  #ifdef CONFIG_SERIAL_8250_CONSOLE
>  
> -#define BOTH_EMPTY (UART_LSR_TEMT | UART_LSR_THRE)
> -
> -/*
> - *	Wait for transmitter & holding register to empty
> - */
> -static inline void wait_for_xmitr(struct uart_8250_port *up, int bits)
> -{
> -	unsigned int status, tmout = 10000;
> -
> -	/* Wait up to 10ms for the character(s) to be sent. */
> -	do {
> -		status = serial_in(up, UART_LSR);
> -
> -		if (status & UART_LSR_BI)
> -			up->lsr_break_flag = UART_LSR_BI;
> -
> -		if (--tmout == 0)
> -			break;
> -		udelay(1);
> -	} while ((status & bits) != bits);
> -
> -	/* Wait up to 1s for flow control if necessary */
> -	if (up->port.flags & UPF_CONS_FLOW) {
> -		tmout = 1000000;
> -		while (--tmout &&
> -		       ((serial_in(up, UART_MSR) & UART_MSR_CTS) == 0))
> -			udelay(1);
> -	}
> -}
> -
>  static void serial8250_console_putchar(struct uart_port *port, int ch)
>  {
>  	struct uart_8250_port *up = (struct uart_8250_port *)port;
> 
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 
-- 
Alex Williamson                             HP Open Source & Linux Org.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/