Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1284614 > unrolled thread

Re: [PATCH LINUX v3 08/10] tty: xuartps: Move RX path into helper function

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH LINUX v3 08/10] tty: xuartps: Move RX path into helper  function Peter Hurley <peter@hurleysoftware.com> - 2015-12-05 18:50 +0100
    Re: [PATCH LINUX v3 08/10] tty: xuartps: Move RX path into helper  function Sören Brinkmann <soren.brinkmann@xilinx.com> - 2015-12-05 22:50 +0100

#1284614 — Re: [PATCH LINUX v3 08/10] tty: xuartps: Move RX path into helper function

On 11/21/2015 09:59 PM, Soren Brinkmann wrote:
> Move RX-related IRQ handling into a helper function.
Fixes a problem where every char received after a parity or frame error
in the current isr will also be tagged as a parity or frame error.

Reviewed-by: Peter Hurley <peter@hurleysoftware.com>

NB: the sysrq problem in cdns_uart_isr() with needing to drop the
locks is because cdns_uart_console_write() tries to take the port->lock.
The 8250 driver handles this problem by not trying to take the
port->lock if port->sysrq is non-zero. See serial8250_console_write().


> Signed-off-by: Soren Brinkmann <soren.brinkmann@xilinx.com>
> ---
>  drivers/tty/serial/xilinx_uartps.c | 50 +++++++++++++++++++++-----------------
>  1 file changed, 28 insertions(+), 22 deletions(-)
> 
> diff --git a/drivers/tty/serial/xilinx_uartps.c b/drivers/tty/serial/xilinx_uartps.c
> index f3ac69387b0a..db9e23eaf300 100644
> --- a/drivers/tty/serial/xilinx_uartps.c
> +++ b/drivers/tty/serial/xilinx_uartps.c
> @@ -172,28 +172,8 @@ struct cdns_uart {
>  #define to_cdns_uart(_nb) container_of(_nb, struct cdns_uart, \
>  		clk_rate_change_nb);
>  
> -/**
> - * cdns_uart_isr - Interrupt handler
> - * @irq: Irq number
> - * @dev_id: Id of the port
> - *
> - * Return: IRQHANDLED
> - */
> -static irqreturn_t cdns_uart_isr(int irq, void *dev_id)
> +static void cdns_uart_handle_rx(struct uart_port *port, unsigned int isrstatus)
>  {
> -	struct uart_port *port = (struct uart_port *)dev_id;
> -	unsigned long flags;
> -	unsigned int isrstatus, numbytes;
> -	unsigned int data;
> -	char status = TTY_NORMAL;

See, the previous code never reset the status back to TTY_NORMAL after any other
status.

> -
> -	spin_lock_irqsave(&port->lock, flags);
> -
> -	/* Read the interrupt status register to determine which
> -	 * interrupt(s) is/are active.
> -	 */
> -	isrstatus = readl(port->membase + CDNS_UART_ISR_OFFSET);
> -
>  	/*
>  	 * There is no hardware break detection, so we interpret framing
>  	 * error with all-zeros data as a break sequence. Most of the time,
> @@ -223,6 +203,9 @@ static irqreturn_t cdns_uart_isr(int irq, void *dev_id)
>  		/* Receive Timeout Interrupt */
>  		while (!(readl(port->membase + CDNS_UART_SR_OFFSET) &
>  					CDNS_UART_SR_RXEMPTY)) {
> +			u32 data;
> +			char status = TTY_NORMAL;
> +
>  			data = readl(port->membase + CDNS_UART_FIFO_OFFSET);
>  
>  			/* Non-NULL byte after BREAK is garbage (99%) */
> @@ -263,10 +246,33 @@ static irqreturn_t cdns_uart_isr(int irq, void *dev_id)
>  			}
>  
>  			uart_insert_char(port, isrstatus, CDNS_UART_IXR_OVERRUN,
> -					data, status);
> +					 data, status);
>  		}
>  		tty_flip_buffer_push(&port->state->port);
>  	}
> +}
> +
> +/**
> + * cdns_uart_isr - Interrupt handler
> + * @irq: Irq number
> + * @dev_id: Id of the port
> + *
> + * Return: IRQHANDLED
> + */
> +static irqreturn_t cdns_uart_isr(int irq, void *dev_id)
> +{
> +	struct uart_port *port = (struct uart_port *)dev_id;
> +	unsigned long flags;
> +	unsigned int isrstatus, numbytes;
> +
> +	spin_lock_irqsave(&port->lock, flags);
> +
> +	/* Read the interrupt status register to determine which
> +	 * interrupt(s) is/are active.
> +	 */
> +	isrstatus = readl(port->membase + CDNS_UART_ISR_OFFSET);
> +
> +	cdns_uart_handle_rx(port, isrstatus);
>  
>  	/* Dispatch an appropriate handler */
>  	if ((isrstatus & CDNS_UART_IXR_TXEMPTY) == CDNS_UART_IXR_TXEMPTY) {
> 

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

[toc] | [next] | [standalone]

#1284651

On Sat, 2015-12-05 at 12:40PM -0500, Peter Hurley wrote:
> On 11/21/2015 09:59 PM, Soren Brinkmann wrote:
> > Move RX-related IRQ handling into a helper function.
> Fixes a problem where every char received after a parity or frame error
> in the current isr will also be tagged as a parity or frame error.
> 
> Reviewed-by: Peter Hurley <peter@hurleysoftware.com>

Thanks. I'll add your text to the commit message.

> 
> NB: the sysrq problem in cdns_uart_isr() with needing to drop the
> locks is because cdns_uart_console_write() tries to take the port->lock.
> The 8250 driver handles this problem by not trying to take the
> port->lock if port->sysrq is non-zero. See serial8250_console_write().

I'll look into this. I'll see if I can include that in the next
iteration of this series or do it later.

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

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web