2014-11-30 11:51:52

by Lino Sanfilippo

Subject: [PATCH] skge: Unmask interrupts in case of spurious interrupts

In case of a spurious interrupt dont forget to reenable the interrupts that
have been masked by reading the interrupt source register.

Signed-off-by: Lino Sanfilippo <[email protected]>
---
drivers/net/ethernet/marvell/skge.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/marvell/skge.c b/drivers/net/ethernet/marvell/skge.c
index 264eab7..7173836 100644
--- a/drivers/net/ethernet/marvell/skge.c
+++ b/drivers/net/ethernet/marvell/skge.c
@@ -3433,10 +3433,9 @@ static irqreturn_t skge_intr(int irq, void *dev_id)

if (status & IS_HW_ERR)
skge_error_irq(hw);
-
+out:
skge_write32(hw, B0_IMSK, hw->intr_mask);
skge_read32(hw, B0_IMSK);
-out:
spin_unlock(&hw->hw_lock);

return IRQ_RETVAL(handled);
--
1.9.1

2014-12-01 09:53:44

by Mirko Lindner

Subject: Re: [PATCH] skge: Unmask interrupts in case of spurious interrupts

On 30/11/14 12:51, Lino Sanfilippo wrote:
> In case of a spurious interrupt dont forget to reenable the interrupts that
> have been masked by reading the interrupt source register.
>
> Signed-off-by: Lino Sanfilippo <[email protected]>
> ---
> drivers/net/ethernet/marvell/skge.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/net/ethernet/marvell/skge.c b/drivers/net/ethernet/marvell/skge.c
> index 264eab7..7173836 100644
> --- a/drivers/net/ethernet/marvell/skge.c
> +++ b/drivers/net/ethernet/marvell/skge.c
> @@ -3433,10 +3433,9 @@ static irqreturn_t skge_intr(int irq, void *dev_id)
>
> if (status & IS_HW_ERR)
> skge_error_irq(hw);
> -
> +out:
> skge_write32(hw, B0_IMSK, hw->intr_mask);
> skge_read32(hw, B0_IMSK);
> -out:
> spin_unlock(&hw->hw_lock);
>
> return IRQ_RETVAL(handled);
>

Looks OK.

Acked-by: Mirko Lindner <[email protected]>

2014-12-06 05:29:50

by David Miller

Subject: Re: [PATCH] skge: Unmask interrupts in case of spurious interrupts

From: Lino Sanfilippo <[email protected]>
Date: Sun, 30 Nov 2014 12:51:31 +0100

> In case of a spurious interrupt dont forget to reenable the interrupts that
> have been masked by reading the interrupt source register.
>
> Signed-off-by: Lino Sanfilippo <[email protected]>

Applied.