Spurious timeouts in mvmdio

On Tue, Dec 03, 2013 at 11:38:23PM +0000, Leigh Brown wrote:
> On 2013-12-03 23:17, Sebastian Hesselbarth wrote:
>> If you want to ensure timeout > 2, why not then just use:
>>
>> - 	unsigned long timeout = usecs_to_jiffies(MVMDIO_SMI_TIMEOUT);
>> + 	unsigned long timeout = 1 + usecs_to_jiffies(MVMDIO_SMI_TIMEOUT);
>
> This will make it correct when using interrupts but it will make the  
> loop wait one jiffie longer than it should when polling.

Alternatively, code it like this instead.

 drivers/net/ethernet/marvell/mvmdio.c |   32 +++++++++++++++-----------------
 1 files changed, 15 insertions(+), 17 deletions(-)

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On 12/04/2013 12:42 AM, Russell King - ARM Linux wrote:
>> This will make it correct when using interrupts but it will make the
>> loop wait one jiffie longer than it should when polling.
>
> Alternatively, code it like this instead.
>
>   drivers/net/ethernet/marvell/mvmdio.c |   32 +++++++++++++++-----------------
>   1 files changed, 15 insertions(+), 17 deletions(-)
>
> diff --git a/drivers/net/ethernet/marvell/mvmdio.c b/drivers/net/ethernet/marvell/mvmdio.c
> index 7354960b583b..a6f59831fc5b 100644
> --- a/drivers/net/ethernet/marvell/mvmdio.c
> +++ b/drivers/net/ethernet/marvell/mvmdio.c
> @@ -76,31 +76,29 @@ static int orion_mdio_wait_ready(struct mii_bus *bus)
>   {
>   	struct orion_mdio_dev *dev = bus->priv;
>   	unsigned long timeout = usecs_to_jiffies(MVMDIO_SMI_TIMEOUT);
> -	unsigned long end = jiffies + timeout;
> -	int timedout = 0;
>
> -	while (1) {
> -	        if (orion_mdio_smi_is_done(dev))
> +	if (dev->err_interrupt > 0) {
> +		if (wait_event_timeout(dev->smi_busy_wait,
> +				       orion_mdio_smi_is_done(dev),
> +				       1 + timeout))
>   			return 0;
> -	        else if (timedout)
> -			break;
> +	} else {
> +		unsigned long end = jiffies + timeout;
>
> -	        if (dev->err_interrupt <= 0) {
> -			usleep_range(MVMDIO_SMI_POLL_INTERVAL_MIN,
> -				     MVMDIO_SMI_POLL_INTERVAL_MAX);
> +		while (1) {
> +	        	if (orion_mdio_smi_is_done(dev))
> +				return 0;
>
>   			if (time_is_before_jiffies(end))
> -				++timedout;
> -	        } else {
> -			wait_event_timeout(dev->smi_busy_wait,
> -				           orion_mdio_smi_is_done(dev),
> -				           timeout);
> -
> -			++timedout;
> -	        }
> +				break;
> +
> +			usleep_range(MVMDIO_SMI_POLL_INTERVAL_MIN,
> +				     MVMDIO_SMI_POLL_INTERVAL_MAX);
> +		}
>   	}
>
>   	dev_err(bus->parent, "Timeout: SMI busy for too long\n");
> +
>   	return  -ETIMEDOUT;
>   }

Hi Russell,

I have just tested your patch on MV88F6281 and MV88F6282 both in polling and 
irq mode and it works just fine.

I had a similar patch almost ready to submit but you were faster than me.

Feel free to add my:

Tested-by: Nicolas Schichan <nschichan@freebox.fr>

Thanks,

On 12/04/2013 12:42 AM, Russell King - ARM Linux wrote:
> Alternatively, code it like this instead.
>
>   drivers/net/ethernet/marvell/mvmdio.c |   32 +++++++++++++++-----------------
>   1 files changed, 15 insertions(+), 17 deletions(-)
>
> diff --git a/drivers/net/ethernet/marvell/mvmdio.c b/drivers/net/ethernet/marvell/mvmdio.c
> index 7354960b583b..a6f59831fc5b 100644
> --- a/drivers/net/ethernet/marvell/mvmdio.c
> +++ b/drivers/net/ethernet/marvell/mvmdio.c
> @@ -76,31 +76,29 @@ static int orion_mdio_wait_ready(struct mii_bus *bus)
>   {
>   	struct orion_mdio_dev *dev = bus->priv;
>   	unsigned long timeout = usecs_to_jiffies(MVMDIO_SMI_TIMEOUT);
> -	unsigned long end = jiffies + timeout;
> -	int timedout = 0;
>
> -	while (1) {
> -	        if (orion_mdio_smi_is_done(dev))
> +	if (dev->err_interrupt > 0) {
> +		if (wait_event_timeout(dev->smi_busy_wait,
> +				       orion_mdio_smi_is_done(dev),
> +				       1 + timeout))
>   			return 0;
> -	        else if (timedout)
> -			break;
> +	} else {
> +		unsigned long end = jiffies + timeout;
>
> -	        if (dev->err_interrupt <= 0) {
> -			usleep_range(MVMDIO_SMI_POLL_INTERVAL_MIN,
> -				     MVMDIO_SMI_POLL_INTERVAL_MAX);
> +		while (1) {
> +	        	if (orion_mdio_smi_is_done(dev))
> +				return 0;
>
>   			if (time_is_before_jiffies(end))
> -				++timedout;
> -	        } else {
> -			wait_event_timeout(dev->smi_busy_wait,
> -				           orion_mdio_smi_is_done(dev),
> -				           timeout);
> -
> -			++timedout;
> -	        }
> +				break;
> +
> +			usleep_range(MVMDIO_SMI_POLL_INTERVAL_MIN,
> +				     MVMDIO_SMI_POLL_INTERVAL_MAX);
> +		}
>   	}
>
>   	dev_err(bus->parent, "Timeout: SMI busy for too long\n");
> +
>   	return  -ETIMEDOUT;
>   }

Hi Russell,

I did not find any commit for this in 3.13-rc4.

Would you prefer I submit my attempt at refactoring the 
orion_mdio_wait_ready() code ?

Regards,

On Mon, Dec 16, 2013 at 07:07:28PM +0100, Nicolas Schichan wrote:
> I did not find any commit for this in 3.13-rc4.
>
> Would you prefer I submit my attempt at refactoring the  
> orion_mdio_wait_ready() code ?

It's up to Jason.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html