[PATCH 1/3] net: lantiq: Wake TX queue again

[PATCH 1/3] net: lantiq: Wake TX queue again

[Hauke Mehrtens]

The call to netif_wake_queue() when the TX descriptors were freed was
missing. When there are no TX buffers available the TX queue will be
stopped, but it was not started again when they are available again,
this is fixed in this patch.

Fixes: fe1a56420cf2 ("net: lantiq: Add Lantiq / Intel VRX200 Ethernet driver")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
---
 drivers/net/ethernet/lantiq_xrx200.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
index 1645e4e7ebdb..1feb9fc710e0 100644
--- a/drivers/net/ethernet/lantiq_xrx200.c
+++ b/drivers/net/ethernet/lantiq_xrx200.c
@@ -268,6 +268,9 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget)
 	net_dev->stats.tx_bytes += bytes;
 	netdev_completed_queue(ch->priv->net_dev, pkts, bytes);
 
+	if (netif_queue_stopped(net_dev))
+		netif_wake_queue(net_dev);
+
 	if (pkts < budget) {
 		napi_complete(&ch->napi);
 		ltq_dma_enable_irq(&ch->dma);
-- 
2.20.1

[PATCH 2/3] net: lantiq: use netif_tx_napi_add() for TX NAPI

[Hauke Mehrtens]

netif_tx_napi_add() should be used for NAPI in the TX direction instead
of the netif_napi_add() function.

Fixes: fe1a56420cf2 ("net: lantiq: Add Lantiq / Intel VRX200 Ethernet driver")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
---
 drivers/net/ethernet/lantiq_xrx200.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
index 1feb9fc710e0..f34e4dc8c661 100644
--- a/drivers/net/ethernet/lantiq_xrx200.c
+++ b/drivers/net/ethernet/lantiq_xrx200.c
@@ -502,7 +502,7 @@ static int xrx200_probe(struct platform_device *pdev)
 
 	/* setup NAPI */
 	netif_napi_add(net_dev, &priv->chan_rx.napi, xrx200_poll_rx, 32);
-	netif_napi_add(net_dev, &priv->chan_tx.napi, xrx200_tx_housekeeping, 32);
+	netif_tx_napi_add(net_dev, &priv->chan_tx.napi, xrx200_tx_housekeeping, 32);
 
 	platform_set_drvdata(pdev, priv);
 
-- 
2.20.1

[PATCH 3/3] net: lantiq: Use napi_complete_done()

[Hauke Mehrtens]

Use napi_complete_done() and activate the interrupts when this function
returns true. This way the generic NAPI code can take care of activating
the interrupts.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
---
 drivers/net/ethernet/lantiq_xrx200.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
index f34e4dc8c661..674ffb2ecd9a 100644
--- a/drivers/net/ethernet/lantiq_xrx200.c
+++ b/drivers/net/ethernet/lantiq_xrx200.c
@@ -229,10 +229,8 @@ static int xrx200_poll_rx(struct napi_struct *napi, int budget)
 		}
 	}
 
-	if (rx < budget) {
-		napi_complete(&ch->napi);
+	if (napi_complete_done(&ch->napi, rx))
 		ltq_dma_enable_irq(&ch->dma);
-	}
 
 	return rx;
 }
@@ -271,10 +269,8 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget)
 	if (netif_queue_stopped(net_dev))
 		netif_wake_queue(net_dev);
 
-	if (pkts < budget) {
-		napi_complete(&ch->napi);
+	if (napi_complete_done(&ch->napi, pkts))
 		ltq_dma_enable_irq(&ch->dma);
-	}
 
 	return pkts;
 }
-- 
2.20.1

Re: [PATCH 2/3] net: lantiq: use netif_tx_napi_add() for TX NAPI

[Eric Dumazet]

On 8/15/20 11:33 AM, Hauke Mehrtens wrote:
> netif_tx_napi_add() should be used for NAPI in the TX direction instead
> of the netif_napi_add() function.
> 
> Fixes: fe1a56420cf2 ("net: lantiq: Add Lantiq / Intel VRX200 Ethernet driver")
> Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
> ---
>  drivers/net/ethernet/lantiq_xrx200.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
> index 1feb9fc710e0..f34e4dc8c661 100644
> --- a/drivers/net/ethernet/lantiq_xrx200.c
> +++ b/drivers/net/ethernet/lantiq_xrx200.c
> @@ -502,7 +502,7 @@ static int xrx200_probe(struct platform_device *pdev)
>  
>  	/* setup NAPI */
>  	netif_napi_add(net_dev, &priv->chan_rx.napi, xrx200_poll_rx, 32);
> -	netif_napi_add(net_dev, &priv->chan_tx.napi, xrx200_tx_housekeeping, 32);
> +	netif_tx_napi_add(net_dev, &priv->chan_tx.napi, xrx200_tx_housekeeping, 32);
>  
>  	platform_set_drvdata(pdev, priv);
>  
> 


This is not a must, simply a matter of (small) optimization in
the very rare case of busy polling users.

The Fixes: tag here is not necessary.

Re: [PATCH 3/3] net: lantiq: Use napi_complete_done()

[Eric Dumazet]

On 8/15/20 11:33 AM, Hauke Mehrtens wrote:
> Use napi_complete_done() and activate the interrupts when this function
> returns true. This way the generic NAPI code can take care of activating
> the interrupts.
> 
> Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
> ---
>  drivers/net/ethernet/lantiq_xrx200.c | 8 ++------
>  1 file changed, 2 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
> index f34e4dc8c661..674ffb2ecd9a 100644
> --- a/drivers/net/ethernet/lantiq_xrx200.c
> +++ b/drivers/net/ethernet/lantiq_xrx200.c
> @@ -229,10 +229,8 @@ static int xrx200_poll_rx(struct napi_struct *napi, int budget)
>  		}
>  	}
>  
> -	if (rx < budget) {
> -		napi_complete(&ch->napi);
> +	if (napi_complete_done(&ch->napi, rx))
>  		ltq_dma_enable_irq(&ch->dma);
> -	}
>  
>  	return rx;
>  }
> @@ -271,10 +269,8 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget)
>  	if (netif_queue_stopped(net_dev))
>  		netif_wake_queue(net_dev);
>  
> -	if (pkts < budget) {
> -		napi_complete(&ch->napi);
> +	if (napi_complete_done(&ch->napi, pkts))
>  		ltq_dma_enable_irq(&ch->dma);
> -	}
>  
>  	return pkts;
>  }
> 


This looks buggy to me.

Please look again to other implementations for a correct usage.

Re: [PATCH 3/3] net: lantiq: Use napi_complete_done()

[Hauke Mehrtens]

[-- Attachment #1.1: Type: text/plain, Size: 1956 bytes --]

On 8/16/20 8:07 PM, Eric Dumazet wrote:
> 
> 
> On 8/15/20 11:33 AM, Hauke Mehrtens wrote:
>> Use napi_complete_done() and activate the interrupts when this function
>> returns true. This way the generic NAPI code can take care of activating
>> the interrupts.
>>
>> Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
>> ---
>>  drivers/net/ethernet/lantiq_xrx200.c | 8 ++------
>>  1 file changed, 2 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
>> index f34e4dc8c661..674ffb2ecd9a 100644
>> --- a/drivers/net/ethernet/lantiq_xrx200.c
>> +++ b/drivers/net/ethernet/lantiq_xrx200.c
>> @@ -229,10 +229,8 @@ static int xrx200_poll_rx(struct napi_struct *napi, int budget)
>>  		}
>>  	}
>>  
>> -	if (rx < budget) {
>> -		napi_complete(&ch->napi);
>> +	if (napi_complete_done(&ch->napi, rx))
>>  		ltq_dma_enable_irq(&ch->dma);
>> -	}
>>  
>>  	return rx;
>>  }
>> @@ -271,10 +269,8 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget)
>>  	if (netif_queue_stopped(net_dev))
>>  		netif_wake_queue(net_dev);
>>  
>> -	if (pkts < budget) {
>> -		napi_complete(&ch->napi);
>> +	if (napi_complete_done(&ch->napi, pkts))
>>  		ltq_dma_enable_irq(&ch->dma);
>> -	}
>>  
>>  	return pkts;
>>  }
>>
> 
> 
> This looks buggy to me.

Hi Eric,

Thanks for looking at the patch.

What exactly looks buggy to you?

We see with and also without this patch problems in the TX path, it
looks like the netif tx queue is not started again.

> Please look again to other implementations for a correct usage.

I looked at multiple drivers and they look similar in this area.
For example microchip/lan743x_main.c is looking similar to me.

The hardware has two interrupts one for the RX and one for TX complete.

Can you please suggest a driver which uses the NAPI in a good way and is
not very complex.

Hauke


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH 3/3] net: lantiq: Use napi_complete_done()

[Eric Dumazet]

On 8/17/20 2:17 PM, Hauke Mehrtens wrote:
> On 8/16/20 8:07 PM, Eric Dumazet wrote:
>>
>>
>> On 8/15/20 11:33 AM, Hauke Mehrtens wrote:
>>> Use napi_complete_done() and activate the interrupts when this function
>>> returns true. This way the generic NAPI code can take care of activating
>>> the interrupts.
>>>
>>> Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
>>> ---
>>>  drivers/net/ethernet/lantiq_xrx200.c | 8 ++------
>>>  1 file changed, 2 insertions(+), 6 deletions(-)
>>>
>>> diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
>>> index f34e4dc8c661..674ffb2ecd9a 100644
>>> --- a/drivers/net/ethernet/lantiq_xrx200.c
>>> +++ b/drivers/net/ethernet/lantiq_xrx200.c
>>> @@ -229,10 +229,8 @@ static int xrx200_poll_rx(struct napi_struct *napi, int budget)
>>>  		}
>>>  	}
>>>  
>>> -	if (rx < budget) {
>>> -		napi_complete(&ch->napi);
>>> +	if (napi_complete_done(&ch->napi, rx))
>>>  		ltq_dma_enable_irq(&ch->dma);
>>> -	}
>>>  
>>>  	return rx;
>>>  }
>>> @@ -271,10 +269,8 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget)
>>>  	if (netif_queue_stopped(net_dev))
>>>  		netif_wake_queue(net_dev);
>>>  
>>> -	if (pkts < budget) {
>>> -		napi_complete(&ch->napi);
>>> +	if (napi_complete_done(&ch->napi, pkts))
>>>  		ltq_dma_enable_irq(&ch->dma);
>>> -	}
>>>  
>>>  	return pkts;
>>>  }
>>>
>>
>>
>> This looks buggy to me.
> 
> Hi Eric,
> 
> Thanks for looking at the patch.
> 
> What exactly looks buggy to you?

You removed the " if (rx < budget) "

But you must not.

Drivers have to keep the test.

Something like that seems more correct :

diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
index 1645e4e7ebdbb3c7abff8fe4207273df20f123d4..e3d617d387ed0f5593c3ba81d1d531d463bb5a6e 100644
--- a/drivers/net/ethernet/lantiq_xrx200.c
+++ b/drivers/net/ethernet/lantiq_xrx200.c
@@ -230,8 +230,8 @@ static int xrx200_poll_rx(struct napi_struct *napi, int budget)
        }
 
        if (rx < budget) {
-               napi_complete(&ch->napi);
-               ltq_dma_enable_irq(&ch->dma);
+               if (napi_complete(&ch->napi, rx))
+                       ltq_dma_enable_irq(&ch->dma);
        }
 
        return rx;
@@ -269,8 +269,8 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget)
        netdev_completed_queue(ch->priv->net_dev, pkts, bytes);
 
        if (pkts < budget) {
-               napi_complete(&ch->napi);
-               ltq_dma_enable_irq(&ch->dma);
+               if (napi_complete(&ch->napi, pkts))
+                       ltq_dma_enable_irq(&ch->dma);
        }
 
        return pkts;

Re: [PATCH 3/3] net: lantiq: Use napi_complete_done()

[Eric Dumazet]

On 8/17/20 5:33 PM, Eric Dumazet wrote:
> 
> 
> On 8/17/20 2:17 PM, Hauke Mehrtens wrote:
>> On 8/16/20 8:07 PM, Eric Dumazet wrote:
>>>
>>>
>>> On 8/15/20 11:33 AM, Hauke Mehrtens wrote:
>>>> Use napi_complete_done() and activate the interrupts when this function
>>>> returns true. This way the generic NAPI code can take care of activating
>>>> the interrupts.
>>>>
>>>> Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
>>>> ---
>>>>  drivers/net/ethernet/lantiq_xrx200.c | 8 ++------
>>>>  1 file changed, 2 insertions(+), 6 deletions(-)
>>>>
>>>> diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
>>>> index f34e4dc8c661..674ffb2ecd9a 100644
>>>> --- a/drivers/net/ethernet/lantiq_xrx200.c
>>>> +++ b/drivers/net/ethernet/lantiq_xrx200.c
>>>> @@ -229,10 +229,8 @@ static int xrx200_poll_rx(struct napi_struct *napi, int budget)
>>>>  		}
>>>>  	}
>>>>  
>>>> -	if (rx < budget) {
>>>> -		napi_complete(&ch->napi);
>>>> +	if (napi_complete_done(&ch->napi, rx))
>>>>  		ltq_dma_enable_irq(&ch->dma);
>>>> -	}
>>>>  
>>>>  	return rx;
>>>>  }
>>>> @@ -271,10 +269,8 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget)
>>>>  	if (netif_queue_stopped(net_dev))
>>>>  		netif_wake_queue(net_dev);
>>>>  
>>>> -	if (pkts < budget) {
>>>> -		napi_complete(&ch->napi);
>>>> +	if (napi_complete_done(&ch->napi, pkts))
>>>>  		ltq_dma_enable_irq(&ch->dma);
>>>> -	}
>>>>  
>>>>  	return pkts;
>>>>  }
>>>>
>>>
>>>
>>> This looks buggy to me.
>>
>> Hi Eric,
>>
>> Thanks for looking at the patch.
>>
>> What exactly looks buggy to you?
> 
> You removed the " if (rx < budget) "
> 
> But you must not.
> 
> Drivers have to keep the test.
> 
> Something like that seems more correct :
> 
> diff --git a/drivers/net/ethernet/lantiq_xrx200.c b/drivers/net/ethernet/lantiq_xrx200.c
> index 1645e4e7ebdbb3c7abff8fe4207273df20f123d4..e3d617d387ed0f5593c3ba81d1d531d463bb5a6e 100644
> --- a/drivers/net/ethernet/lantiq_xrx200.c
> +++ b/drivers/net/ethernet/lantiq_xrx200.c
> @@ -230,8 +230,8 @@ static int xrx200_poll_rx(struct napi_struct *napi, int budget)
>         }
>  
>         if (rx < budget) {
> -               napi_complete(&ch->napi);
> -               ltq_dma_enable_irq(&ch->dma);
> +               if (napi_complete(&ch->napi, rx))

Obviously : s/napi_complete/napi_complete_done/

> +                       ltq_dma_enable_irq(&ch->dma);
>         }
>  
>         return rx;
> @@ -269,8 +269,8 @@ static int xrx200_tx_housekeeping(struct napi_struct *napi, int budget)
>         netdev_completed_queue(ch->priv->net_dev, pkts, bytes);
>  
>         if (pkts < budget) {
> -               napi_complete(&ch->napi);
> -               ltq_dma_enable_irq(&ch->dma);
> +               if (napi_complete(&ch->napi, pkts))

Same : s/napi_complete/napi_complete_done/


> +                       ltq_dma_enable_irq(&ch->dma);
>         }
>  
>         return pkts;
>