[PATCH] virtio_net: zero is an invald queue_pairs number

[PATCH] virtio_net: zero is an invald queue_pairs number

[Amos Kong]

Execute "ethtool -L eth0 combined 0" in guest, if multiqueue
is enabled, virtnet_send_command() will return -EINVAL error,
there is a validation in QEMU.

But if multiqueue is disabled, virtnet_set_queues() will just
return zero (success). We should return error for this situation.

Signed-off-by: Amos Kong <akong@redhat.com>
---
 drivers/net/virtio_net.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 7b68746..8a852b5 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -1285,7 +1285,7 @@ static int virtnet_set_channels(struct net_device *dev,
 	if (channels->rx_count || channels->tx_count || channels->other_count)
 		return -EINVAL;
 
-	if (queue_pairs > vi->max_queue_pairs)
+	if (queue_pairs > vi->max_queue_pairs || queue_pairs == 0)
 		return -EINVAL;
 
 	get_online_cpus();
-- 
1.9.0

Re: [PATCH] virtio_net: zero is an invald queue_pairs number

[Jason Wang]

On 04/18/2014 01:45 PM, Amos Kong wrote:
> Execute "ethtool -L eth0 combined 0" in guest, if multiqueue
> is enabled, virtnet_send_command() will return -EINVAL error,
> there is a validation in QEMU.
>
> But if multiqueue is disabled, virtnet_set_queues() will just
> return zero (success). We should return error for this situation.
>
> Signed-off-by: Amos Kong <akong@redhat.com>
> ---
>  drivers/net/virtio_net.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index 7b68746..8a852b5 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -1285,7 +1285,7 @@ static int virtnet_set_channels(struct net_device *dev,
>  	if (channels->rx_count || channels->tx_count || channels->other_count)
>  		return -EINVAL;
>  
> -	if (queue_pairs > vi->max_queue_pairs)
> +	if (queue_pairs > vi->max_queue_pairs || queue_pairs == 0)
>  		return -EINVAL;
>  
>  	get_online_cpus();

Acked-by:  Jason Wang <jasowang@redhat.com>

Re: [PATCH] virtio_net: zero is an invald queue_pairs number

[Jason Wang]

On 04/21/2014 02:32 PM, Jason Wang wrote:
> On 04/18/2014 01:45 PM, Amos Kong wrote:
>> Execute "ethtool -L eth0 combined 0" in guest, if multiqueue
>> is enabled, virtnet_send_command() will return -EINVAL error,
>> there is a validation in QEMU.
>>
>> But if multiqueue is disabled, virtnet_set_queues() will just
>> return zero (success). We should return error for this situation.
>>
>> Signed-off-by: Amos Kong <akong@redhat.com>
>> ---
>>  drivers/net/virtio_net.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
>> index 7b68746..8a852b5 100644
>> --- a/drivers/net/virtio_net.c
>> +++ b/drivers/net/virtio_net.c
>> @@ -1285,7 +1285,7 @@ static int virtnet_set_channels(struct net_device *dev,
>>  	if (channels->rx_count || channels->tx_count || channels->other_count)
>>  		return -EINVAL;
>>  
>> -	if (queue_pairs > vi->max_queue_pairs)
>> +	if (queue_pairs > vi->max_queue_pairs || queue_pairs == 0)
>>  		return -EINVAL;
>>  
>>  	get_online_cpus();
> Acked-by:  Jason Wang <jasowang@redhat.com>

Looking at virtnet_set_queues(), it will return success even if no cvq
and no mq support. This is wrong. We need return error in this case.
We'd better fix it and check the its return value in virtnet_restore().

With this fix, there's no need for us to check queue_pairs against zero
like what this patch does.

Re: [PATCH] virtio_net: zero is an invald queue_pairs number

[Amos Kong]

On Mon, Apr 21, 2014 at 02:45:25PM +0800, Jason Wang wrote:
> On 04/21/2014 02:32 PM, Jason Wang wrote:
> > On 04/18/2014 01:45 PM, Amos Kong wrote:
> >> Execute "ethtool -L eth0 combined 0" in guest, if multiqueue
> >> is enabled, virtnet_send_command() will return -EINVAL error,
> >> there is a validation in QEMU.
> >>
> >> But if multiqueue is disabled, virtnet_set_queues() will just
> >> return zero (success). We should return error for this situation.
> >>
> >> Signed-off-by: Amos Kong <akong@redhat.com>
> >> ---
> >>  drivers/net/virtio_net.c | 2 +-
> >>  1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> >> index 7b68746..8a852b5 100644
> >> --- a/drivers/net/virtio_net.c
> >> +++ b/drivers/net/virtio_net.c
> >> @@ -1285,7 +1285,7 @@ static int virtnet_set_channels(struct net_device *dev,
> >>  	if (channels->rx_count || channels->tx_count || channels->other_count)
> >>  		return -EINVAL;
> >>  
> >> -	if (queue_pairs > vi->max_queue_pairs)
> >> +	if (queue_pairs > vi->max_queue_pairs || queue_pairs == 0)
> >>  		return -EINVAL;
> >>  
> >>  	get_online_cpus();
> > Acked-by:  Jason Wang <jasowang@redhat.com>
> 
> Looking at virtnet_set_queues(), it will return success even if no cvq
> and no mq support. This is wrong. We need return error in this case.
> We'd better fix it and check the its return value in virtnet_restore().
> 
> With this fix, there's no need for us to check queue_pairs against zero
> like what this patch does.
 
As we talked in IRC, the problem I tried to fix is actually caused by
that we don't return error when MQ isn't support. I will post another
patch to fix it.

However, this fix is also useful, it will catch the error early.

-- 
			Amos.

Re: [PATCH] virtio_net: zero is an invald queue_pairs number

[David Miller]

From: Amos Kong <akong@redhat.com>
Date: Fri, 18 Apr 2014 13:45:41 +0800

> Execute "ethtool -L eth0 combined 0" in guest, if multiqueue
> is enabled, virtnet_send_command() will return -EINVAL error,
> there is a validation in QEMU.
> 
> But if multiqueue is disabled, virtnet_set_queues() will just
> return zero (success). We should return error for this situation.
> 
> Signed-off-by: Amos Kong <akong@redhat.com>

Applied, thanks.