On Fri, 25 May 2018 16:06:58 -0700
"Samudrala, Sridhar" wrote:
> On 5/25/2018 3:38 PM, Stephen Hemminger wrote:
> > On Thu, 24 May 2018 09:55:13 -0700
> > Sridhar Samudrala wrote:
> >
> >> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
> >> index 03ed492c4e14..0f4ba52b641d
Sat, May 26, 2018 at 12:37:44AM CEST, step...@networkplumber.org wrote:
>On Thu, 24 May 2018 09:55:13 -0700
>Sridhar Samudrala wrote:
>
>
>> +spin_lock(&failover_lock);
>
>Since register is not in fast path, this should be a mutex?
I don't get it. Why would you prefer mutex over spinlock here
On 5/25/2018 3:38 PM, Stephen Hemminger wrote:
On Thu, 24 May 2018 09:55:13 -0700
Sridhar Samudrala wrote:
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 03ed492c4e14..0f4ba52b641d 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1421,6 +142
On Thu, 24 May 2018 09:55:13 -0700
Sridhar Samudrala wrote:
> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
> index 03ed492c4e14..0f4ba52b641d 100644
> --- a/include/linux/netdevice.h
> +++ b/include/linux/netdevice.h
> @@ -1421,6 +1421,8 @@ struct net_device_ops {
> * en
On Thu, 24 May 2018 09:55:13 -0700
Sridhar Samudrala wrote:
> + spin_lock(&failover_lock);
Since register is not in fast path, this should be a mutex?
> +int failover_slave_unregister(struct net_device *slave_dev)
> +{
> + struct net_device *failover_dev;
> + struct failover_ops *
The failover module provides a generic interface for paravirtual drivers
to register a netdev and a set of ops with a failover instance. The ops
are used as event handlers that get called to handle netdev register/
unregister/link change/name change events on slave pci ethernet devices
with the sam
