Re: [PATCH RFC] ndo: ndo_queue_xmit/ndo_flush_xmit (was Re: [RFC] [ver3 PATCH 0/6] Implement multiqueue virtio-net)

["Michael S. Tsirkin" <mst@redhat.com>]

On Sun, Nov 13, 2011 at 07:48:28PM +0200, Michael S. Tsirkin wrote:
> @@ -863,6 +872,9 @@ struct net_device_ops {
>  	int			(*ndo_stop)(struct net_device *dev);
>  	netdev_tx_t		(*ndo_start_xmit) (struct sk_buff *skb,
>  						   struct net_device *dev);
> +	netdev_tx_t		(*ndo_queue_xmit)(struct sk_buff *skb,
> +						  struct net_device *dev);
> +	void			(*ndo_flush_xmit)(struct net_device *dev);
>  	u16			(*ndo_select_queue)(struct net_device *dev,
>  						    struct sk_buff *skb);
>  	void			(*ndo_change_rx_flags)(struct net_device *dev,
> @@ -2099,6 +2111,10 @@ extern int		dev_set_mac_address(struct net_device *,

An alternative I considered was to add a boolean flag to ndo_start_xmit
'bool queue' or something like this, plus ndo_flush_xmit.
This will lead to cleaner code I think but will require all drivers
to be changed, so for a proof of concept I decided to go for one
that is less work.

Let me know what looks more palatable ... 

-- 
MST