On 5/30/2018 7:52 PM, Stephen Hemminger wrote:

On Fri, 25 May 2018 16:06:58 -0700
"Samudrala, Sridhar" [off-list ref] wrote:

quoted

On 5/25/2018 3:38 PM, Stephen Hemminger wrote:

quoted

On Thu, 24 May 2018 09:55:13 -0700
Sridhar Samudrala [off-list ref] wrote:
  

quoted

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 {
    *	entity (i.e. the master device for bridged veth)
    * @IFF_MACSEC: device is a MACsec device
    * @IFF_NO_RX_HANDLER: device doesn't support the rx_handler hook
+ * @IFF_FAILOVER: device is a failover master device
+ * @IFF_FAILOVER_SLAVE: device is lower dev of a failover master device
    */
   enum netdev_priv_flags {
   	IFF_802_1Q_VLAN			= 1<<0,

@@ -1450,6 +1452,8 @@ enum netdev_priv_flags {
   	IFF_PHONY_HEADROOM		= 1<<24,
   	IFF_MACSEC			= 1<<25,
   	IFF_NO_RX_HANDLER		= 1<<26,
+	IFF_FAILOVER			= 1<<27,
+	IFF_FAILOVER_SLAVE		= 1<<28,
   };

Why is FAILOVER any different than other master/slave relationships.
I don't think you need to take up precious netdev flag bits for this.

These are netdev priv flags.
Jiri says that IFF_MASTER/IFF_SLAVE are bonding specific flags and cannot be used
with other failover mechanisms. Team also doesn't use this flags and it has its own
priv_flags.

This change breaks userspace.
We already have worked with partners to ignore devices marked as IFF_SLAVE,
and IFF_SLAVE is visible to user space API's.

I specifically made sure not to remove IFF_SLAVE in the netvsc patch.

NAK