On Tue, 2016-02-23 at 11:20 -0800, pravin shelar wrote:

On Tue, Feb 23, 2016 at 4:53 AM, Paolo Abeni [off-list ref] wrote:

quoted

This method allows the controlling device (i.e. the bridge) to specify
additional headroom to be allocated for skb head on frame reception.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 include/linux/netdevice.h | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 47671ce0..fb74277 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h

@@ -1093,6 +1093,12 @@ struct tc_to_netdev {
  *     This function is used to get egress tunnel information for given skb.
  *     This is useful for retrieving outer tunnel header parameters while
  *     sampling packet.

...

quoted

 /**

@@ -1315,6 +1323,8 @@ struct net_device_ops {
  * @IFF_L3MDEV_SLAVE: device is enslaved to an L3 master device
  * @IFF_TEAM: device is a team device
  * @IFF_RXFH_CONFIGURED: device has had Rx Flow indirection table configured
+ * @IFF_PHONY_HEADROOM: the headroom value is controlled by an external
+ *     entity (i.e. the master device for bridged veth)
  */
 enum netdev_priv_flags {
        IFF_802_1Q_VLAN                 = 1<<0,

@@ -1343,6 +1353,7 @@ enum netdev_priv_flags {
        IFF_L3MDEV_SLAVE                = 1<<23,
        IFF_TEAM                        = 1<<24,
        IFF_RXFH_CONFIGURED             = 1<<25,
+       IFF_PHONY_HEADROOM              = 1<<26,
 };

 #define IFF_802_1Q_VLAN                        IFF_802_1Q_VLAN

@@ -1937,6 +1948,15 @@ struct netdev_queue *netdev_pick_tx(struct net_device *dev,
                                    struct sk_buff *skb,
                                    void *accel_priv);

+/* returns the headroom that the master device needs to take in account
+ * when forwarding to this dev
+ */
+static inline unsigned netdev_get_fwd_headroom(struct net_device *dev)
+{
+       return (dev->priv_flags && IFF_PHONY_HEADROOM) ? dev->needed_headroom :
+               0;
+}
+

This looks like typo. It should '&' to check for the bitfield.

You are right! Thank you for spotting it.

There is also another error here; the code should be:

return (dev->priv_flags & IFF_PHONY_HEADROOM) ? 0 : dev->needed_headroom;

i.e. when computing the rx/forwarding headroom, we want to ignore
devices with phony headroom, since they actually will not requires
additional skb headroom on xmit.

I'll fix this in v2.

Paolo