[net,1/2] net/bonding: emit address change event in more places

From: Shlomo Pongratz <shlomop@mellanox.com>

commit 7d26bb103c4 "bonding: emit event when bonding changes MAC" didn't
take care to emit the NETDEV_CHANGEADDR event in other places where bonding
actually changes the mac address (to all zeroes), in bond_release, and
bond_release_all. As a result the neighbours aren't deleted by the core
networking code (which does so upon getting that event).

Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com>
---
 drivers/net/bonding/bond_main.c |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

Or Gerlitz <ogerlitz@mellanox.com> wrote:

>From: Shlomo Pongratz <shlomop@mellanox.com>
>
>commit 7d26bb103c4 "bonding: emit event when bonding changes MAC" didn't
>take care to emit the NETDEV_CHANGEADDR event in other places where bonding
>actually changes the mac address (to all zeroes), in bond_release, and
>bond_release_all. As a result the neighbours aren't deleted by the core
>networking code (which does so upon getting that event).

>Signed-off-by: Shlomo Pongratz <shlomop@mellanox.com>

	Is the bond_release_all notifier call actually necessary in your
testing?

	The bond itself is destroyed immediately after bond_release_all
returns, so I would expect the neighbours to be deleted then.  For that
path we probably don't need to zero the mac at all (or compute features,
for that matter).

	-J


>---
> drivers/net/bonding/bond_main.c |    5 +++++
> 1 files changed, 5 insertions(+), 0 deletions(-)
>
>diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
>index a20b585..b0a278d 100644
>--- a/drivers/net/bonding/bond_main.c
>+++ b/drivers/net/bonding/bond_main.c
>@@ -2035,6 +2035,9 @@ int bond_release(struct net_device *bond_dev, struct net_device *slave_dev)
> 	write_unlock_bh(&bond->lock);
> 	unblock_netpoll_tx();
>
>+	if (bond->slave_cnt == 0)
>+		call_netdevice_notifiers(NETDEV_CHANGEADDR, bond->dev);
>+
> 	bond_compute_features(bond);
> 	if (!(bond_dev->features & NETIF_F_VLAN_CHALLENGED) &&
> 	    (old_features & NETIF_F_VLAN_CHALLENGED))
>@@ -2218,6 +2221,8 @@ static int bond_release_all(struct net_device *bond_dev)
> out:
> 	write_unlock_bh(&bond->lock);
>
>+	call_netdevice_notifiers(NETDEV_CHANGEADDR, bond->dev);
>+
> 	bond_compute_features(bond);
>
> 	return 0;
>-- 
>1.7.1

---
	-Jay Vosburgh, IBM Linux Technology Center, fubar@us.ibm.com

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On 4/4/2012 1:28 AM, Jay Vosburgh wrote:
> Is the bond_release_all notifier call actually necessary in your 
> testing? The bond itself is destroyed immediately after 
> bond_release_all returns, so I would expect the neighbours to be 
> deleted then. For that path we probably don't need to zero the mac at 
> all (or compute features, for that matter).

yep, you're right, will remove this extra emitting of event and resend, 
thanks.

Or.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html