On Thu, Apr 14, 2011 at 07:09:36PM -0700, Ethan Jackson wrote:
> Before this patch, the bonding code had taken over responsibility
> for running the LACP module. However, the bonding code only needs
> the LACP module for some basic status queries. LACP and bonding
> are actually logically parallel modules and do not really have a
> parent child relationship. Furthermore, we need to be able to run
> LACP on non-bonded interfaces which the existing approach
> prevented. This patch gives control of the LACP module back to the
> bridge.
This adds a blank line at the beginning of port_run().
I don't see anything that destroys the lacp object when a port is
destroyed.
I think you missed adding one "->aux":
diff --git a/lib/bond.c b/lib/bond.c
index b1ca5d5..3ec1ba2 100644
--- a/lib/bond.c
+++ b/lib/bond.c
@@ -1483,7 +1483,7 @@ bond_choose_slave(const struct bond *bond)
best = NULL;
HMAP_FOR_EACH (slave, hmap_node, &bond->slaves) {
if (slave->delay_expires != LLONG_MAX
- && lacp_slave_may_enable(bond->lacp, slave)
+ && lacp_slave_may_enable(bond->lacp, slave->aux)
&& (!best || slave->delay_expires < best->delay_expires))
{
best = slave;
}
Otherwise looks good. I'd be concerned about keeping the "struct lacp
*" pointer in bond except that you're taking care of that in later
patches.
_______________________________________________
dev mailing list
[email protected]
http://openvswitch.org/mailman/listinfo/dev
