Re: [PATCH v2 net-next 05/12] bonding: extend arp_validate to be able to receive unvalidated arp-only traffic

On Fri, Jan 17, 2014 at 12:41:29PM -0800, Jay Vosburgh wrote:

Veaceslav Falico [off-list ref] wrote:

...snip...

quoted

--- a/drivers/net/bonding/bonding.h
+++ b/drivers/net/bonding/bonding.h

@@ -318,6 +318,11 @@ static inline bool bond_is_active_slave(struct slave *slave)

#define BOND_ARP_VALIDATE_BACKUP	(1 << BOND_STATE_BACKUP)
#define BOND_ARP_VALIDATE_ALL		(BOND_ARP_VALIDATE_ACTIVE | \
					 BOND_ARP_VALIDATE_BACKUP)
+#define BOND_ARP_VALIDATE_ARP		(BOND_ARP_VALIDATE_ALL + 1)
+#define BOND_ARP_VALIDATE_ACTIVE_ARP	(BOND_ARP_VALIDATE_ACTIVE | \
+					 BOND_ARP_VALIDATE_ARP)
+#define BOND_ARP_VALIDATE_BACKUP_ARP	(BOND_ARP_VALIDATE_BACKUP | \
+					 BOND_ARP_VALIDATE_ARP)

If you go with my suggestion to call the new thing "filtering,"
I'd change these option names, labels, and the function
"slave_do_arp_validate_only" names.  The function name seems kind of
confusing in particular. I think it'd be clearer to replace the
"validate" stuff with "filter."

Hi Jay,

Sorry for the delay. Yep, completely agree, I'll change the doc/rename the
functions/defines to use "filtering" instead of double-arp.

I'll re-send it once I'll fix the current bonding state...

Thank you!

-J

quoted

static inline int slave_do_arp_validate(struct bonding *bond,
					struct slave *slave)

@@ -325,6 +330,12 @@ static inline int slave_do_arp_validate(struct bonding *bond,

	return bond->params.arp_validate & (1 << bond_slave_state(slave));
}

+static inline int slave_do_arp_validate_only(struct bonding *bond,
+					     struct slave *slave)
+{
+	return bond->params.arp_validate & BOND_ARP_VALIDATE_ARP;
+}
+
/* Get the oldest arp which we've received on this slave for bond's
 * arp_targets.
 */
--
1.8.4

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