[dpdk-dev] [PATCH] net/bonding: don't early mark eth device as bonded

Chas Williams 3chas3 at gmail.com
Wed Jan 17 17:44:44 CET 2018


From: "Charles (Chas) Williams" <ciwillia at mail.eng.vyatta.net>

bonding immediately marks the incoming eth device as bonded and doesn't
clear this in later error paths.  Delay marking the dev until we are
certain that we are going to add this eth device to the bond group.

Signed-off-by: Chas Williams <chas3 at att.com>
---
 drivers/net/bonding/rte_eth_bond_api.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/net/bonding/rte_eth_bond_api.c b/drivers/net/bonding/rte_eth_bond_api.c
index 534a890..e9c446b 100644
--- a/drivers/net/bonding/rte_eth_bond_api.c
+++ b/drivers/net/bonding/rte_eth_bond_api.c
@@ -242,9 +242,6 @@ __eth_bond_slave_add_lock_free(uint16_t bonded_port_id, uint16_t slave_port_id)
 		return -1;
 	}
 
-	/* Add slave details to bonded device */
-	slave_eth_dev->data->dev_flags |= RTE_ETH_DEV_BONDED_SLAVE;
-
 	rte_eth_dev_info_get(slave_port_id, &dev_info);
 	if (dev_info.max_rx_pktlen < internals->max_rx_pktlen) {
 		RTE_BOND_LOG(ERR, "Slave (port %u) max_rx_pktlen too small",
@@ -316,18 +313,21 @@ __eth_bond_slave_add_lock_free(uint16_t bonded_port_id, uint16_t slave_port_id)
 
 	internals->slave_count++;
 
-	/* Update all slave devices MACs*/
-	mac_address_slaves_update(bonded_eth_dev);
-
 	if (bonded_eth_dev->data->dev_started) {
 		if (slave_configure(bonded_eth_dev, slave_eth_dev) != 0) {
-			slave_eth_dev->data->dev_flags &= (~RTE_ETH_DEV_BONDED_SLAVE);
+			internals->slave_count--;
 			RTE_BOND_LOG(ERR, "rte_bond_slaves_configure: port=%d",
 					slave_port_id);
 			return -1;
 		}
 	}
 
+	/* Add slave details to bonded device */
+	slave_eth_dev->data->dev_flags |= RTE_ETH_DEV_BONDED_SLAVE;
+
+	/* Update all slave devices MACs*/
+	mac_address_slaves_update(bonded_eth_dev);
+
 	/* Register link status change callback with bonded device pointer as
 	 * argument*/
 	rte_eth_dev_callback_register(slave_port_id, RTE_ETH_EVENT_INTR_LSC,
-- 
2.9.5



More information about the dev mailing list