[PATCH net-next] net: netdev_pick_tx: use get_xps_q if xps map is set

[govindarajulu.v]

From: "govindarajulu.v" <govindarajulu90@xxxxxxxxx>

netdev_pick_tx ignores the xps map configuration if netdev->ndo_select_queue
is defined. Most of the drivers define ndo_select_queue. The problem with this
is, if admin wants kernel to pick tx queue based on xps map (instead of driver
defined ndo_select_queue), he has to netdev->ndo_select_queue = NULL, compile
and reload.

This patch fixes it by checking if dev->xps_maps is defined. If yes it
proceeds with get_xps_queue. If not it proceeds with netdev->ndo_select_queue
(if defined).

Compile test only.

Signed-off-by: govindarajulu.v <govindarajulu90@xxxxxxxxx>
---
 net/core/flow_dissector.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
index 00ee068..b2e3c81 100644
--- a/net/core/flow_dissector.c
+++ b/net/core/flow_dissector.c
@@ -368,10 +368,11 @@ struct netdev_queue *netdev_pick_tx(struct net_device *dev,
 
 	if (dev->real_num_tx_queues != 1) {
 		const struct net_device_ops *ops = dev->netdev_ops;
-		if (ops->ndo_select_queue)
-			queue_index = ops->ndo_select_queue(dev, skb);
-		else
+		if (rcu_access_pointer(dev->xps_maps) ||
+		    !ops->ndo_select_queue)
 			queue_index = __netdev_pick_tx(dev, skb);
+		else
+			queue_index = ops->ndo_select_queue(dev, skb);
 		queue_index = dev_cap_txqueue(dev, queue_index);
 	}
 
-- 
1.8.2.3

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/