Merge branch 'net-refactor-ndo_select_queue'

Paolo Abeni says:

====================
net: refactor ndo_select_queue()

Currently, on most devices implementing ndo_select_queue(), we get 2
indirect calls per xmit packet, at least in some scenarios.

We can avoid one of such indirect calls refactoring the ndo_select_queue()
usage so that we don't need anymore the 'fallback' argument.

The first patch renames a helper used later as a public API, the second one
changes the af packet implementation so that it uses the common infrastructure
to select the xmit queue, and the second patch drops the now unneeded argument
from ndo_select_queue().

Alternatively we could use the INDIRECT_CALL_WRAPPER infrastructure to avoid
the fallback indirect call in the common case, but this solution allows also
for some code cleanup.

 v1 -> v2:
  - renamed select queue helpers, as per Eric's and David's suggestions
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 23 insertions, 29 deletions

diff --git a/net/core/dev.c b/net/core/dev.c
index 2b67f2aa59dd..357111431ec9 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3689,23 +3689,21 @@ get_cpus_map:
 }
 
 u16 dev_pick_tx_zero(struct net_device *dev, struct sk_buff *skb,
-		     struct net_device *sb_dev,
-		     select_queue_fallback_t fallback)
+		     struct net_device *sb_dev)
 {
 	return 0;
 }
 EXPORT_SYMBOL(dev_pick_tx_zero);
 
 u16 dev_pick_tx_cpu_id(struct net_device *dev, struct sk_buff *skb,
-		       struct net_device *sb_dev,
-		       select_queue_fallback_t fallback)
+		       struct net_device *sb_dev)
 {
 	return (u16)raw_smp_processor_id() % dev->real_num_tx_queues;
 }
 EXPORT_SYMBOL(dev_pick_tx_cpu_id);
 
-static u16 __netdev_pick_tx(struct net_device *dev, struct sk_buff *skb,
-			    struct net_device *sb_dev)
+u16 netdev_pick_tx(struct net_device *dev, struct sk_buff *skb,
+		     struct net_device *sb_dev)
 {
 	struct sock *sk = skb->sk;
 	int queue_index = sk_tx_queue_get(sk);
@@ -3729,10 +3727,11 @@ static u16 __netdev_pick_tx(struct net_device *dev, struct sk_buff *skb,
 
 	return queue_index;
 }
+EXPORT_SYMBOL(netdev_pick_tx);
 
-struct netdev_queue *netdev_pick_tx(struct net_device *dev,
-				    struct sk_buff *skb,
-				    struct net_device *sb_dev)
+struct netdev_queue *netdev_core_pick_tx(struct net_device *dev,
+					 struct sk_buff *skb,
+					 struct net_device *sb_dev)
 {
 	int queue_index = 0;
 
@@ -3747,10 +3746,9 @@ struct netdev_queue *netdev_pick_tx(struct net_device *dev,
 		const struct net_device_ops *ops = dev->netdev_ops;
 
 		if (ops->ndo_select_queue)
-			queue_index = ops->ndo_select_queue(dev, skb, sb_dev,
-							    __netdev_pick_tx);
+			queue_index = ops->ndo_select_queue(dev, skb, sb_dev);
 		else
-			queue_index = __netdev_pick_tx(dev, skb, sb_dev);
+			queue_index = netdev_pick_tx(dev, skb, sb_dev);
 
 		queue_index = netdev_cap_txqueue(dev, queue_index);
 	}
@@ -3824,7 +3822,7 @@ static int __dev_queue_xmit(struct sk_buff *skb, struct net_device *sb_dev)
 	else
 		skb_dst_force(skb);
 
-	txq = netdev_pick_tx(dev, skb, sb_dev);
+	txq = netdev_core_pick_tx(dev, skb, sb_dev);
 	q = rcu_dereference_bh(txq->qdisc);
 
 	trace_net_dev_queue(skb);
@@ -4429,7 +4427,7 @@ void generic_xdp_tx(struct sk_buff *skb, struct bpf_prog *xdp_prog)
 	bool free_skb = true;
 	int cpu, rc;
 
-	txq = netdev_pick_tx(dev, skb, NULL);
+	txq = netdev_core_pick_tx(dev, skb, NULL);
 	cpu = smp_processor_id();
 	HARD_TX_LOCK(dev, txq, cpu);
 	if (!netif_xmit_stopped(txq)) {
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index 361aabffb8c0..e365e8fb1c40 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -323,7 +323,7 @@ void netpoll_send_skb_on_dev(struct netpoll *np, struct sk_buff *skb,
 	if (skb_queue_len(&npinfo->txq) == 0 && !netpoll_owner_active(dev)) {
 		struct netdev_queue *txq;
 
-		txq = netdev_pick_tx(dev, skb, NULL);
+		txq = netdev_core_pick_tx(dev, skb, NULL);
 
 		/* try until next clock tick */
 		for (tries = jiffies_to_usecs(1)/USEC_PER_POLL;
diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c
index 4a6ff1482a9f..f0d97eba250b 100644
--- a/net/mac80211/iface.c
+++ b/net/mac80211/iface.c
@@ -1133,8 +1133,7 @@ static void ieee80211_uninit(struct net_device *dev)
 
 static u16 ieee80211_netdev_select_queue(struct net_device *dev,
 					 struct sk_buff *skb,
-					 struct net_device *sb_dev,
-					 select_queue_fallback_t fallback)
+					 struct net_device *sb_dev)
 {
 	return ieee80211_select_queue(IEEE80211_DEV_TO_SUB_IF(dev), skb);
 }
@@ -1179,8 +1178,7 @@ static const struct net_device_ops ieee80211_dataif_ops = {
 
 static u16 ieee80211_monitor_select_queue(struct net_device *dev,
 					  struct sk_buff *skb,
-					  struct net_device *sb_dev,
-					  select_queue_fallback_t fallback)
+					  struct net_device *sb_dev)
 {
 	struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev);
 	struct ieee80211_local *local = sdata->local;
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 323655a25674..741953b42f44 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -275,24 +275,22 @@ static bool packet_use_direct_xmit(const struct packet_sock *po)
 	return po->xmit == packet_direct_xmit;
 }
 
-static u16 __packet_pick_tx_queue(struct net_device *dev, struct sk_buff *skb,
-				  struct net_device *sb_dev)
-{
-	return dev_pick_tx_cpu_id(dev, skb, sb_dev, NULL);
-}
-
 static u16 packet_pick_tx_queue(struct sk_buff *skb)
 {
 	struct net_device *dev = skb->dev;
 	const struct net_device_ops *ops = dev->netdev_ops;
+	int cpu = raw_smp_processor_id();
 	u16 queue_index;
 
+#ifdef CONFIG_XPS
+	skb->sender_cpu = cpu + 1;
+#endif
+	skb_record_rx_queue(skb, cpu % dev->real_num_tx_queues);
 	if (ops->ndo_select_queue) {
-		queue_index = ops->ndo_select_queue(dev, skb, NULL,
-						    __packet_pick_tx_queue);
+		queue_index = ops->ndo_select_queue(dev, skb, NULL);
 		queue_index = netdev_cap_txqueue(dev, queue_index);
 	} else {
-		queue_index = __packet_pick_tx_queue(dev, skb, NULL);
+		queue_index = netdev_pick_tx(dev, skb, NULL);
 	}
 
 	return queue_index;
diff --git a/net/xfrm/xfrm_device.c b/net/xfrm/xfrm_device.c
index b8736f56e7f7..2db1626557c5 100644
--- a/net/xfrm/xfrm_device.c
+++ b/net/xfrm/xfrm_device.c
@@ -247,7 +247,7 @@ void xfrm_dev_resume(struct sk_buff *skb)
 	unsigned long flags;
 
 	rcu_read_lock();
-	txq = netdev_pick_tx(dev, skb, NULL);
+	txq = netdev_core_pick_tx(dev, skb, NULL);
 
 	HARD_TX_LOCK(dev, txq, smp_processor_id());
 	if (!netif_xmit_frozen_or_stopped(txq))