Merge branch '10GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue

Jeff Kirsher says:

====================
L2 Fwd Offload & 10GbE Intel Driver Updates 2018-07-09

This patch series is meant to allow support for the L2 forward offload, aka
MACVLAN offload without the need for using ndo_select_queue.

The existing solution currently requires that we use ndo_select_queue in
the transmit path if we want to associate specific Tx queues with a given
MACVLAN interface. In order to get away from this we need to repurpose the
tc_to_txq array and XPS pointer for the MACVLAN interface and use those as
a means of accessing the queues on the lower device. As a result we cannot
offload a device that is configured as multiqueue, however it doesn't
really make sense to configure a macvlan interfaced as being multiqueue
anyway since it doesn't really have a qdisc of its own in the first place.

The big changes in this set are:
  Allow lower device to update tc_to_txq and XPS map of offloaded MACVLAN
  Disable XPS for single queue devices
  Replace accel_priv with sb_dev in ndo_select_queue
  Add sb_dev parameter to fallback function for ndo_select_queue
  Consolidated ndo_select_queue functions that appeared to be duplicates
====================

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

4 files changed, 173 insertions, 31 deletions

diff --git a/net/core/dev.c b/net/core/dev.c
index d13cddcac41f..1c3f0997e857 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2067,11 +2067,13 @@ int netdev_txq_to_tc(struct net_device *dev, unsigned int txq)
 		struct netdev_tc_txq *tc = &dev->tc_to_txq[0];
 		int i;
 
+		/* walk through the TCs and see if it falls into any of them */
 		for (i = 0; i < TC_MAX_QUEUE; i++, tc++) {
 			if ((txq - tc->offset) < tc->count)
 				return i;
 		}
 
+		/* didn't find it, just return -1 to indicate no match */
 		return -1;
 	}
 
@@ -2260,7 +2262,14 @@ int __netif_set_xps_queue(struct net_device *dev, const unsigned long *mask,
 	unsigned int nr_ids;
 
 	if (dev->num_tc) {
+		/* Do not allow XPS on subordinate device directly */
 		num_tc = dev->num_tc;
+		if (num_tc < 0)
+			return -EINVAL;
+
+		/* If queue belongs to subordinate dev use its map */
+		dev = netdev_get_tx_queue(dev, index)->sb_dev ? : dev;
+
 		tc = netdev_txq_to_tc(dev, index);
 		if (tc < 0)
 			return -EINVAL;
@@ -2448,11 +2457,25 @@ int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
 EXPORT_SYMBOL(netif_set_xps_queue);
 
 #endif
+static void netdev_unbind_all_sb_channels(struct net_device *dev)
+{
+	struct netdev_queue *txq = &dev->_tx[dev->num_tx_queues];
+
+	/* Unbind any subordinate channels */
+	while (txq-- != &dev->_tx[0]) {
+		if (txq->sb_dev)
+			netdev_unbind_sb_channel(dev, txq->sb_dev);
+	}
+}
+
 void netdev_reset_tc(struct net_device *dev)
 {
 #ifdef CONFIG_XPS
 	netif_reset_xps_queues_gt(dev, 0);
 #endif
+	netdev_unbind_all_sb_channels(dev);
+
+	/* Reset TC configuration of device */
 	dev->num_tc = 0;
 	memset(dev->tc_to_txq, 0, sizeof(dev->tc_to_txq));
 	memset(dev->prio_tc_map, 0, sizeof(dev->prio_tc_map));
@@ -2481,11 +2504,77 @@ int netdev_set_num_tc(struct net_device *dev, u8 num_tc)
 #ifdef CONFIG_XPS
 	netif_reset_xps_queues_gt(dev, 0);
 #endif
+	netdev_unbind_all_sb_channels(dev);
+
 	dev->num_tc = num_tc;
 	return 0;
 }
 EXPORT_SYMBOL(netdev_set_num_tc);
 
+void netdev_unbind_sb_channel(struct net_device *dev,
+			      struct net_device *sb_dev)
+{
+	struct netdev_queue *txq = &dev->_tx[dev->num_tx_queues];
+
+#ifdef CONFIG_XPS
+	netif_reset_xps_queues_gt(sb_dev, 0);
+#endif
+	memset(sb_dev->tc_to_txq, 0, sizeof(sb_dev->tc_to_txq));
+	memset(sb_dev->prio_tc_map, 0, sizeof(sb_dev->prio_tc_map));
+
+	while (txq-- != &dev->_tx[0]) {
+		if (txq->sb_dev == sb_dev)
+			txq->sb_dev = NULL;
+	}
+}
+EXPORT_SYMBOL(netdev_unbind_sb_channel);
+
+int netdev_bind_sb_channel_queue(struct net_device *dev,
+				 struct net_device *sb_dev,
+				 u8 tc, u16 count, u16 offset)
+{
+	/* Make certain the sb_dev and dev are already configured */
+	if (sb_dev->num_tc >= 0 || tc >= dev->num_tc)
+		return -EINVAL;
+
+	/* We cannot hand out queues we don't have */
+	if ((offset + count) > dev->real_num_tx_queues)
+		return -EINVAL;
+
+	/* Record the mapping */
+	sb_dev->tc_to_txq[tc].count = count;
+	sb_dev->tc_to_txq[tc].offset = offset;
+
+	/* Provide a way for Tx queue to find the tc_to_txq map or
+	 * XPS map for itself.
+	 */
+	while (count--)
+		netdev_get_tx_queue(dev, count + offset)->sb_dev = sb_dev;
+
+	return 0;
+}
+EXPORT_SYMBOL(netdev_bind_sb_channel_queue);
+
+int netdev_set_sb_channel(struct net_device *dev, u16 channel)
+{
+	/* Do not use a multiqueue device to represent a subordinate channel */
+	if (netif_is_multiqueue(dev))
+		return -ENODEV;
+
+	/* We allow channels 1 - 32767 to be used for subordinate channels.
+	 * Channel 0 is meant to be "native" mode and used only to represent
+	 * the main root device. We allow writing 0 to reset the device back
+	 * to normal mode after being used as a subordinate channel.
+	 */
+	if (channel > S16_MAX)
+		return -EINVAL;
+
+	dev->num_tc = -channel;
+
+	return 0;
+}
+EXPORT_SYMBOL(netdev_set_sb_channel);
+
 /*
  * Routine to help set real_num_tx_queues. To avoid skbs mapped to queues
  * greater than real_num_tx_queues stale skbs on the qdisc must be flushed.
@@ -2697,24 +2786,26 @@ EXPORT_SYMBOL(netif_device_attach);
  * Returns a Tx hash based on the given packet descriptor a Tx queues' number
  * to be used as a distribution range.
  */
-static u16 skb_tx_hash(const struct net_device *dev, struct sk_buff *skb)
+static u16 skb_tx_hash(const struct net_device *dev,
+		       const struct net_device *sb_dev,
+		       struct sk_buff *skb)
 {
 	u32 hash;
 	u16 qoffset = 0;
 	u16 qcount = dev->real_num_tx_queues;
 
+	if (dev->num_tc) {
+		u8 tc = netdev_get_prio_tc_map(dev, skb->priority);
+
+		qoffset = sb_dev->tc_to_txq[tc].offset;
+		qcount = sb_dev->tc_to_txq[tc].count;
+	}
+
 	if (skb_rx_queue_recorded(skb)) {
 		hash = skb_get_rx_queue(skb);
 		while (unlikely(hash >= qcount))
 			hash -= qcount;
-		return hash;
-	}
-
-	if (dev->num_tc) {
-		u8 tc = netdev_get_prio_tc_map(dev, skb->priority);
-
-		qoffset = dev->tc_to_txq[tc].offset;
-		qcount = dev->tc_to_txq[tc].count;
+		return hash + qoffset;
 	}
 
 	return (u16) reciprocal_scale(skb_get_hash(skb), qcount) + qoffset;
@@ -3484,7 +3575,8 @@ static int __get_xps_queue_idx(struct net_device *dev, struct sk_buff *skb,
 }
 #endif
 
-static int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
+static int get_xps_queue(struct net_device *dev, struct net_device *sb_dev,
+			 struct sk_buff *skb)
 {
 #ifdef CONFIG_XPS
 	struct xps_dev_maps *dev_maps;
@@ -3498,7 +3590,7 @@ static int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
 	if (!static_key_false(&xps_rxqs_needed))
 		goto get_cpus_map;
 
-	dev_maps = rcu_dereference(dev->xps_rxqs_map);
+	dev_maps = rcu_dereference(sb_dev->xps_rxqs_map);
 	if (dev_maps) {
 		int tci = sk_rx_queue_get(sk);
 
@@ -3509,7 +3601,7 @@ static int get_xps_queue(struct net_device *dev, struct sk_buff *skb)
 
 get_cpus_map:
 	if (queue_index < 0) {
-		dev_maps = rcu_dereference(dev->xps_cpus_map);
+		dev_maps = rcu_dereference(sb_dev->xps_cpus_map);
 		if (dev_maps) {
 			unsigned int tci = skb->sender_cpu - 1;
 
@@ -3525,17 +3617,36 @@ get_cpus_map:
 #endif
 }
 
-static u16 __netdev_pick_tx(struct net_device *dev, struct sk_buff *skb)
+u16 dev_pick_tx_zero(struct net_device *dev, struct sk_buff *skb,
+		     struct net_device *sb_dev,
+		     select_queue_fallback_t fallback)
+{
+	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)
+{
+	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)
 {
 	struct sock *sk = skb->sk;
 	int queue_index = sk_tx_queue_get(sk);
 
+	sb_dev = sb_dev ? : dev;
+
 	if (queue_index < 0 || skb->ooo_okay ||
 	    queue_index >= dev->real_num_tx_queues) {
-		int new_index = get_xps_queue(dev, skb);
+		int new_index = get_xps_queue(dev, sb_dev, skb);
 
 		if (new_index < 0)
-			new_index = skb_tx_hash(dev, skb);
+			new_index = skb_tx_hash(dev, sb_dev, skb);
 
 		if (queue_index != new_index && sk &&
 		    sk_fullsock(sk) &&
@@ -3550,7 +3661,7 @@ static u16 __netdev_pick_tx(struct net_device *dev, struct sk_buff *skb)
 
 struct netdev_queue *netdev_pick_tx(struct net_device *dev,
 				    struct sk_buff *skb,
-				    void *accel_priv)
+				    struct net_device *sb_dev)
 {
 	int queue_index = 0;
 
@@ -3565,10 +3676,10 @@ 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, accel_priv,
+			queue_index = ops->ndo_select_queue(dev, skb, sb_dev,
 							    __netdev_pick_tx);
 		else
-			queue_index = __netdev_pick_tx(dev, skb);
+			queue_index = __netdev_pick_tx(dev, skb, sb_dev);
 
 		queue_index = netdev_cap_txqueue(dev, queue_index);
 	}
@@ -3580,7 +3691,7 @@ struct netdev_queue *netdev_pick_tx(struct net_device *dev,
 /**
  *	__dev_queue_xmit - transmit a buffer
  *	@skb: buffer to transmit
- *	@accel_priv: private data used for L2 forwarding offload
+ *	@sb_dev: suboordinate device used for L2 forwarding offload
  *
  *	Queue a buffer for transmission to a network device. The caller must
  *	have set the device and priority and built the buffer before calling
@@ -3603,7 +3714,7 @@ struct netdev_queue *netdev_pick_tx(struct net_device *dev,
  *      the BH enable code must have IRQs enabled so that it will not deadlock.
  *          --BLG
  */
-static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv)
+static int __dev_queue_xmit(struct sk_buff *skb, struct net_device *sb_dev)
 {
 	struct net_device *dev = skb->dev;
 	struct netdev_queue *txq;
@@ -3642,7 +3753,7 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv)
 	else
 		skb_dst_force(skb);
 
-	txq = netdev_pick_tx(dev, skb, accel_priv);
+	txq = netdev_pick_tx(dev, skb, sb_dev);
 	q = rcu_dereference_bh(txq->qdisc);
 
 	trace_net_dev_queue(skb);
@@ -3716,9 +3827,9 @@ int dev_queue_xmit(struct sk_buff *skb)
 }
 EXPORT_SYMBOL(dev_queue_xmit);
 
-int dev_queue_xmit_accel(struct sk_buff *skb, void *accel_priv)
+int dev_queue_xmit_accel(struct sk_buff *skb, struct net_device *sb_dev)
 {
-	return __dev_queue_xmit(skb, accel_priv);
+	return __dev_queue_xmit(skb, sb_dev);
 }
 EXPORT_SYMBOL(dev_queue_xmit_accel);
 
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index f25ac5ff48a6..ffa1d18f2c2c 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -1047,13 +1047,30 @@ static ssize_t traffic_class_show(struct netdev_queue *queue,
 				  char *buf)
 {
 	struct net_device *dev = queue->dev;
-	int index = get_netdev_queue_index(queue);
-	int tc = netdev_txq_to_tc(dev, index);
+	int index;
+	int tc;
 
+	if (!netif_is_multiqueue(dev))
+		return -ENOENT;
+
+	index = get_netdev_queue_index(queue);
+
+	/* If queue belongs to subordinate dev use its TC mapping */
+	dev = netdev_get_tx_queue(dev, index)->sb_dev ? : dev;
+
+	tc = netdev_txq_to_tc(dev, index);
 	if (tc < 0)
 		return -EINVAL;
 
-	return sprintf(buf, "%u\n", tc);
+	/* We can report the traffic class one of two ways:
+	 * Subordinate device traffic classes are reported with the traffic
+	 * class first, and then the subordinate class so for example TC0 on
+	 * subordinate device 2 will be reported as "0-2". If the queue
+	 * belongs to the root device it will be reported with just the
+	 * traffic class, so just "0" for TC 0 for example.
+	 */
+	return dev->num_tc < 0 ? sprintf(buf, "%u%d\n", tc, dev->num_tc) :
+				 sprintf(buf, "%u\n", tc);
 }
 
 #ifdef CONFIG_XPS
@@ -1214,10 +1231,20 @@ static ssize_t xps_cpus_show(struct netdev_queue *queue,
 	cpumask_var_t mask;
 	unsigned long index;
 
+	if (!netif_is_multiqueue(dev))
+		return -ENOENT;
+
 	index = get_netdev_queue_index(queue);
 
 	if (dev->num_tc) {
+		/* Do not allow XPS on subordinate device directly */
 		num_tc = dev->num_tc;
+		if (num_tc < 0)
+			return -EINVAL;
+
+		/* If queue belongs to subordinate dev use its map */
+		dev = netdev_get_tx_queue(dev, index)->sb_dev ? : dev;
+
 		tc = netdev_txq_to_tc(dev, index);
 		if (tc < 0)
 			return -EINVAL;
@@ -1260,6 +1287,9 @@ static ssize_t xps_cpus_store(struct netdev_queue *queue,
 	cpumask_var_t mask;
 	int err;
 
+	if (!netif_is_multiqueue(dev))
+		return -ENOENT;
+
 	if (!capable(CAP_NET_ADMIN))
 		return -EPERM;
 
diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c
index 555e389b7dfa..5e6cf2cee965 100644
--- a/net/mac80211/iface.c
+++ b/net/mac80211/iface.c
@@ -1130,7 +1130,7 @@ static void ieee80211_uninit(struct net_device *dev)
 
 static u16 ieee80211_netdev_select_queue(struct net_device *dev,
 					 struct sk_buff *skb,
-					 void *accel_priv,
+					 struct net_device *sb_dev,
 					 select_queue_fallback_t fallback)
 {
 	return ieee80211_select_queue(IEEE80211_DEV_TO_SUB_IF(dev), skb);
@@ -1176,7 +1176,7 @@ static const struct net_device_ops ieee80211_dataif_ops = {
 
 static u16 ieee80211_monitor_select_queue(struct net_device *dev,
 					  struct sk_buff *skb,
-					  void *accel_priv,
+					  struct net_device *sb_dev,
 					  select_queue_fallback_t fallback)
 {
 	struct ieee80211_sub_if_data *sdata = IEEE80211_DEV_TO_SUB_IF(dev);
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 47931ebfaef3..00189a3b07f2 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -275,9 +275,10 @@ 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)
+static u16 __packet_pick_tx_queue(struct net_device *dev, struct sk_buff *skb,
+				  struct net_device *sb_dev)
 {
-	return (u16) raw_smp_processor_id() % dev->real_num_tx_queues;
+	return dev_pick_tx_cpu_id(dev, skb, sb_dev, NULL);
 }
 
 static u16 packet_pick_tx_queue(struct sk_buff *skb)
@@ -291,7 +292,7 @@ static u16 packet_pick_tx_queue(struct sk_buff *skb)
 						    __packet_pick_tx_queue);
 		queue_index = netdev_cap_txqueue(dev, queue_index);
 	} else {
-		queue_index = __packet_pick_tx_queue(dev, skb);
+		queue_index = __packet_pick_tx_queue(dev, skb, NULL);
 	}
 
 	return queue_index;