Merge tag 'mac80211-next-for-davem-2015-12-07' into next

This pull request got a bit bigger than I wanted, due to
needing to reshuffle and fix some bugs. I merged mac80211
to get the right base for some of these changes.

 * new mac80211 API for upcoming driver changes: EOSP handling,
   key iteration
 * scan abort changes allowing to cancel an ongoing scan
 * VHT IBSS 80+80 MHz support
 * re-enable full AP client state tracking after fixes
 * various small fixes (that weren't relevant for mac80211)
 * various cleanups

13 files changed, 243 insertions, 313 deletions

diff --git a/drivers/net/ethernet/mellanox/mlx4/en_cq.c b/drivers/net/ethernet/mellanox/mlx4/en_cq.c
index eb8a4988de63..af975a2b74c6 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_cq.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_cq.c
@@ -155,13 +155,11 @@ int mlx4_en_activate_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq,
 	cq->mcq.comp  = cq->is_tx ? mlx4_en_tx_irq : mlx4_en_rx_irq;
 	cq->mcq.event = mlx4_en_cq_event;
 
-	if (cq->is_tx) {
-		netif_napi_add(cq->dev, &cq->napi, mlx4_en_poll_tx_cq,
-			       NAPI_POLL_WEIGHT);
-	} else {
+	if (cq->is_tx)
+		netif_tx_napi_add(cq->dev, &cq->napi, mlx4_en_poll_tx_cq,
+				  NAPI_POLL_WEIGHT);
+	else
 		netif_napi_add(cq->dev, &cq->napi, mlx4_en_poll_rx_cq, 64);
-		napi_hash_add(&cq->napi);
-	}
 
 	napi_enable(&cq->napi);
 
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
index ddb5541882f5..dd84cabb2a51 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
@@ -337,11 +337,7 @@ static int mlx4_en_get_sset_count(struct net_device *dev, int sset)
 	case ETH_SS_STATS:
 		return bitmap_iterator_count(&it) +
 			(priv->tx_ring_num * 2) +
-#ifdef CONFIG_NET_RX_BUSY_POLL
-			(priv->rx_ring_num * 5);
-#else
 			(priv->rx_ring_num * 2);
-#endif
 	case ETH_SS_TEST:
 		return MLX4_EN_NUM_SELF_TEST - !(priv->mdev->dev->caps.flags
 					& MLX4_DEV_CAP_FLAG_UC_LOOPBACK) * 2;
@@ -408,11 +404,6 @@ static void mlx4_en_get_ethtool_stats(struct net_device *dev,
 	for (i = 0; i < priv->rx_ring_num; i++) {
 		data[index++] = priv->rx_ring[i]->packets;
 		data[index++] = priv->rx_ring[i]->bytes;
-#ifdef CONFIG_NET_RX_BUSY_POLL
-		data[index++] = priv->rx_ring[i]->yields;
-		data[index++] = priv->rx_ring[i]->misses;
-		data[index++] = priv->rx_ring[i]->cleaned;
-#endif
 	}
 	spin_unlock_bh(&priv->stats_lock);
 
@@ -486,14 +477,6 @@ static void mlx4_en_get_strings(struct net_device *dev,
 				"rx%d_packets", i);
 			sprintf(data + (index++) * ETH_GSTRING_LEN,
 				"rx%d_bytes", i);
-#ifdef CONFIG_NET_RX_BUSY_POLL
-			sprintf(data + (index++) * ETH_GSTRING_LEN,
-				"rx%d_napi_yield", i);
-			sprintf(data + (index++) * ETH_GSTRING_LEN,
-				"rx%d_misses", i);
-			sprintf(data + (index++) * ETH_GSTRING_LEN,
-				"rx%d_cleaned", i);
-#endif
 		}
 		break;
 	case ETH_SS_PRIV_FLAGS:
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index 886e1bc86374..659209ff7af6 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -69,34 +69,6 @@ int mlx4_en_setup_tc(struct net_device *dev, u8 up)
 	return 0;
 }
 
-#ifdef CONFIG_NET_RX_BUSY_POLL
-/* must be called with local_bh_disable()d */
-static int mlx4_en_low_latency_recv(struct napi_struct *napi)
-{
-	struct mlx4_en_cq *cq = container_of(napi, struct mlx4_en_cq, napi);
-	struct net_device *dev = cq->dev;
-	struct mlx4_en_priv *priv = netdev_priv(dev);
-	struct mlx4_en_rx_ring *rx_ring = priv->rx_ring[cq->ring];
-	int done;
-
-	if (!priv->port_up)
-		return LL_FLUSH_FAILED;
-
-	if (!mlx4_en_cq_lock_poll(cq))
-		return LL_FLUSH_BUSY;
-
-	done = mlx4_en_process_rx_cq(dev, cq, 4);
-	if (likely(done))
-		rx_ring->cleaned += done;
-	else
-		rx_ring->misses++;
-
-	mlx4_en_cq_unlock_poll(cq);
-
-	return done;
-}
-#endif	/* CONFIG_NET_RX_BUSY_POLL */
-
 #ifdef CONFIG_RFS_ACCEL
 
 struct mlx4_en_filter {
@@ -1561,8 +1533,6 @@ int mlx4_en_start_port(struct net_device *dev)
 	for (i = 0; i < priv->rx_ring_num; i++) {
 		cq = priv->rx_cq[i];
 
-		mlx4_en_cq_init_lock(cq);
-
 		err = mlx4_en_init_affinity_hint(priv, i);
 		if (err) {
 			en_err(priv, "Failed preparing IRQ affinity hint\n");
@@ -1859,13 +1829,6 @@ void mlx4_en_stop_port(struct net_device *dev, int detach)
 	for (i = 0; i < priv->rx_ring_num; i++) {
 		struct mlx4_en_cq *cq = priv->rx_cq[i];
 
-		local_bh_disable();
-		while (!mlx4_en_cq_lock_napi(cq)) {
-			pr_info("CQ %d locked\n", i);
-			mdelay(1);
-		}
-		local_bh_enable();
-
 		napi_synchronize(&cq->napi);
 		mlx4_en_deactivate_rx_ring(priv, priv->rx_ring[i]);
 		mlx4_en_deactivate_cq(priv, cq);
@@ -2504,9 +2467,6 @@ static const struct net_device_ops mlx4_netdev_ops = {
 #ifdef CONFIG_RFS_ACCEL
 	.ndo_rx_flow_steer	= mlx4_en_filter_rfs,
 #endif
-#ifdef CONFIG_NET_RX_BUSY_POLL
-	.ndo_busy_poll		= mlx4_en_low_latency_recv,
-#endif
 	.ndo_get_phys_port_id	= mlx4_en_get_phys_port_id,
 #ifdef CONFIG_MLX4_EN_VXLAN
 	.ndo_add_vxlan_port	= mlx4_en_add_vxlan_port,
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
index e7a5000aa12c..41440b2b20a3 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
@@ -873,10 +873,8 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
 		 * - TCP/IP (v4)
 		 * - without IP options
 		 * - not an IP fragment
-		 * - no LLS polling in progress
 		 */
-		if (!mlx4_en_cq_busy_polling(cq) &&
-		    (dev->features & NETIF_F_GRO)) {
+		if (dev->features & NETIF_F_GRO) {
 			struct sk_buff *gro_skb = napi_get_frags(&cq->napi);
 			if (!gro_skb)
 				goto next;
@@ -927,7 +925,6 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
 						PKT_HASH_TYPE_L3);
 
 			skb_record_rx_queue(gro_skb, cq->ring);
-			skb_mark_napi_id(gro_skb, &cq->napi);
 
 			if (ring->hwtstamp_rx_filter == HWTSTAMP_FILTER_ALL) {
 				timestamp = mlx4_en_get_cqe_ts(cqe);
@@ -990,13 +987,7 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
 					       timestamp);
 		}
 
-		skb_mark_napi_id(skb, &cq->napi);
-
-		if (!mlx4_en_cq_busy_polling(cq))
-			napi_gro_receive(&cq->napi, skb);
-		else
-			netif_receive_skb(skb);
-
+		napi_gro_receive(&cq->napi, skb);
 next:
 		for (nr = 0; nr < priv->num_frags; nr++)
 			mlx4_en_free_frag(priv, frags, nr);
@@ -1038,13 +1029,8 @@ int mlx4_en_poll_rx_cq(struct napi_struct *napi, int budget)
 	struct mlx4_en_priv *priv = netdev_priv(dev);
 	int done;
 
-	if (!mlx4_en_cq_lock_napi(cq))
-		return budget;
-
 	done = mlx4_en_process_rx_cq(dev, cq, budget);
 
-	mlx4_en_cq_unlock_napi(cq);
-
 	/* If we used up all the quota - we're probably not done yet... */
 	if (done == budget) {
 		const struct cpumask *aff;
diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c
index 85f1b1e7e505..31c491e02e69 100644
--- a/drivers/net/ethernet/mellanox/mlx4/main.c
+++ b/drivers/net/ethernet/mellanox/mlx4/main.c
@@ -892,9 +892,10 @@ static int mlx4_slave_cap(struct mlx4_dev *dev)
 		dev->caps.qp1_proxy[i - 1] = func_cap.qp1_proxy_qpn;
 		dev->caps.port_mask[i] = dev->caps.port_type[i];
 		dev->caps.phys_port_id[i] = func_cap.phys_port_id;
-		if (mlx4_get_slave_pkey_gid_tbl_len(dev, i,
-						    &dev->caps.gid_table_len[i],
-						    &dev->caps.pkey_table_len[i]))
+		err = mlx4_get_slave_pkey_gid_tbl_len(dev, i,
+						      &dev->caps.gid_table_len[i],
+						      &dev->caps.pkey_table_len[i]);
+		if (err)
 			goto err_mem;
 	}
 
@@ -906,6 +907,7 @@ static int mlx4_slave_cap(struct mlx4_dev *dev)
 			 dev->caps.uar_page_size * dev->caps.num_uars,
 			 (unsigned long long)
 			 pci_resource_len(dev->persist->pdev, 2));
+		err = -ENOMEM;
 		goto err_mem;
 	}
 
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
index c41f15102ae0..35de7d2e6b34 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
@@ -320,11 +320,6 @@ struct mlx4_en_rx_ring {
 	void *rx_info;
 	unsigned long bytes;
 	unsigned long packets;
-#ifdef CONFIG_NET_RX_BUSY_POLL
-	unsigned long yields;
-	unsigned long misses;
-	unsigned long cleaned;
-#endif
 	unsigned long csum_ok;
 	unsigned long csum_none;
 	unsigned long csum_complete;
@@ -347,18 +342,6 @@ struct mlx4_en_cq {
 	struct mlx4_cqe *buf;
 #define MLX4_EN_OPCODE_ERROR	0x1e
 
-#ifdef CONFIG_NET_RX_BUSY_POLL
-	unsigned int state;
-#define MLX4_EN_CQ_STATE_IDLE        0
-#define MLX4_EN_CQ_STATE_NAPI     1    /* NAPI owns this CQ */
-#define MLX4_EN_CQ_STATE_POLL     2    /* poll owns this CQ */
-#define MLX4_CQ_LOCKED (MLX4_EN_CQ_STATE_NAPI | MLX4_EN_CQ_STATE_POLL)
-#define MLX4_EN_CQ_STATE_NAPI_YIELD  4    /* NAPI yielded this CQ */
-#define MLX4_EN_CQ_STATE_POLL_YIELD  8    /* poll yielded this CQ */
-#define CQ_YIELD (MLX4_EN_CQ_STATE_NAPI_YIELD | MLX4_EN_CQ_STATE_POLL_YIELD)
-#define CQ_USER_PEND (MLX4_EN_CQ_STATE_POLL | MLX4_EN_CQ_STATE_POLL_YIELD)
-	spinlock_t poll_lock; /* protects from LLS/napi conflicts */
-#endif  /* CONFIG_NET_RX_BUSY_POLL */
 	struct irq_desc *irq_desc;
 };
 
@@ -622,115 +605,6 @@ static inline struct mlx4_cqe *mlx4_en_get_cqe(void *buf, int idx, int cqe_sz)
 	return buf + idx * cqe_sz;
 }
 
-#ifdef CONFIG_NET_RX_BUSY_POLL
-static inline void mlx4_en_cq_init_lock(struct mlx4_en_cq *cq)
-{
-	spin_lock_init(&cq->poll_lock);
-	cq->state = MLX4_EN_CQ_STATE_IDLE;
-}
-
-/* called from the device poll rutine to get ownership of a cq */
-static inline bool mlx4_en_cq_lock_napi(struct mlx4_en_cq *cq)
-{
-	int rc = true;
-	spin_lock(&cq->poll_lock);
-	if (cq->state & MLX4_CQ_LOCKED) {
-		WARN_ON(cq->state & MLX4_EN_CQ_STATE_NAPI);
-		cq->state |= MLX4_EN_CQ_STATE_NAPI_YIELD;
-		rc = false;
-	} else
-		/* we don't care if someone yielded */
-		cq->state = MLX4_EN_CQ_STATE_NAPI;
-	spin_unlock(&cq->poll_lock);
-	return rc;
-}
-
-/* returns true is someone tried to get the cq while napi had it */
-static inline bool mlx4_en_cq_unlock_napi(struct mlx4_en_cq *cq)
-{
-	int rc = false;
-	spin_lock(&cq->poll_lock);
-	WARN_ON(cq->state & (MLX4_EN_CQ_STATE_POLL |
-			       MLX4_EN_CQ_STATE_NAPI_YIELD));
-
-	if (cq->state & MLX4_EN_CQ_STATE_POLL_YIELD)
-		rc = true;
-	cq->state = MLX4_EN_CQ_STATE_IDLE;
-	spin_unlock(&cq->poll_lock);
-	return rc;
-}
-
-/* called from mlx4_en_low_latency_poll() */
-static inline bool mlx4_en_cq_lock_poll(struct mlx4_en_cq *cq)
-{
-	int rc = true;
-	spin_lock_bh(&cq->poll_lock);
-	if ((cq->state & MLX4_CQ_LOCKED)) {
-		struct net_device *dev = cq->dev;
-		struct mlx4_en_priv *priv = netdev_priv(dev);
-		struct mlx4_en_rx_ring *rx_ring = priv->rx_ring[cq->ring];
-
-		cq->state |= MLX4_EN_CQ_STATE_POLL_YIELD;
-		rc = false;
-		rx_ring->yields++;
-	} else
-		/* preserve yield marks */
-		cq->state |= MLX4_EN_CQ_STATE_POLL;
-	spin_unlock_bh(&cq->poll_lock);
-	return rc;
-}
-
-/* returns true if someone tried to get the cq while it was locked */
-static inline bool mlx4_en_cq_unlock_poll(struct mlx4_en_cq *cq)
-{
-	int rc = false;
-	spin_lock_bh(&cq->poll_lock);
-	WARN_ON(cq->state & (MLX4_EN_CQ_STATE_NAPI));
-
-	if (cq->state & MLX4_EN_CQ_STATE_POLL_YIELD)
-		rc = true;
-	cq->state = MLX4_EN_CQ_STATE_IDLE;
-	spin_unlock_bh(&cq->poll_lock);
-	return rc;
-}
-
-/* true if a socket is polling, even if it did not get the lock */
-static inline bool mlx4_en_cq_busy_polling(struct mlx4_en_cq *cq)
-{
-	WARN_ON(!(cq->state & MLX4_CQ_LOCKED));
-	return cq->state & CQ_USER_PEND;
-}
-#else
-static inline void mlx4_en_cq_init_lock(struct mlx4_en_cq *cq)
-{
-}
-
-static inline bool mlx4_en_cq_lock_napi(struct mlx4_en_cq *cq)
-{
-	return true;
-}
-
-static inline bool mlx4_en_cq_unlock_napi(struct mlx4_en_cq *cq)
-{
-	return false;
-}
-
-static inline bool mlx4_en_cq_lock_poll(struct mlx4_en_cq *cq)
-{
-	return false;
-}
-
-static inline bool mlx4_en_cq_unlock_poll(struct mlx4_en_cq *cq)
-{
-	return false;
-}
-
-static inline bool mlx4_en_cq_busy_polling(struct mlx4_en_cq *cq)
-{
-	return false;
-}
-#endif /* CONFIG_NET_RX_BUSY_POLL */
-
 #define MLX4_EN_WOL_DO_MODIFY (1ULL << 63)
 
 void mlx4_en_update_loopback_state(struct net_device *dev,
diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
index 9813d34f3e5b..6fec3e993d02 100644
--- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
+++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
@@ -4952,26 +4952,41 @@ static void rem_slave_counters(struct mlx4_dev *dev, int slave)
 	struct res_counter *counter;
 	struct res_counter *tmp;
 	int err;
-	int index;
+	int *counters_arr = NULL;
+	int i, j;
 
 	err = move_all_busy(dev, slave, RES_COUNTER);
 	if (err)
 		mlx4_warn(dev, "rem_slave_counters: Could not move all counters - too busy for slave %d\n",
 			  slave);
 
-	spin_lock_irq(mlx4_tlock(dev));
-	list_for_each_entry_safe(counter, tmp, counter_list, com.list) {
-		if (counter->com.owner == slave) {
-			index = counter->com.res_id;
-			rb_erase(&counter->com.node,
-				 &tracker->res_tree[RES_COUNTER]);
-			list_del(&counter->com.list);
-			kfree(counter);
-			__mlx4_counter_free(dev, index);
+	counters_arr = kmalloc_array(dev->caps.max_counters,
+				     sizeof(*counters_arr), GFP_KERNEL);
+	if (!counters_arr)
+		return;
+
+	do {
+		i = 0;
+		j = 0;
+		spin_lock_irq(mlx4_tlock(dev));
+		list_for_each_entry_safe(counter, tmp, counter_list, com.list) {
+			if (counter->com.owner == slave) {
+				counters_arr[i++] = counter->com.res_id;
+				rb_erase(&counter->com.node,
+					 &tracker->res_tree[RES_COUNTER]);
+				list_del(&counter->com.list);
+				kfree(counter);
+			}
+		}
+		spin_unlock_irq(mlx4_tlock(dev));
+
+		while (j < i) {
+			__mlx4_counter_free(dev, counters_arr[j++]);
 			mlx4_release_resource(dev, slave, RES_COUNTER, 1, 0);
 		}
-	}
-	spin_unlock_irq(mlx4_tlock(dev));
+	} while (i);
+
+	kfree(counters_arr);
 }
 
 static void rem_slave_xrcdns(struct mlx4_dev *dev, int slave)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index f2ae62dd8c09..69f1c1a412b4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -334,9 +334,15 @@ struct mlx5e_tx_skb_cb {
 
 #define MLX5E_TX_SKB_CB(__skb) ((struct mlx5e_tx_skb_cb *)__skb->cb)
 
+enum mlx5e_dma_map_type {
+	MLX5E_DMA_MAP_SINGLE,
+	MLX5E_DMA_MAP_PAGE
+};
+
 struct mlx5e_sq_dma {
-	dma_addr_t addr;
-	u32        size;
+	dma_addr_t              addr;
+	u32                     size;
+	enum mlx5e_dma_map_type type;
 };
 
 enum {
@@ -558,7 +564,7 @@ void mlx5e_completion_event(struct mlx5_core_cq *mcq);
 void mlx5e_cq_error_event(struct mlx5_core_cq *mcq, enum mlx5_event event);
 int mlx5e_napi_poll(struct napi_struct *napi, int budget);
 bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq);
-bool mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget);
+int mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget);
 bool mlx5e_post_rx_wqes(struct mlx5e_rq *rq);
 struct mlx5_cqe64 *mlx5e_get_cqe(struct mlx5e_cq *cq);
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 5fc4d2d78cdf..f6a8cc787603 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -1020,6 +1020,7 @@ err_close_tx_cqs:
 
 err_napi_del:
 	netif_napi_del(&c->napi);
+	napi_hash_del(&c->napi);
 	kfree(c);
 
 	return err;
@@ -1033,6 +1034,10 @@ static void mlx5e_close_channel(struct mlx5e_channel *c)
 	mlx5e_close_cq(&c->rq.cq);
 	mlx5e_close_tx_cqs(c);
 	netif_napi_del(&c->napi);
+
+	napi_hash_del(&c->napi);
+	synchronize_rcu();
+
 	kfree(c);
 }
 
@@ -1332,6 +1337,42 @@ static int mlx5e_modify_tir_lro(struct mlx5e_priv *priv, int tt)
 	return err;
 }
 
+static int mlx5e_refresh_tir_self_loopback_enable(struct mlx5_core_dev *mdev,
+						  u32 tirn)
+{
+	void *in;
+	int inlen;
+	int err;
+
+	inlen = MLX5_ST_SZ_BYTES(modify_tir_in);
+	in = mlx5_vzalloc(inlen);
+	if (!in)
+		return -ENOMEM;
+
+	MLX5_SET(modify_tir_in, in, bitmask.self_lb_en, 1);
+
+	err = mlx5_core_modify_tir(mdev, tirn, in, inlen);
+
+	kvfree(in);
+
+	return err;
+}
+
+static int mlx5e_refresh_tirs_self_loopback_enable(struct mlx5e_priv *priv)
+{
+	int err;
+	int i;
+
+	for (i = 0; i < MLX5E_NUM_TT; i++) {
+		err = mlx5e_refresh_tir_self_loopback_enable(priv->mdev,
+							     priv->tirn[i]);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
 static int mlx5e_set_dev_port_mtu(struct net_device *netdev)
 {
 	struct mlx5e_priv *priv = netdev_priv(netdev);
@@ -1376,6 +1417,13 @@ int mlx5e_open_locked(struct net_device *netdev)
 		goto err_clear_state_opened_flag;
 	}
 
+	err = mlx5e_refresh_tirs_self_loopback_enable(priv);
+	if (err) {
+		netdev_err(netdev, "%s: mlx5e_refresh_tirs_self_loopback_enable failed, %d\n",
+			   __func__, err);
+		goto err_close_channels;
+	}
+
 	mlx5e_update_carrier(priv);
 	mlx5e_redirect_rqts(priv);
 
@@ -1383,6 +1431,8 @@ int mlx5e_open_locked(struct net_device *netdev)
 
 	return 0;
 
+err_close_channels:
+	mlx5e_close_channels(priv);
 err_clear_state_opened_flag:
 	clear_bit(MLX5E_STATE_OPENED, &priv->state);
 	return err;
@@ -1856,6 +1906,8 @@ static int mlx5e_change_mtu(struct net_device *netdev, int new_mtu)
 
 	mlx5_query_port_max_mtu(mdev, &max_mtu, 1);
 
+	max_mtu = MLX5E_HW2SW_MTU(max_mtu);
+
 	if (new_mtu > max_mtu) {
 		netdev_err(netdev,
 			   "%s: Bad MTU (%d) > (%d) Max\n",
@@ -1909,6 +1961,9 @@ static int mlx5e_check_required_hca_cap(struct mlx5_core_dev *mdev)
 			       "Not creating net device, some required device capabilities are missing\n");
 		return -ENOTSUPP;
 	}
+	if (!MLX5_CAP_ETH(mdev, self_lb_en_modifiable))
+		mlx5_core_warn(mdev, "Self loop back prevention is not supported\n");
+
 	return 0;
 }
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index cf0098596e85..7c8c4088d1be 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -33,6 +33,7 @@
 #include <linux/ip.h>
 #include <linux/ipv6.h>
 #include <linux/tcp.h>
+#include <net/busy_poll.h>
 #include "en.h"
 
 static inline int mlx5e_alloc_rx_wqe(struct mlx5e_rq *rq,
@@ -215,16 +216,16 @@ static inline void mlx5e_build_rx_skb(struct mlx5_cqe64 *cqe,
 				       be16_to_cpu(cqe->vlan_info));
 }
 
-bool mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget)
+int mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget)
 {
 	struct mlx5e_rq *rq = container_of(cq, struct mlx5e_rq, cq);
-	int i;
+	int work_done;
 
 	/* avoid accessing cq (dma coherent memory) if not needed */
 	if (!test_and_clear_bit(MLX5E_CQ_HAS_CQES, &cq->flags))
-		return false;
+		return 0;
 
-	for (i = 0; i < budget; i++) {
+	for (work_done = 0; work_done < budget; work_done++) {
 		struct mlx5e_rx_wqe *wqe;
 		struct mlx5_cqe64 *cqe;
 		struct sk_buff *skb;
@@ -269,10 +270,8 @@ wq_ll_pop:
 	/* ensure cq space is freed before enabling more cqes */
 	wmb();
 
-	if (i == budget) {
+	if (work_done == budget)
 		set_bit(MLX5E_CQ_HAS_CQES, &cq->flags);
-		return true;
-	}
 
-	return false;
+	return work_done;
 }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
index cd8f85a251d7..1341b1d3c421 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
@@ -61,39 +61,47 @@ void mlx5e_send_nop(struct mlx5e_sq *sq, bool notify_hw)
 	}
 }
 
-static void mlx5e_dma_pop_last_pushed(struct mlx5e_sq *sq, dma_addr_t *addr,
-				      u32 *size)
+static inline void mlx5e_tx_dma_unmap(struct device *pdev,
+				      struct mlx5e_sq_dma *dma)
 {
-	sq->dma_fifo_pc--;
-	*addr = sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].addr;
-	*size = sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].size;
-}
-
-static void mlx5e_dma_unmap_wqe_err(struct mlx5e_sq *sq, struct sk_buff *skb)
-{
-	dma_addr_t addr;
-	u32 size;
-	int i;
-
-	for (i = 0; i < MLX5E_TX_SKB_CB(skb)->num_dma; i++) {
-		mlx5e_dma_pop_last_pushed(sq, &addr, &size);
-		dma_unmap_single(sq->pdev, addr, size, DMA_TO_DEVICE);
+	switch (dma->type) {
+	case MLX5E_DMA_MAP_SINGLE:
+		dma_unmap_single(pdev, dma->addr, dma->size, DMA_TO_DEVICE);
+		break;
+	case MLX5E_DMA_MAP_PAGE:
+		dma_unmap_page(pdev, dma->addr, dma->size, DMA_TO_DEVICE);
+		break;
+	default:
+		WARN_ONCE(true, "mlx5e_tx_dma_unmap unknown DMA type!\n");
 	}
 }
 
-static inline void mlx5e_dma_push(struct mlx5e_sq *sq, dma_addr_t addr,
-				  u32 size)
+static inline void mlx5e_dma_push(struct mlx5e_sq *sq,
+				  dma_addr_t addr,
+				  u32 size,
+				  enum mlx5e_dma_map_type map_type)
 {
 	sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].addr = addr;
 	sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].size = size;
+	sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].type = map_type;
 	sq->dma_fifo_pc++;
 }
 
-static inline void mlx5e_dma_get(struct mlx5e_sq *sq, u32 i, dma_addr_t *addr,
-				 u32 *size)
+static inline struct mlx5e_sq_dma *mlx5e_dma_get(struct mlx5e_sq *sq, u32 i)
 {
-	*addr = sq->dma_fifo[i & sq->dma_fifo_mask].addr;
-	*size = sq->dma_fifo[i & sq->dma_fifo_mask].size;
+	return &sq->dma_fifo[i & sq->dma_fifo_mask];
+}
+
+static void mlx5e_dma_unmap_wqe_err(struct mlx5e_sq *sq, struct sk_buff *skb)
+{
+	int i;
+
+	for (i = 0; i < MLX5E_TX_SKB_CB(skb)->num_dma; i++) {
+		struct mlx5e_sq_dma *last_pushed_dma =
+			mlx5e_dma_get(sq, --sq->dma_fifo_pc);
+
+		mlx5e_tx_dma_unmap(sq->pdev, last_pushed_dma);
+	}
 }
 
 u16 mlx5e_select_queue(struct net_device *dev, struct sk_buff *skb,
@@ -118,8 +126,15 @@ static inline u16 mlx5e_get_inline_hdr_size(struct mlx5e_sq *sq,
 	 */
 #define MLX5E_MIN_INLINE ETH_HLEN
 
-	if (bf && (skb_headlen(skb) <= sq->max_inline))
-		return skb_headlen(skb);
+	if (bf) {
+		u16 ihs = skb_headlen(skb);
+
+		if (skb_vlan_tag_present(skb))
+			ihs += VLAN_HLEN;
+
+		if (ihs <= sq->max_inline)
+			return skb_headlen(skb);
+	}
 
 	return MLX5E_MIN_INLINE;
 }
@@ -218,7 +233,7 @@ static netdev_tx_t mlx5e_sq_xmit(struct mlx5e_sq *sq, struct sk_buff *skb)
 		dseg->lkey       = sq->mkey_be;
 		dseg->byte_count = cpu_to_be32(headlen);
 
-		mlx5e_dma_push(sq, dma_addr, headlen);
+		mlx5e_dma_push(sq, dma_addr, headlen, MLX5E_DMA_MAP_SINGLE);
 		MLX5E_TX_SKB_CB(skb)->num_dma++;
 
 		dseg++;
@@ -237,7 +252,7 @@ static netdev_tx_t mlx5e_sq_xmit(struct mlx5e_sq *sq, struct sk_buff *skb)
 		dseg->lkey       = sq->mkey_be;
 		dseg->byte_count = cpu_to_be32(fsz);
 
-		mlx5e_dma_push(sq, dma_addr, fsz);
+		mlx5e_dma_push(sq, dma_addr, fsz, MLX5E_DMA_MAP_PAGE);
 		MLX5E_TX_SKB_CB(skb)->num_dma++;
 
 		dseg++;
@@ -353,13 +368,10 @@ bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq)
 			}
 
 			for (j = 0; j < MLX5E_TX_SKB_CB(skb)->num_dma; j++) {
-				dma_addr_t addr;
-				u32 size;
+				struct mlx5e_sq_dma *dma =
+					mlx5e_dma_get(sq, dma_fifo_cc++);
 
-				mlx5e_dma_get(sq, dma_fifo_cc, &addr, &size);
-				dma_fifo_cc++;
-				dma_unmap_single(sq->pdev, addr, size,
-						 DMA_TO_DEVICE);
+				mlx5e_tx_dma_unmap(sq->pdev, dma);
 			}
 
 			npkts++;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
index 2c7cb6755d1d..4ac8d716dbdd 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
@@ -54,6 +54,7 @@ int mlx5e_napi_poll(struct napi_struct *napi, int budget)
 	struct mlx5e_channel *c = container_of(napi, struct mlx5e_channel,
 					       napi);
 	bool busy = false;
+	int work_done;
 	int i;
 
 	clear_bit(MLX5E_CHANNEL_NAPI_SCHED, &c->flags);
@@ -61,26 +62,26 @@ int mlx5e_napi_poll(struct napi_struct *napi, int budget)
 	for (i = 0; i < c->num_tc; i++)
 		busy |= mlx5e_poll_tx_cq(&c->sq[i].cq);
 
-	busy |= mlx5e_poll_rx_cq(&c->rq.cq, budget);
-
+	work_done = mlx5e_poll_rx_cq(&c->rq.cq, budget);
+	busy |= work_done == budget;
 	busy |= mlx5e_post_rx_wqes(&c->rq);
 
 	if (busy)
 		return budget;
 
-	napi_complete(napi);
+	napi_complete_done(napi, work_done);
 
 	/* avoid losing completion event during/after polling cqs */
 	if (test_bit(MLX5E_CHANNEL_NAPI_SCHED, &c->flags)) {
 		napi_schedule(napi);
-		return 0;
+		return work_done;
 	}
 
 	for (i = 0; i < c->num_tc; i++)
 		mlx5e_cq_arm(&c->sq[i].cq);
 	mlx5e_cq_arm(&c->rq.cq);
 
-	return 0;
+	return work_done;
 }
 
 void mlx5e_completion_event(struct mlx5_core_cq *mcq)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c
index 617fb22b5d81..f21e23983a1a 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c
@@ -342,14 +342,35 @@ err_port_add_vid:
 	return err;
 }
 
+static int __mlxsw_sp_port_vlans_set(struct mlxsw_sp_port *mlxsw_sp_port,
+				     u16 vid_begin, u16 vid_end, bool is_member,
+				     bool untagged)
+{
+	u16 vid, vid_e;
+	int err;
+
+	for (vid = vid_begin; vid <= vid_end;
+	     vid += MLXSW_REG_SPVM_REC_MAX_COUNT) {
+		vid_e = min((u16) (vid + MLXSW_REG_SPVM_REC_MAX_COUNT - 1),
+			    vid_end);
+
+		err = mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid_e,
+					     is_member, untagged);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
 static int __mlxsw_sp_port_vlans_add(struct mlxsw_sp_port *mlxsw_sp_port,
 				     u16 vid_begin, u16 vid_end,
 				     bool flag_untagged, bool flag_pvid)
 {
 	struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp;
 	struct net_device *dev = mlxsw_sp_port->dev;
+	u16 vid, last_visited_vid, old_pvid;
 	enum mlxsw_reg_svfa_mt mt;
-	u16 vid, vid_e;
 	int err;
 
 	/* In case this is invoked with BRIDGE_FLAGS_SELF and port is
@@ -377,15 +398,18 @@ static int __mlxsw_sp_port_vlans_add(struct mlxsw_sp_port *mlxsw_sp_port,
 			if (err) {
 				netdev_err(dev, "Failed to create FID=VID=%d mapping\n",
 					   vid);
-				return err;
+				goto err_port_vid_to_fid_set;
 			}
 		}
+	}
 
-		/* Set FID mapping according to port's mode */
+	/* Set FID mapping according to port's mode */
+	for (vid = vid_begin; vid <= vid_end; vid++) {
 		err = mlxsw_sp_port_fid_map(mlxsw_sp_port, vid);
 		if (err) {
 			netdev_err(dev, "Failed to map FID=%d", vid);
-			return err;
+			last_visited_vid = --vid;
+			goto err_port_fid_map;
 		}
 	}
 
@@ -393,40 +417,62 @@ static int __mlxsw_sp_port_vlans_add(struct mlxsw_sp_port *mlxsw_sp_port,
 					true, false);
 	if (err) {
 		netdev_err(dev, "Failed to configure flooding\n");
-		return err;
+		goto err_port_flood_set;
 	}
 
-	for (vid = vid_begin; vid <= vid_end;
-	     vid += MLXSW_REG_SPVM_REC_MAX_COUNT) {
-		vid_e = min((u16) (vid + MLXSW_REG_SPVM_REC_MAX_COUNT - 1),
-			    vid_end);
-
-		err = mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid_e, true,
-					     flag_untagged);
-		if (err) {
-			netdev_err(mlxsw_sp_port->dev, "Unable to add VIDs %d-%d\n",
-				   vid, vid_e);
-			return err;
-		}
+	err = __mlxsw_sp_port_vlans_set(mlxsw_sp_port, vid_begin, vid_end,
+					true, flag_untagged);
+	if (err) {
+		netdev_err(dev, "Unable to add VIDs %d-%d\n", vid_begin,
+			   vid_end);
+		goto err_port_vlans_set;
 	}
 
-	vid = vid_begin;
-	if (flag_pvid && mlxsw_sp_port->pvid != vid) {
-		err = mlxsw_sp_port_pvid_set(mlxsw_sp_port, vid);
+	old_pvid = mlxsw_sp_port->pvid;
+	if (flag_pvid && old_pvid != vid_begin) {
+		err = mlxsw_sp_port_pvid_set(mlxsw_sp_port, vid_begin);
 		if (err) {
-			netdev_err(mlxsw_sp_port->dev, "Unable to add PVID %d\n",
-				   vid);
-			return err;
+			netdev_err(dev, "Unable to add PVID %d\n", vid_begin);
+			goto err_port_pvid_set;
 		}
-		mlxsw_sp_port->pvid = vid;
+		mlxsw_sp_port->pvid = vid_begin;
 	}
 
 	/* Changing activity bits only if HW operation succeded */
 	for (vid = vid_begin; vid <= vid_end; vid++)
 		set_bit(vid, mlxsw_sp_port->active_vlans);
 
-	return mlxsw_sp_port_stp_state_set(mlxsw_sp_port,
-					   mlxsw_sp_port->stp_state);
+	/* STP state change must be done after we set active VLANs */
+	err = mlxsw_sp_port_stp_state_set(mlxsw_sp_port,
+					  mlxsw_sp_port->stp_state);
+	if (err) {
+		netdev_err(dev, "Failed to set STP state\n");
+		goto err_port_stp_state_set;
+	}
+
+	return 0;
+
+err_port_vid_to_fid_set:
+	mlxsw_sp_fid_destroy(mlxsw_sp, vid);
+	return err;
+
+err_port_stp_state_set:
+	for (vid = vid_begin; vid <= vid_end; vid++)
+		clear_bit(vid, mlxsw_sp_port->active_vlans);
+	if (old_pvid != mlxsw_sp_port->pvid)
+		mlxsw_sp_port_pvid_set(mlxsw_sp_port, old_pvid);
+err_port_pvid_set:
+	__mlxsw_sp_port_vlans_set(mlxsw_sp_port, vid_begin, vid_end, false,
+				  false);
+err_port_vlans_set:
+	__mlxsw_sp_port_flood_set(mlxsw_sp_port, vid_begin, vid_end, false,
+				  false);
+err_port_flood_set:
+	last_visited_vid = vid_end;
+err_port_fid_map:
+	for (vid = last_visited_vid; vid >= vid_begin; vid--)
+		mlxsw_sp_port_fid_unmap(mlxsw_sp_port, vid);
+	return err;
 }
 
 static int mlxsw_sp_port_vlans_add(struct mlxsw_sp_port *mlxsw_sp_port,
@@ -532,7 +578,7 @@ static int __mlxsw_sp_port_vlans_del(struct mlxsw_sp_port *mlxsw_sp_port,
 				     u16 vid_begin, u16 vid_end, bool init)
 {
 	struct net_device *dev = mlxsw_sp_port->dev;
-	u16 vid, vid_e;
+	u16 vid, pvid;
 	int err;
 
 	/* In case this is invoked with BRIDGE_FLAGS_SELF and port is
@@ -542,30 +588,23 @@ static int __mlxsw_sp_port_vlans_del(struct mlxsw_sp_port *mlxsw_sp_port,
 	if (!init && !mlxsw_sp_port->bridged)
 		return mlxsw_sp_port_kill_vids(dev, vid_begin, vid_end);
 
-	for (vid = vid_begin; vid <= vid_end;
-	     vid += MLXSW_REG_SPVM_REC_MAX_COUNT) {
-		vid_e = min((u16) (vid + MLXSW_REG_SPVM_REC_MAX_COUNT - 1),
-			    vid_end);
-		err = mlxsw_sp_port_vlan_set(mlxsw_sp_port, vid, vid_e, false,
-					     false);
-		if (err) {
-			netdev_err(mlxsw_sp_port->dev, "Unable to del VIDs %d-%d\n",
-				   vid, vid_e);
-			return err;
-		}
+	err = __mlxsw_sp_port_vlans_set(mlxsw_sp_port, vid_begin, vid_end,
+					false, false);
+	if (err) {
+		netdev_err(dev, "Unable to del VIDs %d-%d\n", vid_begin,
+			   vid_end);
+		return err;
 	}
 
-	if ((mlxsw_sp_port->pvid >= vid_begin) &&
-	    (mlxsw_sp_port->pvid <= vid_end)) {
+	pvid = mlxsw_sp_port->pvid;
+	if (pvid >= vid_begin && pvid <= vid_end && pvid != 1) {
 		/* Default VLAN is always 1 */
-		mlxsw_sp_port->pvid = 1;
-		err = mlxsw_sp_port_pvid_set(mlxsw_sp_port,
-					     mlxsw_sp_port->pvid);
+		err = mlxsw_sp_port_pvid_set(mlxsw_sp_port, 1);
 		if (err) {
-			netdev_err(mlxsw_sp_port->dev, "Unable to del PVID %d\n",
-				   vid);
+			netdev_err(dev, "Unable to del PVID %d\n", pvid);
 			return err;
 		}
+		mlxsw_sp_port->pvid = 1;
 	}
 
 	if (init)