5 files changed, 536 insertions, 186 deletions

diff --git a/drivers/net/ethernet/marvell/mv643xx_eth.c b/drivers/net/ethernet/marvell/mv643xx_eth.c
index b7b8d74c22d9..b151a949f352 100644
--- a/drivers/net/ethernet/marvell/mv643xx_eth.c
+++ b/drivers/net/ethernet/marvell/mv643xx_eth.c
@@ -42,6 +42,7 @@
 #include <linux/dma-mapping.h>
 #include <linux/in.h>
 #include <linux/ip.h>
+#include <net/tso.h>
 #include <linux/tcp.h>
 #include <linux/udp.h>
 #include <linux/etherdevice.h>
@@ -179,10 +180,18 @@ static char mv643xx_eth_driver_version[] = "1.4";
  * Misc definitions.
  */
 #define DEFAULT_RX_QUEUE_SIZE	128
-#define DEFAULT_TX_QUEUE_SIZE	256
+#define DEFAULT_TX_QUEUE_SIZE	512
 #define SKB_DMA_REALIGN		((PAGE_SIZE - NET_SKB_PAD) % SMP_CACHE_BYTES)
 
+#define TSO_HEADER_SIZE		128
 
+/* Max number of allowed TCP segments for software TSO */
+#define MV643XX_MAX_TSO_SEGS 100
+#define MV643XX_MAX_SKB_DESCS (MV643XX_MAX_TSO_SEGS * 2 + MAX_SKB_FRAGS)
+
+#define IS_TSO_HEADER(txq, addr) \
+	((addr >= txq->tso_hdrs_dma) && \
+	 (addr < txq->tso_hdrs_dma + txq->tx_ring_size * TSO_HEADER_SIZE))
 /*
  * RX/TX descriptors.
  */
@@ -250,6 +259,7 @@ struct tx_desc {
 #define GEN_TCP_UDP_CHECKSUM		0x00020000
 #define UDP_FRAME			0x00010000
 #define MAC_HDR_EXTRA_4_BYTES		0x00008000
+#define GEN_TCP_UDP_CHK_FULL		0x00000400
 #define MAC_HDR_EXTRA_8_BYTES		0x00000200
 
 #define TX_IHL_SHIFT			11
@@ -345,6 +355,12 @@ struct tx_queue {
 	int tx_curr_desc;
 	int tx_used_desc;
 
+	int tx_stop_threshold;
+	int tx_wake_threshold;
+
+	char *tso_hdrs;
+	dma_addr_t tso_hdrs_dma;
+
 	struct tx_desc *tx_desc_area;
 	dma_addr_t tx_desc_dma;
 	int tx_desc_area_size;
@@ -491,7 +507,7 @@ static void txq_maybe_wake(struct tx_queue *txq)
 
 	if (netif_tx_queue_stopped(nq)) {
 		__netif_tx_lock(nq, smp_processor_id());
-		if (txq->tx_ring_size - txq->tx_desc_count >= MAX_SKB_FRAGS + 1)
+		if (txq->tx_desc_count <= txq->tx_wake_threshold)
 			netif_tx_wake_queue(nq);
 		__netif_tx_unlock(nq);
 	}
@@ -661,6 +677,198 @@ static inline unsigned int has_tiny_unaligned_frags(struct sk_buff *skb)
 	return 0;
 }
 
+static inline __be16 sum16_as_be(__sum16 sum)
+{
+	return (__force __be16)sum;
+}
+
+static int skb_tx_csum(struct mv643xx_eth_private *mp, struct sk_buff *skb,
+		       u16 *l4i_chk, u32 *command, int length)
+{
+	int ret;
+	u32 cmd = 0;
+
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
+		int hdr_len;
+		int tag_bytes;
+
+		BUG_ON(skb->protocol != htons(ETH_P_IP) &&
+		       skb->protocol != htons(ETH_P_8021Q));
+
+		hdr_len = (void *)ip_hdr(skb) - (void *)skb->data;
+		tag_bytes = hdr_len - ETH_HLEN;
+
+		if (length - hdr_len > mp->shared->tx_csum_limit ||
+		    unlikely(tag_bytes & ~12)) {
+			ret = skb_checksum_help(skb);
+			if (!ret)
+				goto no_csum;
+			return ret;
+		}
+
+		if (tag_bytes & 4)
+			cmd |= MAC_HDR_EXTRA_4_BYTES;
+		if (tag_bytes & 8)
+			cmd |= MAC_HDR_EXTRA_8_BYTES;
+
+		cmd |= GEN_TCP_UDP_CHECKSUM | GEN_TCP_UDP_CHK_FULL |
+			   GEN_IP_V4_CHECKSUM   |
+			   ip_hdr(skb)->ihl << TX_IHL_SHIFT;
+
+		/* TODO: Revisit this. With the usage of GEN_TCP_UDP_CHK_FULL
+		 * it seems we don't need to pass the initial checksum. */
+		switch (ip_hdr(skb)->protocol) {
+		case IPPROTO_UDP:
+			cmd |= UDP_FRAME;
+			*l4i_chk = 0;
+			break;
+		case IPPROTO_TCP:
+			*l4i_chk = 0;
+			break;
+		default:
+			WARN(1, "protocol not supported");
+		}
+	} else {
+no_csum:
+		/* Errata BTS #50, IHL must be 5 if no HW checksum */
+		cmd |= 5 << TX_IHL_SHIFT;
+	}
+	*command = cmd;
+	return 0;
+}
+
+static inline int
+txq_put_data_tso(struct net_device *dev, struct tx_queue *txq,
+		 struct sk_buff *skb, char *data, int length,
+		 bool last_tcp, bool is_last)
+{
+	int tx_index;
+	u32 cmd_sts;
+	struct tx_desc *desc;
+
+	tx_index = txq->tx_curr_desc++;
+	if (txq->tx_curr_desc == txq->tx_ring_size)
+		txq->tx_curr_desc = 0;
+	desc = &txq->tx_desc_area[tx_index];
+
+	desc->l4i_chk = 0;
+	desc->byte_cnt = length;
+	desc->buf_ptr = dma_map_single(dev->dev.parent, data,
+				       length, DMA_TO_DEVICE);
+	if (unlikely(dma_mapping_error(dev->dev.parent, desc->buf_ptr))) {
+		WARN(1, "dma_map_single failed!\n");
+		return -ENOMEM;
+	}
+
+	cmd_sts = BUFFER_OWNED_BY_DMA;
+	if (last_tcp) {
+		/* last descriptor in the TCP packet */
+		cmd_sts |= ZERO_PADDING | TX_LAST_DESC;
+		/* last descriptor in SKB */
+		if (is_last)
+			cmd_sts |= TX_ENABLE_INTERRUPT;
+	}
+	desc->cmd_sts = cmd_sts;
+	return 0;
+}
+
+static inline void
+txq_put_hdr_tso(struct sk_buff *skb, struct tx_queue *txq, int length)
+{
+	struct mv643xx_eth_private *mp = txq_to_mp(txq);
+	int hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
+	int tx_index;
+	struct tx_desc *desc;
+	int ret;
+	u32 cmd_csum = 0;
+	u16 l4i_chk = 0;
+
+	tx_index = txq->tx_curr_desc;
+	desc = &txq->tx_desc_area[tx_index];
+
+	ret = skb_tx_csum(mp, skb, &l4i_chk, &cmd_csum, length);
+	if (ret)
+		WARN(1, "failed to prepare checksum!");
+
+	/* Should we set this? Can't use the value from skb_tx_csum()
+	 * as it's not the correct initial L4 checksum to use. */
+	desc->l4i_chk = 0;
+
+	desc->byte_cnt = hdr_len;
+	desc->buf_ptr = txq->tso_hdrs_dma +
+			txq->tx_curr_desc * TSO_HEADER_SIZE;
+	desc->cmd_sts = cmd_csum | BUFFER_OWNED_BY_DMA  | TX_FIRST_DESC |
+				   GEN_CRC;
+
+	txq->tx_curr_desc++;
+	if (txq->tx_curr_desc == txq->tx_ring_size)
+		txq->tx_curr_desc = 0;
+}
+
+static int txq_submit_tso(struct tx_queue *txq, struct sk_buff *skb,
+			  struct net_device *dev)
+{
+	struct mv643xx_eth_private *mp = txq_to_mp(txq);
+	int total_len, data_left, ret;
+	int desc_count = 0;
+	struct tso_t tso;
+	int hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
+
+	/* Count needed descriptors */
+	if ((txq->tx_desc_count + tso_count_descs(skb)) >= txq->tx_ring_size) {
+		netdev_dbg(dev, "not enough descriptors for TSO!\n");
+		return -EBUSY;
+	}
+
+	/* Initialize the TSO handler, and prepare the first payload */
+	tso_start(skb, &tso);
+
+	total_len = skb->len - hdr_len;
+	while (total_len > 0) {
+		char *hdr;
+
+		data_left = min_t(int, skb_shinfo(skb)->gso_size, total_len);
+		total_len -= data_left;
+		desc_count++;
+
+		/* prepare packet headers: MAC + IP + TCP */
+		hdr = txq->tso_hdrs + txq->tx_curr_desc * TSO_HEADER_SIZE;
+		tso_build_hdr(skb, hdr, &tso, data_left, total_len == 0);
+		txq_put_hdr_tso(skb, txq, data_left);
+
+		while (data_left > 0) {
+			int size;
+			desc_count++;
+
+			size = min_t(int, tso.size, data_left);
+			ret = txq_put_data_tso(dev, txq, skb, tso.data, size,
+					       size == data_left,
+					       total_len == 0);
+			if (ret)
+				goto err_release;
+			data_left -= size;
+			tso_build_data(skb, &tso, size);
+		}
+	}
+
+	__skb_queue_tail(&txq->tx_skb, skb);
+	skb_tx_timestamp(skb);
+
+	/* clear TX_END status */
+	mp->work_tx_end &= ~(1 << txq->index);
+
+	/* ensure all descriptors are written before poking hardware */
+	wmb();
+	txq_enable(txq);
+	txq->tx_desc_count += desc_count;
+	return 0;
+err_release:
+	/* TODO: Release all used data descriptors; header descriptors must not
+	 * be DMA-unmapped.
+	 */
+	return ret;
+}
+
 static void txq_submit_frag_skb(struct tx_queue *txq, struct sk_buff *skb)
 {
 	struct mv643xx_eth_private *mp = txq_to_mp(txq);
@@ -671,8 +879,10 @@ static void txq_submit_frag_skb(struct tx_queue *txq, struct sk_buff *skb)
 		skb_frag_t *this_frag;
 		int tx_index;
 		struct tx_desc *desc;
+		void *addr;
 
 		this_frag = &skb_shinfo(skb)->frags[frag];
+		addr = page_address(this_frag->page.p) + this_frag->page_offset;
 		tx_index = txq->tx_curr_desc++;
 		if (txq->tx_curr_desc == txq->tx_ring_size)
 			txq->tx_curr_desc = 0;
@@ -692,19 +902,13 @@ static void txq_submit_frag_skb(struct tx_queue *txq, struct sk_buff *skb)
 
 		desc->l4i_chk = 0;
 		desc->byte_cnt = skb_frag_size(this_frag);
-		desc->buf_ptr = skb_frag_dma_map(mp->dev->dev.parent,
-						 this_frag, 0,
-						 skb_frag_size(this_frag),
-						 DMA_TO_DEVICE);
+		desc->buf_ptr = dma_map_single(mp->dev->dev.parent, addr,
+					       desc->byte_cnt, DMA_TO_DEVICE);
 	}
 }
 
-static inline __be16 sum16_as_be(__sum16 sum)
-{
-	return (__force __be16)sum;
-}
-
-static int txq_submit_skb(struct tx_queue *txq, struct sk_buff *skb)
+static int txq_submit_skb(struct tx_queue *txq, struct sk_buff *skb,
+			  struct net_device *dev)
 {
 	struct mv643xx_eth_private *mp = txq_to_mp(txq);
 	int nr_frags = skb_shinfo(skb)->nr_frags;
@@ -712,54 +916,22 @@ static int txq_submit_skb(struct tx_queue *txq, struct sk_buff *skb)
 	struct tx_desc *desc;
 	u32 cmd_sts;
 	u16 l4i_chk;
-	int length;
+	int length, ret;
 
-	cmd_sts = TX_FIRST_DESC | GEN_CRC | BUFFER_OWNED_BY_DMA;
+	cmd_sts = 0;
 	l4i_chk = 0;
 
-	if (skb->ip_summed == CHECKSUM_PARTIAL) {
-		int hdr_len;
-		int tag_bytes;
-
-		BUG_ON(skb->protocol != htons(ETH_P_IP) &&
-		       skb->protocol != htons(ETH_P_8021Q));
-
-		hdr_len = (void *)ip_hdr(skb) - (void *)skb->data;
-		tag_bytes = hdr_len - ETH_HLEN;
-		if (skb->len - hdr_len > mp->shared->tx_csum_limit ||
-		    unlikely(tag_bytes & ~12)) {
-			if (skb_checksum_help(skb) == 0)
-				goto no_csum;
-			dev_kfree_skb_any(skb);
-			return 1;
-		}
-
-		if (tag_bytes & 4)
-			cmd_sts |= MAC_HDR_EXTRA_4_BYTES;
-		if (tag_bytes & 8)
-			cmd_sts |= MAC_HDR_EXTRA_8_BYTES;
-
-		cmd_sts |= GEN_TCP_UDP_CHECKSUM |
-			   GEN_IP_V4_CHECKSUM   |
-			   ip_hdr(skb)->ihl << TX_IHL_SHIFT;
-
-		switch (ip_hdr(skb)->protocol) {
-		case IPPROTO_UDP:
-			cmd_sts |= UDP_FRAME;
-			l4i_chk = ntohs(sum16_as_be(udp_hdr(skb)->check));
-			break;
-		case IPPROTO_TCP:
-			l4i_chk = ntohs(sum16_as_be(tcp_hdr(skb)->check));
-			break;
-		default:
-			BUG();
-		}
-	} else {
-no_csum:
-		/* Errata BTS #50, IHL must be 5 if no HW checksum */
-		cmd_sts |= 5 << TX_IHL_SHIFT;
+	if (txq->tx_ring_size - txq->tx_desc_count < MAX_SKB_FRAGS + 1) {
+		if (net_ratelimit())
+			netdev_err(dev, "tx queue full?!\n");
+		return -EBUSY;
 	}
 
+	ret = skb_tx_csum(mp, skb, &l4i_chk, &cmd_sts, skb->len);
+	if (ret)
+		return ret;
+	cmd_sts |= TX_FIRST_DESC | GEN_CRC | BUFFER_OWNED_BY_DMA;
+
 	tx_index = txq->tx_curr_desc++;
 	if (txq->tx_curr_desc == txq->tx_ring_size)
 		txq->tx_curr_desc = 0;
@@ -801,7 +973,7 @@ no_csum:
 static netdev_tx_t mv643xx_eth_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct mv643xx_eth_private *mp = netdev_priv(dev);
-	int length, queue;
+	int length, queue, ret;
 	struct tx_queue *txq;
 	struct netdev_queue *nq;
 
@@ -810,30 +982,26 @@ static netdev_tx_t mv643xx_eth_xmit(struct sk_buff *skb, struct net_device *dev)
 	nq = netdev_get_tx_queue(dev, queue);
 
 	if (has_tiny_unaligned_frags(skb) && __skb_linearize(skb)) {
-		txq->tx_dropped++;
 		netdev_printk(KERN_DEBUG, dev,
 			      "failed to linearize skb with tiny unaligned fragment\n");
 		return NETDEV_TX_BUSY;
 	}
 
-	if (txq->tx_ring_size - txq->tx_desc_count < MAX_SKB_FRAGS + 1) {
-		if (net_ratelimit())
-			netdev_err(dev, "tx queue full?!\n");
-		dev_kfree_skb_any(skb);
-		return NETDEV_TX_OK;
-	}
-
 	length = skb->len;
 
-	if (!txq_submit_skb(txq, skb)) {
-		int entries_left;
-
+	if (skb_is_gso(skb))
+		ret = txq_submit_tso(txq, skb, dev);
+	else
+		ret = txq_submit_skb(txq, skb, dev);
+	if (!ret) {
 		txq->tx_bytes += length;
 		txq->tx_packets++;
 
-		entries_left = txq->tx_ring_size - txq->tx_desc_count;
-		if (entries_left < MAX_SKB_FRAGS + 1)
+		if (txq->tx_desc_count >= txq->tx_stop_threshold)
 			netif_tx_stop_queue(nq);
+	} else {
+		txq->tx_dropped++;
+		dev_kfree_skb_any(skb);
 	}
 
 	return NETDEV_TX_OK;
@@ -907,14 +1075,9 @@ static int txq_reclaim(struct tx_queue *txq, int budget, int force)
 			mp->dev->stats.tx_errors++;
 		}
 
-		if (cmd_sts & TX_FIRST_DESC) {
+		if (!IS_TSO_HEADER(txq, desc->buf_ptr))
 			dma_unmap_single(mp->dev->dev.parent, desc->buf_ptr,
 					 desc->byte_cnt, DMA_TO_DEVICE);
-		} else {
-			dma_unmap_page(mp->dev->dev.parent, desc->buf_ptr,
-				       desc->byte_cnt, DMA_TO_DEVICE);
-		}
-
 		dev_kfree_skb(skb);
 	}
 
@@ -1010,8 +1173,9 @@ static void txq_set_fixed_prio_mode(struct tx_queue *txq)
 
 
 /* mii management interface *************************************************/
-static void mv643xx_adjust_pscr(struct mv643xx_eth_private *mp)
+static void mv643xx_eth_adjust_link(struct net_device *dev)
 {
+	struct mv643xx_eth_private *mp = netdev_priv(dev);
 	u32 pscr = rdlp(mp, PORT_SERIAL_CONTROL);
 	u32 autoneg_disable = FORCE_LINK_PASS |
 	             DISABLE_AUTO_NEG_SPEED_GMII |
@@ -1387,7 +1551,7 @@ mv643xx_eth_set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
 
 	ret = phy_ethtool_sset(mp->phy, cmd);
 	if (!ret)
-		mv643xx_adjust_pscr(mp);
+		mv643xx_eth_adjust_link(dev);
 	return ret;
 }
 
@@ -1456,7 +1620,11 @@ mv643xx_eth_set_ringparam(struct net_device *dev, struct ethtool_ringparam *er)
 		return -EINVAL;
 
 	mp->rx_ring_size = er->rx_pending < 4096 ? er->rx_pending : 4096;
-	mp->tx_ring_size = er->tx_pending < 4096 ? er->tx_pending : 4096;
+	mp->tx_ring_size = clamp_t(unsigned int, er->tx_pending,
+				   MV643XX_MAX_SKB_DESCS * 2, 4096);
+	if (mp->tx_ring_size != er->tx_pending)
+		netdev_warn(dev, "TX queue size set to %u (requested %u)\n",
+			    mp->tx_ring_size, er->tx_pending);
 
 	if (netif_running(dev)) {
 		mv643xx_eth_stop(dev);
@@ -1832,6 +2000,13 @@ static int txq_init(struct mv643xx_eth_private *mp, int index)
 
 	txq->tx_ring_size = mp->tx_ring_size;
 
+	/* A queue must always have room for at least one skb.
+	 * Therefore, stop the queue when the free entries reaches
+	 * the maximum number of descriptors per skb.
+	 */
+	txq->tx_stop_threshold = txq->tx_ring_size - MV643XX_MAX_SKB_DESCS;
+	txq->tx_wake_threshold = txq->tx_stop_threshold / 2;
+
 	txq->tx_desc_count = 0;
 	txq->tx_curr_desc = 0;
 	txq->tx_used_desc = 0;
@@ -1871,6 +2046,15 @@ static int txq_init(struct mv643xx_eth_private *mp, int index)
 					nexti * sizeof(struct tx_desc);
 	}
 
+	/* Allocate DMA buffers for TSO MAC/IP/TCP headers */
+	txq->tso_hdrs = dma_alloc_coherent(mp->dev->dev.parent,
+					   txq->tx_ring_size * TSO_HEADER_SIZE,
+					   &txq->tso_hdrs_dma, GFP_KERNEL);
+	if (txq->tso_hdrs == NULL) {
+		dma_free_coherent(mp->dev->dev.parent, txq->tx_desc_area_size,
+				  txq->tx_desc_area, txq->tx_desc_dma);
+		return -ENOMEM;
+	}
 	skb_queue_head_init(&txq->tx_skb);
 
 	return 0;
@@ -1891,6 +2075,10 @@ static void txq_deinit(struct tx_queue *txq)
 	else
 		dma_free_coherent(mp->dev->dev.parent, txq->tx_desc_area_size,
 				  txq->tx_desc_area, txq->tx_desc_dma);
+	if (txq->tso_hdrs)
+		dma_free_coherent(mp->dev->dev.parent,
+				  txq->tx_ring_size * TSO_HEADER_SIZE,
+				  txq->tso_hdrs, txq->tso_hdrs_dma);
 }
 
 
@@ -2303,7 +2491,7 @@ static int mv643xx_eth_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
 
 	ret = phy_mii_ioctl(mp->phy, ifr, cmd);
 	if (!ret)
-		mv643xx_adjust_pscr(mp);
+		mv643xx_eth_adjust_link(dev);
 	return ret;
 }
 
@@ -2678,6 +2866,7 @@ static void set_params(struct mv643xx_eth_private *mp,
 		       struct mv643xx_eth_platform_data *pd)
 {
 	struct net_device *dev = mp->dev;
+	unsigned int tx_ring_size;
 
 	if (is_valid_ether_addr(pd->mac_addr))
 		memcpy(dev->dev_addr, pd->mac_addr, ETH_ALEN);
@@ -2692,22 +2881,22 @@ static void set_params(struct mv643xx_eth_private *mp,
 
 	mp->rxq_count = pd->rx_queue_count ? : 1;
 
-	mp->tx_ring_size = DEFAULT_TX_QUEUE_SIZE;
+	tx_ring_size = DEFAULT_TX_QUEUE_SIZE;
 	if (pd->tx_queue_size)
-		mp->tx_ring_size = pd->tx_queue_size;
+		tx_ring_size = pd->tx_queue_size;
+
+	mp->tx_ring_size = clamp_t(unsigned int, tx_ring_size,
+				   MV643XX_MAX_SKB_DESCS * 2, 4096);
+	if (mp->tx_ring_size != tx_ring_size)
+		netdev_warn(dev, "TX queue size set to %u (requested %u)\n",
+			    mp->tx_ring_size, tx_ring_size);
+
 	mp->tx_desc_sram_addr = pd->tx_sram_addr;
 	mp->tx_desc_sram_size = pd->tx_sram_size;
 
 	mp->txq_count = pd->tx_queue_count ? : 1;
 }
 
-static void mv643xx_eth_adjust_link(struct net_device *dev)
-{
-	struct mv643xx_eth_private *mp = netdev_priv(dev);
-
-	mv643xx_adjust_pscr(mp);
-}
-
 static struct phy_device *phy_scan(struct mv643xx_eth_private *mp,
 				   int phy_addr)
 {
@@ -2889,7 +3078,7 @@ static int mv643xx_eth_probe(struct platform_device *pdev)
 	if (err)
 		goto out;
 
-	SET_ETHTOOL_OPS(dev, &mv643xx_eth_ethtool_ops);
+	dev->ethtool_ops = &mv643xx_eth_ethtool_ops;
 
 	init_pscr(mp, pd->speed, pd->duplex);
 
@@ -2921,11 +3110,14 @@ static int mv643xx_eth_probe(struct platform_device *pdev)
 	dev->watchdog_timeo = 2 * HZ;
 	dev->base_addr = 0;
 
-	dev->hw_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_RXCSUM;
-	dev->features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_RXCSUM;
-	dev->vlan_features = NETIF_F_SG | NETIF_F_IP_CSUM;
+	dev->features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO;
+	dev->vlan_features = dev->features;
+
+	dev->features |= NETIF_F_RXCSUM;
+	dev->hw_features = dev->features;
 
 	dev->priv_flags |= IFF_UNICAST_FLT;
+	dev->gso_max_segs = MV643XX_MAX_TSO_SEGS;
 
 	SET_NETDEV_DEV(dev, &pdev->dev);
 
diff --git a/drivers/net/ethernet/marvell/mvmdio.c b/drivers/net/ethernet/marvell/mvmdio.c
index 9d5ced263a5e..fc2fb25343f4 100644
--- a/drivers/net/ethernet/marvell/mvmdio.c
+++ b/drivers/net/ethernet/marvell/mvmdio.c
@@ -195,11 +195,10 @@ static int orion_mdio_probe(struct platform_device *pdev)
 		return -ENODEV;
 	}
 
-	bus = mdiobus_alloc_size(sizeof(struct orion_mdio_dev));
-	if (!bus) {
-		dev_err(&pdev->dev, "Cannot allocate MDIO bus\n");
+	bus = devm_mdiobus_alloc_size(&pdev->dev,
+				      sizeof(struct orion_mdio_dev));
+	if (!bus)
 		return -ENOMEM;
-	}
 
 	bus->name = "orion_mdio_bus";
 	bus->read = orion_mdio_read;
@@ -208,11 +207,10 @@ static int orion_mdio_probe(struct platform_device *pdev)
 		 dev_name(&pdev->dev));
 	bus->parent = &pdev->dev;
 
-	bus->irq = kmalloc(sizeof(int) * PHY_MAX_ADDR, GFP_KERNEL);
-	if (!bus->irq) {
-		mdiobus_free(bus);
+	bus->irq = devm_kmalloc_array(&pdev->dev, PHY_MAX_ADDR, sizeof(int),
+				      GFP_KERNEL);
+	if (!bus->irq)
 		return -ENOMEM;
-	}
 
 	for (i = 0; i < PHY_MAX_ADDR; i++)
 		bus->irq[i] = PHY_POLL;
@@ -264,8 +262,6 @@ static int orion_mdio_probe(struct platform_device *pdev)
 out_mdio:
 	if (!IS_ERR(dev->clk))
 		clk_disable_unprepare(dev->clk);
-	kfree(bus->irq);
-	mdiobus_free(bus);
 	return ret;
 }
 
@@ -276,8 +272,6 @@ static int orion_mdio_remove(struct platform_device *pdev)
 
 	writel(0, dev->regs + MVMDIO_ERR_INT_MASK);
 	mdiobus_unregister(bus);
-	kfree(bus->irq);
-	mdiobus_free(bus);
 	if (!IS_ERR(dev->clk))
 		clk_disable_unprepare(dev->clk);
 
diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c
index 14786c8bf99e..45beca17fa50 100644
--- a/drivers/net/ethernet/marvell/mvneta.c
+++ b/drivers/net/ethernet/marvell/mvneta.c
@@ -23,6 +23,7 @@
 #include <net/ip.h>
 #include <net/ipv6.h>
 #include <linux/io.h>
+#include <net/tso.h>
 #include <linux/of.h>
 #include <linux/of_irq.h>
 #include <linux/of_mdio.h>
@@ -218,9 +219,6 @@
 #define MVNETA_RX_COAL_PKTS		32
 #define MVNETA_RX_COAL_USEC		100
 
-/* Napi polling weight */
-#define MVNETA_RX_POLL_WEIGHT		64
-
 /* The two bytes Marvell header. Either contains a special value used
  * by Marvell switches when a specific hardware mode is enabled (not
  * supported by this driver) or is filled automatically by zeroes on
@@ -244,12 +242,20 @@
 
 #define MVNETA_TX_MTU_MAX		0x3ffff
 
+/* TSO header size */
+#define TSO_HEADER_SIZE 128
+
 /* Max number of Rx descriptors */
 #define MVNETA_MAX_RXD 128
 
 /* Max number of Tx descriptors */
 #define MVNETA_MAX_TXD 532
 
+/* Max number of allowed TCP segments for software TSO */
+#define MVNETA_MAX_TSO_SEGS 100
+
+#define MVNETA_MAX_SKB_DESCS (MVNETA_MAX_TSO_SEGS * 2 + MAX_SKB_FRAGS)
+
 /* descriptor aligned size */
 #define MVNETA_DESC_ALIGNED_SIZE	32
 
@@ -258,6 +264,10 @@
 	      ETH_HLEN + ETH_FCS_LEN,			     \
 	      MVNETA_CPU_D_CACHE_LINE_SIZE)
 
+#define IS_TSO_HEADER(txq, addr) \
+	((addr >= txq->tso_hdrs_phys) && \
+	 (addr < txq->tso_hdrs_phys + txq->size * TSO_HEADER_SIZE))
+
 #define MVNETA_RX_BUF_SIZE(pkt_size)   ((pkt_size) + NET_SKB_PAD)
 
 struct mvneta_pcpu_stats {
@@ -279,9 +289,6 @@ struct mvneta_port {
 	u32 cause_rx_tx;
 	struct napi_struct napi;
 
-	/* Napi weight */
-	int weight;
-
 	/* Core clock */
 	struct clk *clk;
 	u8 mcast_count[256];
@@ -390,6 +397,8 @@ struct mvneta_tx_queue {
 	 * descriptor ring
 	 */
 	int count;
+	int tx_stop_threshold;
+	int tx_wake_threshold;
 
 	/* Array of transmitted skb */
 	struct sk_buff **tx_skb;
@@ -413,6 +422,12 @@ struct mvneta_tx_queue {
 
 	/* Index of the next TX DMA descriptor to process */
 	int next_desc_to_proc;
+
+	/* DMA buffers for TSO headers */
+	char *tso_hdrs;
+
+	/* DMA address of TSO headers */
+	dma_addr_t tso_hdrs_phys;
 };
 
 struct mvneta_rx_queue {
@@ -441,7 +456,10 @@ struct mvneta_rx_queue {
 	int next_desc_to_proc;
 };
 
-static int rxq_number = 8;
+/* The hardware supports eight (8) rx queues, but we are only allowing
+ * the first one to be used. Therefore, let's just allocate one queue.
+ */
+static int rxq_number = 1;
 static int txq_number = 8;
 
 static int rxq_def;
@@ -1277,11 +1295,12 @@ static void mvneta_txq_bufs_free(struct mvneta_port *pp,
 
 		mvneta_txq_inc_get(txq);
 
+		if (!IS_TSO_HEADER(txq, tx_desc->buf_phys_addr))
+			dma_unmap_single(pp->dev->dev.parent,
+					 tx_desc->buf_phys_addr,
+					 tx_desc->data_size, DMA_TO_DEVICE);
 		if (!skb)
 			continue;
-
-		dma_unmap_single(pp->dev->dev.parent, tx_desc->buf_phys_addr,
-				 tx_desc->data_size, DMA_TO_DEVICE);
 		dev_kfree_skb_any(skb);
 	}
 }
@@ -1302,7 +1321,7 @@ static void mvneta_txq_done(struct mvneta_port *pp,
 	txq->count -= tx_done;
 
 	if (netif_tx_queue_stopped(nq)) {
-		if (txq->size - txq->count >= MAX_SKB_FRAGS + 1)
+		if (txq->count <= txq->tx_wake_threshold)
 			netif_tx_wake_queue(nq);
 	}
 }
@@ -1519,14 +1538,134 @@ static int mvneta_rx(struct mvneta_port *pp, int rx_todo,
 	return rx_done;
 }
 
+static inline void
+mvneta_tso_put_hdr(struct sk_buff *skb,
+		   struct mvneta_port *pp, struct mvneta_tx_queue *txq)
+{
+	struct mvneta_tx_desc *tx_desc;
+	int hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
+
+	txq->tx_skb[txq->txq_put_index] = NULL;
+	tx_desc = mvneta_txq_next_desc_get(txq);
+	tx_desc->data_size = hdr_len;
+	tx_desc->command = mvneta_skb_tx_csum(pp, skb);
+	tx_desc->command |= MVNETA_TXD_F_DESC;
+	tx_desc->buf_phys_addr = txq->tso_hdrs_phys +
+				 txq->txq_put_index * TSO_HEADER_SIZE;
+	mvneta_txq_inc_put(txq);
+}
+
+static inline int
+mvneta_tso_put_data(struct net_device *dev, struct mvneta_tx_queue *txq,
+		    struct sk_buff *skb, char *data, int size,
+		    bool last_tcp, bool is_last)
+{
+	struct mvneta_tx_desc *tx_desc;
+
+	tx_desc = mvneta_txq_next_desc_get(txq);
+	tx_desc->data_size = size;
+	tx_desc->buf_phys_addr = dma_map_single(dev->dev.parent, data,
+						size, DMA_TO_DEVICE);
+	if (unlikely(dma_mapping_error(dev->dev.parent,
+		     tx_desc->buf_phys_addr))) {
+		mvneta_txq_desc_put(txq);
+		return -ENOMEM;
+	}
+
+	tx_desc->command = 0;
+	txq->tx_skb[txq->txq_put_index] = NULL;
+
+	if (last_tcp) {
+		/* last descriptor in the TCP packet */
+		tx_desc->command = MVNETA_TXD_L_DESC;
+
+		/* last descriptor in SKB */
+		if (is_last)
+			txq->tx_skb[txq->txq_put_index] = skb;
+	}
+	mvneta_txq_inc_put(txq);
+	return 0;
+}
+
+static int mvneta_tx_tso(struct sk_buff *skb, struct net_device *dev,
+			 struct mvneta_tx_queue *txq)
+{
+	int total_len, data_left;
+	int desc_count = 0;
+	struct mvneta_port *pp = netdev_priv(dev);
+	struct tso_t tso;
+	int hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
+	int i;
+
+	/* Count needed descriptors */
+	if ((txq->count + tso_count_descs(skb)) >= txq->size)
+		return 0;
+
+	if (skb_headlen(skb) < (skb_transport_offset(skb) + tcp_hdrlen(skb))) {
+		pr_info("*** Is this even  possible???!?!?\n");
+		return 0;
+	}
+
+	/* Initialize the TSO handler, and prepare the first payload */
+	tso_start(skb, &tso);
+
+	total_len = skb->len - hdr_len;
+	while (total_len > 0) {
+		char *hdr;
+
+		data_left = min_t(int, skb_shinfo(skb)->gso_size, total_len);
+		total_len -= data_left;
+		desc_count++;
+
+		/* prepare packet headers: MAC + IP + TCP */
+		hdr = txq->tso_hdrs + txq->txq_put_index * TSO_HEADER_SIZE;
+		tso_build_hdr(skb, hdr, &tso, data_left, total_len == 0);
+
+		mvneta_tso_put_hdr(skb, pp, txq);
+
+		while (data_left > 0) {
+			int size;
+			desc_count++;
+
+			size = min_t(int, tso.size, data_left);
+
+			if (mvneta_tso_put_data(dev, txq, skb,
+						 tso.data, size,
+						 size == data_left,
+						 total_len == 0))
+				goto err_release;
+			data_left -= size;
+
+			tso_build_data(skb, &tso, size);
+		}
+	}
+
+	return desc_count;
+
+err_release:
+	/* Release all used data descriptors; header descriptors must not
+	 * be DMA-unmapped.
+	 */
+	for (i = desc_count - 1; i >= 0; i--) {
+		struct mvneta_tx_desc *tx_desc = txq->descs + i;
+		if (!IS_TSO_HEADER(txq, tx_desc->buf_phys_addr))
+			dma_unmap_single(pp->dev->dev.parent,
+					 tx_desc->buf_phys_addr,
+					 tx_desc->data_size,
+					 DMA_TO_DEVICE);
+		mvneta_txq_desc_put(txq);
+	}
+	return 0;
+}
+
 /* Handle tx fragmentation processing */
 static int mvneta_tx_frag_process(struct mvneta_port *pp, struct sk_buff *skb,
 				  struct mvneta_tx_queue *txq)
 {
 	struct mvneta_tx_desc *tx_desc;
-	int i;
+	int i, nr_frags = skb_shinfo(skb)->nr_frags;
 
-	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
+	for (i = 0; i < nr_frags; i++) {
 		skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
 		void *addr = page_address(frag->page.p) + frag->page_offset;
 
@@ -1543,20 +1682,16 @@ static int mvneta_tx_frag_process(struct mvneta_port *pp, struct sk_buff *skb,
 			goto error;
 		}
 
-		if (i == (skb_shinfo(skb)->nr_frags - 1)) {
+		if (i == nr_frags - 1) {
 			/* Last descriptor */
 			tx_desc->command = MVNETA_TXD_L_DESC | MVNETA_TXD_Z_PAD;
-
 			txq->tx_skb[txq->txq_put_index] = skb;
-
-			mvneta_txq_inc_put(txq);
 		} else {
 			/* Descriptor in the middle: Not First, Not Last */
 			tx_desc->command = 0;
-
 			txq->tx_skb[txq->txq_put_index] = NULL;
-			mvneta_txq_inc_put(txq);
 		}
+		mvneta_txq_inc_put(txq);
 	}
 
 	return 0;
@@ -1584,15 +1719,18 @@ static int mvneta_tx(struct sk_buff *skb, struct net_device *dev)
 	u16 txq_id = skb_get_queue_mapping(skb);
 	struct mvneta_tx_queue *txq = &pp->txqs[txq_id];
 	struct mvneta_tx_desc *tx_desc;
-	struct netdev_queue *nq;
 	int frags = 0;
 	u32 tx_cmd;
 
 	if (!netif_running(dev))
 		goto out;
 
+	if (skb_is_gso(skb)) {
+		frags = mvneta_tx_tso(skb, dev, txq);
+		goto out;
+	}
+
 	frags = skb_shinfo(skb)->nr_frags + 1;
-	nq    = netdev_get_tx_queue(dev, txq_id);
 
 	/* Get a descriptor for the first part of the packet */
 	tx_desc = mvneta_txq_next_desc_get(txq);
@@ -1635,15 +1773,16 @@ static int mvneta_tx(struct sk_buff *skb, struct net_device *dev)
 		}
 	}
 
-	txq->count += frags;
-	mvneta_txq_pend_desc_add(pp, txq, frags);
-
-	if (txq->size - txq->count < MAX_SKB_FRAGS + 1)
-		netif_tx_stop_queue(nq);
-
 out:
 	if (frags > 0) {
 		struct mvneta_pcpu_stats *stats = this_cpu_ptr(pp->stats);
+		struct netdev_queue *nq = netdev_get_tx_queue(dev, txq_id);
+
+		txq->count += frags;
+		mvneta_txq_pend_desc_add(pp, txq, frags);
+
+		if (txq->count >= txq->tx_stop_threshold)
+			netif_tx_stop_queue(nq);
 
 		u64_stats_update_begin(&stats->syncp);
 		stats->tx_packets++;
@@ -2003,7 +2142,7 @@ static void mvneta_tx_reset(struct mvneta_port *pp)
 {
 	int queue;
 
-	/* free the skb's in the hal tx ring */
+	/* free the skb's in the tx ring */
 	for (queue = 0; queue < txq_number; queue++)
 		mvneta_txq_done_force(pp, &pp->txqs[queue]);
 
@@ -2081,6 +2220,14 @@ static int mvneta_txq_init(struct mvneta_port *pp,
 {
 	txq->size = pp->tx_ring_size;
 
+	/* A queue must always have room for at least one skb.
+	 * Therefore, stop the queue when the free entries reaches
+	 * the maximum number of descriptors per skb.
+	 */
+	txq->tx_stop_threshold = txq->size - MVNETA_MAX_SKB_DESCS;
+	txq->tx_wake_threshold = txq->tx_stop_threshold / 2;
+
+
 	/* Allocate memory for TX descriptors */
 	txq->descs = dma_alloc_coherent(pp->dev->dev.parent,
 					txq->size * MVNETA_DESC_ALIGNED_SIZE,
@@ -2109,6 +2256,18 @@ static int mvneta_txq_init(struct mvneta_port *pp,
 				  txq->descs, txq->descs_phys);
 		return -ENOMEM;
 	}
+
+	/* Allocate DMA buffers for TSO MAC/IP/TCP headers */
+	txq->tso_hdrs = dma_alloc_coherent(pp->dev->dev.parent,
+					   txq->size * TSO_HEADER_SIZE,
+					   &txq->tso_hdrs_phys, GFP_KERNEL);
+	if (txq->tso_hdrs == NULL) {
+		kfree(txq->tx_skb);
+		dma_free_coherent(pp->dev->dev.parent,
+				  txq->size * MVNETA_DESC_ALIGNED_SIZE,
+				  txq->descs, txq->descs_phys);
+		return -ENOMEM;
+	}
 	mvneta_tx_done_pkts_coal_set(pp, txq, txq->done_pkts_coal);
 
 	return 0;
@@ -2120,6 +2279,10 @@ static void mvneta_txq_deinit(struct mvneta_port *pp,
 {
 	kfree(txq->tx_skb);
 
+	if (txq->tso_hdrs)
+		dma_free_coherent(pp->dev->dev.parent,
+				  txq->size * TSO_HEADER_SIZE,
+				  txq->tso_hdrs, txq->tso_hdrs_phys);
 	if (txq->descs)
 		dma_free_coherent(pp->dev->dev.parent,
 				  txq->size * MVNETA_DESC_ALIGNED_SIZE,
@@ -2279,24 +2442,28 @@ static int mvneta_change_mtu(struct net_device *dev, int mtu)
 		return 0;
 
 	/* The interface is running, so we have to force a
-	 * reallocation of the RXQs
+	 * reallocation of the queues
 	 */
 	mvneta_stop_dev(pp);
 
 	mvneta_cleanup_txqs(pp);
 	mvneta_cleanup_rxqs(pp);
 
-	pp->pkt_size = MVNETA_RX_PKT_SIZE(pp->dev->mtu);
+	pp->pkt_size = MVNETA_RX_PKT_SIZE(dev->mtu);
 	pp->frag_size = SKB_DATA_ALIGN(MVNETA_RX_BUF_SIZE(pp->pkt_size)) +
 	                SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
 
 	ret = mvneta_setup_rxqs(pp);
 	if (ret) {
-		netdev_err(pp->dev, "unable to setup rxqs after MTU change\n");
+		netdev_err(dev, "unable to setup rxqs after MTU change\n");
 		return ret;
 	}
 
-	mvneta_setup_txqs(pp);
+	ret = mvneta_setup_txqs(pp);
+	if (ret) {
+		netdev_err(dev, "unable to setup txqs after MTU change\n");
+		return ret;
+	}
 
 	mvneta_start_dev(pp);
 	mvneta_port_up(pp);
@@ -2323,22 +2490,19 @@ static void mvneta_get_mac_addr(struct mvneta_port *pp, unsigned char *addr)
 static int mvneta_set_mac_addr(struct net_device *dev, void *addr)
 {
 	struct mvneta_port *pp = netdev_priv(dev);
-	u8 *mac = addr + 2;
-	int i;
-
-	if (netif_running(dev))
-		return -EBUSY;
+	struct sockaddr *sockaddr = addr;
+	int ret;
 
+	ret = eth_prepare_mac_addr_change(dev, addr);
+	if (ret < 0)
+		return ret;
 	/* Remove previous address table entry */
 	mvneta_mac_addr_set(pp, dev->dev_addr, -1);
 
 	/* Set new addr in hw */
-	mvneta_mac_addr_set(pp, mac, rxq_def);
-
-	/* Set addr in the device */
-	for (i = 0; i < ETH_ALEN; i++)
-		dev->dev_addr[i] = mac[i];
+	mvneta_mac_addr_set(pp, sockaddr->sa_data, rxq_def);
 
+	eth_commit_mac_addr_change(dev, addr);
 	return 0;
 }
 
@@ -2433,8 +2597,6 @@ static int mvneta_open(struct net_device *dev)
 	struct mvneta_port *pp = netdev_priv(dev);
 	int ret;
 
-	mvneta_mac_addr_set(pp, dev->dev_addr, rxq_def);
-
 	pp->pkt_size = MVNETA_RX_PKT_SIZE(pp->dev->mtu);
 	pp->frag_size = SKB_DATA_ALIGN(MVNETA_RX_BUF_SIZE(pp->pkt_size)) +
 	                SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
@@ -2600,8 +2762,12 @@ static int mvneta_ethtool_set_ringparam(struct net_device *dev,
 		return -EINVAL;
 	pp->rx_ring_size = ring->rx_pending < MVNETA_MAX_RXD ?
 		ring->rx_pending : MVNETA_MAX_RXD;
-	pp->tx_ring_size = ring->tx_pending < MVNETA_MAX_TXD ?
-		ring->tx_pending : MVNETA_MAX_TXD;
+
+	pp->tx_ring_size = clamp_t(u16, ring->tx_pending,
+				   MVNETA_MAX_SKB_DESCS * 2, MVNETA_MAX_TXD);
+	if (pp->tx_ring_size != ring->tx_pending)
+		netdev_warn(dev, "TX queue size set to %u (requested %u)\n",
+			    pp->tx_ring_size, ring->tx_pending);
 
 	if (netif_running(dev)) {
 		mvneta_stop(dev);
@@ -2638,7 +2804,7 @@ const struct ethtool_ops mvneta_eth_tool_ops = {
 };
 
 /* Initialize hw */
-static int mvneta_init(struct mvneta_port *pp, int phy_addr)
+static int mvneta_init(struct device *dev, struct mvneta_port *pp)
 {
 	int queue;
 
@@ -2648,8 +2814,8 @@ static int mvneta_init(struct mvneta_port *pp, int phy_addr)
 	/* Set port default values */
 	mvneta_defaults_set(pp);
 
-	pp->txqs = kzalloc(txq_number * sizeof(struct mvneta_tx_queue),
-			   GFP_KERNEL);
+	pp->txqs = devm_kcalloc(dev, txq_number, sizeof(struct mvneta_tx_queue),
+				GFP_KERNEL);
 	if (!pp->txqs)
 		return -ENOMEM;
 
@@ -2661,12 +2827,10 @@ static int mvneta_init(struct mvneta_port *pp, int phy_addr)
 		txq->done_pkts_coal = MVNETA_TXDONE_COAL_PKTS;
 	}
 
-	pp->rxqs = kzalloc(rxq_number * sizeof(struct mvneta_rx_queue),
-			   GFP_KERNEL);
-	if (!pp->rxqs) {
-		kfree(pp->txqs);
+	pp->rxqs = devm_kcalloc(dev, rxq_number, sizeof(struct mvneta_rx_queue),
+				GFP_KERNEL);
+	if (!pp->rxqs)
 		return -ENOMEM;
-	}
 
 	/* Create Rx descriptor rings */
 	for (queue = 0; queue < rxq_number; queue++) {
@@ -2680,12 +2844,6 @@ static int mvneta_init(struct mvneta_port *pp, int phy_addr)
 	return 0;
 }
 
-static void mvneta_deinit(struct mvneta_port *pp)
-{
-	kfree(pp->txqs);
-	kfree(pp->rxqs);
-}
-
 /* platform glue : initialize decoding windows */
 static void mvneta_conf_mbus_windows(struct mvneta_port *pp,
 				     const struct mbus_dram_target_info *dram)
@@ -2768,7 +2926,6 @@ static int mvneta_probe(struct platform_device *pdev)
 	struct resource *res;
 	struct device_node *dn = pdev->dev.of_node;
 	struct device_node *phy_node;
-	u32 phy_addr;
 	struct mvneta_port *pp;
 	struct net_device *dev;
 	const char *dt_mac_addr;
@@ -2797,9 +2954,22 @@ static int mvneta_probe(struct platform_device *pdev)
 
 	phy_node = of_parse_phandle(dn, "phy", 0);
 	if (!phy_node) {
-		dev_err(&pdev->dev, "no associated PHY\n");
-		err = -ENODEV;
-		goto err_free_irq;
+		if (!of_phy_is_fixed_link(dn)) {
+			dev_err(&pdev->dev, "no PHY specified\n");
+			err = -ENODEV;
+			goto err_free_irq;
+		}
+
+		err = of_phy_register_fixed_link(dn);
+		if (err < 0) {
+			dev_err(&pdev->dev, "cannot register fixed PHY\n");
+			goto err_free_irq;
+		}
+
+		/* In the case of a fixed PHY, the DT node associated
+		 * to the PHY is the Ethernet MAC DT node.
+		 */
+		phy_node = dn;
 	}
 
 	phy_mode = of_get_phy_mode(dn);
@@ -2813,11 +2983,9 @@ static int mvneta_probe(struct platform_device *pdev)
 	dev->watchdog_timeo = 5 * HZ;
 	dev->netdev_ops = &mvneta_netdev_ops;
 
-	SET_ETHTOOL_OPS(dev, &mvneta_eth_tool_ops);
+	dev->ethtool_ops = &mvneta_eth_tool_ops;
 
 	pp = netdev_priv(dev);
-
-	pp->weight = MVNETA_RX_POLL_WEIGHT;
 	pp->phy_node = phy_node;
 	pp->phy_interface = phy_mode;
 
@@ -2864,33 +3032,32 @@ static int mvneta_probe(struct platform_device *pdev)
 	pp->dev = dev;
 	SET_NETDEV_DEV(dev, &pdev->dev);
 
-	err = mvneta_init(pp, phy_addr);
-	if (err < 0) {
-		dev_err(&pdev->dev, "can't init eth hal\n");
+	err = mvneta_init(&pdev->dev, pp);
+	if (err < 0)
 		goto err_free_stats;
-	}
 
 	err = mvneta_port_power_up(pp, phy_mode);
 	if (err < 0) {
 		dev_err(&pdev->dev, "can't power up port\n");
-		goto err_deinit;
+		goto err_free_stats;
 	}
 
 	dram_target_info = mv_mbus_dram_info();
 	if (dram_target_info)
 		mvneta_conf_mbus_windows(pp, dram_target_info);
 
-	netif_napi_add(dev, &pp->napi, mvneta_poll, pp->weight);
+	netif_napi_add(dev, &pp->napi, mvneta_poll, NAPI_POLL_WEIGHT);
 
-	dev->features = NETIF_F_SG | NETIF_F_IP_CSUM;
-	dev->hw_features |= NETIF_F_SG | NETIF_F_IP_CSUM;
-	dev->vlan_features |= NETIF_F_SG | NETIF_F_IP_CSUM;
+	dev->features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO;
+	dev->hw_features |= dev->features;
+	dev->vlan_features |= dev->features;
 	dev->priv_flags |= IFF_UNICAST_FLT;
+	dev->gso_max_segs = MVNETA_MAX_TSO_SEGS;
 
 	err = register_netdev(dev);
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to register\n");
-		goto err_deinit;
+		goto err_free_stats;
 	}
 
 	netdev_info(dev, "Using %s mac address %pM\n", mac_from,
@@ -2900,8 +3067,6 @@ static int mvneta_probe(struct platform_device *pdev)
 
 	return 0;
 
-err_deinit:
-	mvneta_deinit(pp);
 err_free_stats:
 	free_percpu(pp->stats);
 err_clk:
@@ -2920,7 +3085,6 @@ static int mvneta_remove(struct platform_device *pdev)
 	struct mvneta_port *pp = netdev_priv(dev);
 
 	unregister_netdev(dev);
-	mvneta_deinit(pp);
 	clk_disable_unprepare(pp->clk);
 	free_percpu(pp->stats);
 	irq_dispose_mapping(dev->irq);
diff --git a/drivers/net/ethernet/marvell/pxa168_eth.c b/drivers/net/ethernet/marvell/pxa168_eth.c
index b358c2f6f4bd..8f5aa7c62b18 100644
--- a/drivers/net/ethernet/marvell/pxa168_eth.c
+++ b/drivers/net/ethernet/marvell/pxa168_eth.c
@@ -1488,7 +1488,7 @@ static int pxa168_eth_probe(struct platform_device *pdev)
 	dev->netdev_ops = &pxa168_eth_netdev_ops;
 	dev->watchdog_timeo = 2 * HZ;
 	dev->base_addr = 0;
-	SET_ETHTOOL_OPS(dev, &pxa168_ethtool_ops);
+	dev->ethtool_ops = &pxa168_ethtool_ops;
 
 	INIT_WORK(&pep->tx_timeout_task, pxa168_eth_tx_timeout_task);
 
diff --git a/drivers/net/ethernet/marvell/sky2.c b/drivers/net/ethernet/marvell/sky2.c
index b81106451a0a..69693384b58c 100644
--- a/drivers/net/ethernet/marvell/sky2.c
+++ b/drivers/net/ethernet/marvell/sky2.c
@@ -4760,7 +4760,7 @@ static struct net_device *sky2_init_netdev(struct sky2_hw *hw, unsigned port,
 
 	SET_NETDEV_DEV(dev, &hw->pdev->dev);
 	dev->irq = hw->pdev->irq;
-	SET_ETHTOOL_OPS(dev, &sky2_ethtool_ops);
+	dev->ethtool_ops = &sky2_ethtool_ops;
 	dev->watchdog_timeo = TX_WATCHDOG;
 	dev->netdev_ops = &sky2_netdev_ops[port];