Merge branch 'linus' into idle-test

22 files changed, 822 insertions, 732 deletions

diff --git a/drivers/net/sfc/efx.c b/drivers/net/sfc/efx.c
index fb83cdd9464..711449c6e67 100644
--- a/drivers/net/sfc/efx.c
+++ b/drivers/net/sfc/efx.c
@@ -23,7 +23,6 @@
 #include <linux/gfp.h>
 #include "net_driver.h"
 #include "efx.h"
-#include "mdio_10g.h"
 #include "nic.h"
 
 #include "mcdi.h"
@@ -462,9 +461,6 @@ efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel)
 		}
 	}
 
-	spin_lock_init(&channel->tx_stop_lock);
-	atomic_set(&channel->tx_stop_count, 1);
-
 	rx_queue = &channel->rx_queue;
 	rx_queue->efx = efx;
 	setup_timer(&rx_queue->slow_fill, efx_rx_slow_fill,
@@ -921,6 +917,7 @@ static void efx_mac_work(struct work_struct *data)
 
 static int efx_probe_port(struct efx_nic *efx)
 {
+	unsigned char *perm_addr;
 	int rc;
 
 	netif_dbg(efx, probe, efx->net_dev, "create port\n");
@@ -934,11 +931,12 @@ static int efx_probe_port(struct efx_nic *efx)
 		return rc;
 
 	/* Sanity check MAC address */
-	if (is_valid_ether_addr(efx->mac_address)) {
-		memcpy(efx->net_dev->dev_addr, efx->mac_address, ETH_ALEN);
+	perm_addr = efx->net_dev->perm_addr;
+	if (is_valid_ether_addr(perm_addr)) {
+		memcpy(efx->net_dev->dev_addr, perm_addr, ETH_ALEN);
 	} else {
 		netif_err(efx, probe, efx->net_dev, "invalid MAC address %pM\n",
-			  efx->mac_address);
+			  perm_addr);
 		if (!allow_bad_hwaddr) {
 			rc = -EINVAL;
 			goto err;
@@ -1405,11 +1403,11 @@ static void efx_start_all(struct efx_nic *efx)
 	 * restart the transmit interface early so the watchdog timer stops */
 	efx_start_port(efx);
 
-	efx_for_each_channel(channel, efx) {
-		if (efx_dev_registered(efx))
-			efx_wake_queue(channel);
+	if (efx_dev_registered(efx))
+		netif_tx_wake_all_queues(efx->net_dev);
+
+	efx_for_each_channel(channel, efx)
 		efx_start_channel(channel);
-	}
 
 	if (efx->legacy_irq)
 		efx->legacy_irq_enabled = true;
@@ -1497,9 +1495,7 @@ static void efx_stop_all(struct efx_nic *efx)
 	/* Stop the kernel transmit interface late, so the watchdog
 	 * timer isn't ticking over the flush */
 	if (efx_dev_registered(efx)) {
-		struct efx_channel *channel;
-		efx_for_each_channel(channel, efx)
-			efx_stop_queue(channel);
+		netif_tx_stop_all_queues(efx->net_dev);
 		netif_tx_lock_bh(efx->net_dev);
 		netif_tx_unlock_bh(efx->net_dev);
 	}
@@ -1895,6 +1891,7 @@ static DEVICE_ATTR(phy_type, 0644, show_phy_type, NULL);
 static int efx_register_netdev(struct efx_nic *efx)
 {
 	struct net_device *net_dev = efx->net_dev;
+	struct efx_channel *channel;
 	int rc;
 
 	net_dev->watchdog_timeo = 5 * HZ;
@@ -1917,6 +1914,14 @@ static int efx_register_netdev(struct efx_nic *efx)
 	if (rc)
 		goto fail_locked;
 
+	efx_for_each_channel(channel, efx) {
+		struct efx_tx_queue *tx_queue;
+		efx_for_each_channel_tx_queue(tx_queue, channel) {
+			tx_queue->core_txq = netdev_get_tx_queue(
+				efx->net_dev, tx_queue->queue / EFX_TXQ_TYPES);
+		}
+	}
+
 	/* Always start with carrier off; PHY events will detect the link */
 	netif_carrier_off(efx->net_dev);
 
@@ -1980,7 +1985,6 @@ void efx_reset_down(struct efx_nic *efx, enum reset_type method)
 
 	efx_stop_all(efx);
 	mutex_lock(&efx->mac_lock);
-	mutex_lock(&efx->spi_lock);
 
 	efx_fini_channels(efx);
 	if (efx->port_initialized && method != RESET_TYPE_INVISIBLE)
@@ -2022,7 +2026,6 @@ int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok)
 	efx_init_channels(efx);
 	efx_restore_filters(efx);
 
-	mutex_unlock(&efx->spi_lock);
 	mutex_unlock(&efx->mac_lock);
 
 	efx_start_all(efx);
@@ -2032,7 +2035,6 @@ int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok)
 fail:
 	efx->port_initialized = false;
 
-	mutex_unlock(&efx->spi_lock);
 	mutex_unlock(&efx->mac_lock);
 
 	return rc;
@@ -2220,8 +2222,6 @@ static int efx_init_struct(struct efx_nic *efx, struct efx_nic_type *type,
 	/* Initialise common structures */
 	memset(efx, 0, sizeof(*efx));
 	spin_lock_init(&efx->biu_lock);
-	mutex_init(&efx->mdio_lock);
-	mutex_init(&efx->spi_lock);
 #ifdef CONFIG_SFC_MTD
 	INIT_LIST_HEAD(&efx->mtd_list);
 #endif
diff --git a/drivers/net/sfc/efx.h b/drivers/net/sfc/efx.h
index 10a1bf40da9..d43a7e5212b 100644
--- a/drivers/net/sfc/efx.h
+++ b/drivers/net/sfc/efx.h
@@ -36,8 +36,6 @@ efx_hard_start_xmit(struct sk_buff *skb, struct net_device *net_dev);
 extern netdev_tx_t
 efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb);
 extern void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index);
-extern void efx_stop_queue(struct efx_channel *channel);
-extern void efx_wake_queue(struct efx_channel *channel);
 
 /* RX */
 extern int efx_probe_rx_queue(struct efx_rx_queue *rx_queue);
@@ -74,9 +72,8 @@ extern int efx_filter_insert_filter(struct efx_nic *efx,
 				    bool replace);
 extern int efx_filter_remove_filter(struct efx_nic *efx,
 				    struct efx_filter_spec *spec);
-extern void efx_filter_table_clear(struct efx_nic *efx,
-				   enum efx_filter_table_id table_id,
-				   enum efx_filter_priority priority);
+extern void efx_filter_clear_rx(struct efx_nic *efx,
+				enum efx_filter_priority priority);
 
 /* Channels */
 extern void efx_process_channel_now(struct efx_channel *channel);
diff --git a/drivers/net/sfc/ethtool.c b/drivers/net/sfc/ethtool.c
index edb9d16b8b4..0e8bb19ed60 100644
--- a/drivers/net/sfc/ethtool.c
+++ b/drivers/net/sfc/ethtool.c
@@ -11,14 +11,13 @@
 #include <linux/netdevice.h>
 #include <linux/ethtool.h>
 #include <linux/rtnetlink.h>
+#include <linux/in.h>
 #include "net_driver.h"
 #include "workarounds.h"
 #include "selftest.h"
 #include "efx.h"
 #include "filter.h"
 #include "nic.h"
-#include "spi.h"
-#include "mdio_10g.h"
 
 struct ethtool_string {
 	char name[ETH_GSTRING_LEN];
@@ -560,12 +559,8 @@ static int efx_ethtool_set_flags(struct net_device *net_dev, u32 data)
 	if (rc)
 		return rc;
 
-	if (!(data & ETH_FLAG_NTUPLE)) {
-		efx_filter_table_clear(efx, EFX_FILTER_TABLE_RX_IP,
-				       EFX_FILTER_PRI_MANUAL);
-		efx_filter_table_clear(efx, EFX_FILTER_TABLE_RX_MAC,
-				       EFX_FILTER_PRI_MANUAL);
-	}
+	if (!(data & ETH_FLAG_NTUPLE))
+		efx_filter_clear_rx(efx, EFX_FILTER_PRI_MANUAL);
 
 	return 0;
 }
@@ -584,6 +579,9 @@ static void efx_ethtool_self_test(struct net_device *net_dev,
 		goto fail1;
 	}
 
+	netif_info(efx, drv, efx->net_dev, "starting %sline testing\n",
+		   (test->flags & ETH_TEST_FL_OFFLINE) ? "off" : "on");
+
 	/* We need rx buffers and interrupts. */
 	already_up = (efx->net_dev->flags & IFF_UP);
 	if (!already_up) {
@@ -602,9 +600,9 @@ static void efx_ethtool_self_test(struct net_device *net_dev,
 	if (!already_up)
 		dev_close(efx->net_dev);
 
-	netif_dbg(efx, drv, efx->net_dev, "%s %sline self-tests\n",
-		  rc == 0 ? "passed" : "failed",
-		  (test->flags & ETH_TEST_FL_OFFLINE) ? "off" : "on");
+	netif_info(efx, drv, efx->net_dev, "%s %sline self-tests\n",
+		   rc == 0 ? "passed" : "failed",
+		   (test->flags & ETH_TEST_FL_OFFLINE) ? "off" : "on");
 
  fail2:
  fail1:
@@ -622,68 +620,6 @@ static int efx_ethtool_nway_reset(struct net_device *net_dev)
 	return mdio45_nway_restart(&efx->mdio);
 }
 
-static u32 efx_ethtool_get_link(struct net_device *net_dev)
-{
-	struct efx_nic *efx = netdev_priv(net_dev);
-
-	return efx->link_state.up;
-}
-
-static int efx_ethtool_get_eeprom_len(struct net_device *net_dev)
-{
-	struct efx_nic *efx = netdev_priv(net_dev);
-	struct efx_spi_device *spi = efx->spi_eeprom;
-
-	if (!spi)
-		return 0;
-	return min(spi->size, EFX_EEPROM_BOOTCONFIG_END) -
-		min(spi->size, EFX_EEPROM_BOOTCONFIG_START);
-}
-
-static int efx_ethtool_get_eeprom(struct net_device *net_dev,
-				  struct ethtool_eeprom *eeprom, u8 *buf)
-{
-	struct efx_nic *efx = netdev_priv(net_dev);
-	struct efx_spi_device *spi = efx->spi_eeprom;
-	size_t len;
-	int rc;
-
-	rc = mutex_lock_interruptible(&efx->spi_lock);
-	if (rc)
-		return rc;
-	rc = falcon_spi_read(efx, spi,
-			     eeprom->offset + EFX_EEPROM_BOOTCONFIG_START,
-			     eeprom->len, &len, buf);
-	mutex_unlock(&efx->spi_lock);
-
-	eeprom->magic = EFX_ETHTOOL_EEPROM_MAGIC;
-	eeprom->len = len;
-	return rc;
-}
-
-static int efx_ethtool_set_eeprom(struct net_device *net_dev,
-				  struct ethtool_eeprom *eeprom, u8 *buf)
-{
-	struct efx_nic *efx = netdev_priv(net_dev);
-	struct efx_spi_device *spi = efx->spi_eeprom;
-	size_t len;
-	int rc;
-
-	if (eeprom->magic != EFX_ETHTOOL_EEPROM_MAGIC)
-		return -EINVAL;
-
-	rc = mutex_lock_interruptible(&efx->spi_lock);
-	if (rc)
-		return rc;
-	rc = falcon_spi_write(efx, spi,
-			      eeprom->offset + EFX_EEPROM_BOOTCONFIG_START,
-			      eeprom->len, &len, buf);
-	mutex_unlock(&efx->spi_lock);
-
-	eeprom->len = len;
-	return rc;
-}
-
 static int efx_ethtool_get_coalesce(struct net_device *net_dev,
 				    struct ethtool_coalesce *coalesce)
 {
@@ -978,6 +914,7 @@ static int efx_ethtool_set_rx_ntuple(struct net_device *net_dev,
 	struct ethhdr *mac_entry = &ntuple->fs.h_u.ether_spec;
 	struct ethhdr *mac_mask = &ntuple->fs.m_u.ether_spec;
 	struct efx_filter_spec filter;
+	int rc;
 
 	/* Range-check action */
 	if (ntuple->fs.action < ETHTOOL_RXNTUPLE_ACTION_CLEAR ||
@@ -987,9 +924,16 @@ static int efx_ethtool_set_rx_ntuple(struct net_device *net_dev,
 	if (~ntuple->fs.data_mask)
 		return -EINVAL;
 
+	efx_filter_init_rx(&filter, EFX_FILTER_PRI_MANUAL, 0,
+			   (ntuple->fs.action == ETHTOOL_RXNTUPLE_ACTION_DROP) ?
+			   0xfff : ntuple->fs.action);
+
 	switch (ntuple->fs.flow_type) {
 	case TCP_V4_FLOW:
-	case UDP_V4_FLOW:
+	case UDP_V4_FLOW: {
+		u8 proto = (ntuple->fs.flow_type == TCP_V4_FLOW ?
+			    IPPROTO_TCP : IPPROTO_UDP);
+
 		/* Must match all of destination, */
 		if (ip_mask->ip4dst | ip_mask->pdst)
 			return -EINVAL;
@@ -1001,7 +945,22 @@ static int efx_ethtool_set_rx_ntuple(struct net_device *net_dev,
 		/* and nothing else */
 		if ((u8)~ip_mask->tos | (u16)~ntuple->fs.vlan_tag_mask)
 			return -EINVAL;
+
+		if (!ip_mask->ip4src)
+			rc = efx_filter_set_ipv4_full(&filter, proto,
+						      ip_entry->ip4dst,
+						      ip_entry->pdst,
+						      ip_entry->ip4src,
+						      ip_entry->psrc);
+		else
+			rc = efx_filter_set_ipv4_local(&filter, proto,
+						       ip_entry->ip4dst,
+						       ip_entry->pdst);
+		if (rc)
+			return rc;
 		break;
+	}
+
 	case ETHER_FLOW:
 		/* Must match all of destination, */
 		if (!is_zero_ether_addr(mac_mask->h_dest))
@@ -1014,58 +973,24 @@ static int efx_ethtool_set_rx_ntuple(struct net_device *net_dev,
 		if (!is_broadcast_ether_addr(mac_mask->h_source) ||
 		    mac_mask->h_proto != htons(0xffff))
 			return -EINVAL;
+
+		rc = efx_filter_set_eth_local(
+			&filter,
+			(ntuple->fs.vlan_tag_mask == 0xf000) ?
+			ntuple->fs.vlan_tag : EFX_FILTER_VID_UNSPEC,
+			mac_entry->h_dest);
+		if (rc)
+			return rc;
 		break;
+
 	default:
 		return -EINVAL;
 	}
 
-	filter.priority = EFX_FILTER_PRI_MANUAL;
-	filter.flags = 0;
-
-	switch (ntuple->fs.flow_type) {
-	case TCP_V4_FLOW:
-		if (!ip_mask->ip4src)
-			efx_filter_set_rx_tcp_full(&filter,
-						   htonl(ip_entry->ip4src),
-						   htons(ip_entry->psrc),
-						   htonl(ip_entry->ip4dst),
-						   htons(ip_entry->pdst));
-		else
-			efx_filter_set_rx_tcp_wild(&filter,
-						   htonl(ip_entry->ip4dst),
-						   htons(ip_entry->pdst));
-		break;
-	case UDP_V4_FLOW:
-		if (!ip_mask->ip4src)
-			efx_filter_set_rx_udp_full(&filter,
-						   htonl(ip_entry->ip4src),
-						   htons(ip_entry->psrc),
-						   htonl(ip_entry->ip4dst),
-						   htons(ip_entry->pdst));
-		else
-			efx_filter_set_rx_udp_wild(&filter,
-						   htonl(ip_entry->ip4dst),
-						   htons(ip_entry->pdst));
-		break;
-	case ETHER_FLOW:
-		if (ntuple->fs.vlan_tag_mask == 0xf000)
-			efx_filter_set_rx_mac_full(&filter,
-						   ntuple->fs.vlan_tag & 0xfff,
-						   mac_entry->h_dest);
-		else
-			efx_filter_set_rx_mac_wild(&filter, mac_entry->h_dest);
-		break;
-	}
-
-	if (ntuple->fs.action == ETHTOOL_RXNTUPLE_ACTION_CLEAR) {
+	if (ntuple->fs.action == ETHTOOL_RXNTUPLE_ACTION_CLEAR)
 		return efx_filter_remove_filter(efx, &filter);
-	} else {
-		if (ntuple->fs.action == ETHTOOL_RXNTUPLE_ACTION_DROP)
-			filter.dmaq_id = 0xfff;
-		else
-			filter.dmaq_id = ntuple->fs.action;
+	else
 		return efx_filter_insert_filter(efx, &filter, true);
-	}
 }
 
 static int efx_ethtool_get_rxfh_indir(struct net_device *net_dev,
@@ -1115,10 +1040,7 @@ const struct ethtool_ops efx_ethtool_ops = {
 	.get_msglevel		= efx_ethtool_get_msglevel,
 	.set_msglevel		= efx_ethtool_set_msglevel,
 	.nway_reset		= efx_ethtool_nway_reset,
-	.get_link		= efx_ethtool_get_link,
-	.get_eeprom_len		= efx_ethtool_get_eeprom_len,
-	.get_eeprom		= efx_ethtool_get_eeprom,
-	.set_eeprom		= efx_ethtool_set_eeprom,
+	.get_link		= ethtool_op_get_link,
 	.get_coalesce		= efx_ethtool_get_coalesce,
 	.set_coalesce		= efx_ethtool_set_coalesce,
 	.get_ringparam		= efx_ethtool_get_ringparam,
diff --git a/drivers/net/sfc/falcon.c b/drivers/net/sfc/falcon.c
index 267019bb2b1..70e4f7dcce8 100644
--- a/drivers/net/sfc/falcon.c
+++ b/drivers/net/sfc/falcon.c
@@ -24,7 +24,6 @@
 #include "nic.h"
 #include "regs.h"
 #include "io.h"
-#include "mdio_10g.h"
 #include "phy.h"
 #include "workarounds.h"
 
@@ -255,7 +254,6 @@ int falcon_spi_cmd(struct efx_nic *efx, const struct efx_spi_device *spi,
 	/* Input validation */
 	if (len > FALCON_SPI_MAX_LEN)
 		return -EINVAL;
-	BUG_ON(!mutex_is_locked(&efx->spi_lock));
 
 	/* Check that previous command is not still running */
 	rc = falcon_spi_poll(efx);
@@ -719,6 +717,7 @@ static int falcon_mdio_write(struct net_device *net_dev,
 			     int prtad, int devad, u16 addr, u16 value)
 {
 	struct efx_nic *efx = netdev_priv(net_dev);
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	efx_oword_t reg;
 	int rc;
 
@@ -726,7 +725,7 @@ static int falcon_mdio_write(struct net_device *net_dev,
 		   "writing MDIO %d register %d.%d with 0x%04x\n",
 		    prtad, devad, addr, value);
 
-	mutex_lock(&efx->mdio_lock);
+	mutex_lock(&nic_data->mdio_lock);
 
 	/* Check MDIO not currently being accessed */
 	rc = falcon_gmii_wait(efx);
@@ -762,7 +761,7 @@ static int falcon_mdio_write(struct net_device *net_dev,
 	}
 
 out:
-	mutex_unlock(&efx->mdio_lock);
+	mutex_unlock(&nic_data->mdio_lock);
 	return rc;
 }
 
@@ -771,10 +770,11 @@ static int falcon_mdio_read(struct net_device *net_dev,
 			    int prtad, int devad, u16 addr)
 {
 	struct efx_nic *efx = netdev_priv(net_dev);
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	efx_oword_t reg;
 	int rc;
 
-	mutex_lock(&efx->mdio_lock);
+	mutex_lock(&nic_data->mdio_lock);
 
 	/* Check MDIO not currently being accessed */
 	rc = falcon_gmii_wait(efx);
@@ -813,7 +813,7 @@ static int falcon_mdio_read(struct net_device *net_dev,
 	}
 
 out:
-	mutex_unlock(&efx->mdio_lock);
+	mutex_unlock(&nic_data->mdio_lock);
 	return rc;
 }
 
@@ -841,6 +841,7 @@ static int falcon_probe_port(struct efx_nic *efx)
 	}
 
 	/* Fill out MDIO structure and loopback modes */
+	mutex_init(&nic_data->mdio_lock);
 	efx->mdio.mdio_read = falcon_mdio_read;
 	efx->mdio.mdio_write = falcon_mdio_write;
 	rc = efx->phy_op->probe(efx);
@@ -880,6 +881,41 @@ static void falcon_remove_port(struct efx_nic *efx)
 	efx_nic_free_buffer(efx, &efx->stats_buffer);
 }
 
+/* Global events are basically PHY events */
+static bool
+falcon_handle_global_event(struct efx_channel *channel, efx_qword_t *event)
+{
+	struct efx_nic *efx = channel->efx;
+	struct falcon_nic_data *nic_data = efx->nic_data;
+
+	if (EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_G_PHY0_INTR) ||
+	    EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_XG_PHY0_INTR) ||
+	    EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_XFP_PHY0_INTR))
+		/* Ignored */
+		return true;
+
+	if ((efx_nic_rev(efx) == EFX_REV_FALCON_B0) &&
+	    EFX_QWORD_FIELD(*event, FSF_BB_GLB_EV_XG_MGT_INTR)) {
+		nic_data->xmac_poll_required = true;
+		return true;
+	}
+
+	if (efx_nic_rev(efx) <= EFX_REV_FALCON_A1 ?
+	    EFX_QWORD_FIELD(*event, FSF_AA_GLB_EV_RX_RECOVERY) :
+	    EFX_QWORD_FIELD(*event, FSF_BB_GLB_EV_RX_RECOVERY)) {
+		netif_err(efx, rx_err, efx->net_dev,
+			  "channel %d seen global RX_RESET event. Resetting.\n",
+			  channel->channel);
+
+		atomic_inc(&efx->rx_reset);
+		efx_schedule_reset(efx, EFX_WORKAROUND_6555(efx) ?
+				   RESET_TYPE_RX_RECOVERY : RESET_TYPE_DISABLE);
+		return true;
+	}
+
+	return false;
+}
+
 /**************************************************************************
  *
  * Falcon test code
@@ -889,6 +925,7 @@ static void falcon_remove_port(struct efx_nic *efx)
 static int
 falcon_read_nvram(struct efx_nic *efx, struct falcon_nvconfig *nvconfig_out)
 {
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	struct falcon_nvconfig *nvconfig;
 	struct efx_spi_device *spi;
 	void *region;
@@ -896,8 +933,11 @@ falcon_read_nvram(struct efx_nic *efx, struct falcon_nvconfig *nvconfig_out)
 	__le16 *word, *limit;
 	u32 csum;
 
-	spi = efx->spi_flash ? efx->spi_flash : efx->spi_eeprom;
-	if (!spi)
+	if (efx_spi_present(&nic_data->spi_flash))
+		spi = &nic_data->spi_flash;
+	else if (efx_spi_present(&nic_data->spi_eeprom))
+		spi = &nic_data->spi_eeprom;
+	else
 		return -EINVAL;
 
 	region = kmalloc(FALCON_NVCONFIG_END, GFP_KERNEL);
@@ -905,12 +945,13 @@ falcon_read_nvram(struct efx_nic *efx, struct falcon_nvconfig *nvconfig_out)
 		return -ENOMEM;
 	nvconfig = region + FALCON_NVCONFIG_OFFSET;
 
-	mutex_lock(&efx->spi_lock);
+	mutex_lock(&nic_data->spi_lock);
 	rc = falcon_spi_read(efx, spi, 0, FALCON_NVCONFIG_END, NULL, region);
-	mutex_unlock(&efx->spi_lock);
+	mutex_unlock(&nic_data->spi_lock);
 	if (rc) {
 		netif_err(efx, hw, efx->net_dev, "Failed to read %s\n",
-			  efx->spi_flash ? "flash" : "EEPROM");
+			  efx_spi_present(&nic_data->spi_flash) ?
+			  "flash" : "EEPROM");
 		rc = -EIO;
 		goto out;
 	}
@@ -1012,7 +1053,7 @@ static int falcon_b0_test_registers(struct efx_nic *efx)
 
 /* Resets NIC to known state.  This routine must be called in process
  * context and is allowed to sleep. */
-static int falcon_reset_hw(struct efx_nic *efx, enum reset_type method)
+static int __falcon_reset_hw(struct efx_nic *efx, enum reset_type method)
 {
 	struct falcon_nic_data *nic_data = efx->nic_data;
 	efx_oword_t glb_ctl_reg_ker;
@@ -1108,6 +1149,18 @@ fail5:
 	return rc;
 }
 
+static int falcon_reset_hw(struct efx_nic *efx, enum reset_type method)
+{
+	struct falcon_nic_data *nic_data = efx->nic_data;
+	int rc;
+
+	mutex_lock(&nic_data->spi_lock);
+	rc = __falcon_reset_hw(efx, method);
+	mutex_unlock(&nic_data->spi_lock);
+
+	return rc;
+}
+
 static void falcon_monitor(struct efx_nic *efx)
 {
 	bool link_changed;
@@ -1189,16 +1242,11 @@ static int falcon_reset_sram(struct efx_nic *efx)
 	return -ETIMEDOUT;
 }
 
-static int falcon_spi_device_init(struct efx_nic *efx,
-				  struct efx_spi_device **spi_device_ret,
+static void falcon_spi_device_init(struct efx_nic *efx,
+				  struct efx_spi_device *spi_device,
 				  unsigned int device_id, u32 device_type)
 {
-	struct efx_spi_device *spi_device;
-
 	if (device_type != 0) {
-		spi_device = kzalloc(sizeof(*spi_device), GFP_KERNEL);
-		if (!spi_device)
-			return -ENOMEM;
 		spi_device->device_id = device_id;
 		spi_device->size =
 			1 << SPI_DEV_TYPE_FIELD(device_type, SPI_DEV_TYPE_SIZE);
@@ -1215,27 +1263,15 @@ static int falcon_spi_device_init(struct efx_nic *efx,
 			1 << SPI_DEV_TYPE_FIELD(device_type,
 						SPI_DEV_TYPE_BLOCK_SIZE);
 	} else {
-		spi_device = NULL;
+		spi_device->size = 0;
 	}
-
-	kfree(*spi_device_ret);
-	*spi_device_ret = spi_device;
-	return 0;
-}
-
-static void falcon_remove_spi_devices(struct efx_nic *efx)
-{
-	kfree(efx->spi_eeprom);
-	efx->spi_eeprom = NULL;
-	kfree(efx->spi_flash);
-	efx->spi_flash = NULL;
 }
 
 /* Extract non-volatile configuration */
 static int falcon_probe_nvconfig(struct efx_nic *efx)
 {
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	struct falcon_nvconfig *nvconfig;
-	int board_rev;
 	int rc;
 
 	nvconfig = kmalloc(sizeof(*nvconfig), GFP_KERNEL);
@@ -1243,55 +1279,32 @@ static int falcon_probe_nvconfig(struct efx_nic *efx)
 		return -ENOMEM;
 
 	rc = falcon_read_nvram(efx, nvconfig);
-	if (rc == -EINVAL) {
-		netif_err(efx, probe, efx->net_dev,
-			  "NVRAM is invalid therefore using defaults\n");
-		efx->phy_type = PHY_TYPE_NONE;
-		efx->mdio.prtad = MDIO_PRTAD_NONE;
-		board_rev = 0;
-		rc = 0;
-	} else if (rc) {
-		goto fail1;
-	} else {
-		struct falcon_nvconfig_board_v2 *v2 = &nvconfig->board_v2;
-		struct falcon_nvconfig_board_v3 *v3 = &nvconfig->board_v3;
-
-		efx->phy_type = v2->port0_phy_type;
-		efx->mdio.prtad = v2->port0_phy_addr;
-		board_rev = le16_to_cpu(v2->board_revision);
-
-		if (le16_to_cpu(nvconfig->board_struct_ver) >= 3) {
-			rc = falcon_spi_device_init(
-				efx, &efx->spi_flash, FFE_AB_SPI_DEVICE_FLASH,
-				le32_to_cpu(v3->spi_device_type
-					    [FFE_AB_SPI_DEVICE_FLASH]));
-			if (rc)
-				goto fail2;
-			rc = falcon_spi_device_init(
-				efx, &efx->spi_eeprom, FFE_AB_SPI_DEVICE_EEPROM,
-				le32_to_cpu(v3->spi_device_type
-					    [FFE_AB_SPI_DEVICE_EEPROM]));
-			if (rc)
-				goto fail2;
-		}
+	if (rc)
+		goto out;
+
+	efx->phy_type = nvconfig->board_v2.port0_phy_type;
+	efx->mdio.prtad = nvconfig->board_v2.port0_phy_addr;
+
+	if (le16_to_cpu(nvconfig->board_struct_ver) >= 3) {
+		falcon_spi_device_init(
+			efx, &nic_data->spi_flash, FFE_AB_SPI_DEVICE_FLASH,
+			le32_to_cpu(nvconfig->board_v3
+				    .spi_device_type[FFE_AB_SPI_DEVICE_FLASH]));
+		falcon_spi_device_init(
+			efx, &nic_data->spi_eeprom, FFE_AB_SPI_DEVICE_EEPROM,
+			le32_to_cpu(nvconfig->board_v3
+				    .spi_device_type[FFE_AB_SPI_DEVICE_EEPROM]));
 	}
 
 	/* Read the MAC addresses */
-	memcpy(efx->mac_address, nvconfig->mac_address[0], ETH_ALEN);
+	memcpy(efx->net_dev->perm_addr, nvconfig->mac_address[0], ETH_ALEN);
 
 	netif_dbg(efx, probe, efx->net_dev, "PHY is %d phy_id %d\n",
 		  efx->phy_type, efx->mdio.prtad);
 
-	rc = falcon_probe_board(efx, board_rev);
-	if (rc)
-		goto fail2;
-
-	kfree(nvconfig);
-	return 0;
-
- fail2:
-	falcon_remove_spi_devices(efx);
- fail1:
+	rc = falcon_probe_board(efx,
+				le16_to_cpu(nvconfig->board_v2.board_revision));
+out:
 	kfree(nvconfig);
 	return rc;
 }
@@ -1299,6 +1312,7 @@ static int falcon_probe_nvconfig(struct efx_nic *efx)
 /* Probe all SPI devices on the NIC */
 static void falcon_probe_spi_devices(struct efx_nic *efx)
 {
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	efx_oword_t nic_stat, gpio_ctl, ee_vpd_cfg;
 	int boot_dev;
 
@@ -1327,12 +1341,14 @@ static void falcon_probe_spi_devices(struct efx_nic *efx)
 		efx_writeo(efx, &ee_vpd_cfg, FR_AB_EE_VPD_CFG0);
 	}
 
+	mutex_init(&nic_data->spi_lock);
+
 	if (boot_dev == FFE_AB_SPI_DEVICE_FLASH)
-		falcon_spi_device_init(efx, &efx->spi_flash,
+		falcon_spi_device_init(efx, &nic_data->spi_flash,
 				       FFE_AB_SPI_DEVICE_FLASH,
 				       default_flash_type);
 	if (boot_dev == FFE_AB_SPI_DEVICE_EEPROM)
-		falcon_spi_device_init(efx, &efx->spi_eeprom,
+		falcon_spi_device_init(efx, &nic_data->spi_eeprom,
 				       FFE_AB_SPI_DEVICE_EEPROM,
 				       large_eeprom_type);
 }
@@ -1397,7 +1413,7 @@ static int falcon_probe_nic(struct efx_nic *efx)
 	}
 
 	/* Now we can reset the NIC */
-	rc = falcon_reset_hw(efx, RESET_TYPE_ALL);
+	rc = __falcon_reset_hw(efx, RESET_TYPE_ALL);
 	if (rc) {
 		netif_err(efx, probe, efx->net_dev, "failed to reset NIC\n");
 		goto fail3;
@@ -1419,8 +1435,11 @@ static int falcon_probe_nic(struct efx_nic *efx)
 
 	/* Read in the non-volatile configuration */
 	rc = falcon_probe_nvconfig(efx);
-	if (rc)
+	if (rc) {
+		if (rc == -EINVAL)
+			netif_err(efx, probe, efx->net_dev, "NVRAM is invalid\n");
 		goto fail5;
+	}
 
 	/* Initialise I2C adapter */
 	board = falcon_board(efx);
@@ -1452,7 +1471,6 @@ static int falcon_probe_nic(struct efx_nic *efx)
 	BUG_ON(i2c_del_adapter(&board->i2c_adap));
 	memset(&board->i2c_adap, 0, sizeof(board->i2c_adap));
  fail5:
-	falcon_remove_spi_devices(efx);
 	efx_nic_free_buffer(efx, &efx->irq_status);
  fail4:
  fail3:
@@ -1606,10 +1624,9 @@ static void falcon_remove_nic(struct efx_nic *efx)
 	BUG_ON(rc);
 	memset(&board->i2c_adap, 0, sizeof(board->i2c_adap));
 
-	falcon_remove_spi_devices(efx);
 	efx_nic_free_buffer(efx, &efx->irq_status);
 
-	falcon_reset_hw(efx, RESET_TYPE_ALL);
+	__falcon_reset_hw(efx, RESET_TYPE_ALL);
 
 	/* Release the second function after the reset */
 	if (nic_data->pci_dev2) {
@@ -1720,6 +1737,7 @@ struct efx_nic_type falcon_a1_nic_type = {
 	.reset = falcon_reset_hw,
 	.probe_port = falcon_probe_port,
 	.remove_port = falcon_remove_port,
+	.handle_global_event = falcon_handle_global_event,
 	.prepare_flush = falcon_prepare_flush,
 	.update_stats = falcon_update_nic_stats,
 	.start_stats = falcon_start_nic_stats,
@@ -1760,6 +1778,7 @@ struct efx_nic_type falcon_b0_nic_type = {
 	.reset = falcon_reset_hw,
 	.probe_port = falcon_probe_port,
 	.remove_port = falcon_remove_port,
+	.handle_global_event = falcon_handle_global_event,
 	.prepare_flush = falcon_prepare_flush,
 	.update_stats = falcon_update_nic_stats,
 	.start_stats = falcon_start_nic_stats,
diff --git a/drivers/net/sfc/falcon_boards.c b/drivers/net/sfc/falcon_boards.c
index cfc6a5b5a47..2dd16f0b3ce 100644
--- a/drivers/net/sfc/falcon_boards.c
+++ b/drivers/net/sfc/falcon_boards.c
@@ -13,8 +13,6 @@
 #include "phy.h"
 #include "efx.h"
 #include "nic.h"
-#include "regs.h"
-#include "io.h"
 #include "workarounds.h"
 
 /* Macros for unpacking the board revision */
@@ -30,17 +28,28 @@
 #define FALCON_BOARD_SFN4112F 0x52
 
 /* Board temperature is about 15°C above ambient when air flow is
- * limited. */
+ * limited.  The maximum acceptable ambient temperature varies
+ * depending on the PHY specifications but the critical temperature
+ * above which we should shut down to avoid damage is 80°C. */
 #define FALCON_BOARD_TEMP_BIAS	15
+#define FALCON_BOARD_TEMP_CRIT	(80 + FALCON_BOARD_TEMP_BIAS)
 
 /* SFC4000 datasheet says: 'The maximum permitted junction temperature
  * is 125°C; the thermal design of the environment for the SFC4000
  * should aim to keep this well below 100°C.' */
+#define FALCON_JUNC_TEMP_MIN	0
 #define FALCON_JUNC_TEMP_MAX	90
+#define FALCON_JUNC_TEMP_CRIT	125
 
 /*****************************************************************************
  * Support for LM87 sensor chip used on several boards
  */
+#define LM87_REG_TEMP_HW_INT_LOCK	0x13
+#define LM87_REG_TEMP_HW_EXT_LOCK	0x14
+#define LM87_REG_TEMP_HW_INT		0x17
+#define LM87_REG_TEMP_HW_EXT		0x18
+#define LM87_REG_TEMP_EXT1		0x26
+#define LM87_REG_TEMP_INT		0x27
 #define LM87_REG_ALARMS1		0x41
 #define LM87_REG_ALARMS2		0x42
 #define LM87_IN_LIMITS(nr, _min, _max)			\
@@ -57,6 +66,27 @@
 
 #if defined(CONFIG_SENSORS_LM87) || defined(CONFIG_SENSORS_LM87_MODULE)
 
+static int efx_poke_lm87(struct i2c_client *client, const u8 *reg_values)
+{
+	while (*reg_values) {
+		u8 reg = *reg_values++;
+		u8 value = *reg_values++;
+		int rc = i2c_smbus_write_byte_data(client, reg, value);
+		if (rc)
+			return rc;
+	}
+	return 0;
+}
+
+static const u8 falcon_lm87_common_regs[] = {
+	LM87_REG_TEMP_HW_INT_LOCK, FALCON_BOARD_TEMP_CRIT,
+	LM87_REG_TEMP_HW_INT, FALCON_BOARD_TEMP_CRIT,
+	LM87_TEMP_EXT1_LIMITS(FALCON_JUNC_TEMP_MIN, FALCON_JUNC_TEMP_MAX),
+	LM87_REG_TEMP_HW_EXT_LOCK, FALCON_JUNC_TEMP_CRIT,
+	LM87_REG_TEMP_HW_EXT, FALCON_JUNC_TEMP_CRIT,
+	0
+};
+
 static int efx_init_lm87(struct efx_nic *efx, struct i2c_board_info *info,
 			 const u8 *reg_values)
 {
@@ -67,13 +97,16 @@ static int efx_init_lm87(struct efx_nic *efx, struct i2c_board_info *info,
 	if (!client)
 		return -EIO;
 
-	while (*reg_values) {
-		u8 reg = *reg_values++;
-		u8 value = *reg_values++;
-		rc = i2c_smbus_write_byte_data(client, reg, value);
-		if (rc)
-			goto err;
-	}
+	/* Read-to-clear alarm/interrupt status */
+	i2c_smbus_read_byte_data(client, LM87_REG_ALARMS1);
+	i2c_smbus_read_byte_data(client, LM87_REG_ALARMS2);
+
+	rc = efx_poke_lm87(client, reg_values);
+	if (rc)
+		goto err;
+	rc = efx_poke_lm87(client, falcon_lm87_common_regs);
+	if (rc)
+		goto err;
 
 	board->hwmon_client = client;
 	return 0;
@@ -91,36 +124,56 @@ static void efx_fini_lm87(struct efx_nic *efx)
 static int efx_check_lm87(struct efx_nic *efx, unsigned mask)
 {
 	struct i2c_client *client = falcon_board(efx)->hwmon_client;
-	s32 alarms1, alarms2;
+	bool temp_crit, elec_fault, is_failure;
+	u16 alarms;
+	s32 reg;
 
 	/* If link is up then do not monitor temperature */
 	if (EFX_WORKAROUND_7884(efx) && efx->link_state.up)
 		return 0;
 
-	alarms1 = i2c_smbus_read_byte_data(client, LM87_REG_ALARMS1);
-	alarms2 = i2c_smbus_read_byte_data(client, LM87_REG_ALARMS2);
-	if (alarms1 < 0)
-		return alarms1;
-	if (alarms2 < 0)
-		return alarms2;
-	alarms1 &= mask;
-	alarms2 &= mask >> 8;
-	if (alarms1 || alarms2) {
+	reg = i2c_smbus_read_byte_data(client, LM87_REG_ALARMS1);
+	if (reg < 0)
+		return reg;
+	alarms = reg;
+	reg = i2c_smbus_read_byte_data(client, LM87_REG_ALARMS2);
+	if (reg < 0)
+		return reg;
+	alarms |= reg << 8;
+	alarms &= mask;
+
+	temp_crit = false;
+	if (alarms & LM87_ALARM_TEMP_INT) {
+		reg = i2c_smbus_read_byte_data(client, LM87_REG_TEMP_INT);
+		if (reg < 0)
+			return reg;
+		if (reg > FALCON_BOARD_TEMP_CRIT)
+			temp_crit = true;
+	}
+	if (alarms & LM87_ALARM_TEMP_EXT1) {
+		reg = i2c_smbus_read_byte_data(client, LM87_REG_TEMP_EXT1);
+		if (reg < 0)
+			return reg;
+		if (reg > FALCON_JUNC_TEMP_CRIT)
+			temp_crit = true;
+	}
+	elec_fault = alarms & ~(LM87_ALARM_TEMP_INT | LM87_ALARM_TEMP_EXT1);
+	is_failure = temp_crit || elec_fault;
+
+	if (alarms)
 		netif_err(efx, hw, efx->net_dev,
-			  "LM87 detected a hardware failure (status %02x:%02x)"
-			  "%s%s%s\n",
-			  alarms1, alarms2,
-			  (alarms1 & LM87_ALARM_TEMP_INT) ?
+			  "LM87 detected a hardware %s (status %02x:%02x)"
+			  "%s%s%s%s\n",
+			  is_failure ? "failure" : "problem",
+			  alarms & 0xff, alarms >> 8,
+			  (alarms & LM87_ALARM_TEMP_INT) ?
 			  "; board is overheating" : "",
-			  (alarms1 & LM87_ALARM_TEMP_EXT1) ?
+			  (alarms & LM87_ALARM_TEMP_EXT1) ?
 			  "; controller is overheating" : "",
-			  (alarms1 & ~(LM87_ALARM_TEMP_INT | LM87_ALARM_TEMP_EXT1)
-			   || alarms2) ?
-			  "; electrical fault" : "");
-		return -ERANGE;
-	}
+			  temp_crit ? "; reached critical temperature" : "",
+			  elec_fault ? "; electrical fault" : "");
 
-	return 0;
+	return is_failure ? -ERANGE : 0;
 }
 
 #else /* !CONFIG_SENSORS_LM87 */
@@ -325,7 +378,7 @@ static ssize_t set_phy_flash_cfg(struct device *dev,
 		new_mode = old_mode & ~PHY_MODE_SPECIAL;
 	else
 		new_mode = PHY_MODE_SPECIAL;
-	if (old_mode == new_mode) {
+	if (!((old_mode ^ new_mode) & PHY_MODE_SPECIAL)) {
 		err = 0;
 	} else if (efx->state != STATE_RUNNING || netif_running(efx->net_dev)) {
 		err = -EBUSY;
@@ -362,10 +415,11 @@ static void sfe4001_fini(struct efx_nic *efx)
 
 static int sfe4001_check_hw(struct efx_nic *efx)
 {
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	s32 status;
 
 	/* If XAUI link is up then do not monitor */
-	if (EFX_WORKAROUND_7884(efx) && !efx->xmac_poll_required)
+	if (EFX_WORKAROUND_7884(efx) && !nic_data->xmac_poll_required)
 		return 0;
 
 	/* Check the powered status of the PHY. Lack of power implies that
diff --git a/drivers/net/sfc/falcon_xmac.c b/drivers/net/sfc/falcon_xmac.c
index b31f595ebb5..b49e8439464 100644
--- a/drivers/net/sfc/falcon_xmac.c
+++ b/drivers/net/sfc/falcon_xmac.c
@@ -16,7 +16,6 @@
 #include "io.h"
 #include "mac.h"
 #include "mdio_10g.h"
-#include "phy.h"
 #include "workarounds.h"
 
 /**************************************************************************
@@ -88,6 +87,7 @@ int falcon_reset_xaui(struct efx_nic *efx)
 
 static void falcon_ack_status_intr(struct efx_nic *efx)
 {
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	efx_oword_t reg;
 
 	if ((efx_nic_rev(efx) != EFX_REV_FALCON_B0) || LOOPBACK_INTERNAL(efx))
@@ -99,7 +99,7 @@ static void falcon_ack_status_intr(struct efx_nic *efx)
 
 	/* We can only use this interrupt to signal the negative edge of
 	 * xaui_align [we have to poll the positive edge]. */
-	if (efx->xmac_poll_required)
+	if (nic_data->xmac_poll_required)
 		return;
 
 	efx_reado(efx, &reg, FR_AB_XM_MGT_INT_MSK);
@@ -277,12 +277,14 @@ static bool falcon_xmac_check_fault(struct efx_nic *efx)
 
 static int falcon_reconfigure_xmac(struct efx_nic *efx)
 {
+	struct falcon_nic_data *nic_data = efx->nic_data;
+
 	falcon_reconfigure_xgxs_core(efx);
 	falcon_reconfigure_xmac_core(efx);
 
 	falcon_reconfigure_mac_wrapper(efx);
 
-	efx->xmac_poll_required = !falcon_xmac_link_ok_retry(efx, 5);
+	nic_data->xmac_poll_required = !falcon_xmac_link_ok_retry(efx, 5);
 	falcon_ack_status_intr(efx);
 
 	return 0;
@@ -350,11 +352,13 @@ static void falcon_update_stats_xmac(struct efx_nic *efx)
 
 void falcon_poll_xmac(struct efx_nic *efx)
 {
+	struct falcon_nic_data *nic_data = efx->nic_data;
+
 	if (!EFX_WORKAROUND_5147(efx) || !efx->link_state.up ||
-	    !efx->xmac_poll_required)
+	    !nic_data->xmac_poll_required)
 		return;
 
-	efx->xmac_poll_required = !falcon_xmac_link_ok_retry(efx, 1);
+	nic_data->xmac_poll_required = !falcon_xmac_link_ok_retry(efx, 1);
 	falcon_ack_status_intr(efx);
 }
 
diff --git a/drivers/net/sfc/filter.c b/drivers/net/sfc/filter.c
index 52cb6082b91..d4722c41c4c 100644
--- a/drivers/net/sfc/filter.c
+++ b/drivers/net/sfc/filter.c
@@ -7,6 +7,7 @@
  * by the Free Software Foundation, incorporated herein by reference.
  */
 
+#include <linux/in.h>
 #include "efx.h"
 #include "filter.h"
 #include "io.h"
@@ -26,19 +27,26 @@
  */
 #define FILTER_CTL_SRCH_MAX 200
 
+enum efx_filter_table_id {
+	EFX_FILTER_TABLE_RX_IP = 0,
+	EFX_FILTER_TABLE_RX_MAC,
+	EFX_FILTER_TABLE_COUNT,
+};
+
 struct efx_filter_table {
+	enum efx_filter_table_id id;
 	u32		offset;		/* address of table relative to BAR */
 	unsigned	size;		/* number of entries */
 	unsigned	step;		/* step between entries */
 	unsigned	used;		/* number currently used */
 	unsigned long	*used_bitmap;
 	struct efx_filter_spec *spec;
+	unsigned	search_depth[EFX_FILTER_TYPE_COUNT];
 };
 
 struct efx_filter_state {
 	spinlock_t	lock;
 	struct efx_filter_table table[EFX_FILTER_TABLE_COUNT];
-	unsigned	search_depth[EFX_FILTER_TYPE_COUNT];
 };
 
 /* The filter hash function is LFSR polynomial x^16 + x^3 + 1 of a 32-bit
@@ -65,68 +73,203 @@ static u16 efx_filter_increment(u32 key)
 }
 
 static enum efx_filter_table_id
-efx_filter_type_table_id(enum efx_filter_type type)
+efx_filter_spec_table_id(const struct efx_filter_spec *spec)
+{
+	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_IP != (EFX_FILTER_TCP_FULL >> 2));
+	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_IP != (EFX_FILTER_TCP_WILD >> 2));
+	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_IP != (EFX_FILTER_UDP_FULL >> 2));
+	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_IP != (EFX_FILTER_UDP_WILD >> 2));
+	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_MAC != (EFX_FILTER_MAC_FULL >> 2));
+	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_MAC != (EFX_FILTER_MAC_WILD >> 2));
+	EFX_BUG_ON_PARANOID(spec->type == EFX_FILTER_UNSPEC);
+	return spec->type >> 2;
+}
+
+static struct efx_filter_table *
+efx_filter_spec_table(struct efx_filter_state *state,
+		      const struct efx_filter_spec *spec)
 {
-	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_IP != (EFX_FILTER_RX_TCP_FULL >> 2));
-	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_IP != (EFX_FILTER_RX_TCP_WILD >> 2));
-	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_IP != (EFX_FILTER_RX_UDP_FULL >> 2));
-	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_IP != (EFX_FILTER_RX_UDP_WILD >> 2));
-	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_MAC != (EFX_FILTER_RX_MAC_FULL >> 2));
-	BUILD_BUG_ON(EFX_FILTER_TABLE_RX_MAC != (EFX_FILTER_RX_MAC_WILD >> 2));
-	return type >> 2;
+	if (spec->type == EFX_FILTER_UNSPEC)
+		return NULL;
+	else
+		return &state->table[efx_filter_spec_table_id(spec)];
 }
 
-static void
-efx_filter_table_reset_search_depth(struct efx_filter_state *state,
-				    enum efx_filter_table_id table_id)
+static void efx_filter_table_reset_search_depth(struct efx_filter_table *table)
 {
-	memset(state->search_depth + (table_id << 2), 0,
-	       sizeof(state->search_depth[0]) << 2);
+	memset(table->search_depth, 0, sizeof(table->search_depth));
 }
 
 static void efx_filter_push_rx_limits(struct efx_nic *efx)
 {
 	struct efx_filter_state *state = efx->filter_state;
+	struct efx_filter_table *table;
 	efx_oword_t filter_ctl;
 
 	efx_reado(efx, &filter_ctl, FR_BZ_RX_FILTER_CTL);
 
+	table = &state->table[EFX_FILTER_TABLE_RX_IP];
 	EFX_SET_OWORD_FIELD(filter_ctl, FRF_BZ_TCP_FULL_SRCH_LIMIT,
-			    state->search_depth[EFX_FILTER_RX_TCP_FULL] +
+			    table->search_depth[EFX_FILTER_TCP_FULL] +
 			    FILTER_CTL_SRCH_FUDGE_FULL);
 	EFX_SET_OWORD_FIELD(filter_ctl, FRF_BZ_TCP_WILD_SRCH_LIMIT,
-			    state->search_depth[EFX_FILTER_RX_TCP_WILD] +
+			    table->search_depth[EFX_FILTER_TCP_WILD] +
 			    FILTER_CTL_SRCH_FUDGE_WILD);
 	EFX_SET_OWORD_FIELD(filter_ctl, FRF_BZ_UDP_FULL_SRCH_LIMIT,
-			    state->search_depth[EFX_FILTER_RX_UDP_FULL] +
+			    table->search_depth[EFX_FILTER_UDP_FULL] +
 			    FILTER_CTL_SRCH_FUDGE_FULL);
 	EFX_SET_OWORD_FIELD(filter_ctl, FRF_BZ_UDP_WILD_SRCH_LIMIT,
-			    state->search_depth[EFX_FILTER_RX_UDP_WILD] +
+			    table->search_depth[EFX_FILTER_UDP_WILD] +
 			    FILTER_CTL_SRCH_FUDGE_WILD);
 
-	if (state->table[EFX_FILTER_TABLE_RX_MAC].size) {
+	table = &state->table[EFX_FILTER_TABLE_RX_MAC];
+	if (table->size) {
 		EFX_SET_OWORD_FIELD(
 			filter_ctl, FRF_CZ_ETHERNET_FULL_SEARCH_LIMIT,
-			state->search_depth[EFX_FILTER_RX_MAC_FULL] +
+			table->search_depth[EFX_FILTER_MAC_FULL] +
 			FILTER_CTL_SRCH_FUDGE_FULL);
 		EFX_SET_OWORD_FIELD(
 			filter_ctl, FRF_CZ_ETHERNET_WILDCARD_SEARCH_LIMIT,
-			state->search_depth[EFX_FILTER_RX_MAC_WILD] +
+			table->search_depth[EFX_FILTER_MAC_WILD] +
 			FILTER_CTL_SRCH_FUDGE_WILD);
 	}
 
 	efx_writeo(efx, &filter_ctl, FR_BZ_RX_FILTER_CTL);
 }
 
+static inline void __efx_filter_set_ipv4(struct efx_filter_spec *spec,
+					 __be32 host1, __be16 port1,
+					 __be32 host2, __be16 port2)
+{
+	spec->data[0] = ntohl(host1) << 16 | ntohs(port1);
+	spec->data[1] = ntohs(port2) << 16 | ntohl(host1) >> 16;
+	spec->data[2] = ntohl(host2);
+}
+
+/**
+ * efx_filter_set_ipv4_local - specify IPv4 host, transport protocol and port
+ * @spec: Specification to initialise
+ * @proto: Transport layer protocol number
+ * @host: Local host address (network byte order)
+ * @port: Local port (network byte order)
+ */
+int efx_filter_set_ipv4_local(struct efx_filter_spec *spec, u8 proto,
+			      __be32 host, __be16 port)
+{
+	__be32 host1;
+	__be16 port1;
+
+	EFX_BUG_ON_PARANOID(!(spec->flags & EFX_FILTER_FLAG_RX));
+
+	/* This cannot currently be combined with other filtering */
+	if (spec->type != EFX_FILTER_UNSPEC)
+		return -EPROTONOSUPPORT;
+
+	if (port == 0)
+		return -EINVAL;
+
+	switch (proto) {
+	case IPPROTO_TCP:
+		spec->type = EFX_FILTER_TCP_WILD;
+		break;
+	case IPPROTO_UDP:
+		spec->type = EFX_FILTER_UDP_WILD;
+		break;
+	default:
+		return -EPROTONOSUPPORT;
+	}
+
+	/* Filter is constructed in terms of source and destination,
+	 * with the odd wrinkle that the ports are swapped in a UDP
+	 * wildcard filter.  We need to convert from local and remote
+	 * (= zero for wildcard) addresses.
+	 */
+	host1 = 0;
+	if (proto != IPPROTO_UDP) {
+		port1 = 0;
+	} else {
+		port1 = port;
+		port = 0;
+	}
+
+	__efx_filter_set_ipv4(spec, host1, port1, host, port);
+	return 0;
+}
+
+/**
+ * efx_filter_set_ipv4_full - specify IPv4 hosts, transport protocol and ports
+ * @spec: Specification to initialise
+ * @proto: Transport layer protocol number
+ * @host: Local host address (network byte order)
+ * @port: Local port (network byte order)
+ * @rhost: Remote host address (network byte order)
+ * @rport: Remote port (network byte order)
+ */
+int efx_filter_set_ipv4_full(struct efx_filter_spec *spec, u8 proto,
+			     __be32 host, __be16 port,
+			     __be32 rhost, __be16 rport)
+{
+	EFX_BUG_ON_PARANOID(!(spec->flags & EFX_FILTER_FLAG_RX));
+
+	/* This cannot currently be combined with other filtering */
+	if (spec->type != EFX_FILTER_UNSPEC)
+		return -EPROTONOSUPPORT;
+
+	if (port == 0 || rport == 0)
+		return -EINVAL;
+
+	switch (proto) {
+	case IPPROTO_TCP:
+		spec->type = EFX_FILTER_TCP_FULL;
+		break;
+	case IPPROTO_UDP:
+		spec->type = EFX_FILTER_UDP_FULL;
+		break;
+	default:
+		return -EPROTONOSUPPORT;
+	}
+
+	__efx_filter_set_ipv4(spec, rhost, rport, host, port);
+	return 0;
+}
+
+/**
+ * efx_filter_set_eth_local - specify local Ethernet address and optional VID
+ * @spec: Specification to initialise
+ * @vid: VLAN ID to match, or %EFX_FILTER_VID_UNSPEC
+ * @addr: Local Ethernet MAC address
+ */
+int efx_filter_set_eth_local(struct efx_filter_spec *spec,
+			     u16 vid, const u8 *addr)
+{
+	EFX_BUG_ON_PARANOID(!(spec->flags & EFX_FILTER_FLAG_RX));
+
+	/* This cannot currently be combined with other filtering */
+	if (spec->type != EFX_FILTER_UNSPEC)
+		return -EPROTONOSUPPORT;
+
+	if (vid == EFX_FILTER_VID_UNSPEC) {
+		spec->type = EFX_FILTER_MAC_WILD;
+		spec->data[0] = 0;
+	} else {
+		spec->type = EFX_FILTER_MAC_FULL;
+		spec->data[0] = vid;
+	}
+
+	spec->data[1] = addr[2] << 24 | addr[3] << 16 | addr[4] << 8 | addr[5];
+	spec->data[2] = addr[0] << 8 | addr[1];
+	return 0;
+}
+
 /* Build a filter entry and return its n-tuple key. */
 static u32 efx_filter_build(efx_oword_t *filter, struct efx_filter_spec *spec)
 {
 	u32 data3;
 
-	switch (efx_filter_type_table_id(spec->type)) {
+	switch (efx_filter_spec_table_id(spec)) {
 	case EFX_FILTER_TABLE_RX_IP: {
-		bool is_udp = (spec->type == EFX_FILTER_RX_UDP_FULL ||
-			       spec->type == EFX_FILTER_RX_UDP_WILD);
+		bool is_udp = (spec->type == EFX_FILTER_UDP_FULL ||
+			       spec->type == EFX_FILTER_UDP_WILD);
 		EFX_POPULATE_OWORD_7(
 			*filter,
 			FRF_BZ_RSS_EN,
@@ -143,7 +286,7 @@ static u32 efx_filter_build(efx_oword_t *filter, struct efx_filter_spec *spec)
 	}
 
 	case EFX_FILTER_TABLE_RX_MAC: {
-		bool is_wild = spec->type == EFX_FILTER_RX_MAC_WILD;
+		bool is_wild = spec->type == EFX_FILTER_MAC_WILD;
 		EFX_POPULATE_OWORD_8(
 			*filter,
 			FRF_CZ_RMFT_RSS_EN,
@@ -206,6 +349,14 @@ found:
 	return filter_idx;
 }
 
+/* Construct/deconstruct external filter IDs */
+
+static inline int
+efx_filter_make_id(enum efx_filter_table_id table_id, unsigned index)
+{
+	return table_id << 16 | index;
+}
+
 /**
  * efx_filter_insert_filter - add or replace a filter
  * @efx: NIC in which to insert the filter
@@ -213,30 +364,28 @@ found:
  * @replace: Flag for whether the specified filter may replace a filter
  *	with an identical match expression and equal or lower priority
  *
- * On success, return the filter index within its table.
+ * On success, return the filter ID.
  * On failure, return a negative error code.
  */
 int efx_filter_insert_filter(struct efx_nic *efx, struct efx_filter_spec *spec,
 			     bool replace)
 {
 	struct efx_filter_state *state = efx->filter_state;
-	enum efx_filter_table_id table_id =
-		efx_filter_type_table_id(spec->type);
-	struct efx_filter_table *table = &state->table[table_id];
+	struct efx_filter_table *table = efx_filter_spec_table(state, spec);
 	struct efx_filter_spec *saved_spec;
 	efx_oword_t filter;
 	int filter_idx, depth;
 	u32 key;
 	int rc;
 
-	if (table->size == 0)
+	if (!table || table->size == 0)
 		return -EINVAL;
 
 	key = efx_filter_build(&filter, spec);
 
 	netif_vdbg(efx, hw, efx->net_dev,
 		   "%s: type %d search_depth=%d", __func__, spec->type,
-		   state->search_depth[spec->type]);
+		   table->search_depth[spec->type]);
 
 	spin_lock_bh(&state->lock);
 
@@ -263,8 +412,8 @@ int efx_filter_insert_filter(struct efx_nic *efx, struct efx_filter_spec *spec,
 	}
 	*saved_spec = *spec;
 
-	if (state->search_depth[spec->type] < depth) {
-		state->search_depth[spec->type] = depth;
+	if (table->search_depth[spec->type] < depth) {
+		table->search_depth[spec->type] = depth;
 		efx_filter_push_rx_limits(efx);
 	}
 
@@ -273,6 +422,7 @@ int efx_filter_insert_filter(struct efx_nic *efx, struct efx_filter_spec *spec,
 	netif_vdbg(efx, hw, efx->net_dev,
 		   "%s: filter type %d index %d rxq %u set",
 		   __func__, spec->type, filter_idx, spec->dmaq_id);
+	rc = efx_filter_make_id(table->id, filter_idx);
 
 out:
 	spin_unlock_bh(&state->lock);
@@ -306,15 +456,16 @@ static void efx_filter_table_clear_entry(struct efx_nic *efx,
 int efx_filter_remove_filter(struct efx_nic *efx, struct efx_filter_spec *spec)
 {
 	struct efx_filter_state *state = efx->filter_state;
-	enum efx_filter_table_id table_id =
-		efx_filter_type_table_id(spec->type);
-	struct efx_filter_table *table = &state->table[table_id];
+	struct efx_filter_table *table = efx_filter_spec_table(state, spec);
 	struct efx_filter_spec *saved_spec;
 	efx_oword_t filter;
 	int filter_idx, depth;
 	u32 key;
 	int rc;
 
+	if (!table)
+		return -EINVAL;
+
 	key = efx_filter_build(&filter, spec);
 
 	spin_lock_bh(&state->lock);
@@ -332,7 +483,7 @@ int efx_filter_remove_filter(struct efx_nic *efx, struct efx_filter_spec *spec)
 
 	efx_filter_table_clear_entry(efx, table, filter_idx);
 	if (table->used == 0)
-		efx_filter_table_reset_search_depth(state, table_id);
+		efx_filter_table_reset_search_depth(table);
 	rc = 0;
 
 out:
@@ -340,15 +491,9 @@ out:
 	return rc;
 }
 
-/**
- * efx_filter_table_clear - remove filters from a table by priority
- * @efx: NIC from which to remove the filters
- * @table_id: Table from which to remove the filters
- * @priority: Maximum priority to remove
- */
-void efx_filter_table_clear(struct efx_nic *efx,
-			    enum efx_filter_table_id table_id,
-			    enum efx_filter_priority priority)
+static void efx_filter_table_clear(struct efx_nic *efx,
+				   enum efx_filter_table_id table_id,
+				   enum efx_filter_priority priority)
 {
 	struct efx_filter_state *state = efx->filter_state;
 	struct efx_filter_table *table = &state->table[table_id];
@@ -360,11 +505,22 @@ void efx_filter_table_clear(struct efx_nic *efx,
 		if (table->spec[filter_idx].priority <= priority)
 			efx_filter_table_clear_entry(efx, table, filter_idx);
 	if (table->used == 0)
-		efx_filter_table_reset_search_depth(state, table_id);
+		efx_filter_table_reset_search_depth(table);
 
 	spin_unlock_bh(&state->lock);
 }
 
+/**
+ * efx_filter_clear_rx - remove RX filters by priority
+ * @efx: NIC from which to remove the filters
+ * @priority: Maximum priority to remove
+ */
+void efx_filter_clear_rx(struct efx_nic *efx, enum efx_filter_priority priority)
+{
+	efx_filter_table_clear(efx, EFX_FILTER_TABLE_RX_IP, priority);
+	efx_filter_table_clear(efx, EFX_FILTER_TABLE_RX_MAC, priority);
+}
+
 /* Restore filter stater after reset */
 void efx_restore_filters(struct efx_nic *efx)
 {
@@ -407,6 +563,7 @@ int efx_probe_filters(struct efx_nic *efx)
 
 	if (efx_nic_rev(efx) >= EFX_REV_FALCON_B0) {
 		table = &state->table[EFX_FILTER_TABLE_RX_IP];
+		table->id = EFX_FILTER_TABLE_RX_IP;
 		table->offset = FR_BZ_RX_FILTER_TBL0;
 		table->size = FR_BZ_RX_FILTER_TBL0_ROWS;
 		table->step = FR_BZ_RX_FILTER_TBL0_STEP;
@@ -414,6 +571,7 @@ int efx_probe_filters(struct efx_nic *efx)
 
 	if (efx_nic_rev(efx) >= EFX_REV_SIENA_A0) {
 		table = &state->table[EFX_FILTER_TABLE_RX_MAC];
+		table->id = EFX_FILTER_TABLE_RX_MAC;
 		table->offset = FR_CZ_RX_MAC_FILTER_TBL0;
 		table->size = FR_CZ_RX_MAC_FILTER_TBL0_ROWS;
 		table->step = FR_CZ_RX_MAC_FILTER_TBL0_STEP;
@@ -428,10 +586,9 @@ int efx_probe_filters(struct efx_nic *efx)
 					     GFP_KERNEL);
 		if (!table->used_bitmap)
 			goto fail;
-		table->spec = vmalloc(table->size * sizeof(*table->spec));
+		table->spec = vzalloc(table->size * sizeof(*table->spec));
 		if (!table->spec)
 			goto fail;
-		memset(table->spec, 0, table->size * sizeof(*table->spec));
 	}
 
 	return 0;
diff --git a/drivers/net/sfc/filter.h b/drivers/net/sfc/filter.h
index a53319ded79..872f2132a49 100644
--- a/drivers/net/sfc/filter.h
+++ b/drivers/net/sfc/filter.h
@@ -12,31 +12,27 @@
 
 #include <linux/types.h>
 
-enum efx_filter_table_id {
-	EFX_FILTER_TABLE_RX_IP = 0,
-	EFX_FILTER_TABLE_RX_MAC,
-	EFX_FILTER_TABLE_COUNT,
-};
-
 /**
  * enum efx_filter_type - type of hardware filter
- * @EFX_FILTER_RX_TCP_FULL: RX, matching TCP/IPv4 4-tuple
- * @EFX_FILTER_RX_TCP_WILD: RX, matching TCP/IPv4 destination (host, port)
- * @EFX_FILTER_RX_UDP_FULL: RX, matching UDP/IPv4 4-tuple
- * @EFX_FILTER_RX_UDP_WILD: RX, matching UDP/IPv4 destination (host, port)
- * @EFX_FILTER_RX_MAC_FULL: RX, matching Ethernet destination MAC address, VID
- * @EFX_FILTER_RX_MAC_WILD: RX, matching Ethernet destination MAC address
+ * @EFX_FILTER_TCP_FULL: Matching TCP/IPv4 4-tuple
+ * @EFX_FILTER_TCP_WILD: Matching TCP/IPv4 destination (host, port)
+ * @EFX_FILTER_UDP_FULL: Matching UDP/IPv4 4-tuple
+ * @EFX_FILTER_UDP_WILD: Matching UDP/IPv4 destination (host, port)
+ * @EFX_FILTER_MAC_FULL: Matching Ethernet destination MAC address, VID
+ * @EFX_FILTER_MAC_WILD: Matching Ethernet destination MAC address
+ * @EFX_FILTER_UNSPEC: Match type is unspecified
  *
- * Falcon NICs only support the RX TCP/IPv4 and UDP/IPv4 filter types.
+ * Falcon NICs only support the TCP/IPv4 and UDP/IPv4 filter types.
  */
 enum efx_filter_type {
-	EFX_FILTER_RX_TCP_FULL = 0,
-	EFX_FILTER_RX_TCP_WILD,
-	EFX_FILTER_RX_UDP_FULL,
-	EFX_FILTER_RX_UDP_WILD,
-	EFX_FILTER_RX_MAC_FULL = 4,
-	EFX_FILTER_RX_MAC_WILD,
-	EFX_FILTER_TYPE_COUNT,
+	EFX_FILTER_TCP_FULL = 0,
+	EFX_FILTER_TCP_WILD,
+	EFX_FILTER_UDP_FULL,
+	EFX_FILTER_UDP_WILD,
+	EFX_FILTER_MAC_FULL = 4,
+	EFX_FILTER_MAC_WILD,
+	EFX_FILTER_TYPE_COUNT,		/* number of specific types */
+	EFX_FILTER_UNSPEC = 0xf,
 };
 
 /**
@@ -63,13 +59,13 @@ enum efx_filter_priority {
  * @EFX_FILTER_FLAG_RX_OVERRIDE_IP: Enables a MAC filter to override
  *	any IP filter that matches the same packet.  By default, IP
  *	filters take precedence.
- *
- * Currently, no flags are defined for TX filters.
+ * @EFX_FILTER_FLAG_RX: Filter is for RX
  */
 enum efx_filter_flags {
 	EFX_FILTER_FLAG_RX_RSS = 0x01,
 	EFX_FILTER_FLAG_RX_SCATTER = 0x02,
 	EFX_FILTER_FLAG_RX_OVERRIDE_IP = 0x04,
+	EFX_FILTER_FLAG_RX = 0x08,
 };
 
 /**
@@ -91,99 +87,26 @@ struct efx_filter_spec {
 	u32	data[3];
 };
 
-/**
- * efx_filter_set_rx_tcp_full - specify RX filter with TCP/IPv4 full match
- * @spec: Specification to initialise
- * @shost: Source host address (host byte order)
- * @sport: Source port (host byte order)
- * @dhost: Destination host address (host byte order)
- * @dport: Destination port (host byte order)
- */
-static inline void
-efx_filter_set_rx_tcp_full(struct efx_filter_spec *spec,
-			   u32 shost, u16 sport, u32 dhost, u16 dport)
-{
-	spec->type = EFX_FILTER_RX_TCP_FULL;
-	spec->data[0] = sport | shost << 16;
-	spec->data[1] = dport << 16 | shost >> 16;
-	spec->data[2] = dhost;
-}
-
-/**
- * efx_filter_set_rx_tcp_wild - specify RX filter with TCP/IPv4 wildcard match
- * @spec: Specification to initialise
- * @dhost: Destination host address (host byte order)
- * @dport: Destination port (host byte order)
- */
-static inline void
-efx_filter_set_rx_tcp_wild(struct efx_filter_spec *spec, u32 dhost, u16 dport)
-{
-	spec->type = EFX_FILTER_RX_TCP_WILD;
-	spec->data[0] = 0;
-	spec->data[1] = dport << 16;
-	spec->data[2] = dhost;
-}
-
-/**
- * efx_filter_set_rx_udp_full - specify RX filter with UDP/IPv4 full match
- * @spec: Specification to initialise
- * @shost: Source host address (host byte order)
- * @sport: Source port (host byte order)
- * @dhost: Destination host address (host byte order)
- * @dport: Destination port (host byte order)
- */
-static inline void
-efx_filter_set_rx_udp_full(struct efx_filter_spec *spec,
-			   u32 shost, u16 sport, u32 dhost, u16 dport)
-{
-	spec->type = EFX_FILTER_RX_UDP_FULL;
-	spec->data[0] = sport | shost << 16;
-	spec->data[1] = dport << 16 | shost >> 16;
-	spec->data[2] = dhost;
-}
-
-/**
- * efx_filter_set_rx_udp_wild - specify RX filter with UDP/IPv4 wildcard match
- * @spec: Specification to initialise
- * @dhost: Destination host address (host byte order)
- * @dport: Destination port (host byte order)
- */
-static inline void
-efx_filter_set_rx_udp_wild(struct efx_filter_spec *spec, u32 dhost, u16 dport)
+static inline void efx_filter_init_rx(struct efx_filter_spec *spec,
+				      enum efx_filter_priority priority,
+				      enum efx_filter_flags flags,
+				      unsigned rxq_id)
 {
-	spec->type = EFX_FILTER_RX_UDP_WILD;
-	spec->data[0] = dport;
-	spec->data[1] = 0;
-	spec->data[2] = dhost;
+	spec->type = EFX_FILTER_UNSPEC;
+	spec->priority = priority;
+	spec->flags = EFX_FILTER_FLAG_RX | flags;
+	spec->dmaq_id = rxq_id;
 }
 
-/**
- * efx_filter_set_rx_mac_full - specify RX filter with MAC full match
- * @spec: Specification to initialise
- * @vid: VLAN ID
- * @addr: Destination MAC address
- */
-static inline void efx_filter_set_rx_mac_full(struct efx_filter_spec *spec,
-					      u16 vid, const u8 *addr)
-{
-	spec->type = EFX_FILTER_RX_MAC_FULL;
-	spec->data[0] = vid;
-	spec->data[1] = addr[2] << 24 | addr[3] << 16 | addr[4] << 8 | addr[5];
-	spec->data[2] = addr[0] << 8 | addr[1];
-}
-
-/**
- * efx_filter_set_rx_mac_full - specify RX filter with MAC wildcard match
- * @spec: Specification to initialise
- * @addr: Destination MAC address
- */
-static inline void efx_filter_set_rx_mac_wild(struct efx_filter_spec *spec,
-					      const u8 *addr)
-{
-	spec->type = EFX_FILTER_RX_MAC_WILD;
-	spec->data[0] = 0;
-	spec->data[1] = addr[2] << 24 | addr[3] << 16 | addr[4] << 8 | addr[5];
-	spec->data[2] = addr[0] << 8 | addr[1];
-}
+extern int efx_filter_set_ipv4_local(struct efx_filter_spec *spec, u8 proto,
+				     __be32 host, __be16 port);
+extern int efx_filter_set_ipv4_full(struct efx_filter_spec *spec, u8 proto,
+				    __be32 host, __be16 port,
+				    __be32 rhost, __be16 rport);
+extern int efx_filter_set_eth_local(struct efx_filter_spec *spec,
+				    u16 vid, const u8 *addr);
+enum {
+	EFX_FILTER_VID_UNSPEC = 0xffff,
+};
 
 #endif /* EFX_FILTER_H */
diff --git a/drivers/net/sfc/io.h b/drivers/net/sfc/io.h
index 85a99fe8743..6da4ae20a03 100644
--- a/drivers/net/sfc/io.h
+++ b/drivers/net/sfc/io.h
@@ -22,28 +22,39 @@
  *
  * Notes on locking strategy:
  *
- * Most NIC registers require 16-byte (or 8-byte, for SRAM) atomic writes
- * which necessitates locking.
- * Under normal operation few writes to NIC registers are made and these
- * registers (EVQ_RPTR_REG, RX_DESC_UPD_REG and TX_DESC_UPD_REG) are special
- * cased to allow 4-byte (hence lockless) accesses.
+ * Most CSRs are 128-bit (oword) and therefore cannot be read or
+ * written atomically.  Access from the host is buffered by the Bus
+ * Interface Unit (BIU).  Whenever the host reads from the lowest
+ * address of such a register, or from the address of a different such
+ * register, the BIU latches the register's value.  Subsequent reads
+ * from higher addresses of the same register will read the latched
+ * value.  Whenever the host writes part of such a register, the BIU
+ * collects the written value and does not write to the underlying
+ * register until all 4 dwords have been written.  A similar buffering
+ * scheme applies to host access to the NIC's 64-bit SRAM.
  *
- * It *is* safe to write to these 4-byte registers in the middle of an
- * access to an 8-byte or 16-byte register.  We therefore use a
- * spinlock to protect accesses to the larger registers, but no locks
- * for the 4-byte registers.
+ * Access to different CSRs and 64-bit SRAM words must be serialised,
+ * since interleaved access can result in lost writes or lost
+ * information from read-to-clear fields.  We use efx_nic::biu_lock
+ * for this.  (We could use separate locks for read and write, but
+ * this is not normally a performance bottleneck.)
  *
- * A write barrier is needed to ensure that DW3 is written after DW0/1/2
- * due to the way the 16byte registers are "collected" in the BIU.
+ * The DMA descriptor pointers (RX_DESC_UPD and TX_DESC_UPD) are
+ * 128-bit but are special-cased in the BIU to avoid the need for
+ * locking in the host:
  *
- * We also lock when carrying out reads, to ensure consistency of the
- * data (made possible since the BIU reads all 128 bits into a cache).
- * Reads are very rare, so this isn't a significant performance
- * impact.  (Most data transferred from NIC to host is DMAed directly
- * into host memory).
- *
- * I/O BAR access uses locks for both reads and writes (but is only provided
- * for testing purposes).
+ * - They are write-only.
+ * - The semantics of writing to these registers are such that
+ *   replacing the low 96 bits with zero does not affect functionality.
+ * - If the host writes to the last dword address of such a register
+ *   (i.e. the high 32 bits) the underlying register will always be
+ *   written.  If the collector does not hold values for the low 96
+ *   bits of the register, they will be written as zero.  Writing to
+ *   the last qword does not have this effect and must not be done.
+ * - If the host writes to the address of any other part of such a
+ *   register while the collector already holds values for some other
+ *   register, the write is discarded and the collector maintains its
+ *   current state.
  */
 
 #if BITS_PER_LONG == 64
@@ -72,7 +83,7 @@ static inline __le32 _efx_readd(struct efx_nic *efx, unsigned int reg)
 	return (__force __le32)__raw_readl(efx->membase + reg);
 }
 
-/* Writes to a normal 16-byte Efx register, locking as appropriate. */
+/* Write a normal 128-bit CSR, locking as appropriate. */
 static inline void efx_writeo(struct efx_nic *efx, efx_oword_t *value,
 			      unsigned int reg)
 {
@@ -85,21 +96,18 @@ static inline void efx_writeo(struct efx_nic *efx, efx_oword_t *value,
 	spin_lock_irqsave(&efx->biu_lock, flags);
 #ifdef EFX_USE_QWORD_IO
 	_efx_writeq(efx, value->u64[0], reg + 0);
-	wmb();
 	_efx_writeq(efx, value->u64[1], reg + 8);
 #else
 	_efx_writed(efx, value->u32[0], reg + 0);
 	_efx_writed(efx, value->u32[1], reg + 4);
 	_efx_writed(efx, value->u32[2], reg + 8);
-	wmb();
 	_efx_writed(efx, value->u32[3], reg + 12);
 #endif
 	mmiowb();
 	spin_unlock_irqrestore(&efx->biu_lock, flags);
 }
 
-/* Write an 8-byte NIC SRAM entry through the supplied mapping,
- * locking as appropriate. */
+/* Write 64-bit SRAM through the supplied mapping, locking as appropriate. */
 static inline void efx_sram_writeq(struct efx_nic *efx, void __iomem *membase,
 				   efx_qword_t *value, unsigned int index)
 {
@@ -115,36 +123,25 @@ static inline void efx_sram_writeq(struct efx_nic *efx, void __iomem *membase,
 	__raw_writeq((__force u64)value->u64[0], membase + addr);
 #else
 	__raw_writel((__force u32)value->u32[0], membase + addr);
-	wmb();
 	__raw_writel((__force u32)value->u32[1], membase + addr + 4);
 #endif
 	mmiowb();
 	spin_unlock_irqrestore(&efx->biu_lock, flags);
 }
 
-/* Write dword to NIC register that allows partial writes
- *
- * Some registers (EVQ_RPTR_REG, RX_DESC_UPD_REG and
- * TX_DESC_UPD_REG) can be written to as a single dword.  This allows
- * for lockless writes.
- */
+/* Write a 32-bit CSR or the last dword of a special 128-bit CSR */
 static inline void efx_writed(struct efx_nic *efx, efx_dword_t *value,
 			      unsigned int reg)
 {
 	netif_vdbg(efx, hw, efx->net_dev,
-		   "writing partial register %x with "EFX_DWORD_FMT"\n",
+		   "writing register %x with "EFX_DWORD_FMT"\n",
 		   reg, EFX_DWORD_VAL(*value));
 
 	/* No lock required */
 	_efx_writed(efx, value->u32[0], reg);
 }
 
-/* Read from a NIC register
- *
- * This reads an entire 16-byte register in one go, locking as
- * appropriate.  It is essential to read the first dword first, as this
- * prompts the NIC to load the current value into the shadow register.
- */
+/* Read a 128-bit CSR, locking as appropriate. */
 static inline void efx_reado(struct efx_nic *efx, efx_oword_t *value,
 			     unsigned int reg)
 {
@@ -152,7 +149,6 @@ static inline void efx_reado(struct efx_nic *efx, efx_oword_t *value,
 
 	spin_lock_irqsave(&efx->biu_lock, flags);
 	value->u32[0] = _efx_readd(efx, reg + 0);
-	rmb();
 	value->u32[1] = _efx_readd(efx, reg + 4);
 	value->u32[2] = _efx_readd(efx, reg + 8);
 	value->u32[3] = _efx_readd(efx, reg + 12);
@@ -163,8 +159,7 @@ static inline void efx_reado(struct efx_nic *efx, efx_oword_t *value,
 		   EFX_OWORD_VAL(*value));
 }
 
-/* Read an 8-byte SRAM entry through supplied mapping,
- * locking as appropriate. */
+/* Read 64-bit SRAM through the supplied mapping, locking as appropriate. */
 static inline void efx_sram_readq(struct efx_nic *efx, void __iomem *membase,
 				  efx_qword_t *value, unsigned int index)
 {
@@ -176,7 +171,6 @@ static inline void efx_sram_readq(struct efx_nic *efx, void __iomem *membase,
 	value->u64[0] = (__force __le64)__raw_readq(membase + addr);
 #else
 	value->u32[0] = (__force __le32)__raw_readl(membase + addr);
-	rmb();
 	value->u32[1] = (__force __le32)__raw_readl(membase + addr + 4);
 #endif
 	spin_unlock_irqrestore(&efx->biu_lock, flags);
@@ -186,7 +180,7 @@ static inline void efx_sram_readq(struct efx_nic *efx, void __iomem *membase,
 		   addr, EFX_QWORD_VAL(*value));
 }
 
-/* Read dword from register that allows partial writes (sic) */
+/* Read a 32-bit CSR or SRAM */
 static inline void efx_readd(struct efx_nic *efx, efx_dword_t *value,
 				unsigned int reg)
 {
@@ -196,28 +190,28 @@ static inline void efx_readd(struct efx_nic *efx, efx_dword_t *value,
 		   reg, EFX_DWORD_VAL(*value));
 }
 
-/* Write to a register forming part of a table */
+/* Write a 128-bit CSR forming part of a table */
 static inline void efx_writeo_table(struct efx_nic *efx, efx_oword_t *value,
 				      unsigned int reg, unsigned int index)
 {
 	efx_writeo(efx, value, reg + index * sizeof(efx_oword_t));
 }
 
-/* Read to a register forming part of a table */
+/* Read a 128-bit CSR forming part of a table */
 static inline void efx_reado_table(struct efx_nic *efx, efx_oword_t *value,
 				     unsigned int reg, unsigned int index)
 {
 	efx_reado(efx, value, reg + index * sizeof(efx_oword_t));
 }
 
-/* Write to a dword register forming part of a table */
+/* Write a 32-bit CSR forming part of a table, or 32-bit SRAM */
 static inline void efx_writed_table(struct efx_nic *efx, efx_dword_t *value,
 				       unsigned int reg, unsigned int index)
 {
 	efx_writed(efx, value, reg + index * sizeof(efx_oword_t));
 }
 
-/* Read from a dword register forming part of a table */
+/* Read a 32-bit CSR forming part of a table, or 32-bit SRAM */
 static inline void efx_readd_table(struct efx_nic *efx, efx_dword_t *value,
 				   unsigned int reg, unsigned int index)
 {
@@ -231,29 +225,54 @@ static inline void efx_readd_table(struct efx_nic *efx, efx_dword_t *value,
 #define EFX_PAGED_REG(page, reg) \
 	((page) * EFX_PAGE_BLOCK_SIZE + (reg))
 
-/* As for efx_writeo(), but for a page-mapped register. */
-static inline void efx_writeo_page(struct efx_nic *efx, efx_oword_t *value,
-				   unsigned int reg, unsigned int page)
+/* Write the whole of RX_DESC_UPD or TX_DESC_UPD */
+static inline void _efx_writeo_page(struct efx_nic *efx, efx_oword_t *value,
+				    unsigned int reg, unsigned int page)
 {
-	efx_writeo(efx, value, EFX_PAGED_REG(page, reg));
-}
+	reg = EFX_PAGED_REG(page, reg);
 
-/* As for efx_writed(), but for a page-mapped register. */
-static inline void efx_writed_page(struct efx_nic *efx, efx_dword_t *value,
-				   unsigned int reg, unsigned int page)
+	netif_vdbg(efx, hw, efx->net_dev,
+		   "writing register %x with " EFX_OWORD_FMT "\n", reg,
+		   EFX_OWORD_VAL(*value));
+
+#ifdef EFX_USE_QWORD_IO
+	_efx_writeq(efx, value->u64[0], reg + 0);
+#else
+	_efx_writed(efx, value->u32[0], reg + 0);
+	_efx_writed(efx, value->u32[1], reg + 4);
+#endif
+	_efx_writed(efx, value->u32[2], reg + 8);
+	_efx_writed(efx, value->u32[3], reg + 12);
+}
+#define efx_writeo_page(efx, value, reg, page)				\
+	_efx_writeo_page(efx, value,					\
+			 reg +						\
+			 BUILD_BUG_ON_ZERO((reg) != 0x830 && (reg) != 0xa10), \
+			 page)
+
+/* Write a page-mapped 32-bit CSR (EVQ_RPTR or the high bits of
+ * RX_DESC_UPD or TX_DESC_UPD)
+ */
+static inline void _efx_writed_page(struct efx_nic *efx, efx_dword_t *value,
+				    unsigned int reg, unsigned int page)
 {
 	efx_writed(efx, value, EFX_PAGED_REG(page, reg));
 }
-
-/* Write dword to page-mapped register with an extra lock.
- *
- * As for efx_writed_page(), but for a register that suffers from
- * SFC bug 3181. Take out a lock so the BIU collector cannot be
- * confused. */
-static inline void efx_writed_page_locked(struct efx_nic *efx,
-					  efx_dword_t *value,
-					  unsigned int reg,
-					  unsigned int page)
+#define efx_writed_page(efx, value, reg, page)				\
+	_efx_writed_page(efx, value,					\
+			 reg +						\
+			 BUILD_BUG_ON_ZERO((reg) != 0x400 && (reg) != 0x83c \
+					   && (reg) != 0xa1c),		\
+			 page)
+
+/* Write TIMER_COMMAND.  This is a page-mapped 32-bit CSR, but a bug
+ * in the BIU means that writes to TIMER_COMMAND[0] invalidate the
+ * collector register.
+ */
+static inline void _efx_writed_page_locked(struct efx_nic *efx,
+					   efx_dword_t *value,
+					   unsigned int reg,
+					   unsigned int page)
 {
 	unsigned long flags __attribute__ ((unused));
 
@@ -265,5 +284,9 @@ static inline void efx_writed_page_locked(struct efx_nic *efx,
 		efx_writed(efx, value, EFX_PAGED_REG(page, reg));
 	}
 }
+#define efx_writed_page_locked(efx, value, reg, page)			\
+	_efx_writed_page_locked(efx, value,				\
+				reg + BUILD_BUG_ON_ZERO((reg) != 0x420), \
+				page)
 
 #endif /* EFX_IO_H */
diff --git a/drivers/net/sfc/mcdi.c b/drivers/net/sfc/mcdi.c
index 12cf910c2ce..b716e827b29 100644
--- a/drivers/net/sfc/mcdi.c
+++ b/drivers/net/sfc/mcdi.c
@@ -381,7 +381,7 @@ int efx_mcdi_rpc(struct efx_nic *efx, unsigned cmd,
 				  -rc);
 			efx_schedule_reset(efx, RESET_TYPE_MC_FAILURE);
 		} else
-			netif_err(efx, hw, efx->net_dev,
+			netif_dbg(efx, hw, efx->net_dev,
 				  "MC command 0x%x inlen %d failed rc=%d\n",
 				  cmd, (int)inlen, -rc);
 	}
@@ -463,6 +463,7 @@ static void efx_mcdi_ev_death(struct efx_nic *efx, int rc)
 		if (mcdi->mode == MCDI_MODE_EVENTS) {
 			mcdi->resprc = rc;
 			mcdi->resplen = 0;
+			++mcdi->credits;
 		}
 	} else
 		/* Nobody was waiting for an MCDI request, so trigger a reset */
diff --git a/drivers/net/sfc/mcdi_phy.c b/drivers/net/sfc/mcdi_phy.c
index c992742446b..0e97eed663c 100644
--- a/drivers/net/sfc/mcdi_phy.c
+++ b/drivers/net/sfc/mcdi_phy.c
@@ -16,7 +16,6 @@
 #include "phy.h"
 #include "mcdi.h"
 #include "mcdi_pcol.h"
-#include "mdio_10g.h"
 #include "nic.h"
 #include "selftest.h"
 
diff --git a/drivers/net/sfc/mdio_10g.c b/drivers/net/sfc/mdio_10g.c
index 98d94602042..56b0266b441 100644
--- a/drivers/net/sfc/mdio_10g.c
+++ b/drivers/net/sfc/mdio_10g.c
@@ -15,7 +15,6 @@
 #include "net_driver.h"
 #include "mdio_10g.h"
 #include "workarounds.h"
-#include "nic.h"
 
 unsigned efx_mdio_id_oui(u32 id)
 {
diff --git a/drivers/net/sfc/mtd.c b/drivers/net/sfc/mtd.c
index 02e54b4f701..d38627448c2 100644
--- a/drivers/net/sfc/mtd.c
+++ b/drivers/net/sfc/mtd.c
@@ -321,14 +321,15 @@ static int falcon_mtd_read(struct mtd_info *mtd, loff_t start,
 	struct efx_mtd *efx_mtd = mtd->priv;
 	const struct efx_spi_device *spi = efx_mtd->spi;
 	struct efx_nic *efx = efx_mtd->efx;
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	int rc;
 
-	rc = mutex_lock_interruptible(&efx->spi_lock);
+	rc = mutex_lock_interruptible(&nic_data->spi_lock);
 	if (rc)
 		return rc;
 	rc = falcon_spi_read(efx, spi, part->offset + start, len,
 			     retlen, buffer);
-	mutex_unlock(&efx->spi_lock);
+	mutex_unlock(&nic_data->spi_lock);
 	return rc;
 }
 
@@ -337,13 +338,14 @@ static int falcon_mtd_erase(struct mtd_info *mtd, loff_t start, size_t len)
 	struct efx_mtd_partition *part = to_efx_mtd_partition(mtd);
 	struct efx_mtd *efx_mtd = mtd->priv;
 	struct efx_nic *efx = efx_mtd->efx;
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	int rc;
 
-	rc = mutex_lock_interruptible(&efx->spi_lock);
+	rc = mutex_lock_interruptible(&nic_data->spi_lock);
 	if (rc)
 		return rc;
 	rc = efx_spi_erase(part, part->offset + start, len);
-	mutex_unlock(&efx->spi_lock);
+	mutex_unlock(&nic_data->spi_lock);
 	return rc;
 }
 
@@ -354,14 +356,15 @@ static int falcon_mtd_write(struct mtd_info *mtd, loff_t start,
 	struct efx_mtd *efx_mtd = mtd->priv;
 	const struct efx_spi_device *spi = efx_mtd->spi;
 	struct efx_nic *efx = efx_mtd->efx;
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	int rc;
 
-	rc = mutex_lock_interruptible(&efx->spi_lock);
+	rc = mutex_lock_interruptible(&nic_data->spi_lock);
 	if (rc)
 		return rc;
 	rc = falcon_spi_write(efx, spi, part->offset + start, len,
 			      retlen, buffer);
-	mutex_unlock(&efx->spi_lock);
+	mutex_unlock(&nic_data->spi_lock);
 	return rc;
 }
 
@@ -370,11 +373,12 @@ static int falcon_mtd_sync(struct mtd_info *mtd)
 	struct efx_mtd_partition *part = to_efx_mtd_partition(mtd);
 	struct efx_mtd *efx_mtd = mtd->priv;
 	struct efx_nic *efx = efx_mtd->efx;
+	struct falcon_nic_data *nic_data = efx->nic_data;
 	int rc;
 
-	mutex_lock(&efx->spi_lock);
+	mutex_lock(&nic_data->spi_lock);
 	rc = efx_spi_slow_wait(part, true);
-	mutex_unlock(&efx->spi_lock);
+	mutex_unlock(&nic_data->spi_lock);
 	return rc;
 }
 
@@ -387,35 +391,67 @@ static struct efx_mtd_ops falcon_mtd_ops = {
 
 static int falcon_mtd_probe(struct efx_nic *efx)
 {
-	struct efx_spi_device *spi = efx->spi_flash;
+	struct falcon_nic_data *nic_data = efx->nic_data;
+	struct efx_spi_device *spi;
 	struct efx_mtd *efx_mtd;
-	int rc;
+	int rc = -ENODEV;
 
 	ASSERT_RTNL();
 
-	if (!spi || spi->size <= FALCON_FLASH_BOOTCODE_START)
-		return -ENODEV;
-
-	efx_mtd = kzalloc(sizeof(*efx_mtd) + sizeof(efx_mtd->part[0]),
-			  GFP_KERNEL);
-	if (!efx_mtd)
-		return -ENOMEM;
-
-	efx_mtd->spi = spi;
-	efx_mtd->name = "flash";
-	efx_mtd->ops = &falcon_mtd_ops;
+	spi = &nic_data->spi_flash;
+	if (efx_spi_present(spi) && spi->size > FALCON_FLASH_BOOTCODE_START) {
+		efx_mtd = kzalloc(sizeof(*efx_mtd) + sizeof(efx_mtd->part[0]),
+				  GFP_KERNEL);
+		if (!efx_mtd)
+			return -ENOMEM;
+
+		efx_mtd->spi = spi;
+		efx_mtd->name = "flash";
+		efx_mtd->ops = &falcon_mtd_ops;
+
+		efx_mtd->n_parts = 1;
+		efx_mtd->part[0].mtd.type = MTD_NORFLASH;
+		efx_mtd->part[0].mtd.flags = MTD_CAP_NORFLASH;
+		efx_mtd->part[0].mtd.size = spi->size - FALCON_FLASH_BOOTCODE_START;
+		efx_mtd->part[0].mtd.erasesize = spi->erase_size;
+		efx_mtd->part[0].offset = FALCON_FLASH_BOOTCODE_START;
+		efx_mtd->part[0].type_name = "sfc_flash_bootrom";
+
+		rc = efx_mtd_probe_device(efx, efx_mtd);
+		if (rc) {
+			kfree(efx_mtd);
+			return rc;
+		}
+	}
 
-	efx_mtd->n_parts = 1;
-	efx_mtd->part[0].mtd.type = MTD_NORFLASH;
-	efx_mtd->part[0].mtd.flags = MTD_CAP_NORFLASH;
-	efx_mtd->part[0].mtd.size = spi->size - FALCON_FLASH_BOOTCODE_START;
-	efx_mtd->part[0].mtd.erasesize = spi->erase_size;
-	efx_mtd->part[0].offset = FALCON_FLASH_BOOTCODE_START;
-	efx_mtd->part[0].type_name = "sfc_flash_bootrom";
+	spi = &nic_data->spi_eeprom;
+	if (efx_spi_present(spi) && spi->size > EFX_EEPROM_BOOTCONFIG_START) {
+		efx_mtd = kzalloc(sizeof(*efx_mtd) + sizeof(efx_mtd->part[0]),
+				  GFP_KERNEL);
+		if (!efx_mtd)
+			return -ENOMEM;
+
+		efx_mtd->spi = spi;
+		efx_mtd->name = "EEPROM";
+		efx_mtd->ops = &falcon_mtd_ops;
+
+		efx_mtd->n_parts = 1;
+		efx_mtd->part[0].mtd.type = MTD_RAM;
+		efx_mtd->part[0].mtd.flags = MTD_CAP_RAM;
+		efx_mtd->part[0].mtd.size =
+			min(spi->size, EFX_EEPROM_BOOTCONFIG_END) -
+			EFX_EEPROM_BOOTCONFIG_START;
+		efx_mtd->part[0].mtd.erasesize = spi->erase_size;
+		efx_mtd->part[0].offset = EFX_EEPROM_BOOTCONFIG_START;
+		efx_mtd->part[0].type_name = "sfc_bootconfig";
+
+		rc = efx_mtd_probe_device(efx, efx_mtd);
+		if (rc) {
+			kfree(efx_mtd);
+			return rc;
+		}
+	}
 
-	rc = efx_mtd_probe_device(efx, efx_mtd);
-	if (rc)
-		kfree(efx_mtd);
 	return rc;
 }
 
diff --git a/drivers/net/sfc/net_driver.h b/drivers/net/sfc/net_driver.h
index b137c889152..bdce66ddf93 100644
--- a/drivers/net/sfc/net_driver.h
+++ b/drivers/net/sfc/net_driver.h
@@ -136,14 +136,19 @@ struct efx_tx_buffer {
  * @efx: The associated Efx NIC
  * @queue: DMA queue number
  * @channel: The associated channel
+ * @core_txq: The networking core TX queue structure
  * @buffer: The software buffer ring
  * @txd: The hardware descriptor ring
  * @ptr_mask: The size of the ring minus 1.
  * @flushed: Used when handling queue flushing
  * @read_count: Current read pointer.
  *	This is the number of buffers that have been removed from both rings.
- * @stopped: Stopped count.
- *	Set if this TX queue is currently stopping its port.
+ * @old_write_count: The value of @write_count when last checked.
+ *	This is here for performance reasons.  The xmit path will
+ *	only get the up-to-date value of @write_count if this
+ *	variable indicates that the queue is empty.  This is to
+ *	avoid cache-line ping-pong between the xmit path and the
+ *	completion path.
  * @insert_count: Current insert pointer
  *	This is the number of buffers that have been added to the
  *	software ring.
@@ -163,13 +168,17 @@ struct efx_tx_buffer {
  * @tso_long_headers: Number of packets with headers too long for standard
  *	blocks
  * @tso_packets: Number of packets via the TSO xmit path
+ * @pushes: Number of times the TX push feature has been used
+ * @empty_read_count: If the completion path has seen the queue as empty
+ *	and the transmission path has not yet checked this, the value of
+ *	@read_count bitwise-added to %EFX_EMPTY_COUNT_VALID; otherwise 0.
  */
 struct efx_tx_queue {
 	/* Members which don't change on the fast path */
 	struct efx_nic *efx ____cacheline_aligned_in_smp;
 	unsigned queue;
 	struct efx_channel *channel;
-	struct efx_nic *nic;
+	struct netdev_queue *core_txq;
 	struct efx_tx_buffer *buffer;
 	struct efx_special_buffer txd;
 	unsigned int ptr_mask;
@@ -177,7 +186,7 @@ struct efx_tx_queue {
 
 	/* Members used mainly on the completion path */
 	unsigned int read_count ____cacheline_aligned_in_smp;
-	int stopped;
+	unsigned int old_write_count;
 
 	/* Members used only on the xmit path */
 	unsigned int insert_count ____cacheline_aligned_in_smp;
@@ -187,6 +196,11 @@ struct efx_tx_queue {
 	unsigned int tso_bursts;
 	unsigned int tso_long_headers;
 	unsigned int tso_packets;
+	unsigned int pushes;
+
+	/* Members shared between paths and sometimes updated */
+	unsigned int empty_read_count ____cacheline_aligned_in_smp;
+#define EFX_EMPTY_COUNT_VALID 0x80000000
 };
 
 /**
@@ -305,7 +319,6 @@ enum efx_rx_alloc_method {
  * @irq_moderation: IRQ moderation value (in hardware ticks)
  * @napi_dev: Net device used with NAPI
  * @napi_str: NAPI control structure
- * @reset_work: Scheduled reset work thread
  * @work_pending: Is work pending via NAPI?
  * @eventq: Event queue buffer
  * @eventq_mask: Event queue pointer mask
@@ -326,8 +339,6 @@ enum efx_rx_alloc_method {
  * @n_rx_overlength: Count of RX_OVERLENGTH errors
  * @n_skbuff_leaks: Count of skbuffs leaked due to RX overrun
  * @rx_queue: RX queue for this channel
- * @tx_stop_count: Core TX queue stop count
- * @tx_stop_lock: Core TX queue stop lock
  * @tx_queue: TX queues for this channel
  */
 struct efx_channel {
@@ -366,10 +377,6 @@ struct efx_channel {
 	bool rx_pkt_csummed;
 
 	struct efx_rx_queue rx_queue;
-
-	atomic_t tx_stop_count;
-	spinlock_t tx_stop_lock;
-
 	struct efx_tx_queue tx_queue[2];
 };
 
@@ -626,10 +633,8 @@ struct efx_filter_state;
  *	Work items do not hold and must not acquire RTNL.
  * @workqueue_name: Name of workqueue
  * @reset_work: Scheduled reset workitem
- * @monitor_work: Hardware monitor workitem
  * @membase_phys: Memory BAR value as physical address
  * @membase: Memory BAR value
- * @biu_lock: BIU (bus interface unit) lock
  * @interrupt_mode: Interrupt mode
  * @irq_rx_adaptive: Adaptive IRQ moderation enabled for RX event queues
  * @irq_rx_moderation: IRQ moderation time for RX event queues
@@ -648,23 +653,14 @@ struct efx_filter_state;
  * @n_tx_channels: Number of channels used for TX
  * @rx_buffer_len: RX buffer length
  * @rx_buffer_order: Order (log2) of number of pages for each RX buffer
+ * @rx_hash_key: Toeplitz hash key for RSS
  * @rx_indir_table: Indirection table for RSS
  * @int_error_count: Number of internal errors seen recently
  * @int_error_expire: Time at which error count will be expired
  * @irq_status: Interrupt status buffer
- * @last_irq_cpu: Last CPU to handle interrupt.
- *	This register is written with the SMP processor ID whenever an
- *	interrupt is handled.  It is used by efx_nic_test_interrupt()
- *	to verify that an interrupt has occurred.
  * @irq_zero_count: Number of legacy IRQs seen with queue flags == 0
  * @fatal_irq_level: IRQ level (bit number) used for serious errors
- * @spi_flash: SPI flash device
- *	This field will be %NULL if no flash device is present (or for Siena).
- * @spi_eeprom: SPI EEPROM device
- *	This field will be %NULL if no EEPROM device is present (or for Siena).
- * @spi_lock: SPI bus lock
  * @mtd_list: List of MTDs attached to the NIC
- * @n_rx_nodesc_drop_cnt: RX no descriptor drop count
  * @nic_data: Hardware dependant state
  * @mac_lock: MAC access lock. Protects @port_enabled, @phy_mode,
  *	@port_inhibited, efx_monitor() and efx_reconfigure_port()
@@ -677,21 +673,14 @@ struct efx_filter_state;
  * @port_initialized: Port initialized?
  * @net_dev: Operating system network device. Consider holding the rtnl lock
  * @rx_checksum_enabled: RX checksumming enabled
- * @mac_stats: MAC statistics. These include all statistics the MACs
- *	can provide.  Generic code converts these into a standard
- *	&struct net_device_stats.
  * @stats_buffer: DMA buffer for statistics
- * @stats_lock: Statistics update lock. Serialises statistics fetches
  * @mac_op: MAC interface
- * @mac_address: Permanent MAC address
  * @phy_type: PHY type
- * @mdio_lock: MDIO lock
  * @phy_op: PHY interface
  * @phy_data: PHY private data (including PHY-specific stats)
  * @mdio: PHY MDIO interface
  * @mdio_bus: PHY MDIO bus ID (only used by Siena)
  * @phy_mode: PHY operating mode. Serialised by @mac_lock.
- * @xmac_poll_required: XMAC link state needs polling
  * @link_advertising: Autonegotiation advertising flags
  * @link_state: Current state of the link
  * @n_link_state_changes: Number of times the link has changed state
@@ -702,10 +691,23 @@ struct efx_filter_state;
  * @loopback_mode: Loopback status
  * @loopback_modes: Supported loopback mode bitmask
  * @loopback_selftest: Offline self-test private state
+ * @monitor_work: Hardware monitor workitem
+ * @biu_lock: BIU (bus interface unit) lock
+ * @last_irq_cpu: Last CPU to handle interrupt.
+ *	This register is written with the SMP processor ID whenever an
+ *	interrupt is handled.  It is used by efx_nic_test_interrupt()
+ *	to verify that an interrupt has occurred.
+ * @n_rx_nodesc_drop_cnt: RX no descriptor drop count
+ * @mac_stats: MAC statistics. These include all statistics the MACs
+ *	can provide.  Generic code converts these into a standard
+ *	&struct net_device_stats.
+ * @stats_lock: Statistics update lock. Serialises statistics fetches
  *
  * This is stored in the private area of the &struct net_device.
  */
 struct efx_nic {
+	/* The following fields should be written very rarely */
+
 	char name[IFNAMSIZ];
 	struct pci_dev *pci_dev;
 	const struct efx_nic_type *type;
@@ -714,10 +716,9 @@ struct efx_nic {
 	struct workqueue_struct *workqueue;
 	char workqueue_name[16];
 	struct work_struct reset_work;
-	struct delayed_work monitor_work;
 	resource_size_t membase_phys;
 	void __iomem *membase;
-	spinlock_t biu_lock;
+
 	enum efx_int_mode interrupt_mode;
 	bool irq_rx_adaptive;
 	unsigned int irq_rx_moderation;
@@ -744,19 +745,13 @@ struct efx_nic {
 	unsigned long int_error_expire;
 
 	struct efx_buffer irq_status;
-	volatile signed int last_irq_cpu;
 	unsigned irq_zero_count;
 	unsigned fatal_irq_level;
 
-	struct efx_spi_device *spi_flash;
-	struct efx_spi_device *spi_eeprom;
-	struct mutex spi_lock;
 #ifdef CONFIG_SFC_MTD
 	struct list_head mtd_list;
 #endif
 
-	unsigned n_rx_nodesc_drop_cnt;
-
 	void *nic_data;
 
 	struct mutex mac_lock;
@@ -768,22 +763,17 @@ struct efx_nic {
 	struct net_device *net_dev;
 	bool rx_checksum_enabled;
 
-	struct efx_mac_stats mac_stats;
 	struct efx_buffer stats_buffer;
-	spinlock_t stats_lock;
 
 	struct efx_mac_operations *mac_op;
-	unsigned char mac_address[ETH_ALEN];
 
 	unsigned int phy_type;
-	struct mutex mdio_lock;
 	struct efx_phy_operations *phy_op;
 	void *phy_data;
 	struct mdio_if_info mdio;
 	unsigned int mdio_bus;
 	enum efx_phy_mode phy_mode;
 
-	bool xmac_poll_required;
 	u32 link_advertising;
 	struct efx_link_state link_state;
 	unsigned int n_link_state_changes;
@@ -799,6 +789,15 @@ struct efx_nic {
 	void *loopback_selftest;
 
 	struct efx_filter_state *filter_state;
+
+	/* The following fields may be written more often */
+
+	struct delayed_work monitor_work ____cacheline_aligned_in_smp;
+	spinlock_t biu_lock;
+	volatile signed int last_irq_cpu;
+	unsigned n_rx_nodesc_drop_cnt;
+	struct efx_mac_stats mac_stats;
+	spinlock_t stats_lock;
 };
 
 static inline int efx_dev_registered(struct efx_nic *efx)
@@ -831,6 +830,7 @@ static inline unsigned int efx_port_num(struct efx_nic *efx)
  *	be called while the controller is uninitialised.
  * @probe_port: Probe the MAC and PHY
  * @remove_port: Free resources allocated by probe_port()
+ * @handle_global_event: Handle a "global" event (may be %NULL)
  * @prepare_flush: Prepare the hardware for flushing the DMA queues
  * @update_stats: Update statistics not provided by event handling
  * @start_stats: Start the regular fetching of statistics
@@ -875,6 +875,7 @@ struct efx_nic_type {
 	int (*reset)(struct efx_nic *efx, enum reset_type method);
 	int (*probe_port)(struct efx_nic *efx);
 	void (*remove_port)(struct efx_nic *efx);
+	bool (*handle_global_event)(struct efx_channel *channel, efx_qword_t *);
 	void (*prepare_flush)(struct efx_nic *efx);
 	void (*update_stats)(struct efx_nic *efx);
 	void (*start_stats)(struct efx_nic *efx);
diff --git a/drivers/net/sfc/nic.c b/drivers/net/sfc/nic.c
index 67cb0c96838..da386599ab6 100644
--- a/drivers/net/sfc/nic.c
+++ b/drivers/net/sfc/nic.c
@@ -362,6 +362,35 @@ static inline void efx_notify_tx_desc(struct efx_tx_queue *tx_queue)
 			FR_AZ_TX_DESC_UPD_DWORD_P0, tx_queue->queue);
 }
 
+/* Write pointer and first descriptor for TX descriptor ring */
+static inline void efx_push_tx_desc(struct efx_tx_queue *tx_queue,
+				    const efx_qword_t *txd)
+{
+	unsigned write_ptr;
+	efx_oword_t reg;
+
+	BUILD_BUG_ON(FRF_AZ_TX_DESC_LBN != 0);
+	BUILD_BUG_ON(FR_AA_TX_DESC_UPD_KER != FR_BZ_TX_DESC_UPD_P0);
+
+	write_ptr = tx_queue->write_count & tx_queue->ptr_mask;
+	EFX_POPULATE_OWORD_2(reg, FRF_AZ_TX_DESC_PUSH_CMD, true,
+			     FRF_AZ_TX_DESC_WPTR, write_ptr);
+	reg.qword[0] = *txd;
+	efx_writeo_page(tx_queue->efx, &reg,
+			FR_BZ_TX_DESC_UPD_P0, tx_queue->queue);
+}
+
+static inline bool
+efx_may_push_tx_desc(struct efx_tx_queue *tx_queue, unsigned int write_count)
+{
+	unsigned empty_read_count = ACCESS_ONCE(tx_queue->empty_read_count);
+
+	if (empty_read_count == 0)
+		return false;
+
+	tx_queue->empty_read_count = 0;
+	return ((empty_read_count ^ write_count) & ~EFX_EMPTY_COUNT_VALID) == 0;
+}
 
 /* For each entry inserted into the software descriptor ring, create a
  * descriptor in the hardware TX descriptor ring (in host memory), and
@@ -373,6 +402,7 @@ void efx_nic_push_buffers(struct efx_tx_queue *tx_queue)
 	struct efx_tx_buffer *buffer;
 	efx_qword_t *txd;
 	unsigned write_ptr;
+	unsigned old_write_count = tx_queue->write_count;
 
 	BUG_ON(tx_queue->write_count == tx_queue->insert_count);
 
@@ -391,7 +421,15 @@ void efx_nic_push_buffers(struct efx_tx_queue *tx_queue)
 	} while (tx_queue->write_count != tx_queue->insert_count);
 
 	wmb(); /* Ensure descriptors are written before they are fetched */
-	efx_notify_tx_desc(tx_queue);
+
+	if (efx_may_push_tx_desc(tx_queue, old_write_count)) {
+		txd = efx_tx_desc(tx_queue,
+				  old_write_count & tx_queue->ptr_mask);
+		efx_push_tx_desc(tx_queue, txd);
+		++tx_queue->pushes;
+	} else {
+		efx_notify_tx_desc(tx_queue);
+	}
 }
 
 /* Allocate hardware resources for a TX queue */
@@ -894,46 +932,6 @@ efx_handle_generated_event(struct efx_channel *channel, efx_qword_t *event)
 			  channel->channel, EFX_QWORD_VAL(*event));
 }
 
-/* Global events are basically PHY events */
-static void
-efx_handle_global_event(struct efx_channel *channel, efx_qword_t *event)
-{
-	struct efx_nic *efx = channel->efx;
-	bool handled = false;
-
-	if (EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_G_PHY0_INTR) ||
-	    EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_XG_PHY0_INTR) ||
-	    EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_XFP_PHY0_INTR)) {
-		/* Ignored */
-		handled = true;
-	}
-
-	if ((efx_nic_rev(efx) >= EFX_REV_FALCON_B0) &&
-	    EFX_QWORD_FIELD(*event, FSF_BB_GLB_EV_XG_MGT_INTR)) {
-		efx->xmac_poll_required = true;
-		handled = true;
-	}
-
-	if (efx_nic_rev(efx) <= EFX_REV_FALCON_A1 ?
-	    EFX_QWORD_FIELD(*event, FSF_AA_GLB_EV_RX_RECOVERY) :
-	    EFX_QWORD_FIELD(*event, FSF_BB_GLB_EV_RX_RECOVERY)) {
-		netif_err(efx, rx_err, efx->net_dev,
-			  "channel %d seen global RX_RESET event. Resetting.\n",
-			  channel->channel);
-
-		atomic_inc(&efx->rx_reset);
-		efx_schedule_reset(efx, EFX_WORKAROUND_6555(efx) ?
-				   RESET_TYPE_RX_RECOVERY : RESET_TYPE_DISABLE);
-		handled = true;
-	}
-
-	if (!handled)
-		netif_err(efx, hw, efx->net_dev,
-			  "channel %d unknown global event "
-			  EFX_QWORD_FMT "\n", channel->channel,
-			  EFX_QWORD_VAL(*event));
-}
-
 static void
 efx_handle_driver_event(struct efx_channel *channel, efx_qword_t *event)
 {
@@ -1050,15 +1048,17 @@ int efx_nic_process_eventq(struct efx_channel *channel, int budget)
 		case FSE_AZ_EV_CODE_DRV_GEN_EV:
 			efx_handle_generated_event(channel, &event);
 			break;
-		case FSE_AZ_EV_CODE_GLOBAL_EV:
-			efx_handle_global_event(channel, &event);
-			break;
 		case FSE_AZ_EV_CODE_DRIVER_EV:
 			efx_handle_driver_event(channel, &event);
 			break;
 		case FSE_CZ_EV_CODE_MCDI_EV:
 			efx_mcdi_process_event(channel, &event);
 			break;
+		case FSE_AZ_EV_CODE_GLOBAL_EV:
+			if (efx->type->handle_global_event &&
+			    efx->type->handle_global_event(channel, &event))
+				break;
+			/* else fall through */
 		default:
 			netif_err(channel->efx, hw, channel->efx->net_dev,
 				  "channel %d unknown event type %d (data "
@@ -1670,7 +1670,7 @@ void efx_nic_init_common(struct efx_nic *efx)
 	EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_RX_SPACER, 0xfe);
 	EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_RX_SPACER_EN, 1);
 	EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_ONE_PKT_PER_Q, 1);
-	EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_PUSH_EN, 0);
+	EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_PUSH_EN, 1);
 	EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_DIS_NON_IP_EV, 1);
 	/* Enable SW_EV to inherit in char driver - assume harmless here */
 	EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_SOFT_EVT_EN, 1);
diff --git a/drivers/net/sfc/nic.h b/drivers/net/sfc/nic.h
index 0438dc98722..eb0586925b5 100644
--- a/drivers/net/sfc/nic.h
+++ b/drivers/net/sfc/nic.h
@@ -15,6 +15,7 @@
 #include "net_driver.h"
 #include "efx.h"
 #include "mcdi.h"
+#include "spi.h"
 
 /*
  * Falcon hardware control
@@ -113,6 +114,11 @@ struct falcon_board {
  * @stats_pending: Is there a pending DMA of MAC statistics.
  * @stats_timer: A timer for regularly fetching MAC statistics.
  * @stats_dma_done: Pointer to the flag which indicates DMA completion.
+ * @spi_flash: SPI flash device
+ * @spi_eeprom: SPI EEPROM device
+ * @spi_lock: SPI bus lock
+ * @mdio_lock: MDIO bus lock
+ * @xmac_poll_required: XMAC link state needs polling
  */
 struct falcon_nic_data {
 	struct pci_dev *pci_dev2;
@@ -121,6 +127,11 @@ struct falcon_nic_data {
 	bool stats_pending;
 	struct timer_list stats_timer;
 	u32 *stats_dma_done;
+	struct efx_spi_device spi_flash;
+	struct efx_spi_device spi_eeprom;
+	struct mutex spi_lock;
+	struct mutex mdio_lock;
+	bool xmac_poll_required;
 };
 
 static inline struct falcon_board *falcon_board(struct efx_nic *efx)
@@ -135,7 +146,6 @@ static inline struct falcon_board *falcon_board(struct efx_nic *efx)
  * @fw_build: Firmware build number
  * @mcdi: Management-Controller-to-Driver Interface
  * @wol_filter_id: Wake-on-LAN packet filter id
- * @ipv6_rss_key: Toeplitz hash key for IPv6 RSS
  */
 struct siena_nic_data {
 	u64 fw_version;
diff --git a/drivers/net/sfc/qt202x_phy.c b/drivers/net/sfc/qt202x_phy.c
index 68813d1d85f..ea3ae008931 100644
--- a/drivers/net/sfc/qt202x_phy.c
+++ b/drivers/net/sfc/qt202x_phy.c
@@ -41,6 +41,8 @@
 #define PCS_UC_STATUS_LBN	0
 #define PCS_UC_STATUS_WIDTH	8
 #define PCS_UC_STATUS_FW_SAVE	0x20
+#define PMA_PMD_MODE_REG	0xc301
+#define PMA_PMD_RXIN_SEL_LBN	6
 #define PMA_PMD_FTX_CTRL2_REG	0xc309
 #define PMA_PMD_FTX_STATIC_LBN	13
 #define PMA_PMD_VEND1_REG	0xc001
@@ -282,6 +284,10 @@ static int qt2025c_select_phy_mode(struct efx_nic *efx)
 	 * slow) reload of the firmware image (the microcontroller's code
 	 * memory is not affected by the microcontroller reset). */
 	efx_mdio_write(efx, 1, 0xc317, 0x00ff);
+	/* PMA/PMD loopback sets RXIN to inverse polarity and the firmware
+	 * restart doesn't reset it. We need to do that ourselves. */
+	efx_mdio_set_flag(efx, 1, PMA_PMD_MODE_REG,
+			  1 << PMA_PMD_RXIN_SEL_LBN, false);
 	efx_mdio_write(efx, 1, 0xc300, 0x0002);
 	msleep(20);
 
diff --git a/drivers/net/sfc/rx.c b/drivers/net/sfc/rx.c
index 6d0959b5158..3925fd62117 100644
--- a/drivers/net/sfc/rx.c
+++ b/drivers/net/sfc/rx.c
@@ -37,7 +37,7 @@
  * This driver supports two methods for allocating and using RX buffers:
  * each RX buffer may be backed by an skb or by an order-n page.
  *
- * When LRO is in use then the second method has a lower overhead,
+ * When GRO is in use then the second method has a lower overhead,
  * since we don't have to allocate then free skbs on reassembled frames.
  *
  * Values:
@@ -50,25 +50,25 @@
  *
  *   - Since pushing and popping descriptors are separated by the rx_queue
  *     size, so the watermarks should be ~rxd_size.
- *   - The performance win by using page-based allocation for LRO is less
- *     than the performance hit of using page-based allocation of non-LRO,
+ *   - The performance win by using page-based allocation for GRO is less
+ *     than the performance hit of using page-based allocation of non-GRO,
  *     so the watermarks should reflect this.
  *
  * Per channel we maintain a single variable, updated by each channel:
  *
- *   rx_alloc_level += (lro_performed ? RX_ALLOC_FACTOR_LRO :
+ *   rx_alloc_level += (gro_performed ? RX_ALLOC_FACTOR_GRO :
  *                      RX_ALLOC_FACTOR_SKB)
  * Per NAPI poll interval, we constrain rx_alloc_level to 0..MAX (which
  * limits the hysteresis), and update the allocation strategy:
  *
- *   rx_alloc_method = (rx_alloc_level > RX_ALLOC_LEVEL_LRO ?
+ *   rx_alloc_method = (rx_alloc_level > RX_ALLOC_LEVEL_GRO ?
  *                      RX_ALLOC_METHOD_PAGE : RX_ALLOC_METHOD_SKB)
  */
 static int rx_alloc_method = RX_ALLOC_METHOD_AUTO;
 
-#define RX_ALLOC_LEVEL_LRO 0x2000
+#define RX_ALLOC_LEVEL_GRO 0x2000
 #define RX_ALLOC_LEVEL_MAX 0x3000
-#define RX_ALLOC_FACTOR_LRO 1
+#define RX_ALLOC_FACTOR_GRO 1
 #define RX_ALLOC_FACTOR_SKB (-2)
 
 /* This is the percentage fill level below which new RX descriptors
@@ -441,19 +441,19 @@ static void efx_rx_packet__check_len(struct efx_rx_queue *rx_queue,
 	efx_rx_queue_channel(rx_queue)->n_rx_overlength++;
 }
 
-/* Pass a received packet up through the generic LRO stack
+/* Pass a received packet up through the generic GRO stack
  *
  * Handles driverlink veto, and passes the fragment up via
- * the appropriate LRO method
+ * the appropriate GRO method
  */
-static void efx_rx_packet_lro(struct efx_channel *channel,
+static void efx_rx_packet_gro(struct efx_channel *channel,
 			      struct efx_rx_buffer *rx_buf,
 			      bool checksummed)
 {
 	struct napi_struct *napi = &channel->napi_str;
 	gro_result_t gro_result;
 
-	/* Pass the skb/page into the LRO engine */
+	/* Pass the skb/page into the GRO engine */
 	if (rx_buf->page) {
 		struct efx_nic *efx = channel->efx;
 		struct page *page = rx_buf->page;
@@ -499,7 +499,7 @@ static void efx_rx_packet_lro(struct efx_channel *channel,
 	if (gro_result == GRO_NORMAL) {
 		channel->rx_alloc_level += RX_ALLOC_FACTOR_SKB;
 	} else if (gro_result != GRO_DROP) {
-		channel->rx_alloc_level += RX_ALLOC_FACTOR_LRO;
+		channel->rx_alloc_level += RX_ALLOC_FACTOR_GRO;
 		channel->irq_mod_score += 2;
 	}
 }
@@ -605,7 +605,7 @@ void __efx_rx_packet(struct efx_channel *channel,
 	}
 
 	if (likely(checksummed || rx_buf->page)) {
-		efx_rx_packet_lro(channel, rx_buf, checksummed);
+		efx_rx_packet_gro(channel, rx_buf, checksummed);
 		return;
 	}
 
@@ -628,7 +628,7 @@ void efx_rx_strategy(struct efx_channel *channel)
 {
 	enum efx_rx_alloc_method method = rx_alloc_method;
 
-	/* Only makes sense to use page based allocation if LRO is enabled */
+	/* Only makes sense to use page based allocation if GRO is enabled */
 	if (!(channel->efx->net_dev->features & NETIF_F_GRO)) {
 		method = RX_ALLOC_METHOD_SKB;
 	} else if (method == RX_ALLOC_METHOD_AUTO) {
@@ -639,7 +639,7 @@ void efx_rx_strategy(struct efx_channel *channel)
 			channel->rx_alloc_level = RX_ALLOC_LEVEL_MAX;
 
 		/* Decide on the allocation method */
-		method = ((channel->rx_alloc_level > RX_ALLOC_LEVEL_LRO) ?
+		method = ((channel->rx_alloc_level > RX_ALLOC_LEVEL_GRO) ?
 			  RX_ALLOC_METHOD_PAGE : RX_ALLOC_METHOD_SKB);
 	}
 
diff --git a/drivers/net/sfc/siena.c b/drivers/net/sfc/siena.c
index 45236f58a25..bf845617644 100644
--- a/drivers/net/sfc/siena.c
+++ b/drivers/net/sfc/siena.c
@@ -194,13 +194,7 @@ static int siena_reset_hw(struct efx_nic *efx, enum reset_type method)
 
 static int siena_probe_nvconfig(struct efx_nic *efx)
 {
-	int rc;
-
-	rc = efx_mcdi_get_board_cfg(efx, efx->mac_address, NULL);
-	if (rc)
-		return rc;
-
-	return 0;
+	return efx_mcdi_get_board_cfg(efx, efx->net_dev->perm_addr, NULL);
 }
 
 static int siena_probe_nic(struct efx_nic *efx)
@@ -562,7 +556,7 @@ static int siena_set_wol(struct efx_nic *efx, u32 type)
 		if (nic_data->wol_filter_id != -1)
 			efx_mcdi_wol_filter_remove(efx,
 						   nic_data->wol_filter_id);
-		rc = efx_mcdi_wol_filter_set_magic(efx, efx->mac_address,
+		rc = efx_mcdi_wol_filter_set_magic(efx, efx->net_dev->dev_addr,
 						   &nic_data->wol_filter_id);
 		if (rc)
 			goto fail;
diff --git a/drivers/net/sfc/spi.h b/drivers/net/sfc/spi.h
index 8bf4fce0813..879b7f6bde3 100644
--- a/drivers/net/sfc/spi.h
+++ b/drivers/net/sfc/spi.h
@@ -61,6 +61,11 @@ struct efx_spi_device {
 	unsigned int block_size;
 };
 
+static inline bool efx_spi_present(const struct efx_spi_device *spi)
+{
+	return spi->size != 0;
+}
+
 int falcon_spi_cmd(struct efx_nic *efx,
 		   const struct efx_spi_device *spi, unsigned int command,
 		   int address, const void* in, void *out, size_t len);
diff --git a/drivers/net/sfc/tenxpress.c b/drivers/net/sfc/tenxpress.c
index 1bc6c48c96e..f102912eba9 100644
--- a/drivers/net/sfc/tenxpress.c
+++ b/drivers/net/sfc/tenxpress.c
@@ -15,9 +15,7 @@
 #include "mdio_10g.h"
 #include "nic.h"
 #include "phy.h"
-#include "regs.h"
 #include "workarounds.h"
-#include "selftest.h"
 
 /* We expect these MMDs to be in the package. */
 #define TENXPRESS_REQUIRED_DEVS (MDIO_DEVS_PMAPMD	| \
diff --git a/drivers/net/sfc/tx.c b/drivers/net/sfc/tx.c
index 11726989fe2..2f5e9da657b 100644
--- a/drivers/net/sfc/tx.c
+++ b/drivers/net/sfc/tx.c
@@ -30,50 +30,6 @@
  */
 #define EFX_TXQ_THRESHOLD(_efx) ((_efx)->txq_entries / 2u)
 
-/* We need to be able to nest calls to netif_tx_stop_queue(), partly
- * because of the 2 hardware queues associated with each core queue,
- * but also so that we can inhibit TX for reasons other than a full
- * hardware queue. */
-void efx_stop_queue(struct efx_channel *channel)
-{
-	struct efx_nic *efx = channel->efx;
-	struct efx_tx_queue *tx_queue = efx_channel_get_tx_queue(channel, 0);
-
-	if (!tx_queue)
-		return;
-
-	spin_lock_bh(&channel->tx_stop_lock);
-	netif_vdbg(efx, tx_queued, efx->net_dev, "stop TX queue\n");
-
-	atomic_inc(&channel->tx_stop_count);
-	netif_tx_stop_queue(
-		netdev_get_tx_queue(efx->net_dev,
-				    tx_queue->queue / EFX_TXQ_TYPES));
-
-	spin_unlock_bh(&channel->tx_stop_lock);
-}
-
-/* Decrement core TX queue stop count and wake it if the count is 0 */
-void efx_wake_queue(struct efx_channel *channel)
-{
-	struct efx_nic *efx = channel->efx;
-	struct efx_tx_queue *tx_queue = efx_channel_get_tx_queue(channel, 0);
-
-	if (!tx_queue)
-		return;
-
-	local_bh_disable();
-	if (atomic_dec_and_lock(&channel->tx_stop_count,
-				&channel->tx_stop_lock)) {
-		netif_vdbg(efx, tx_queued, efx->net_dev, "waking TX queue\n");
-		netif_tx_wake_queue(
-			netdev_get_tx_queue(efx->net_dev,
-					    tx_queue->queue / EFX_TXQ_TYPES));
-		spin_unlock(&channel->tx_stop_lock);
-	}
-	local_bh_enable();
-}
-
 static void efx_dequeue_buffer(struct efx_tx_queue *tx_queue,
 			       struct efx_tx_buffer *buffer)
 {
@@ -234,21 +190,22 @@ netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
 				 * checked.  Update the xmit path's
 				 * copy of read_count.
 				 */
-				++tx_queue->stopped;
+				netif_tx_stop_queue(tx_queue->core_txq);
 				/* This memory barrier protects the
-				 * change of stopped from the access
+				 * change of queue state from the access
 				 * of read_count. */
 				smp_mb();
 				tx_queue->old_read_count =
-					*(volatile unsigned *)
-					&tx_queue->read_count;
+					ACCESS_ONCE(tx_queue->read_count);
 				fill_level = (tx_queue->insert_count
 					      - tx_queue->old_read_count);
 				q_space = efx->txq_entries - 1 - fill_level;
-				if (unlikely(q_space-- <= 0))
-					goto stop;
+				if (unlikely(q_space-- <= 0)) {
+					rc = NETDEV_TX_BUSY;
+					goto unwind;
+				}
 				smp_mb();
-				--tx_queue->stopped;
+				netif_tx_start_queue(tx_queue->core_txq);
 			}
 
 			insert_ptr = tx_queue->insert_count & tx_queue->ptr_mask;
@@ -308,13 +265,6 @@ netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
 
 	/* Mark the packet as transmitted, and free the SKB ourselves */
 	dev_kfree_skb_any(skb);
-	goto unwind;
-
- stop:
-	rc = NETDEV_TX_BUSY;
-
-	if (tx_queue->stopped == 1)
-		efx_stop_queue(tx_queue->channel);
 
  unwind:
 	/* Work backwards until we hit the original insert pointer value */
@@ -407,22 +357,25 @@ void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index)
 	efx_dequeue_buffers(tx_queue, index);
 
 	/* See if we need to restart the netif queue.  This barrier
-	 * separates the update of read_count from the test of
-	 * stopped. */
+	 * separates the update of read_count from the test of the
+	 * queue state. */
 	smp_mb();
-	if (unlikely(tx_queue->stopped) && likely(efx->port_enabled)) {
+	if (unlikely(netif_tx_queue_stopped(tx_queue->core_txq)) &&
+	    likely(efx->port_enabled)) {
 		fill_level = tx_queue->insert_count - tx_queue->read_count;
 		if (fill_level < EFX_TXQ_THRESHOLD(efx)) {
 			EFX_BUG_ON_PARANOID(!efx_dev_registered(efx));
+			netif_tx_wake_queue(tx_queue->core_txq);
+		}
+	}
 
-			/* Do this under netif_tx_lock(), to avoid racing
-			 * with efx_xmit(). */
-			netif_tx_lock(efx->net_dev);
-			if (tx_queue->stopped) {
-				tx_queue->stopped = 0;
-				efx_wake_queue(tx_queue->channel);
-			}
-			netif_tx_unlock(efx->net_dev);
+	/* Check whether the hardware queue is now empty */
+	if ((int)(tx_queue->read_count - tx_queue->old_write_count) >= 0) {
+		tx_queue->old_write_count = ACCESS_ONCE(tx_queue->write_count);
+		if (tx_queue->read_count == tx_queue->old_write_count) {
+			smp_mb();
+			tx_queue->empty_read_count =
+				tx_queue->read_count | EFX_EMPTY_COUNT_VALID;
 		}
 	}
 }
@@ -470,9 +423,10 @@ void efx_init_tx_queue(struct efx_tx_queue *tx_queue)
 
 	tx_queue->insert_count = 0;
 	tx_queue->write_count = 0;
+	tx_queue->old_write_count = 0;
 	tx_queue->read_count = 0;
 	tx_queue->old_read_count = 0;
-	BUG_ON(tx_queue->stopped);
+	tx_queue->empty_read_count = 0 | EFX_EMPTY_COUNT_VALID;
 
 	/* Set up TX descriptor ring */
 	efx_nic_init_tx(tx_queue);
@@ -508,12 +462,6 @@ void efx_fini_tx_queue(struct efx_tx_queue *tx_queue)
 
 	/* Free up TSO header cache */
 	efx_fini_tso(tx_queue);
-
-	/* Release queue's stop on port, if any */
-	if (tx_queue->stopped) {
-		tx_queue->stopped = 0;
-		efx_wake_queue(tx_queue->channel);
-	}
 }
 
 void efx_remove_tx_queue(struct efx_tx_queue *tx_queue)
@@ -755,12 +703,12 @@ static int efx_tx_queue_insert(struct efx_tx_queue *tx_queue,
 			 * since the xmit path last checked.  Update
 			 * the xmit path's copy of read_count.
 			 */
-			++tx_queue->stopped;
+			netif_tx_stop_queue(tx_queue->core_txq);
 			/* This memory barrier protects the change of
-			 * stopped from the access of read_count. */
+			 * queue state from the access of read_count. */
 			smp_mb();
 			tx_queue->old_read_count =
-				*(volatile unsigned *)&tx_queue->read_count;
+				ACCESS_ONCE(tx_queue->read_count);
 			fill_level = (tx_queue->insert_count
 				      - tx_queue->old_read_count);
 			q_space = efx->txq_entries - 1 - fill_level;
@@ -769,7 +717,7 @@ static int efx_tx_queue_insert(struct efx_tx_queue *tx_queue,
 				return 1;
 			}
 			smp_mb();
-			--tx_queue->stopped;
+			netif_tx_start_queue(tx_queue->core_txq);
 		}
 
 		insert_ptr = tx_queue->insert_count & tx_queue->ptr_mask;
@@ -1109,8 +1057,10 @@ static int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue,
 
 	while (1) {
 		rc = tso_fill_packet_with_fragment(tx_queue, skb, &state);
-		if (unlikely(rc))
-			goto stop;
+		if (unlikely(rc)) {
+			rc2 = NETDEV_TX_BUSY;
+			goto unwind;
+		}
 
 		/* Move onto the next fragment? */
 		if (state.in_len == 0) {
@@ -1139,14 +1089,6 @@ static int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue,
 	netif_err(efx, tx_err, efx->net_dev,
 		  "Out of memory for TSO headers, or PCI mapping error\n");
 	dev_kfree_skb_any(skb);
-	goto unwind;
-
- stop:
-	rc2 = NETDEV_TX_BUSY;
-
-	/* Stop the queue if it wasn't stopped before. */
-	if (tx_queue->stopped == 1)
-		efx_stop_queue(tx_queue->channel);
 
  unwind:
 	/* Free the DMA mapping we were in the process of writing out */