Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next

Jeff Kirsher says:

====================
Intel Wired LAN Driver Updates

This series contains updates to i40e only (again).

Jesse provides a fix for when tx_rings structure is NULL and we do not want
to panic. Then refactors the flow control set up and disables L2 flow control
by default.  Provides some trivial fixes as well as prevent compiler warnings.
Then to align to similar behaviour in ixgbe, use the total number of CPUs in
the system to suggest the number of transmit and receive queue pairs.

Shannon provides a i40e ethtool fix to get some more reasonable information
reports back out to the ethtool.  In addition, fixes PF reset after offline
test, where it reorders the test to put the register test last as it is the
only one that needs a reset, and we wait to trigger the reset until after we
clear the testing bit.  Lastly provides basic support for handling suspend
and resume for now, later on Wake-On-LAN support will be added.

Anjali provides changes to tell the stack about our actual number of queues
in order for RFS/RPS/XFS to work correctly.  Then provides several patches to
implement dynamically changing the queue count for the main VSI.  Adds
basic support for get/set channels for RSS so that the number of receive and
transmit queue pair can be changed via ethtool.  Cleans up the use of
rtnl_lock in the reset patch since it runs from a work time.

Neerav Parikh cleans up the VF interface to remove FCoE code as this
feature will not be supported on VF interfaces.

v2:
  - submitted patch 1 to net (since it was a fix needed for net), so dropped
    from this series (this patch will get added to net-next when Dave syncs
    his trees)
  - Dropped patches 4 & 11 from previous submission because of feedback
    received from Ben Hutchings and Sergei Shtylyov.
====================

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

drivers/net/ethernet/intel/i40e/i40e.h

@@ -133,6 +133,7 @@ enum i40e_state_t {
 	__I40E_GLOBAL_RESET_REQUESTED,
 	__I40E_EMP_RESET_REQUESTED,
 	__I40E_FILTER_OVERFLOW_PROMISC,
+	__I40E_SUSPENDED,
 };
 
 enum i40e_interrupt_policy {
@@ -199,6 +200,7 @@ struct i40e_pf {
 	u16 num_tc_qps;            /* num queue pairs per TC */
 	u16 num_lan_qps;           /* num lan queues this pf has set up */
 	u16 num_lan_msix;          /* num queue vectors for the base pf vsi */
+	int queues_left;           /* queues left unclaimed */
 	u16 rss_size;              /* num queues in the RSS array */
 	u16 rss_size_max;          /* HW defined max RSS queues */
 	u16 fdir_pf_filter_count;  /* num of guaranteed filters for this PF */
@@ -501,6 +503,7 @@ int i40e_up(struct i40e_vsi *vsi);
 void i40e_down(struct i40e_vsi *vsi);
 extern const char i40e_driver_name[];
 extern const char i40e_driver_version_str[];
+void i40e_do_reset_safe(struct i40e_pf *pf, u32 reset_flags);
 void i40e_do_reset(struct i40e_pf *pf, u32 reset_flags);
 void i40e_update_stats(struct i40e_vsi *vsi);
 void i40e_update_eth_stats(struct i40e_vsi *vsi);
@@ -530,6 +533,7 @@ struct i40e_vsi *i40e_vsi_setup(struct i40e_pf *pf, u8 type,
 int i40e_vsi_release(struct i40e_vsi *vsi);
 struct i40e_vsi *i40e_vsi_lookup(struct i40e_pf *pf, enum i40e_vsi_type type,
 				 struct i40e_vsi *start_vsi);
+int i40e_reconfig_rss_queues(struct i40e_pf *pf, int queue_count);
 struct i40e_veb *i40e_veb_setup(struct i40e_pf *pf, u16 flags, u16 uplink_seid,
 				u16 downlink_seid, u8 enabled_tc);
 void i40e_veb_release(struct i40e_veb *veb);

drivers/net/ethernet/intel/i40e/i40e_debugfs.c

@@ -1462,19 +1462,19 @@ static ssize_t i40e_dbg_command_write(struct file *filp,
 		}
 	} else if (strncmp(cmd_buf, "pfr", 3) == 0) {
 		dev_info(&pf->pdev->dev, "forcing PFR\n");
-		i40e_do_reset(pf, (1 << __I40E_PF_RESET_REQUESTED));
+		i40e_do_reset_safe(pf, (1 << __I40E_PF_RESET_REQUESTED));
 
 	} else if (strncmp(cmd_buf, "corer", 5) == 0) {
 		dev_info(&pf->pdev->dev, "forcing CoreR\n");
-		i40e_do_reset(pf, (1 << __I40E_CORE_RESET_REQUESTED));
+		i40e_do_reset_safe(pf, (1 << __I40E_CORE_RESET_REQUESTED));
 
 	} else if (strncmp(cmd_buf, "globr", 5) == 0) {
 		dev_info(&pf->pdev->dev, "forcing GlobR\n");
-		i40e_do_reset(pf, (1 << __I40E_GLOBAL_RESET_REQUESTED));
+		i40e_do_reset_safe(pf, (1 << __I40E_GLOBAL_RESET_REQUESTED));
 
 	} else if (strncmp(cmd_buf, "empr", 4) == 0) {
 		dev_info(&pf->pdev->dev, "forcing EMPR\n");
-		i40e_do_reset(pf, (1 << __I40E_EMP_RESET_REQUESTED));
+		i40e_do_reset_safe(pf, (1 << __I40E_EMP_RESET_REQUESTED));
 
 	} else if (strncmp(cmd_buf, "read", 4) == 0) {
 		u32 address;

drivers/net/ethernet/intel/i40e/i40e_diag.c

@@ -119,7 +119,7 @@ i40e_status i40e_diag_eeprom_test(struct i40e_hw *hw)
 
 	/* read NVM control word and if NVM valid, validate EEPROM checksum*/
 	ret_code = i40e_read_nvm_word(hw, I40E_SR_NVM_CONTROL_WORD, &reg_val);
-	if ((!ret_code) &&
+	if (!ret_code &&
 	    ((reg_val & I40E_SR_CONTROL_WORD_1_MASK) ==
 	     (0x01 << I40E_SR_CONTROL_WORD_1_SHIFT))) {
 		ret_code = i40e_validate_nvm_checksum(hw, NULL);

drivers/net/ethernet/intel/i40e/i40e_ethtool.c

@@ -193,32 +193,48 @@ static int i40e_get_settings(struct net_device *netdev,
 		ecmd->supported = SUPPORTED_10000baseKR_Full;
 		ecmd->advertising = ADVERTISED_10000baseKR_Full;
 		break;
-	case I40E_PHY_TYPE_10GBASE_T:
 	default:
-		ecmd->supported = SUPPORTED_10000baseT_Full;
-		ecmd->advertising = ADVERTISED_10000baseT_Full;
+		if (i40e_is_40G_device(hw->device_id)) {
+			ecmd->supported = SUPPORTED_40000baseSR4_Full;
+			ecmd->advertising = ADVERTISED_40000baseSR4_Full;
+		} else {
+			ecmd->supported = SUPPORTED_10000baseT_Full;
+			ecmd->advertising = ADVERTISED_10000baseT_Full;
+		}
 		break;
 	}
 
-	/* for now just say autoneg all the time */
 	ecmd->supported |= SUPPORTED_Autoneg;
+	ecmd->advertising |= ADVERTISED_Autoneg;
+	ecmd->autoneg = ((hw_link_info->an_info & I40E_AQ_AN_COMPLETED) ?
+			  AUTONEG_ENABLE : AUTONEG_DISABLE);
 
-	if (hw->phy.media_type == I40E_MEDIA_TYPE_BACKPLANE) {
+	switch (hw->phy.media_type) {
+	case I40E_MEDIA_TYPE_BACKPLANE:
 		ecmd->supported |= SUPPORTED_Backplane;
 		ecmd->advertising |= ADVERTISED_Backplane;
 		ecmd->port = PORT_NONE;
-	} else if (hw->phy.media_type == I40E_MEDIA_TYPE_BASET) {
+		break;
+	case I40E_MEDIA_TYPE_BASET:
 		ecmd->supported |= SUPPORTED_TP;
 		ecmd->advertising |= ADVERTISED_TP;
 		ecmd->port = PORT_TP;
-	} else if (hw->phy.media_type == I40E_MEDIA_TYPE_DA) {
+		break;
+	case I40E_MEDIA_TYPE_DA:
+	case I40E_MEDIA_TYPE_CX4:
 		ecmd->supported |= SUPPORTED_FIBRE;
 		ecmd->advertising |= ADVERTISED_FIBRE;
 		ecmd->port = PORT_DA;
-	} else {
+		break;
+	case I40E_MEDIA_TYPE_FIBER:
 		ecmd->supported |= SUPPORTED_FIBRE;
 		ecmd->advertising |= ADVERTISED_FIBRE;
 		ecmd->port = PORT_FIBRE;
+		break;
+	case I40E_MEDIA_TYPE_UNKNOWN:
+	default:
+		ecmd->port = PORT_OTHER;
+		break;
 	}
 
 	ecmd->transceiver = XCVR_EXTERNAL;
@@ -260,12 +276,14 @@ static void i40e_get_pauseparam(struct net_device *netdev,
 		((hw_link_info->an_info & I40E_AQ_AN_COMPLETED) ?
 		  AUTONEG_ENABLE : AUTONEG_DISABLE);
 
-	pause->rx_pause = 0;
-	pause->tx_pause = 0;
-	if (hw_link_info->an_info & I40E_AQ_LINK_PAUSE_RX)
+	if (hw->fc.current_mode == I40E_FC_RX_PAUSE) {
+		pause->rx_pause = 1;
+	} else if (hw->fc.current_mode == I40E_FC_TX_PAUSE) {
+		pause->tx_pause = 1;
+	} else if (hw->fc.current_mode == I40E_FC_FULL) {
 		pause->rx_pause = 1;
-	if (hw_link_info->an_info & I40E_AQ_LINK_PAUSE_TX)
 		pause->tx_pause = 1;
+	}
 }
 
 static u32 i40e_get_msglevel(struct net_device *netdev)
@@ -732,7 +750,6 @@ static int i40e_reg_test(struct net_device *netdev, u64 *data)
 	netif_info(pf, hw, netdev, "register test\n");
 	*data = i40e_diag_reg_test(&pf->hw);
 
-	i40e_do_reset(pf, (1 << __I40E_PF_RESET_REQUESTED));
 	return *data;
 }
 
@@ -780,20 +797,18 @@ static void i40e_diag_test(struct net_device *netdev,
 	struct i40e_netdev_priv *np = netdev_priv(netdev);
 	struct i40e_pf *pf = np->vsi->back;
 
-	set_bit(__I40E_TESTING, &pf->state);
 	if (eth_test->flags == ETH_TEST_FL_OFFLINE) {
 		/* Offline tests */
 		netif_info(pf, drv, netdev, "offline testing starting\n");
 
+		set_bit(__I40E_TESTING, &pf->state);
+
 		/* Link test performed before hardware reset
 		 * so autoneg doesn't interfere with test result
 		 */
 		if (i40e_link_test(netdev, &data[I40E_ETH_TEST_LINK]))
 			eth_test->flags |= ETH_TEST_FL_FAILED;
 
-		if (i40e_reg_test(netdev, &data[I40E_ETH_TEST_REG]))
-			eth_test->flags |= ETH_TEST_FL_FAILED;
-
 		if (i40e_eeprom_test(netdev, &data[I40E_ETH_TEST_EEPROM]))
 			eth_test->flags |= ETH_TEST_FL_FAILED;
 
@@ -803,6 +818,12 @@ static void i40e_diag_test(struct net_device *netdev,
 		if (i40e_loopback_test(netdev, &data[I40E_ETH_TEST_LOOPBACK]))
 			eth_test->flags |= ETH_TEST_FL_FAILED;
 
+		/* run reg test last, a reset is required after it */
+		if (i40e_reg_test(netdev, &data[I40E_ETH_TEST_REG]))
+			eth_test->flags |= ETH_TEST_FL_FAILED;
+
+		clear_bit(__I40E_TESTING, &pf->state);
+		i40e_do_reset(pf, (1 << __I40E_PF_RESET_REQUESTED));
 	} else {
 		/* Online tests */
 		netif_info(pf, drv, netdev, "online testing starting\n");
@@ -816,7 +837,6 @@ static void i40e_diag_test(struct net_device *netdev,
 		data[I40E_ETH_TEST_INTR] = 0;
 		data[I40E_ETH_TEST_LOOPBACK] = 0;
 	}
-	clear_bit(__I40E_TESTING, &pf->state);
 
 	netif_info(pf, drv, netdev, "testing finished\n");
 }
@@ -1452,6 +1472,94 @@ static int i40e_set_rxnfc(struct net_device *netdev, struct ethtool_rxnfc *cmd)
 	return ret;
 }
 
+/**
+ * i40e_max_channels - get Max number of combined channels supported
+ * @vsi: vsi pointer
+ **/
+static unsigned int i40e_max_channels(struct i40e_vsi *vsi)
+{
+	/* TODO: This code assumes DCB and FD is disabled for now. */
+	return vsi->alloc_queue_pairs;
+}
+
+/**
+ * i40e_get_channels - Get the current channels enabled and max supported etc.
+ * @netdev: network interface device structure
+ * @ch: ethtool channels structure
+ *
+ * We don't support separate tx and rx queues as channels. The other count
+ * represents how many queues are being used for control. max_combined counts
+ * how many queue pairs we can support. They may not be mapped 1 to 1 with
+ * q_vectors since we support a lot more queue pairs than q_vectors.
+ **/
+static void i40e_get_channels(struct net_device *dev,
+			       struct ethtool_channels *ch)
+{
+	struct i40e_netdev_priv *np = netdev_priv(dev);
+	struct i40e_vsi *vsi = np->vsi;
+	struct i40e_pf *pf = vsi->back;
+
+	/* report maximum channels */
+	ch->max_combined = i40e_max_channels(vsi);
+
+	/* report info for other vector */
+	ch->other_count = (pf->flags & I40E_FLAG_FDIR_ENABLED) ? 1 : 0;
+	ch->max_other = ch->other_count;
+
+	/* Note: This code assumes DCB is disabled for now. */
+	ch->combined_count = vsi->num_queue_pairs;
+}
+
+/**
+ * i40e_set_channels - Set the new channels count.
+ * @netdev: network interface device structure
+ * @ch: ethtool channels structure
+ *
+ * The new channels count may not be the same as requested by the user
+ * since it gets rounded down to a power of 2 value.
+ **/
+static int i40e_set_channels(struct net_device *dev,
+			      struct ethtool_channels *ch)
+{
+	struct i40e_netdev_priv *np = netdev_priv(dev);
+	unsigned int count = ch->combined_count;
+	struct i40e_vsi *vsi = np->vsi;
+	struct i40e_pf *pf = vsi->back;
+	int new_count;
+
+	/* We do not support setting channels for any other VSI at present */
+	if (vsi->type != I40E_VSI_MAIN)
+		return -EINVAL;
+
+	/* verify they are not requesting separate vectors */
+	if (!count || ch->rx_count || ch->tx_count)
+		return -EINVAL;
+
+	/* verify other_count has not changed */
+	if (ch->other_count != ((pf->flags & I40E_FLAG_FDIR_ENABLED) ? 1 : 0))
+		return -EINVAL;
+
+	/* verify the number of channels does not exceed hardware limits */
+	if (count > i40e_max_channels(vsi))
+		return -EINVAL;
+
+	/* update feature limits from largest to smallest supported values */
+	/* TODO: Flow director limit, DCB etc */
+
+	/* cap RSS limit */
+	if (count > pf->rss_size_max)
+		count = pf->rss_size_max;
+
+	/* use rss_reconfig to rebuild with new queue count and update traffic
+	 * class queue mapping
+	 */
+	new_count = i40e_reconfig_rss_queues(pf, count);
+	if (new_count > 1)
+		return 0;
+	else
+		return -EINVAL;
+}
+
 static const struct ethtool_ops i40e_ethtool_ops = {
 	.get_settings		= i40e_get_settings,
 	.get_drvinfo		= i40e_get_drvinfo,
@@ -1476,6 +1584,8 @@ static const struct ethtool_ops i40e_ethtool_ops = {
 	.get_ethtool_stats	= i40e_get_ethtool_stats,
 	.get_coalesce		= i40e_get_coalesce,
 	.set_coalesce		= i40e_set_coalesce,
+	.get_channels		= i40e_get_channels,
+	.set_channels		= i40e_set_channels,
 	.get_ts_info		= i40e_get_ts_info,
 };
 

drivers/net/ethernet/intel/i40e/i40e_hmc.c

@@ -47,10 +47,10 @@ i40e_status i40e_add_sd_table_entry(struct i40e_hw *hw,
 					      u64 direct_mode_sz)
 {
 	enum i40e_memory_type mem_type __attribute__((unused));
-	i40e_status ret_code = 0;
 	struct i40e_hmc_sd_entry *sd_entry;
 	bool dma_mem_alloc_done = false;
 	struct i40e_dma_mem mem;
+	i40e_status ret_code;
 	u64 alloc_len;
 
 	if (NULL == hmc_info->sd_table.sd_entry) {

drivers/net/ethernet/intel/i40e/i40e_main.c

@@ -48,7 +48,7 @@ static void i40e_vsi_reinit_locked(struct i40e_vsi *vsi);
 static void i40e_handle_reset_warning(struct i40e_pf *pf);
 static int i40e_add_vsi(struct i40e_vsi *vsi);
 static int i40e_add_veb(struct i40e_veb *veb, struct i40e_vsi *vsi);
-static int i40e_setup_pf_switch(struct i40e_pf *pf);
+static int i40e_setup_pf_switch(struct i40e_pf *pf, bool reinit);
 static int i40e_setup_misc_vector(struct i40e_pf *pf);
 static void i40e_determine_queue_usage(struct i40e_pf *pf);
 static int i40e_setup_pf_filter_control(struct i40e_pf *pf);
@@ -354,6 +354,9 @@ static struct rtnl_link_stats64 *i40e_get_netdev_stats_struct(
 	struct rtnl_link_stats64 *vsi_stats = i40e_get_vsi_stats_struct(vsi);
 	int i;
 
+	if (test_bit(__I40E_DOWN, &vsi->state))
+		return stats;
+
 	rcu_read_lock();
 	for (i = 0; i < vsi->num_queue_pairs; i++) {
 		struct i40e_ring *tx_ring, *rx_ring;
@@ -3966,6 +3969,15 @@ static int i40e_open(struct net_device *netdev)
 	if (err)
 		goto err_setup_rx;
 
+	/* Notify the stack of the actual queue counts. */
+	err = netif_set_real_num_tx_queues(netdev, pf->num_tx_queues);
+	if (err)
+		goto err_set_queues;
+
+	err = netif_set_real_num_rx_queues(netdev, pf->num_rx_queues);
+	if (err)
+		goto err_set_queues;
+
 	err = i40e_up_complete(vsi);
 	if (err)
 		goto err_up_complete;
@@ -3982,6 +3994,7 @@ static int i40e_open(struct net_device *netdev)
 
 err_up_complete:
 	i40e_down(vsi);
+err_set_queues:
 	i40e_vsi_free_irq(vsi);
 err_setup_rx:
 	i40e_vsi_free_rx_resources(vsi);
@@ -4118,6 +4131,19 @@ void i40e_do_reset(struct i40e_pf *pf, u32 reset_flags)
 	}
 }
 
+/**
+ * i40e_do_reset_safe - Protected reset path for userland calls.
+ * @pf: board private structure
+ * @reset_flags: which reset is requested
+ *
+ **/
+void i40e_do_reset_safe(struct i40e_pf *pf, u32 reset_flags)
+{
+	rtnl_lock();
+	i40e_do_reset(pf, reset_flags);
+	rtnl_unlock();
+}
+
 /**
  * i40e_handle_lan_overflow_event - Handler for LAN queue overflow event
  * @pf: board private structure
@@ -4363,6 +4389,7 @@ static void i40e_reset_subtask(struct i40e_pf *pf)
 {
 	u32 reset_flags = 0;
 
+	rtnl_lock();
 	if (test_bit(__I40E_REINIT_REQUESTED, &pf->state)) {
 		reset_flags |= (1 << __I40E_REINIT_REQUESTED);
 		clear_bit(__I40E_REINIT_REQUESTED, &pf->state);
@@ -4385,7 +4412,7 @@ static void i40e_reset_subtask(struct i40e_pf *pf)
 	 */
 	if (test_bit(__I40E_RESET_INTR_RECEIVED, &pf->state)) {
 		i40e_handle_reset_warning(pf);
-		return;
+		goto unlock;
 	}
 
 	/* If we're already down or resetting, just bail */
@@ -4393,6 +4420,9 @@ static void i40e_reset_subtask(struct i40e_pf *pf)
 	    !test_bit(__I40E_DOWN, &pf->state) &&
 	    !test_bit(__I40E_CONFIG_BUSY, &pf->state))
 		i40e_do_reset(pf, reset_flags);
+
+unlock:
+	rtnl_unlock();
 }
 
 /**
@@ -4757,8 +4787,9 @@ static int i40e_prep_for_reset(struct i40e_pf *pf)
 /**
  * i40e_reset_and_rebuild - reset and rebuid using a saved config
  * @pf: board private structure
+ * @reinit: if the Main VSI needs to re-initialized.
  **/
-static void i40e_reset_and_rebuild(struct i40e_pf *pf)
+static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit)
 {
 	struct i40e_driver_version dv;
 	struct i40e_hw *hw = &pf->hw;
@@ -4806,7 +4837,7 @@ static void i40e_reset_and_rebuild(struct i40e_pf *pf)
 	}
 
 	/* do basic switch setup */
-	ret = i40e_setup_pf_switch(pf);
+	ret = i40e_setup_pf_switch(pf, reinit);
 	if (ret)
 		goto end_core_reset;
 
@@ -4897,7 +4928,7 @@ static void i40e_handle_reset_warning(struct i40e_pf *pf)
 
 	ret = i40e_prep_for_reset(pf);
 	if (!ret)
-		i40e_reset_and_rebuild(pf);
+		i40e_reset_and_rebuild(pf, false);
 }
 
 /**
@@ -5078,11 +5109,12 @@ static int i40e_set_num_rings_in_vsi(struct i40e_vsi *vsi)
 /**
  * i40e_vsi_alloc_arrays - Allocate queue and vector pointer arrays for the vsi
  * @type: VSI pointer
+ * @alloc_qvectors: a bool to specify if q_vectors need to be allocated.
  *
  * On error: returns error code (negative)
  * On success: returns 0
  **/
-static int i40e_vsi_alloc_arrays(struct i40e_vsi *vsi)
+static int i40e_vsi_alloc_arrays(struct i40e_vsi *vsi, bool alloc_qvectors)
 {
 	int size;
 	int ret = 0;
@@ -5094,12 +5126,14 @@ static int i40e_vsi_alloc_arrays(struct i40e_vsi *vsi)
 		return -ENOMEM;
 	vsi->rx_rings = &vsi->tx_rings[vsi->alloc_queue_pairs];
 
-	/* allocate memory for q_vector pointers */
-	size = sizeof(struct i40e_q_vectors *) * vsi->num_q_vectors;
-	vsi->q_vectors = kzalloc(size, GFP_KERNEL);
-	if (!vsi->q_vectors) {
-		ret = -ENOMEM;
-		goto err_vectors;
+	if (alloc_qvectors) {
+		/* allocate memory for q_vector pointers */
+		size = sizeof(struct i40e_q_vectors *) * vsi->num_q_vectors;
+		vsi->q_vectors = kzalloc(size, GFP_KERNEL);
+		if (!vsi->q_vectors) {
+			ret = -ENOMEM;
+			goto err_vectors;
+		}
 	}
 	return ret;
 
@@ -5169,7 +5203,7 @@ static int i40e_vsi_mem_alloc(struct i40e_pf *pf, enum i40e_vsi_type type)
 	if (ret)
 		goto err_rings;
 
-	ret = i40e_vsi_alloc_arrays(vsi);
+	ret = i40e_vsi_alloc_arrays(vsi, true);
 	if (ret)
 		goto err_rings;
 
@@ -5191,15 +5225,18 @@ unlock_pf:
 /**
  * i40e_vsi_free_arrays - Free queue and vector pointer arrays for the VSI
  * @type: VSI pointer
+ * @free_qvectors: a bool to specify if q_vectors need to be freed.
  *
  * On error: returns error code (negative)
  * On success: returns 0
  **/
-static void i40e_vsi_free_arrays(struct i40e_vsi *vsi)
+static void i40e_vsi_free_arrays(struct i40e_vsi *vsi, bool free_qvectors)
 {
 	/* free the ring and vector containers */
-	kfree(vsi->q_vectors);
-	vsi->q_vectors = NULL;
+	if (free_qvectors) {
+		kfree(vsi->q_vectors);
+		vsi->q_vectors = NULL;
+	}
 	kfree(vsi->tx_rings);
 	vsi->tx_rings = NULL;
 	vsi->rx_rings = NULL;
@@ -5241,7 +5278,7 @@ static int i40e_vsi_clear(struct i40e_vsi *vsi)
 	i40e_put_lump(pf->qp_pile, vsi->base_queue, vsi->idx);
 	i40e_put_lump(pf->irq_pile, vsi->base_vector, vsi->idx);
 
-	i40e_vsi_free_arrays(vsi);
+	i40e_vsi_free_arrays(vsi, true);
 
 	pf->vsi[vsi->idx] = NULL;
 	if (vsi->idx < pf->next_vsi)
@@ -5391,15 +5428,18 @@ static int i40e_init_msix(struct i40e_pf *pf)
 	/* The number of vectors we'll request will be comprised of:
 	 *   - Add 1 for "other" cause for Admin Queue events, etc.
 	 *   - The number of LAN queue pairs
-	 *        already adjusted for the NUMA node
-	 *        assumes symmetric Tx/Rx pairing
+	 *	- Queues being used for RSS.
+	 *		We don't need as many as max_rss_size vectors.
+	 *		use rss_size instead in the calculation since that
+	 *		is governed by number of cpus in the system.
+	 *	- assumes symmetric Tx/Rx pairing
 	 *   - The number of VMDq pairs
 	 * Once we count this up, try the request.
 	 *
 	 * If we can't get what we want, we'll simplify to nearly nothing
 	 * and try again.  If that still fails, we punt.
 	 */
-	pf->num_lan_msix = pf->num_lan_qps;
+	pf->num_lan_msix = pf->num_lan_qps - (pf->rss_size_max - pf->rss_size);
 	pf->num_vmdq_msix = pf->num_vmdq_qps;
 	v_budget = 1 + pf->num_lan_msix;
 	v_budget += (pf->num_vmdq_vsis * pf->num_vmdq_msix);
@@ -5679,6 +5719,42 @@ static int i40e_config_rss(struct i40e_pf *pf)
 	return 0;
 }
 
+/**
+ * i40e_reconfig_rss_queues - change number of queues for rss and rebuild
+ * @pf: board private structure
+ * @queue_count: the requested queue count for rss.
+ *
+ * returns 0 if rss is not enabled, if enabled returns the final rss queue
+ * count which may be different from the requested queue count.
+ **/
+int i40e_reconfig_rss_queues(struct i40e_pf *pf, int queue_count)
+{
+	if (!(pf->flags & I40E_FLAG_RSS_ENABLED))
+		return 0;
+
+	queue_count = min_t(int, queue_count, pf->rss_size_max);
+	queue_count = rounddown_pow_of_two(queue_count);
+
+	if (queue_count != pf->rss_size) {
+		if (pf->queues_left < (queue_count - pf->rss_size)) {
+			dev_info(&pf->pdev->dev,
+				"Not enough queues to do RSS on %d queues: remaining queues %d\n",
+				queue_count, pf->queues_left);
+			return pf->rss_size;
+		}
+		i40e_prep_for_reset(pf);
+
+		pf->num_lan_qps += (queue_count - pf->rss_size);
+		pf->queues_left -= (queue_count - pf->rss_size);
+		pf->rss_size = queue_count;
+
+		i40e_reset_and_rebuild(pf, true);
+		i40e_config_rss(pf);
+	}
+	dev_info(&pf->pdev->dev, "RSS count:  %d\n", pf->rss_size);
+	return pf->rss_size;
+}
+
 /**
  * i40e_sw_init - Initialize general software structures (struct i40e_pf)
  * @pf: board private structure to initialize
@@ -5718,8 +5794,7 @@ static int i40e_sw_init(struct i40e_pf *pf)
 				 pf->hw.func_caps.num_tx_qp);
 	if (pf->hw.func_caps.rss) {
 		pf->flags |= I40E_FLAG_RSS_ENABLED;
-		pf->rss_size = min_t(int, pf->rss_size_max,
-				     nr_cpus_node(numa_node_id()));
+		pf->rss_size = min_t(int, pf->rss_size_max, num_online_cpus());
 	} else {
 		pf->rss_size = 1;
 	}
@@ -5861,7 +5936,7 @@ static int i40e_config_netdev(struct i40e_vsi *vsi)
 	int etherdev_size;
 
 	etherdev_size = sizeof(struct i40e_netdev_priv);
-	netdev = alloc_etherdev_mq(etherdev_size, vsi->num_queue_pairs);
+	netdev = alloc_etherdev_mq(etherdev_size, vsi->alloc_queue_pairs);
 	if (!netdev)
 		return -ENOMEM;
 
@@ -6253,6 +6328,69 @@ vector_setup_out:
 	return ret;
 }
 
+/**
+ * i40e_vsi_reinit_setup - return and reallocate resources for a VSI
+ * @vsi: pointer to the vsi.
+ *
+ * This re-allocates a vsi's queue resources.
+ *
+ * Returns pointer to the successfully allocated and configured VSI sw struct
+ * on success, otherwise returns NULL on failure.
+ **/
+static struct i40e_vsi *i40e_vsi_reinit_setup(struct i40e_vsi *vsi)
+{
+	struct i40e_pf *pf = vsi->back;
+	u8 enabled_tc;
+	int ret;
+
+	i40e_put_lump(pf->qp_pile, vsi->base_queue, vsi->idx);
+	i40e_vsi_clear_rings(vsi);
+
+	i40e_vsi_free_arrays(vsi, false);
+	i40e_set_num_rings_in_vsi(vsi);
+	ret = i40e_vsi_alloc_arrays(vsi, false);
+	if (ret)
+		goto err_vsi;
+
+	ret = i40e_get_lump(pf, pf->qp_pile, vsi->alloc_queue_pairs, vsi->idx);
+	if (ret < 0) {
+		dev_info(&pf->pdev->dev, "VSI %d get_lump failed %d\n",
+			 vsi->seid, ret);
+		goto err_vsi;
+	}
+	vsi->base_queue = ret;
+
+	/* Update the FW view of the VSI. Force a reset of TC and queue
+	 * layout configurations.
+	 */
+	enabled_tc = pf->vsi[pf->lan_vsi]->tc_config.enabled_tc;
+	pf->vsi[pf->lan_vsi]->tc_config.enabled_tc = 0;
+	pf->vsi[pf->lan_vsi]->seid = pf->main_vsi_seid;
+	i40e_vsi_config_tc(pf->vsi[pf->lan_vsi], enabled_tc);
+
+	/* assign it some queues */
+	ret = i40e_alloc_rings(vsi);
+	if (ret)
+		goto err_rings;
+
+	/* map all of the rings to the q_vectors */
+	i40e_vsi_map_rings_to_vectors(vsi);
+	return vsi;
+
+err_rings:
+	i40e_vsi_free_q_vectors(vsi);
+	if (vsi->netdev_registered) {
+		vsi->netdev_registered = false;
+		unregister_netdev(vsi->netdev);
+		free_netdev(vsi->netdev);
+		vsi->netdev = NULL;
+	}
+	i40e_aq_delete_element(&pf->hw, vsi->seid, NULL);
+err_vsi:
+	i40e_vsi_clear(vsi);
+	return NULL;
+}
+
 /**
  * i40e_vsi_setup - Set up a VSI by a given type
  * @pf: board private structure
@@ -6896,11 +7034,13 @@ int i40e_fetch_switch_configuration(struct i40e_pf *pf, bool printconfig)
 /**
  * i40e_setup_pf_switch - Setup the HW switch on startup or after reset
  * @pf: board private structure
+ * @reinit: if the Main VSI needs to re-initialized.
  *
  * Returns 0 on success, negative value on failure
  **/
-static int i40e_setup_pf_switch(struct i40e_pf *pf)
+static int i40e_setup_pf_switch(struct i40e_pf *pf, bool reinit)
 {
+	u32 rxfc = 0, txfc = 0, rxfc_reg;
 	int ret;
 
 	/* find out what's out there already */
@@ -6920,7 +7060,7 @@ static int i40e_setup_pf_switch(struct i40e_pf *pf)
 		i40e_fdir_setup(pf);
 
 	/* first time setup */
-	if (pf->lan_vsi == I40E_NO_VSI) {
+	if (pf->lan_vsi == I40E_NO_VSI || reinit) {
 		struct i40e_vsi *vsi = NULL;
 		u16 uplink_seid;
 
@@ -6931,8 +7071,10 @@ static int i40e_setup_pf_switch(struct i40e_pf *pf)
 			uplink_seid = pf->veb[pf->lan_veb]->seid;
 		else
 			uplink_seid = pf->mac_seid;
-
-		vsi = i40e_vsi_setup(pf, I40E_VSI_MAIN, uplink_seid, 0);
+		if (pf->lan_vsi == I40E_NO_VSI)
+			vsi = i40e_vsi_setup(pf, I40E_VSI_MAIN, uplink_seid, 0);
+		else if (reinit)
+			vsi = i40e_vsi_reinit_setup(pf->vsi[pf->lan_vsi]);
 		if (!vsi) {
 			dev_info(&pf->pdev->dev, "setup of MAIN VSI failed\n");
 			i40e_fdir_teardown(pf);
@@ -6971,20 +7113,65 @@ static int i40e_setup_pf_switch(struct i40e_pf *pf)
 	i40e_aq_get_link_info(&pf->hw, true, NULL, NULL);
 	i40e_link_event(pf);
 
-	/* Initialize user-specifics link properties */
+	/* Initialize user-specific link properties */
 	pf->fc_autoneg_status = ((pf->hw.phy.link_info.an_info &
 				  I40E_AQ_AN_COMPLETED) ? true : false);
-	pf->hw.fc.requested_mode = I40E_FC_DEFAULT;
-	if (pf->hw.phy.link_info.an_info &
-	   (I40E_AQ_LINK_PAUSE_TX | I40E_AQ_LINK_PAUSE_RX))
+	/* requested_mode is set in probe or by ethtool */
+	if (!pf->fc_autoneg_status)
+		goto no_autoneg;
+
+	if ((pf->hw.phy.link_info.an_info & I40E_AQ_LINK_PAUSE_TX) &&
+	    (pf->hw.phy.link_info.an_info & I40E_AQ_LINK_PAUSE_RX))
 		pf->hw.fc.current_mode = I40E_FC_FULL;
 	else if (pf->hw.phy.link_info.an_info & I40E_AQ_LINK_PAUSE_TX)
 		pf->hw.fc.current_mode = I40E_FC_TX_PAUSE;
 	else if (pf->hw.phy.link_info.an_info & I40E_AQ_LINK_PAUSE_RX)
 		pf->hw.fc.current_mode = I40E_FC_RX_PAUSE;
 	else
-		pf->hw.fc.current_mode = I40E_FC_DEFAULT;
+		pf->hw.fc.current_mode = I40E_FC_NONE;
+
+	/* sync the flow control settings with the auto-neg values */
+	switch (pf->hw.fc.current_mode) {
+	case I40E_FC_FULL:
+		txfc = 1;
+		rxfc = 1;
+		break;
+	case I40E_FC_TX_PAUSE:
+		txfc = 1;
+		rxfc = 0;
+		break;
+	case I40E_FC_RX_PAUSE:
+		txfc = 0;
+		rxfc = 1;
+		break;
+	case I40E_FC_NONE:
+	case I40E_FC_DEFAULT:
+		txfc = 0;
+		rxfc = 0;
+		break;
+	case I40E_FC_PFC:
+		/* TBD */
+		break;
+	/* no default case, we have to handle all possibilities here */
+	}
+
+	wr32(&pf->hw, I40E_PRTDCB_FCCFG, txfc << I40E_PRTDCB_FCCFG_TFCE_SHIFT);
+
+	rxfc_reg = rd32(&pf->hw, I40E_PRTDCB_MFLCN) &
+		   ~I40E_PRTDCB_MFLCN_RFCE_MASK;
+	rxfc_reg |= (rxfc << I40E_PRTDCB_MFLCN_RFCE_SHIFT);
 
+	wr32(&pf->hw, I40E_PRTDCB_MFLCN, rxfc_reg);
+
+	goto fc_complete;
+
+no_autoneg:
+	/* disable L2 flow control, user can turn it on if they wish */
+	wr32(&pf->hw, I40E_PRTDCB_FCCFG, 0);
+	wr32(&pf->hw, I40E_PRTDCB_MFLCN, rd32(&pf->hw, I40E_PRTDCB_MFLCN) &
+					 ~I40E_PRTDCB_MFLCN_RFCE_MASK);
+
+fc_complete:
 	return ret;
 }
 
@@ -6998,7 +7185,7 @@ static u16 i40e_set_rss_size(struct i40e_pf *pf, int queues_left)
 	int num_tc0;
 
 	num_tc0 = min_t(int, queues_left, pf->rss_size_max);
-	num_tc0 = min_t(int, num_tc0, nr_cpus_node(numa_node_id()));
+	num_tc0 = min_t(int, num_tc0, num_online_cpus());
 	num_tc0 = rounddown_pow_of_two(num_tc0);
 
 	return num_tc0;
@@ -7049,7 +7236,7 @@ static void i40e_determine_queue_usage(struct i40e_pf *pf)
 		pf->rss_size = i40e_set_rss_size(pf, queues_left);
 
 		queues_left -= pf->rss_size;
-		pf->num_lan_qps = pf->rss_size;
+		pf->num_lan_qps = pf->rss_size_max;
 
 	} else if (pf->flags & I40E_FLAG_RSS_ENABLED	  &&
 		   !(pf->flags & I40E_FLAG_FDIR_ENABLED)  &&
@@ -7068,7 +7255,7 @@ static void i40e_determine_queue_usage(struct i40e_pf *pf)
 			return;
 		}
 
-		pf->num_lan_qps = pf->rss_size + accum_tc_size;
+		pf->num_lan_qps = pf->rss_size_max + accum_tc_size;
 
 	} else if (pf->flags & I40E_FLAG_RSS_ENABLED   &&
 		  (pf->flags & I40E_FLAG_FDIR_ENABLED) &&
@@ -7084,7 +7271,7 @@ static void i40e_determine_queue_usage(struct i40e_pf *pf)
 			return;
 		}
 
-		pf->num_lan_qps = pf->rss_size;
+		pf->num_lan_qps = pf->rss_size_max;
 
 	} else if (pf->flags & I40E_FLAG_RSS_ENABLED   &&
 		  (pf->flags & I40E_FLAG_FDIR_ENABLED) &&
@@ -7104,7 +7291,7 @@ static void i40e_determine_queue_usage(struct i40e_pf *pf)
 			return;
 		}
 
-		pf->num_lan_qps = pf->rss_size + accum_tc_size;
+		pf->num_lan_qps = pf->rss_size_max + accum_tc_size;
 
 	} else {
 		dev_info(&pf->pdev->dev,
@@ -7126,6 +7313,7 @@ static void i40e_determine_queue_usage(struct i40e_pf *pf)
 		queues_left -= (pf->num_vmdq_vsis * pf->num_vmdq_qps);
 	}
 
+	pf->queues_left = queues_left;
 	return;
 }
 
@@ -7278,6 +7466,9 @@ static int i40e_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		goto err_pf_reset;
 	}
 
+	/* set up a default setting for link flow control */
+	pf->hw.fc.requested_mode = I40E_FC_NONE;
+
 	err = i40e_init_adminq(hw);
 	dev_info(&pdev->dev, "%s\n", i40e_fw_version_str(hw));
 	if (((hw->nvm.version & I40E_NVM_VERSION_HI_MASK)
@@ -7355,7 +7546,7 @@ static int i40e_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		goto err_switch_setup;
 	}
 
-	err = i40e_setup_pf_switch(pf);
+	err = i40e_setup_pf_switch(pf, false);
 	if (err) {
 		dev_info(&pdev->dev, "setup_pf_switch failed: %d\n", err);
 		goto err_vsis;
@@ -7558,7 +7749,11 @@ static pci_ers_result_t i40e_pci_error_detected(struct pci_dev *pdev,
 	dev_info(&pdev->dev, "%s: error %d\n", __func__, error);
 
 	/* shutdown all operations */
-	i40e_pf_quiesce_all_vsi(pf);
+	if (!test_bit(__I40E_SUSPENDED, &pf->state)) {
+		rtnl_lock();
+		i40e_prep_for_reset(pf);
+		rtnl_unlock();
+	}
 
 	/* Request a slot reset */
 	return PCI_ERS_RESULT_NEED_RESET;
@@ -7621,9 +7816,95 @@ static void i40e_pci_error_resume(struct pci_dev *pdev)
 	struct i40e_pf *pf = pci_get_drvdata(pdev);
 
 	dev_info(&pdev->dev, "%s\n", __func__);
+	if (test_bit(__I40E_SUSPENDED, &pf->state))
+		return;
+
+	rtnl_lock();
 	i40e_handle_reset_warning(pf);
+	rtnl_lock();
+}
+
+/**
+ * i40e_shutdown - PCI callback for shutting down
+ * @pdev: PCI device information struct
+ **/
+static void i40e_shutdown(struct pci_dev *pdev)
+{
+	struct i40e_pf *pf = pci_get_drvdata(pdev);
+
+	set_bit(__I40E_SUSPENDED, &pf->state);
+	set_bit(__I40E_DOWN, &pf->state);
+	rtnl_lock();
+	i40e_prep_for_reset(pf);
+	rtnl_unlock();
+
+	if (system_state == SYSTEM_POWER_OFF) {
+		pci_wake_from_d3(pdev, false);    /* No WoL support yet */
+		pci_set_power_state(pdev, PCI_D3hot);
+	}
+}
+
+#ifdef CONFIG_PM
+/**
+ * i40e_suspend - PCI callback for moving to D3
+ * @pdev: PCI device information struct
+ **/
+static int i40e_suspend(struct pci_dev *pdev, pm_message_t state)
+{
+	struct i40e_pf *pf = pci_get_drvdata(pdev);
+
+	set_bit(__I40E_SUSPENDED, &pf->state);
+	set_bit(__I40E_DOWN, &pf->state);
+	rtnl_lock();
+	i40e_prep_for_reset(pf);
+	rtnl_unlock();
+
+	pci_wake_from_d3(pdev, false);    /* No WoL support yet */
+	pci_set_power_state(pdev, PCI_D3hot);
+
+	return 0;
+}
+
+/**
+ * i40e_resume - PCI callback for waking up from D3
+ * @pdev: PCI device information struct
+ **/
+static int i40e_resume(struct pci_dev *pdev)
+{
+	struct i40e_pf *pf = pci_get_drvdata(pdev);
+	u32 err;
+
+	pci_set_power_state(pdev, PCI_D0);
+	pci_restore_state(pdev);
+	/* pci_restore_state() clears dev->state_saves, so
+	 * call pci_save_state() again to restore it.
+	 */
+	pci_save_state(pdev);
+
+	err = pci_enable_device_mem(pdev);
+	if (err) {
+		dev_err(&pdev->dev,
+			"%s: Cannot enable PCI device from suspend\n",
+			__func__);
+		return err;
+	}
+	pci_set_master(pdev);
+
+	/* no wakeup events while running */
+	pci_wake_from_d3(pdev, false);
+
+	/* handling the reset will rebuild the device state */
+	if (test_and_clear_bit(__I40E_SUSPENDED, &pf->state)) {
+		clear_bit(__I40E_DOWN, &pf->state);
+		rtnl_lock();
+		i40e_reset_and_rebuild(pf, false);
+		rtnl_unlock();
+	}
+
+	return 0;
 }
 
+#endif
 static const struct pci_error_handlers i40e_err_handler = {
 	.error_detected = i40e_pci_error_detected,
 	.slot_reset = i40e_pci_error_slot_reset,
@@ -7635,6 +7916,11 @@ static struct pci_driver i40e_driver = {
 	.id_table = i40e_pci_tbl,
 	.probe    = i40e_probe,
 	.remove   = i40e_remove,
+#ifdef CONFIG_PM
+	.suspend  = i40e_suspend,
+	.resume   = i40e_resume,
+#endif
+	.shutdown = i40e_shutdown,
 	.err_handler = &i40e_err_handler,
 	.sriov_configure = i40e_pci_sriov_configure,
 };

drivers/net/ethernet/intel/i40e/i40e_type.h

@@ -49,6 +49,10 @@
 #define I40E_VF_DEVICE_ID		0x154C
 #define I40E_VF_HV_DEVICE_ID		0x1571
 
+#define i40e_is_40G_device(d)		((d) == I40E_QSFP_A_DEVICE_ID  || \
+					 (d) == I40E_QSFP_B_DEVICE_ID  || \
+					 (d) == I40E_QSFP_C_DEVICE_ID)
+
 #define I40E_FW_API_VERSION_MAJOR  0x0001
 #define I40E_FW_API_VERSION_MINOR  0x0000
 

drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c

@@ -521,6 +521,7 @@ static int i40e_alloc_vsi_res(struct i40e_vf *vf, enum i40e_vsi_type type)
 		f = i40e_add_filter(vsi, vf->default_lan_addr.addr,
 				    0, true, false);
 	}
+
 	if (!f) {
 		dev_err(&pf->pdev->dev, "Unable to add ucast filter\n");
 		ret = -ENOMEM;
@@ -763,6 +764,7 @@ static void i40e_free_vf_res(struct i40e_vf *vf)
 		vf->lan_vsi_index = 0;
 		vf->lan_vsi_id = 0;
 	}
+
 	/* reset some of the state varibles keeping
 	 * track of the resources
 	 */
@@ -1776,30 +1778,6 @@ error_param:
 	return i40e_vc_send_resp_to_vf(vf, I40E_VIRTCHNL_OP_DEL_VLAN, aq_ret);
 }
 
-/**
- * i40e_vc_fcoe_msg
- * @vf: pointer to the vf info
- * @msg: pointer to the msg buffer
- * @msglen: msg length
- *
- * called from the vf for the fcoe msgs
- **/
-static int i40e_vc_fcoe_msg(struct i40e_vf *vf, u8 *msg, u16 msglen)
-{
-	i40e_status aq_ret = 0;
-
-	if (!test_bit(I40E_VF_STAT_ACTIVE, &vf->vf_states) ||
-	    !test_bit(I40E_VF_STAT_FCOEENA, &vf->vf_states)) {
-		aq_ret = I40E_ERR_PARAM;
-		goto error_param;
-	}
-	aq_ret = I40E_ERR_NOT_IMPLEMENTED;
-
-error_param:
-	/* send the response to the vf */
-	return i40e_vc_send_resp_to_vf(vf, I40E_VIRTCHNL_OP_FCOE, aq_ret);
-}
-
 /**
  * i40e_vc_validate_vf_msg
  * @vf: pointer to the vf info
@@ -1973,9 +1951,6 @@ int i40e_vc_process_vf_msg(struct i40e_pf *pf, u16 vf_id, u32 v_opcode,
 	case I40E_VIRTCHNL_OP_GET_STATS:
 		ret = i40e_vc_get_stats_msg(vf, msg, msglen);
 		break;
-	case I40E_VIRTCHNL_OP_FCOE:
-		ret = i40e_vc_fcoe_msg(vf, msg, msglen);
-		break;
 	case I40E_VIRTCHNL_OP_UNKNOWN:
 	default:
 		dev_err(&pf->pdev->dev,