[net-next v2 00/15][pull request] Intel Wired LAN Driver Updates

[net-next v2 00/15][pull request] Intel Wired LAN Driver Updates

[Jeff Kirsher]

This series contains updates to i40e, i40evf, e1000e, ixgbe and ixgbevf.

Mitch adds support for the VF link state ndo which allows the PF driver
to control the virtual link state of the VF devices.  Added
support for viewing and modifying RSS hash options and RSS hash look-up
table programming through ethtool for i40evf.  Fixed complaint about
the use of min() where min_t() should be used in i40evf.

Anjali adds support for ethtool -k option for NTUPLE control for i40e.

Elizabeth cleans up and refactors i40e_open() to separate out the VSI
code into its own i40e_vsi_open().

Jesse enables the hardware feature head write back to avoid updating the
descriptor ring by marking each descriptor with a DD bit and instead
writes a memory location with an update to where the driver should clean
up to in i40e and i40evf.  Reduces context descriptors for i40e/i40evf
since we do not need context descriptors for every packet, only for
TSO or timesync.

Dan Carpenter fixes a potential array underflow in i40e_vc_process_vf_msg().

Dave fixes an e1000e hardware unit hang where the check for pending Tx work
when link is lost was mistakenly moved to be done only when link is first
detected to be lost.  Fixed a problem with poor network performance on
certain silicon in e1000e when configured for 100M HDX performance.

Carolyn adds register defines needed for time sync functions and the code
to call the updated defines.

Jacob adds the ixgbe function for writing PCI config word and checks
whether the adapter has been removed first.

Mark adds the bit __IXGBEVF_REMOVING to indicate that the module is being
removed because the __IXGBEVF_DOWN bit had been overloaded for this
purpose, but leads to trouble.  ixgbevf_down function can now prevent
multiple executions by doing test_and_set_bit on __IXGBEVF_DOWN.

v2:
- dropped patch Mitch's patch "i40evf: correctly program RSS HLUT table"
  based on feedback from Ben Hutchings so that Mitch can re-work the
  patch solution

The following are changes since commit 8cfad496c4257441710735ccef622f3829870164:
  ieee802154: properly unshare skbs in ieee802154 *_rcv functions
and are available in the git repository at:
  git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next master

Anjali Singhai Jain (1):
  i40e: Patch to enable Ethtool/netdev feature flag for NTUPLE control

Carolyn Wyborny (1):
  igb: Add register defines needed for time sync functions

Catherine Sullivan (1):
  i40e/i40evf: Bump build versions

Dan Carpenter (1):
  i40e: potential array underflow in i40e_vc_process_vf_msg()

David Ertman (2):
  e1000e: Fix Hardware Unit Hang
  e1000e: Fix Explicitly set Transmit Control Register

Elizabeth Kappler (1):
  i40e: Refactor and cleanup i40e_open(), adding i40e_vsi_open()

Jacob Keller (1):
  ixgbe: add ixgbe_write_pci_cfg_word with ixgbe_removed check

Jesse Brandeburg (2):
  i40e/i40evf: enable hardware feature head write back
  i40e/i40evf: reduce context descriptors

Mark Rustad (2):
  ixgbevf: Indicate removal state explicitly
  ixgbevf: Protect ixgbevf_down with __IXGBEVF_DOWN bit

Mitch Williams (3):
  i40e: support VF link state ndo
  i40evf: correctly program RSS HLUT table
  i40evf: use min_t

 drivers/net/ethernet/intel/e1000e/netdev.c         | 42 +++++++---
 drivers/net/ethernet/intel/i40e/i40e.h             |  4 +-
 drivers/net/ethernet/intel/i40e/i40e_main.c        | 89 ++++++++++++++++++---
 drivers/net/ethernet/intel/i40e/i40e_txrx.c        | 48 ++++++++++--
 drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 91 ++++++++++++++++++++--
 drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h |  4 +
 drivers/net/ethernet/intel/i40evf/i40e_txrx.c      | 48 ++++++++++--
 drivers/net/ethernet/intel/i40evf/i40evf_main.c    | 23 +++---
 drivers/net/ethernet/intel/igb/e1000_defines.h     | 70 ++++++++++++++++-
 drivers/net/ethernet/intel/igb/e1000_regs.h        |  9 +++
 drivers/net/ethernet/intel/igb/igb_ptp.c           |  4 +-
 drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c     |  6 +-
 drivers/net/ethernet/intel/ixgbe/ixgbe_common.h    |  1 +
 drivers/net/ethernet/intel/ixgbe/ixgbe_main.c      |  9 +++
 drivers/net/ethernet/intel/ixgbevf/ixgbevf.h       |  5 +-
 drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c  | 18 +++--
 16 files changed, 400 insertions(+), 71 deletions(-)

-- 
1.8.3.1

[net-next v2 01/15] i40e: support VF link state ndo

[Jeff Kirsher]

From: Mitch Williams <mitch.a.williams@intel.com>

This netdev op allows the PF driver to control the virtual link state of
the VF devices. This can be used to deny naughty VF drivers access to
the wire, or to allow VFs (regardless of temperament) to communicate
with each other over the device's internal switch even though external
link is down.

Add the actual ndo function, and modify vc_notify_link_state to check
the link status of each VF before sending a message in the case when
physical link changes state.

Change-ID: Ib5a6924da78c540789f21d26b5e8086d71c29384
Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c        |  1 +
 drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 86 ++++++++++++++++++++--
 drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h |  4 +
 3 files changed, 85 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 3daaf20..9c4f53e 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -6547,6 +6547,7 @@ static const struct net_device_ops i40e_netdev_ops = {
 	.ndo_set_vf_vlan	= i40e_ndo_set_vf_port_vlan,
 	.ndo_set_vf_tx_rate	= i40e_ndo_set_vf_bw,
 	.ndo_get_vf_config	= i40e_ndo_get_vf_config,
+	.ndo_set_vf_link_state	= i40e_ndo_set_vf_link_state,
 #ifdef CONFIG_I40E_VXLAN
 	.ndo_add_vxlan_port	= i40e_add_vxlan_port,
 	.ndo_del_vxlan_port	= i40e_del_vxlan_port,
diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
index 7839343..f018238 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
@@ -1920,15 +1920,28 @@ static void i40e_vc_vf_broadcast(struct i40e_pf *pf,
 void i40e_vc_notify_link_state(struct i40e_pf *pf)
 {
 	struct i40e_virtchnl_pf_event pfe;
+	struct i40e_hw *hw = &pf->hw;
+	struct i40e_vf *vf = pf->vf;
+	struct i40e_link_status *ls = &pf->hw.phy.link_info;
+	int i;
 
 	pfe.event = I40E_VIRTCHNL_EVENT_LINK_CHANGE;
 	pfe.severity = I40E_PF_EVENT_SEVERITY_INFO;
-	pfe.event_data.link_event.link_status =
-	    pf->hw.phy.link_info.link_info & I40E_AQ_LINK_UP;
-	pfe.event_data.link_event.link_speed = pf->hw.phy.link_info.link_speed;
-
-	i40e_vc_vf_broadcast(pf, I40E_VIRTCHNL_OP_EVENT, I40E_SUCCESS,
-			     (u8 *)&pfe, sizeof(struct i40e_virtchnl_pf_event));
+	for (i = 0; i < pf->num_alloc_vfs; i++) {
+		if (vf->link_forced) {
+			pfe.event_data.link_event.link_status = vf->link_up;
+			pfe.event_data.link_event.link_speed =
+				(vf->link_up ? I40E_LINK_SPEED_40GB : 0);
+		} else {
+			pfe.event_data.link_event.link_status =
+				ls->link_info & I40E_AQ_LINK_UP;
+			pfe.event_data.link_event.link_speed = ls->link_speed;
+		}
+		i40e_aq_send_msg_to_vf(hw, vf->vf_id, I40E_VIRTCHNL_OP_EVENT,
+				       0, (u8 *)&pfe, sizeof(pfe),
+				       NULL);
+		vf++;
+	}
 }
 
 /**
@@ -2193,3 +2206,64 @@ int i40e_ndo_get_vf_config(struct net_device *netdev,
 error_param:
 	return ret;
 }
+
+/**
+ * i40e_ndo_set_vf_link_state
+ * @netdev: network interface device structure
+ * @vf_id: vf identifier
+ * @link: required link state
+ *
+ * Set the link state of a specified VF, regardless of physical link state
+ **/
+int i40e_ndo_set_vf_link_state(struct net_device *netdev, int vf_id, int link)
+{
+	struct i40e_netdev_priv *np = netdev_priv(netdev);
+	struct i40e_pf *pf = np->vsi->back;
+	struct i40e_virtchnl_pf_event pfe;
+	struct i40e_hw *hw = &pf->hw;
+	struct i40e_vf *vf;
+	int ret = 0;
+
+	/* validate the request */
+	if (vf_id >= pf->num_alloc_vfs) {
+		dev_err(&pf->pdev->dev, "Invalid VF Identifier %d\n", vf_id);
+		ret = -EINVAL;
+		goto error_out;
+	}
+
+	vf = &(pf->vf[vf_id]);
+
+	pfe.event = I40E_VIRTCHNL_EVENT_LINK_CHANGE;
+	pfe.severity = I40E_PF_EVENT_SEVERITY_INFO;
+
+	switch (link) {
+	case IFLA_VF_LINK_STATE_AUTO:
+		vf->link_forced = false;
+		pfe.event_data.link_event.link_status =
+			pf->hw.phy.link_info.link_info & I40E_AQ_LINK_UP;
+		pfe.event_data.link_event.link_speed =
+			pf->hw.phy.link_info.link_speed;
+		break;
+	case IFLA_VF_LINK_STATE_ENABLE:
+		vf->link_forced = true;
+		vf->link_up = true;
+		pfe.event_data.link_event.link_status = true;
+		pfe.event_data.link_event.link_speed = I40E_LINK_SPEED_40GB;
+		break;
+	case IFLA_VF_LINK_STATE_DISABLE:
+		vf->link_forced = true;
+		vf->link_up = false;
+		pfe.event_data.link_event.link_status = false;
+		pfe.event_data.link_event.link_speed = 0;
+		break;
+	default:
+		ret = -EINVAL;
+		goto error_out;
+	}
+	/* Notify the VF of its new link state */
+	i40e_aq_send_msg_to_vf(hw, vf->vf_id, I40E_VIRTCHNL_OP_EVENT,
+			       0, (u8 *)&pfe, sizeof(pfe), NULL);
+
+error_out:
+	return ret;
+}
diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h
index bedf0ba..389c47f 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h
@@ -98,6 +98,8 @@ struct i40e_vf {
 
 	unsigned long vf_caps;	/* vf's adv. capabilities */
 	unsigned long vf_states;	/* vf's runtime states */
+	bool link_forced;
+	bool link_up;		/* only valid if vf link is forced */
 };
 
 void i40e_free_vfs(struct i40e_pf *pf);
@@ -116,6 +118,8 @@ int i40e_ndo_set_vf_port_vlan(struct net_device *netdev,
 int i40e_ndo_set_vf_bw(struct net_device *netdev, int vf_id, int tx_rate);
 int i40e_ndo_get_vf_config(struct net_device *netdev,
 			   int vf_id, struct ifla_vf_info *ivi);
+int i40e_ndo_set_vf_link_state(struct net_device *netdev, int vf_id, int link);
+
 void i40e_vc_notify_link_state(struct i40e_pf *pf);
 void i40e_vc_notify_reset(struct i40e_pf *pf);
 
-- 
1.8.3.1

[net-next v2 02/15] i40evf: correctly program RSS HLUT table

[Jeff Kirsher]

From: Mitch Williams <mitch.a.williams@intel.com>

The HLUT programming loop in in i40evf_configure_rss was a) overly-
complicated, and b) just plain broken. Most of the entries ended up being
not written at all, so most of the flows ended up at queue zero.

Refactor the HLUT programming loop to simply walk through the registers
and write four values to each one, incrementing through the number of
available queues.

Change-ID: I75766179bc67e4e997187794f3144e28c83fd00d
Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40evf/i40evf_main.c | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
index d62e27f..7cb0cda 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
@@ -1414,6 +1414,7 @@ restart_watchdog:
 	schedule_work(&adapter->adminq_task);
 }
 
+#define NEXT_QUEUE(_j) (++_j >= adapter->vsi_res->num_queue_pairs ? _j = 0 : _j)
 /**
  * i40evf_configure_rss - Prepare for RSS if used
  * @adapter: board private structure
@@ -1444,15 +1445,13 @@ static void i40evf_configure_rss(struct i40evf_adapter *adapter)
 	wr32(hw, I40E_VFQF_HENA(1), (u32)(hena >> 32));
 
 	/* Populate the LUT with max no. of queues in round robin fashion */
-	for (i = 0, j = 0; i < I40E_VFQF_HLUT_MAX_INDEX; i++, j++) {
-		if (j == adapter->vsi_res->num_queue_pairs)
-			j = 0;
-		/* lut = 4-byte sliding window of 4 lut entries */
-		lut = (lut << 8) | (j &
-			 ((0x1 << 8) - 1));
-		/* On i = 3, we have 4 entries in lut; write to the register */
-		if ((i & 3) == 3)
-			wr32(hw, I40E_VFQF_HLUT(i >> 2), lut);
+	j = adapter->vsi_res->num_queue_pairs;
+	for (i = 0; i <= I40E_VFQF_HLUT_MAX_INDEX; i++) {
+		lut = NEXT_QUEUE(j);
+		lut |= NEXT_QUEUE(j) << 8;
+		lut |= NEXT_QUEUE(j) << 16;
+		lut |= NEXT_QUEUE(j) << 24;
+		wr32(hw, I40E_VFQF_HLUT(i), lut);
 	}
 	i40e_flush(hw);
 }
-- 
1.8.3.1

[net-next v2 03/15] i40evf: use min_t

[Jeff Kirsher]

From: Mitch Williams <mitch.a.williams@intel.com>

Checkpatch complained in an earlier patch about using min(), but that
change would have been completely unrelated to the point of that patch.
So fix it here.

Change-ID: I2cd87b39cfd406850d283b88f259757a6bcd14cd
Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40evf/i40evf_main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
index 7cb0cda..963387c 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
@@ -1140,8 +1140,8 @@ static int i40evf_set_interrupt_capability(struct i40evf_adapter *adapter)
 	 * than CPU's.  So let's be conservative and only ask for
 	 * (roughly) twice the number of vectors as there are CPU's.
 	 */
-	v_budget = min(pairs, (int)(num_online_cpus() * 2)) + NONQ_VECS;
-	v_budget = min(v_budget, (int)adapter->vf_res->max_vectors);
+	v_budget = min_t(int, pairs, (int)(num_online_cpus() * 2)) + NONQ_VECS;
+	v_budget = min_t(int, v_budget, (int)adapter->vf_res->max_vectors);
 
 	/* A failure in MSI-X entry allocation isn't fatal, but it does
 	 * mean we disable MSI-X capabilities of the adapter.
-- 
1.8.3.1

[net-next v2 04/15] i40e: Patch to enable Ethtool/netdev feature flag for NTUPLE control

[Jeff Kirsher]

From: Anjali Singhai Jain <anjali.singhai@intel.com>

This enables option '-k/-K' in ethtool for NTUPLE control.
NTUPLE control requires a reset, to take effect. When the feature is
turned off, the SW list of stored FD SB filters gets cleaned up.

Change-ID: I9d564b67a10d4afa11de3b320d601c3d2e6edc1f
Signed-off-by: Anjali Singhai Jain <anjali.singhai@intel.com>
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Tested-by: Kavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e.h      |  1 +
 drivers/net/ethernet/intel/i40e/i40e_main.c | 41 +++++++++++++++++++++++++++++
 2 files changed, 42 insertions(+)

diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index bd1b469..ac04112 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -558,6 +558,7 @@ int i40e_add_del_fdir(struct i40e_vsi *vsi,
 		      struct i40e_fdir_filter *input, bool add);
 void i40e_fdir_check_and_reenable(struct i40e_pf *pf);
 int i40e_get_current_fd_count(struct i40e_pf *pf);
+bool i40e_set_ntuple(struct i40e_pf *pf, netdev_features_t features);
 void i40e_set_ethtool_ops(struct net_device *netdev);
 struct i40e_mac_filter *i40e_add_filter(struct i40e_vsi *vsi,
 					u8 *macaddr, s16 vlan,
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 9c4f53e..5015dad 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -6409,6 +6409,39 @@ sw_init_done:
 }
 
 /**
+ * i40e_set_ntuple - set the ntuple feature flag and take action
+ * @pf: board private structure to initialize
+ * @features: the feature set that the stack is suggesting
+ *
+ * returns a bool to indicate if reset needs to happen
+ **/
+bool i40e_set_ntuple(struct i40e_pf *pf, netdev_features_t features)
+{
+	bool need_reset = false;
+
+	/* Check if Flow Director n-tuple support was enabled or disabled.  If
+	 * the state changed, we need to reset.
+	 */
+	if (features & NETIF_F_NTUPLE) {
+		/* Enable filters and mark for reset */
+		if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
+			need_reset = true;
+		pf->flags |= I40E_FLAG_FD_SB_ENABLED;
+	} else {
+		/* turn off filters, mark for reset and clear SW filter list */
+		if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
+			need_reset = true;
+			i40e_fdir_filter_exit(pf);
+		}
+		pf->flags &= ~I40E_FLAG_FD_SB_ENABLED;
+		/* if ATR was disabled it can be re-enabled. */
+		if (!(pf->flags & I40E_FLAG_FD_ATR_ENABLED))
+			pf->flags |= I40E_FLAG_FD_ATR_ENABLED;
+	}
+	return need_reset;
+}
+
+/**
  * i40e_set_features - set the netdev feature flags
  * @netdev: ptr to the netdev being adjusted
  * @features: the feature set that the stack is suggesting
@@ -6418,12 +6451,19 @@ static int i40e_set_features(struct net_device *netdev,
 {
 	struct i40e_netdev_priv *np = netdev_priv(netdev);
 	struct i40e_vsi *vsi = np->vsi;
+	struct i40e_pf *pf = vsi->back;
+	bool need_reset;
 
 	if (features & NETIF_F_HW_VLAN_CTAG_RX)
 		i40e_vlan_stripping_enable(vsi);
 	else
 		i40e_vlan_stripping_disable(vsi);
 
+	need_reset = i40e_set_ntuple(pf, features);
+
+	if (need_reset)
+		i40e_do_reset(pf, (1 << __I40E_PF_RESET_REQUESTED));
+
 	return 0;
 }
 
@@ -6595,6 +6635,7 @@ static int i40e_config_netdev(struct i40e_vsi *vsi)
 			   NETIF_F_TSO		       |
 			   NETIF_F_TSO6		       |
 			   NETIF_F_RXCSUM	       |
+			   NETIF_F_NTUPLE	       |
 			   NETIF_F_RXHASH	       |
 			   0;
 
-- 
1.8.3.1

[net-next v2 05/15] i40e: Refactor and cleanup i40e_open(), adding i40e_vsi_open()

[Jeff Kirsher]

From: Elizabeth Kappler <elizabeth.m.kappler@intel.com>

This patch cleans up and moves a portion of i40e_open to i40e_vsi_open,
in order to have a shorter vsi_open function that does only that.

Change-ID: I1c418dda94dcfc0eb7d4386a70c330692ef5ecc9
Signed-off-by: Elizabeth Kappler <elizabeth.m.kappler@intel.com>
Signed-off-by: Akeem G Abodunrin <akeem.g.abodunrin@intel.com>
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Tested-by: Kavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e.h      |  3 ++-
 drivers/net/ethernet/intel/i40e/i40e_main.c | 40 +++++++++++++++++++++++------
 2 files changed, 34 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index ac04112..33cd8b6 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -208,7 +208,7 @@ struct i40e_pf {
 	bool fc_autoneg_status;
 
 	u16 eeprom_version;
-	u16 num_vmdq_vsis;         /* num vmdq pools this pf has set up */
+	u16 num_vmdq_vsis;         /* num vmdq vsis this pf has set up */
 	u16 num_vmdq_qps;          /* num queue pairs per vmdq pool */
 	u16 num_vmdq_msix;         /* num queue vectors per vmdq pool */
 	u16 num_req_vfs;           /* num vfs requested for this vf */
@@ -597,6 +597,7 @@ void i40e_irq_dynamic_enable(struct i40e_vsi *vsi, int vector);
 void i40e_irq_dynamic_disable_icr0(struct i40e_pf *pf);
 void i40e_irq_dynamic_enable_icr0(struct i40e_pf *pf);
 int i40e_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd);
+int i40e_vsi_open(struct i40e_vsi *vsi);
 void i40e_vlan_stripping_disable(struct i40e_vsi *vsi);
 int i40e_vsi_add_vlan(struct i40e_vsi *vsi, s16 vid);
 int i40e_vsi_kill_vlan(struct i40e_vsi *vsi, s16 vid);
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 5015dad..661cd42 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -4235,7 +4235,6 @@ static int i40e_open(struct net_device *netdev)
 	struct i40e_netdev_priv *np = netdev_priv(netdev);
 	struct i40e_vsi *vsi = np->vsi;
 	struct i40e_pf *pf = vsi->back;
-	char int_name[IFNAMSIZ];
 	int err;
 
 	/* disallow open during test */
@@ -4244,6 +4243,31 @@ static int i40e_open(struct net_device *netdev)
 
 	netif_carrier_off(netdev);
 
+	err = i40e_vsi_open(vsi);
+	if (err)
+		return err;
+
+#ifdef CONFIG_I40E_VXLAN
+	vxlan_get_rx_port(netdev);
+#endif
+
+	return 0;
+}
+
+/**
+ * i40e_vsi_open -
+ * @vsi: the VSI to open
+ *
+ * Finish initialization of the VSI.
+ *
+ * Returns 0 on success, negative value on failure
+ **/
+int i40e_vsi_open(struct i40e_vsi *vsi)
+{
+	struct i40e_pf *pf = vsi->back;
+	char int_name[IFNAMSIZ];
+	int err;
+
 	/* allocate descriptors */
 	err = i40e_vsi_setup_tx_resources(vsi);
 	if (err)
@@ -4256,18 +4280,22 @@ static int i40e_open(struct net_device *netdev)
 	if (err)
 		goto err_setup_rx;
 
+	if (!vsi->netdev) {
+		err = EINVAL;
+		goto err_setup_rx;
+	}
 	snprintf(int_name, sizeof(int_name) - 1, "%s-%s",
-		 dev_driver_string(&pf->pdev->dev), netdev->name);
+		 dev_driver_string(&pf->pdev->dev), vsi->netdev->name);
 	err = i40e_vsi_request_irq(vsi, int_name);
 	if (err)
 		goto err_setup_rx;
 
 	/* Notify the stack of the actual queue counts. */
-	err = netif_set_real_num_tx_queues(netdev, vsi->num_queue_pairs);
+	err = netif_set_real_num_tx_queues(vsi->netdev, vsi->num_queue_pairs);
 	if (err)
 		goto err_set_queues;
 
-	err = netif_set_real_num_rx_queues(netdev, vsi->num_queue_pairs);
+	err = netif_set_real_num_rx_queues(vsi->netdev, vsi->num_queue_pairs);
 	if (err)
 		goto err_set_queues;
 
@@ -4275,10 +4303,6 @@ static int i40e_open(struct net_device *netdev)
 	if (err)
 		goto err_up_complete;
 
-#ifdef CONFIG_I40E_VXLAN
-	vxlan_get_rx_port(netdev);
-#endif
-
 	return 0;
 
 err_up_complete:
-- 
1.8.3.1

[net-next v2 06/15] i40e/i40evf: enable hardware feature head write back

[Jeff Kirsher]

From: Jesse Brandeburg <jesse.brandeburg@intel.com>

The hardware supports a feature to avoid updating the descriptor
ring by marking each descriptor with a DD bit, and instead
writes a memory location with an update to where the driver
should clean up to.  Enable this feature.

Change-ID: I5da4e0681f0b581a6401c950a81808792267fe57
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Tested-by: Kavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c        |  5 +++
 drivers/net/ethernet/intel/i40e/i40e_txrx.c        | 45 +++++++++++++++++++---
 drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c |  3 ++
 drivers/net/ethernet/intel/i40evf/i40e_txrx.c      | 45 +++++++++++++++++++---
 4 files changed, 86 insertions(+), 12 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 661cd42..95433d4 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -2181,6 +2181,11 @@ static int i40e_configure_tx_ring(struct i40e_ring *ring)
 	tx_ctx.fd_ena = !!(vsi->back->flags & (I40E_FLAG_FD_SB_ENABLED |
 					       I40E_FLAG_FD_ATR_ENABLED));
 	tx_ctx.timesync_ena = !!(vsi->back->flags & I40E_FLAG_PTP);
+	/* FDIR VSI tx ring can still use RS bit and writebacks */
+	if (vsi->type != I40E_VSI_FDIR)
+		tx_ctx.head_wb_ena = 1;
+	tx_ctx.head_wb_addr = ring->dma +
+			      (ring->count * sizeof(struct i40e_tx_desc));
 
 	/* As part of VSI creation/update, FW allocates certain
 	 * Tx arbitration queue sets for each TC enabled for
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index 88666ad..9892a69 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -619,6 +619,19 @@ static bool i40e_check_tx_hang(struct i40e_ring *tx_ring)
 }
 
 /**
+ * i40e_get_head - Retrieve head from head writeback
+ * @tx_ring:  tx ring to fetch head of
+ *
+ * Returns value of Tx ring head based on value stored
+ * in head write-back location
+ **/
+static inline u32 i40e_get_head(struct i40e_ring *tx_ring)
+{
+	void *head = (struct i40e_tx_desc *)tx_ring->desc + tx_ring->count;
+	return le32_to_cpu(*(volatile __le32 *)head);
+}
+
+/**
  * i40e_clean_tx_irq - Reclaim resources after transmit completes
  * @tx_ring:  tx ring to clean
  * @budget:   how many cleans we're allowed
@@ -629,6 +642,7 @@ static bool i40e_clean_tx_irq(struct i40e_ring *tx_ring, int budget)
 {
 	u16 i = tx_ring->next_to_clean;
 	struct i40e_tx_buffer *tx_buf;
+	struct i40e_tx_desc *tx_head;
 	struct i40e_tx_desc *tx_desc;
 	unsigned int total_packets = 0;
 	unsigned int total_bytes = 0;
@@ -637,6 +651,8 @@ static bool i40e_clean_tx_irq(struct i40e_ring *tx_ring, int budget)
 	tx_desc = I40E_TX_DESC(tx_ring, i);
 	i -= tx_ring->count;
 
+	tx_head = I40E_TX_DESC(tx_ring, i40e_get_head(tx_ring));
+
 	do {
 		struct i40e_tx_desc *eop_desc = tx_buf->next_to_watch;
 
@@ -647,9 +663,8 @@ static bool i40e_clean_tx_irq(struct i40e_ring *tx_ring, int budget)
 		/* prevent any other reads prior to eop_desc */
 		read_barrier_depends();
 
-		/* if the descriptor isn't done, no work yet to do */
-		if (!(eop_desc->cmd_type_offset_bsz &
-		      cpu_to_le64(I40E_TX_DESC_DTYPE_DESC_DONE)))
+		/* we have caught up to head, no work left to do */
+		if (tx_head == tx_desc)
 			break;
 
 		/* clear next_to_watch to prevent false hangs */
@@ -905,6 +920,10 @@ int i40e_setup_tx_descriptors(struct i40e_ring *tx_ring)
 
 	/* round up to nearest 4K */
 	tx_ring->size = tx_ring->count * sizeof(struct i40e_tx_desc);
+	/* add u32 for head writeback, align after this takes care of
+	 * guaranteeing this is at least one cache line in size
+	 */
+	tx_ring->size += sizeof(u32);
 	tx_ring->size = ALIGN(tx_ring->size, 4096);
 	tx_ring->desc = dma_alloc_coherent(dev, tx_ring->size,
 					   &tx_ring->dma, GFP_KERNEL);
@@ -2042,9 +2061,23 @@ static void i40e_tx_map(struct i40e_ring *tx_ring, struct sk_buff *skb,
 		tx_bi = &tx_ring->tx_bi[i];
 	}
 
-	tx_desc->cmd_type_offset_bsz =
-		build_ctob(td_cmd, td_offset, size, td_tag) |
-		cpu_to_le64((u64)I40E_TXD_CMD << I40E_TXD_QW1_CMD_SHIFT);
+	/* Place RS bit on last descriptor of any packet that spans across the
+	 * 4th descriptor (WB_STRIDE aka 0x3) in a 64B cacheline.
+	 */
+#define WB_STRIDE 0x3
+	if (((i & WB_STRIDE) != WB_STRIDE) &&
+	    (first <= &tx_ring->tx_bi[i]) &&
+	    (first >= &tx_ring->tx_bi[i & ~WB_STRIDE])) {
+		tx_desc->cmd_type_offset_bsz =
+			build_ctob(td_cmd, td_offset, size, td_tag) |
+			cpu_to_le64((u64)I40E_TX_DESC_CMD_EOP <<
+					 I40E_TXD_QW1_CMD_SHIFT);
+	} else {
+		tx_desc->cmd_type_offset_bsz =
+			build_ctob(td_cmd, td_offset, size, td_tag) |
+			cpu_to_le64((u64)I40E_TXD_CMD <<
+					 I40E_TXD_QW1_CMD_SHIFT);
+	}
 
 	netdev_tx_sent_queue(netdev_get_tx_queue(tx_ring->netdev,
 						 tx_ring->queue_index),
diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
index f018238..46fc2c6 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
@@ -230,6 +230,9 @@ static int i40e_config_vsi_tx_queue(struct i40e_vf *vf, u16 vsi_idx,
 	tx_ctx.qlen = info->ring_len;
 	tx_ctx.rdylist = le16_to_cpu(pf->vsi[vsi_idx]->info.qs_handle[0]);
 	tx_ctx.rdylist_act = 0;
+	tx_ctx.head_wb_ena = 1;
+	tx_ctx.head_wb_addr = info->dma_ring_addr +
+			      (info->ring_len * sizeof(struct i40e_tx_desc));
 
 	/* clear the context in the HMC */
 	ret = i40e_clear_lan_tx_queue_context(hw, pf_queue_id);
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index b1d87c6..d4d9842 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -170,6 +170,19 @@ static bool i40e_check_tx_hang(struct i40e_ring *tx_ring)
 }
 
 /**
+ * i40e_get_head - Retrieve head from head writeback
+ * @tx_ring:  tx ring to fetch head of
+ *
+ * Returns value of Tx ring head based on value stored
+ * in head write-back location
+ **/
+static inline u32 i40e_get_head(struct i40e_ring *tx_ring)
+{
+	void *head = (struct i40e_tx_desc *)tx_ring->desc + tx_ring->count;
+	return le32_to_cpu(*(volatile __le32 *)head);
+}
+
+/**
  * i40e_clean_tx_irq - Reclaim resources after transmit completes
  * @tx_ring:  tx ring to clean
  * @budget:   how many cleans we're allowed
@@ -180,6 +193,7 @@ static bool i40e_clean_tx_irq(struct i40e_ring *tx_ring, int budget)
 {
 	u16 i = tx_ring->next_to_clean;
 	struct i40e_tx_buffer *tx_buf;
+	struct i40e_tx_desc *tx_head;
 	struct i40e_tx_desc *tx_desc;
 	unsigned int total_packets = 0;
 	unsigned int total_bytes = 0;
@@ -188,6 +202,8 @@ static bool i40e_clean_tx_irq(struct i40e_ring *tx_ring, int budget)
 	tx_desc = I40E_TX_DESC(tx_ring, i);
 	i -= tx_ring->count;
 
+	tx_head = I40E_TX_DESC(tx_ring, i40e_get_head(tx_ring));
+
 	do {
 		struct i40e_tx_desc *eop_desc = tx_buf->next_to_watch;
 
@@ -198,9 +214,8 @@ static bool i40e_clean_tx_irq(struct i40e_ring *tx_ring, int budget)
 		/* prevent any other reads prior to eop_desc */
 		read_barrier_depends();
 
-		/* if the descriptor isn't done, no work yet to do */
-		if (!(eop_desc->cmd_type_offset_bsz &
-		      cpu_to_le64(I40E_TX_DESC_DTYPE_DESC_DONE)))
+		/* we have caught up to head, no work left to do */
+		if (tx_head == tx_desc)
 			break;
 
 		/* clear next_to_watch to prevent false hangs */
@@ -432,6 +447,10 @@ int i40evf_setup_tx_descriptors(struct i40e_ring *tx_ring)
 
 	/* round up to nearest 4K */
 	tx_ring->size = tx_ring->count * sizeof(struct i40e_tx_desc);
+	/* add u32 for head writeback, align after this takes care of
+	 * guaranteeing this is at least one cache line in size
+	 */
+	tx_ring->size += sizeof(u32);
 	tx_ring->size = ALIGN(tx_ring->size, 4096);
 	tx_ring->desc = dma_alloc_coherent(dev, tx_ring->size,
 					   &tx_ring->dma, GFP_KERNEL);
@@ -1377,9 +1396,23 @@ static void i40e_tx_map(struct i40e_ring *tx_ring, struct sk_buff *skb,
 		tx_bi = &tx_ring->tx_bi[i];
 	}
 
-	tx_desc->cmd_type_offset_bsz =
-		build_ctob(td_cmd, td_offset, size, td_tag) |
-		cpu_to_le64((u64)I40E_TXD_CMD << I40E_TXD_QW1_CMD_SHIFT);
+	/* Place RS bit on last descriptor of any packet that spans across the
+	 * 4th descriptor (WB_STRIDE aka 0x3) in a 64B cacheline.
+	 */
+#define WB_STRIDE 0x3
+	if (((i & WB_STRIDE) != WB_STRIDE) &&
+	    (first <= &tx_ring->tx_bi[i]) &&
+	    (first >= &tx_ring->tx_bi[i & ~WB_STRIDE])) {
+		tx_desc->cmd_type_offset_bsz =
+			build_ctob(td_cmd, td_offset, size, td_tag) |
+			cpu_to_le64((u64)I40E_TX_DESC_CMD_EOP <<
+					 I40E_TXD_QW1_CMD_SHIFT);
+	} else {
+		tx_desc->cmd_type_offset_bsz =
+			build_ctob(td_cmd, td_offset, size, td_tag) |
+			cpu_to_le64((u64)I40E_TXD_CMD <<
+					 I40E_TXD_QW1_CMD_SHIFT);
+	}
 
 	netdev_tx_sent_queue(netdev_get_tx_queue(tx_ring->netdev,
 						 tx_ring->queue_index),
-- 
1.8.3.1

[net-next v2 07/15] i40e/i40evf: reduce context descriptors

[Jeff Kirsher]

From: Jesse Brandeburg <jesse.brandeburg@intel.com>

We don't need context descriptors for every packet, only tso
or timesync.  This fixes a bug in the driver where it would
always add a context even if all the passed in values
to the context descriptor function were 0/default values.

Change-ID: I0101d2b893380707b5c2de61aab3e16d4310e9a1
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Tested-by: Kavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_txrx.c   | 3 ++-
 drivers/net/ethernet/intel/i40evf/i40e_txrx.c | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index 9892a69..752e680 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -1950,7 +1950,8 @@ static void i40e_create_tx_ctx(struct i40e_ring *tx_ring,
 	struct i40e_tx_context_desc *context_desc;
 	int i = tx_ring->next_to_use;
 
-	if (!cd_type_cmd_tso_mss && !cd_tunneling && !cd_l2tag2)
+	if ((cd_type_cmd_tso_mss == I40E_TX_DESC_DTYPE_CONTEXT) &&
+	    !cd_tunneling && !cd_l2tag2)
 		return;
 
 	/* grab the next descriptor */
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index d4d9842..18f174d 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -1285,7 +1285,8 @@ static void i40e_create_tx_ctx(struct i40e_ring *tx_ring,
 	struct i40e_tx_context_desc *context_desc;
 	int i = tx_ring->next_to_use;
 
-	if (!cd_type_cmd_tso_mss && !cd_tunneling && !cd_l2tag2)
+	if ((cd_type_cmd_tso_mss == I40E_TX_DESC_DTYPE_CONTEXT) &&
+	    !cd_tunneling && !cd_l2tag2)
 		return;
 
 	/* grab the next descriptor */
-- 
1.8.3.1

[net-next v2 08/15] i40e: potential array underflow in i40e_vc_process_vf_msg()

[Jeff Kirsher]

From: Dan Carpenter <dan.carpenter@oracle.com>

If "vf_id" is smaller than hw->func_caps.vf_base_id then it leads to
an array underflow of the pf->vf[] array.  This is unlikely to happen
unless the hardware is bad, but it's a small change and it silences a
static checker warning.

Fixes: 7efa84b7abc1 ('i40e: support VFs on PFs other than 0')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
index 46fc2c6..f0dc71b 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
@@ -1774,7 +1774,7 @@ int i40e_vc_process_vf_msg(struct i40e_pf *pf, u16 vf_id, u32 v_opcode,
 			   u32 v_retval, u8 *msg, u16 msglen)
 {
 	struct i40e_hw *hw = &pf->hw;
-	int local_vf_id = vf_id - hw->func_caps.vf_base_id;
+	unsigned int local_vf_id = vf_id - hw->func_caps.vf_base_id;
 	struct i40e_vf *vf;
 	int ret;
 
-- 
1.8.3.1

[net-next v2 09/15] i40e/i40evf: Bump build versions

[Jeff Kirsher]

From: Catherine Sullivan <catherine.sullivan@intel.com>

Bump to version 0.3.36 for i40e and 0.9.16 for i40evf.

Change-ID: I7b4ff97b32d2825181803c03c316381a7608a618
Signed-off-by: Catherine Sullivan <catherine.sullivan@intel.com>
Tested-by: Kavindya Deegala <kavindya.s.deegala@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c     | 2 +-
 drivers/net/ethernet/intel/i40evf/i40evf_main.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 95433d4..1133542 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -38,7 +38,7 @@ static const char i40e_driver_string[] =
 
 #define DRV_VERSION_MAJOR 0
 #define DRV_VERSION_MINOR 3
-#define DRV_VERSION_BUILD 34
+#define DRV_VERSION_BUILD 36
 #define DRV_VERSION __stringify(DRV_VERSION_MAJOR) "." \
 	     __stringify(DRV_VERSION_MINOR) "." \
 	     __stringify(DRV_VERSION_BUILD)    DRV_KERN
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
index 963387c..21934bf 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
@@ -31,7 +31,7 @@ char i40evf_driver_name[] = "i40evf";
 static const char i40evf_driver_string[] =
 	"Intel(R) XL710 X710 Virtual Function Network Driver";
 
-#define DRV_VERSION "0.9.14"
+#define DRV_VERSION "0.9.16"
 const char i40evf_driver_version[] = DRV_VERSION;
 static const char i40evf_copyright[] =
 	"Copyright (c) 2013 - 2014 Intel Corporation.";
-- 
1.8.3.1

[net-next v2 10/15] e1000e: Fix Hardware Unit Hang

[Jeff Kirsher]

From: David Ertman <davidx.m.ertman@intel.com>

The check for pending Tx work when link is lost was mistakenly moved to be
done only when link is first detected to be lost.  It turns out there is a
small window of opportunity for additional Tx work to get queued up shortly
after link is dropped.

Move the check back to the place it was before in the watchdog task.  Put in
additional debug information for other reset paths and a final catch-all for
false hangs in the scheduled function that prints out the hardware hang
message.

Signed-off-by: Dave Ertman <davidx.m.ertman@intel.com>
Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/e1000e/netdev.c | 32 +++++++++++++++++++++---------
 1 file changed, 23 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index eafad41..d577723 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -1090,8 +1090,14 @@ static void e1000_print_hw_hang(struct work_struct *work)
 		adapter->tx_hang_recheck = true;
 		return;
 	}
-	/* Real hang detected */
 	adapter->tx_hang_recheck = false;
+
+	if (er32(TDH(0)) == er32(TDT(0))) {
+		e_dbg("false hang detected, ignoring\n");
+		return;
+	}
+
+	/* Real hang detected */
 	netif_stop_queue(netdev);
 
 	e1e_rphy(hw, MII_BMSR, &phy_status);
@@ -1121,6 +1127,8 @@ static void e1000_print_hw_hang(struct work_struct *work)
 	      eop, jiffies, eop_desc->upper.fields.status, er32(STATUS),
 	      phy_status, phy_1000t_status, phy_ext_status, pci_status);
 
+	e1000e_dump(adapter);
+
 	/* Suggest workaround for known h/w issue */
 	if ((hw->mac.type == e1000_pchlan) && (er32(CTRL) & E1000_CTRL_TFCE))
 		e_err("Try turning off Tx pause (flow control) via ethtool\n");
@@ -4798,6 +4806,7 @@ static void e1000e_check_82574_phy_workaround(struct e1000_adapter *adapter)
 
 	if (adapter->phy_hang_count > 1) {
 		adapter->phy_hang_count = 0;
+		e_dbg("PHY appears hung - resetting\n");
 		schedule_work(&adapter->reset_task);
 	}
 }
@@ -4956,15 +4965,11 @@ static void e1000_watchdog_task(struct work_struct *work)
 				mod_timer(&adapter->phy_info_timer,
 					  round_jiffies(jiffies + 2 * HZ));
 
-			/* The link is lost so the controller stops DMA.
-			 * If there is queued Tx work that cannot be done
-			 * or if on an 8000ES2LAN which requires a Rx packet
-			 * buffer work-around on link down event, reset the
-			 * controller to flush the Tx/Rx packet buffers.
-			 * (Do the reset outside of interrupt context).
+			/* 8000ES2LAN requires a Rx packet buffer work-around
+			 * on link down event; reset the controller to flush
+			 * the Rx packet buffer.
 			 */
-			if ((adapter->flags & FLAG_RX_NEEDS_RESTART) ||
-			    (e1000_desc_unused(tx_ring) + 1 < tx_ring->count))
+			if (adapter->flags & FLAG_RX_NEEDS_RESTART)
 				adapter->flags |= FLAG_RESTART_NOW;
 			else
 				pm_schedule_suspend(netdev->dev.parent,
@@ -4987,6 +4992,15 @@ link_up:
 	adapter->gotc_old = adapter->stats.gotc;
 	spin_unlock(&adapter->stats64_lock);
 
+	/* If the link is lost the controller stops DMA, but
+	 * if there is queued Tx work it cannot be done.  So
+	 * reset the controller to flush the Tx packet buffers.
+	 */
+	if (!netif_carrier_ok(netdev) &&
+	    (e1000_desc_unused(tx_ring) + 1 < tx_ring->count))
+		adapter->flags |= FLAG_RESTART_NOW;
+
+	/* If reset is necessary, do it outside of interrupt context. */
 	if (adapter->flags & FLAG_RESTART_NOW) {
 		schedule_work(&adapter->reset_task);
 		/* return immediately since reset is imminent */
-- 
1.8.3.1

[net-next v2 11/15] e1000e: Fix Explicitly set Transmit Control Register

[Jeff Kirsher]

From: David Ertman <davidx.m.ertman@intel.com>

This patch causes the TCTL to be explicitly set to fix a problem with
poor network performance (throughput) on certain silicon when configured
for 100M HDX performance.

Cc: Todd Fujinaka <todd.fujinaka@intel.com>
Signed-off-by: Dave Ertman <davidx.m.ertman@intel.com>
Acked-by: Bruce W. Allan <bruce.w.allan@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/e1000e/netdev.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index d577723..6bd1832 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -2898,7 +2898,7 @@ static void e1000_configure_tx(struct e1000_adapter *adapter)
 	struct e1000_hw *hw = &adapter->hw;
 	struct e1000_ring *tx_ring = adapter->tx_ring;
 	u64 tdba;
-	u32 tdlen, tarc;
+	u32 tdlen, tctl, tarc;
 
 	/* Setup the HW Tx Head and Tail descriptor pointers */
 	tdba = tx_ring->dma;
@@ -2935,6 +2935,12 @@ static void e1000_configure_tx(struct e1000_adapter *adapter)
 	/* erratum work around: set txdctl the same for both queues */
 	ew32(TXDCTL(1), er32(TXDCTL(0)));
 
+	/* Program the Transmit Control Register */
+	tctl = er32(TCTL);
+	tctl &= ~E1000_TCTL_CT;
+	tctl |= E1000_TCTL_PSP | E1000_TCTL_RTLC |
+		(E1000_COLLISION_THRESHOLD << E1000_CT_SHIFT);
+
 	if (adapter->flags & FLAG_TARC_SPEED_MODE_BIT) {
 		tarc = er32(TARC(0));
 		/* set the speed mode bit, we'll clear it if we're not at
@@ -2965,6 +2971,8 @@ static void e1000_configure_tx(struct e1000_adapter *adapter)
 	/* enable Report Status bit */
 	adapter->txd_cmd |= E1000_TXD_CMD_RS;
 
+	ew32(TCTL, tctl);
+
 	hw->mac.ops.config_collision_dist(hw);
 }
 
-- 
1.8.3.1

[net-next v2 12/15] igb: Add register defines needed for time sync functions

[Jeff Kirsher]

From: Carolyn Wyborny <carolyn.wyborny@intel.com>

This patch adds defines needed for implementing the auxiliary time sync
functions and also changes code to call the updated defines instead of
the old.

Reported-by: Richard Cochran <ricahrdcochran@gmail.com>
Signed-off-by: Carolyn Wyborny <carolyn.wyborny@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/igb/e1000_defines.h | 70 ++++++++++++++++++++++++--
 drivers/net/ethernet/intel/igb/e1000_regs.h    |  9 ++++
 drivers/net/ethernet/intel/igb/igb_ptp.c       |  4 +-
 3 files changed, 78 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/intel/igb/e1000_defines.h b/drivers/net/ethernet/intel/igb/e1000_defines.h
index 393c896..b05bf92 100644
--- a/drivers/net/ethernet/intel/igb/e1000_defines.h
+++ b/drivers/net/ethernet/intel/igb/e1000_defines.h
@@ -43,7 +43,11 @@
 #define E1000_WUFC_BC   0x00000010 /* Broadcast Wakeup Enable */
 
 /* Extended Device Control */
+#define E1000_CTRL_EXT_SDP2_DATA 0x00000040 /* Value of SW Defineable Pin 2 */
 #define E1000_CTRL_EXT_SDP3_DATA 0x00000080 /* Value of SW Defineable Pin 3 */
+#define E1000_CTRL_EXT_SDP2_DIR  0x00000400 /* SDP2 Data direction */
+#define E1000_CTRL_EXT_SDP3_DIR  0x00000800 /* SDP3 Data direction */
+
 /* Physical Func Reset Done Indication */
 #define E1000_CTRL_EXT_PFRSTD    0x00004000
 #define E1000_CTRL_EXT_LINK_MODE_MASK 0x00C00000
@@ -190,7 +194,8 @@
 /* enable link status from external LINK_0 and LINK_1 pins */
 #define E1000_CTRL_SWDPIN0  0x00040000  /* SWDPIN 0 value */
 #define E1000_CTRL_SWDPIN1  0x00080000  /* SWDPIN 1 value */
-#define E1000_CTRL_SWDPIO0  0x00400000  /* SWDPIN 0 Input or output */
+#define E1000_CTRL_SDP0_DIR 0x00400000  /* SDP0 Data direction */
+#define E1000_CTRL_SDP1_DIR 0x00800000  /* SDP1 Data direction */
 #define E1000_CTRL_RST      0x04000000  /* Global reset */
 #define E1000_CTRL_RFCE     0x08000000  /* Receive Flow Control enable */
 #define E1000_CTRL_TFCE     0x10000000  /* Transmit flow control enable */
@@ -528,8 +533,67 @@
 
 #define E1000_TIMINCA_16NS_SHIFT 24
 
-#define E1000_TSICR_TXTS 0x00000002
-#define E1000_TSIM_TXTS 0x00000002
+/* Time Sync Interrupt Cause/Mask Register Bits */
+
+#define TSINTR_SYS_WRAP  (1 << 0) /* SYSTIM Wrap around. */
+#define TSINTR_TXTS      (1 << 1) /* Transmit Timestamp. */
+#define TSINTR_RXTS      (1 << 2) /* Receive Timestamp. */
+#define TSINTR_TT0       (1 << 3) /* Target Time 0 Trigger. */
+#define TSINTR_TT1       (1 << 4) /* Target Time 1 Trigger. */
+#define TSINTR_AUTT0     (1 << 5) /* Auxiliary Timestamp 0 Taken. */
+#define TSINTR_AUTT1     (1 << 6) /* Auxiliary Timestamp 1 Taken. */
+#define TSINTR_TADJ      (1 << 7) /* Time Adjust Done. */
+
+#define TSYNC_INTERRUPTS TSINTR_TXTS
+#define E1000_TSICR_TXTS TSINTR_TXTS
+
+/* TSAUXC Configuration Bits */
+#define TSAUXC_EN_TT0    (1 << 0)  /* Enable target time 0. */
+#define TSAUXC_EN_TT1    (1 << 1)  /* Enable target time 1. */
+#define TSAUXC_EN_CLK0   (1 << 2)  /* Enable Configurable Frequency Clock 0. */
+#define TSAUXC_SAMP_AUT0 (1 << 3)  /* Latch SYSTIML/H into AUXSTMPL/0. */
+#define TSAUXC_ST0       (1 << 4)  /* Start Clock 0 Toggle on Target Time 0. */
+#define TSAUXC_EN_CLK1   (1 << 5)  /* Enable Configurable Frequency Clock 1. */
+#define TSAUXC_SAMP_AUT1 (1 << 6)  /* Latch SYSTIML/H into AUXSTMPL/1. */
+#define TSAUXC_ST1       (1 << 7)  /* Start Clock 1 Toggle on Target Time 1. */
+#define TSAUXC_EN_TS0    (1 << 8)  /* Enable hardware timestamp 0. */
+#define TSAUXC_AUTT0     (1 << 9)  /* Auxiliary Timestamp Taken. */
+#define TSAUXC_EN_TS1    (1 << 10) /* Enable hardware timestamp 0. */
+#define TSAUXC_AUTT1     (1 << 11) /* Auxiliary Timestamp Taken. */
+#define TSAUXC_PLSG      (1 << 17) /* Generate a pulse. */
+#define TSAUXC_DISABLE   (1 << 31) /* Disable SYSTIM Count Operation. */
+
+/* SDP Configuration Bits */
+#define AUX0_SEL_SDP0    (0 << 0)  /* Assign SDP0 to auxiliary time stamp 0. */
+#define AUX0_SEL_SDP1    (1 << 0)  /* Assign SDP1 to auxiliary time stamp 0. */
+#define AUX0_SEL_SDP2    (2 << 0)  /* Assign SDP2 to auxiliary time stamp 0. */
+#define AUX0_SEL_SDP3    (3 << 0)  /* Assign SDP3 to auxiliary time stamp 0. */
+#define AUX0_TS_SDP_EN   (1 << 2)  /* Enable auxiliary time stamp trigger 0. */
+#define AUX1_SEL_SDP0    (0 << 3)  /* Assign SDP0 to auxiliary time stamp 1. */
+#define AUX1_SEL_SDP1    (1 << 3)  /* Assign SDP1 to auxiliary time stamp 1. */
+#define AUX1_SEL_SDP2    (2 << 3)  /* Assign SDP2 to auxiliary time stamp 1. */
+#define AUX1_SEL_SDP3    (3 << 3)  /* Assign SDP3 to auxiliary time stamp 1. */
+#define AUX1_TS_SDP_EN   (1 << 5)  /* Enable auxiliary time stamp trigger 1. */
+#define TS_SDP0_SEL_TT0  (0 << 6)  /* Target time 0 is output on SDP0. */
+#define TS_SDP0_SEL_TT1  (1 << 6)  /* Target time 1 is output on SDP0. */
+#define TS_SDP0_SEL_FC0  (2 << 6)  /* Freq clock  0 is output on SDP0. */
+#define TS_SDP0_SEL_FC1  (3 << 6)  /* Freq clock  1 is output on SDP0. */
+#define TS_SDP0_EN       (1 << 8)  /* SDP0 is assigned to Tsync. */
+#define TS_SDP1_SEL_TT0  (0 << 9)  /* Target time 0 is output on SDP1. */
+#define TS_SDP1_SEL_TT1  (1 << 9)  /* Target time 1 is output on SDP1. */
+#define TS_SDP1_SEL_FC0  (2 << 9)  /* Freq clock  0 is output on SDP1. */
+#define TS_SDP1_SEL_FC1  (3 << 9)  /* Freq clock  1 is output on SDP1. */
+#define TS_SDP1_EN       (1 << 11) /* SDP1 is assigned to Tsync. */
+#define TS_SDP2_SEL_TT0  (0 << 12) /* Target time 0 is output on SDP2. */
+#define TS_SDP2_SEL_TT1  (1 << 12) /* Target time 1 is output on SDP2. */
+#define TS_SDP2_SEL_FC0  (2 << 12) /* Freq clock  0 is output on SDP2. */
+#define TS_SDP2_SEL_FC1  (3 << 12) /* Freq clock  1 is output on SDP2. */
+#define TS_SDP2_EN       (1 << 14) /* SDP2 is assigned to Tsync. */
+#define TS_SDP3_SEL_TT0  (0 << 15) /* Target time 0 is output on SDP3. */
+#define TS_SDP3_SEL_TT1  (1 << 15) /* Target time 1 is output on SDP3. */
+#define TS_SDP3_SEL_FC0  (2 << 15) /* Freq clock  0 is output on SDP3. */
+#define TS_SDP3_SEL_FC1  (3 << 15) /* Freq clock  1 is output on SDP3. */
+#define TS_SDP3_EN       (1 << 17) /* SDP3 is assigned to Tsync. */
 
 #define E1000_MDICNFG_EXT_MDIO    0x80000000      /* MDI ext/int destination */
 #define E1000_MDICNFG_COM_MDIO    0x40000000      /* MDI shared w/ lan 0 */
diff --git a/drivers/net/ethernet/intel/igb/e1000_regs.h b/drivers/net/ethernet/intel/igb/e1000_regs.h
index abdd935..e9c5fdd 100644
--- a/drivers/net/ethernet/intel/igb/e1000_regs.h
+++ b/drivers/net/ethernet/intel/igb/e1000_regs.h
@@ -40,6 +40,7 @@
 #define E1000_FCT      0x00030  /* Flow Control Type - RW */
 #define E1000_CONNSW   0x00034  /* Copper/Fiber switch control - RW */
 #define E1000_VET      0x00038  /* VLAN Ether Type - RW */
+#define E1000_TSSDP    0x0003C  /* Time Sync SDP Configuration Register - RW */
 #define E1000_ICR      0x000C0  /* Interrupt Cause Read - R/clr */
 #define E1000_ITR      0x000C4  /* Interrupt Throttling Rate - RW */
 #define E1000_ICS      0x000C8  /* Interrupt Cause Set - WO */
@@ -101,6 +102,14 @@
 #define E1000_SYSTIMH    0x0B604 /* System time register High - RO */
 #define E1000_TIMINCA    0x0B608 /* Increment attributes register - RW */
 #define E1000_TSAUXC     0x0B640 /* Timesync Auxiliary Control register */
+#define E1000_TRGTTIML0  0x0B644 /* Target Time Register 0 Low  - RW */
+#define E1000_TRGTTIMH0  0x0B648 /* Target Time Register 0 High - RW */
+#define E1000_TRGTTIML1  0x0B64C /* Target Time Register 1 Low  - RW */
+#define E1000_TRGTTIMH1  0x0B650 /* Target Time Register 1 High - RW */
+#define E1000_AUXSTMPL0  0x0B65C /* Auxiliary Time Stamp 0 Register Low  - RO */
+#define E1000_AUXSTMPH0  0x0B660 /* Auxiliary Time Stamp 0 Register High - RO */
+#define E1000_AUXSTMPL1  0x0B664 /* Auxiliary Time Stamp 1 Register Low  - RO */
+#define E1000_AUXSTMPH1  0x0B668 /* Auxiliary Time Stamp 1 Register High - RO */
 #define E1000_SYSTIMR    0x0B6F8 /* System time register Residue */
 #define E1000_TSICR      0x0B66C /* Interrupt Cause Register */
 #define E1000_TSIM       0x0B674 /* Interrupt Mask Register */
diff --git a/drivers/net/ethernet/intel/igb/igb_ptp.c b/drivers/net/ethernet/intel/igb/igb_ptp.c
index 9c9c141..a894551a 100644
--- a/drivers/net/ethernet/intel/igb/igb_ptp.c
+++ b/drivers/net/ethernet/intel/igb/igb_ptp.c
@@ -799,7 +799,7 @@ void igb_ptp_init(struct igb_adapter *adapter)
 
 	/* Initialize the time sync interrupts for devices that support it. */
 	if (hw->mac.type >= e1000_82580) {
-		wr32(E1000_TSIM, E1000_TSIM_TXTS);
+		wr32(E1000_TSIM, TSYNC_INTERRUPTS);
 		wr32(E1000_IMS, E1000_IMS_TS);
 	}
 
@@ -877,7 +877,7 @@ void igb_ptp_reset(struct igb_adapter *adapter)
 	case e1000_i211:
 		/* Enable the timer functions and interrupts. */
 		wr32(E1000_TSAUXC, 0x0);
-		wr32(E1000_TSIM, E1000_TSIM_TXTS);
+		wr32(E1000_TSIM, TSYNC_INTERRUPTS);
 		wr32(E1000_IMS, E1000_IMS_TS);
 		break;
 	default:
-- 
1.8.3.1

[net-next v2 13/15] ixgbe: add ixgbe_write_pci_cfg_word with ixgbe_removed check

[Jeff Kirsher]

From: Jacob Keller <jacob.e.keller@intel.com>

Inline with the current use for ixgbe_read_pci_cfg_word, create a
similar function for writing PCI config, which checks whether the
adapter has been removed first, if Live Error Recovery has been enabled.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c  | 6 +-----
 drivers/net/ethernet/intel/ixgbe/ixgbe_common.h | 1 +
 drivers/net/ethernet/intel/ixgbe/ixgbe_main.c   | 9 +++++++++
 3 files changed, 11 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c
index f8ebe58..7fe2254 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c
@@ -58,7 +58,6 @@ static s32 ixgbe_read_i2c_eeprom_82598(struct ixgbe_hw *hw, u8 byte_offset,
  **/
 static void ixgbe_set_pcie_completion_timeout(struct ixgbe_hw *hw)
 {
-	struct ixgbe_adapter *adapter = hw->back;
 	u32 gcr = IXGBE_READ_REG(hw, IXGBE_GCR);
 	u16 pcie_devctl2;
 
@@ -84,11 +83,8 @@ static void ixgbe_set_pcie_completion_timeout(struct ixgbe_hw *hw)
 	 * 16ms to 55ms
 	 */
 	pcie_devctl2 = ixgbe_read_pci_cfg_word(hw, IXGBE_PCI_DEVICE_CONTROL2);
-	if (ixgbe_removed(hw->hw_addr))
-		return;
 	pcie_devctl2 |= IXGBE_PCI_DEVICE_CONTROL2_16ms;
-	pci_write_config_word(adapter->pdev,
-	                      IXGBE_PCI_DEVICE_CONTROL2, pcie_devctl2);
+	ixgbe_write_pci_cfg_word(hw, IXGBE_PCI_DEVICE_CONTROL2, pcie_devctl2);
 out:
 	/* disable completion timeout resend */
 	gcr &= ~IXGBE_GCR_CMPL_TMOUT_RESEND;
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
index d1d67ba..afa1cda 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
@@ -133,6 +133,7 @@ s32 ixgbe_init_thermal_sensor_thresh_generic(struct ixgbe_hw *hw);
 #define IXGBE_FAILED_READ_CFG_WORD 0xffffU
 
 u16 ixgbe_read_pci_cfg_word(struct ixgbe_hw *hw, u32 reg);
+void ixgbe_write_pci_cfg_word(struct ixgbe_hw *hw, u32 reg, u16 value);
 
 static inline bool ixgbe_removed(void __iomem *addr)
 {
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
index 18cd8ca..c773d6c 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
@@ -361,6 +361,15 @@ static u32 ixgbe_read_pci_cfg_dword(struct ixgbe_hw *hw, u32 reg)
 }
 #endif /* CONFIG_PCI_IOV */
 
+void ixgbe_write_pci_cfg_word(struct ixgbe_hw *hw, u32 reg, u16 value)
+{
+	struct ixgbe_adapter *adapter = hw->back;
+
+	if (ixgbe_removed(hw->hw_addr))
+		return;
+	pci_write_config_word(adapter->pdev, reg, value);
+}
+
 static void ixgbe_service_event_complete(struct ixgbe_adapter *adapter)
 {
 	BUG_ON(!test_bit(__IXGBE_SERVICE_SCHED, &adapter->state));
-- 
1.8.3.1

[net-next v2 14/15] ixgbevf: Indicate removal state explicitly

[Jeff Kirsher]

From: Mark Rustad <mark.d.rustad@intel.com>

Add a bit, __IXGBEVF_REMOVING, to indicate that the module is being
removed. The __IXGBEVF_DOWN bit had been overloaded for this purpose,
but that leads to trouble. A few places now check both __IXGBEVF_DOWN
and __IXGBEVF_REMOVING.

Signed-off-by: Mark Rustad <mark.d.rustad@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/ixgbevf/ixgbevf.h      |  5 +++--
 drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 14 +++++++++-----
 2 files changed, 12 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h
index 5482932..08fb88a 100644
--- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h
+++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h
@@ -1,7 +1,7 @@
 /*******************************************************************************
 
   Intel 82599 Virtual Function driver
-  Copyright(c) 1999 - 2012 Intel Corporation.
+  Copyright(c) 1999 - 2014 Intel Corporation.
 
   This program is free software; you can redistribute it and/or modify it
   under the terms and conditions of the GNU General Public License,
@@ -412,7 +412,8 @@ struct ixgbevf_adapter {
 enum ixbgevf_state_t {
 	__IXGBEVF_TESTING,
 	__IXGBEVF_RESETTING,
-	__IXGBEVF_DOWN
+	__IXGBEVF_DOWN,
+	__IXGBEVF_REMOVING,
 };
 
 struct ixgbevf_cb {
diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
index 8581079..940d924 100644
--- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
@@ -1,7 +1,7 @@
 /*******************************************************************************
 
   Intel 82599 Virtual Function driver
-  Copyright(c) 1999 - 2012 Intel Corporation.
+  Copyright(c) 1999 - 2014 Intel Corporation.
 
   This program is free software; you can redistribute it and/or modify it
   under the terms and conditions of the GNU General Public License,
@@ -608,7 +608,8 @@ static int ixgbevf_poll(struct napi_struct *napi, int budget)
 	napi_complete(napi);
 	if (adapter->rx_itr_setting & 1)
 		ixgbevf_set_itr(q_vector);
-	if (!test_bit(__IXGBEVF_DOWN, &adapter->state))
+	if (!test_bit(__IXGBEVF_DOWN, &adapter->state) &&
+	    !test_bit(__IXGBEVF_REMOVING, &adapter->state))
 		ixgbevf_irq_enable_queues(adapter,
 					  1 << q_vector->v_idx);
 
@@ -833,7 +834,8 @@ static irqreturn_t ixgbevf_msix_other(int irq, void *data)
 
 	hw->mac.get_link_status = 1;
 
-	if (!test_bit(__IXGBEVF_DOWN, &adapter->state))
+	if (!test_bit(__IXGBEVF_DOWN, &adapter->state) &&
+	    !test_bit(__IXGBEVF_REMOVING, &adapter->state))
 		mod_timer(&adapter->watchdog_timer, jiffies);
 
 	IXGBE_WRITE_REG(hw, IXGBE_VTEIMS, adapter->eims_other);
@@ -2329,6 +2331,7 @@ static void ixgbevf_reset_task(struct work_struct *work)
 
 	/* If we're already down or resetting, just bail */
 	if (test_bit(__IXGBEVF_DOWN, &adapter->state) ||
+	    test_bit(__IXGBEVF_REMOVING, &adapter->state) ||
 	    test_bit(__IXGBEVF_RESETTING, &adapter->state))
 		return;
 
@@ -2413,7 +2416,8 @@ static void ixgbevf_watchdog_task(struct work_struct *work)
 
 pf_has_reset:
 	/* Reset the timer */
-	if (!test_bit(__IXGBEVF_DOWN, &adapter->state))
+	if (!test_bit(__IXGBEVF_DOWN, &adapter->state) &&
+	    !test_bit(__IXGBEVF_REMOVING, &adapter->state))
 		mod_timer(&adapter->watchdog_timer,
 			  round_jiffies(jiffies + (2 * HZ)));
 
@@ -3563,7 +3567,7 @@ static void ixgbevf_remove(struct pci_dev *pdev)
 	struct net_device *netdev = pci_get_drvdata(pdev);
 	struct ixgbevf_adapter *adapter = netdev_priv(netdev);
 
-	set_bit(__IXGBEVF_DOWN, &adapter->state);
+	set_bit(__IXGBEVF_REMOVING, &adapter->state);
 
 	del_timer_sync(&adapter->watchdog_timer);
 
-- 
1.8.3.1

[net-next v2 15/15] ixgbevf: Protect ixgbevf_down with __IXGBEVF_DOWN bit

[Jeff Kirsher]

From: Mark Rustad <mark.d.rustad@intel.com>

The ixgbevf_down function can now prevent multiple executions by
doing test_and_set_bit on __IXGBEVF_DOWN. This did not work before
introduction of the __IXGBEVF_REMOVING bit, because of overloading
of __IXGBEVF_DOWN. Also add smp_mb__before_clear_bit call before
clearing the __IXGBEVF_DOWN bit.

Signed-off-by: Mark Rustad <mark.d.rustad@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
index 940d924..a2cba53 100644
--- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c
@@ -1620,6 +1620,7 @@ static void ixgbevf_up_complete(struct ixgbevf_adapter *adapter)
 
 	spin_unlock_bh(&adapter->mbx_lock);
 
+	smp_mb__before_clear_bit();
 	clear_bit(__IXGBEVF_DOWN, &adapter->state);
 	ixgbevf_napi_enable_all(adapter);
 
@@ -1744,7 +1745,8 @@ void ixgbevf_down(struct ixgbevf_adapter *adapter)
 	int i;
 
 	/* signal that we are down to the interrupt handler */
-	set_bit(__IXGBEVF_DOWN, &adapter->state);
+	if (test_and_set_bit(__IXGBEVF_DOWN, &adapter->state))
+		return; /* do nothing if already down */
 
 	/* disable all enabled rx queues */
 	for (i = 0; i < adapter->num_rx_queues; i++)
-- 
1.8.3.1

Re: [net-next v2 00/15][pull request] Intel Wired LAN Driver Updates

[Jeff Kirsher]

[-- Attachment #1: Type: text/plain, Size: 4838 bytes --]

On Tue, 2014-03-18 at 20:41 -0700, Jeff Kirsher wrote:
> This series contains updates to i40e, i40evf, e1000e, ixgbe and ixgbevf.
> 
> Mitch adds support for the VF link state ndo which allows the PF driver
> to control the virtual link state of the VF devices.  Added
> support for viewing and modifying RSS hash options and RSS hash look-up
> table programming through ethtool for i40evf.  Fixed complaint about
> the use of min() where min_t() should be used in i40evf.
> 
> Anjali adds support for ethtool -k option for NTUPLE control for i40e.
> 
> Elizabeth cleans up and refactors i40e_open() to separate out the VSI
> code into its own i40e_vsi_open().
> 
> Jesse enables the hardware feature head write back to avoid updating the
> descriptor ring by marking each descriptor with a DD bit and instead
> writes a memory location with an update to where the driver should clean
> up to in i40e and i40evf.  Reduces context descriptors for i40e/i40evf
> since we do not need context descriptors for every packet, only for
> TSO or timesync.
> 
> Dan Carpenter fixes a potential array underflow in i40e_vc_process_vf_msg().
> 
> Dave fixes an e1000e hardware unit hang where the check for pending Tx work
> when link is lost was mistakenly moved to be done only when link is first
> detected to be lost.  Fixed a problem with poor network performance on
> certain silicon in e1000e when configured for 100M HDX performance.
> 
> Carolyn adds register defines needed for time sync functions and the code
> to call the updated defines.
> 
> Jacob adds the ixgbe function for writing PCI config word and checks
> whether the adapter has been removed first.
> 
> Mark adds the bit __IXGBEVF_REMOVING to indicate that the module is being
> removed because the __IXGBEVF_DOWN bit had been overloaded for this
> purpose, but leads to trouble.  ixgbevf_down function can now prevent
> multiple executions by doing test_and_set_bit on __IXGBEVF_DOWN.
> 
> v2:
> - dropped patch Mitch's patch "i40evf: correctly program RSS HLUT table"
>   based on feedback from Ben Hutchings so that Mitch can re-work the
>   patch solution

Or noticed I referenced the wrong patch name.  v2 dropped the previous
patch 3 in the series which was "i40evf: Support RSS option in ethtool"
patch from Mitch.

> 
> The following are changes since commit 8cfad496c4257441710735ccef622f3829870164:
>   ieee802154: properly unshare skbs in ieee802154 *_rcv functions
> and are available in the git repository at:
>   git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next master
> 
> Anjali Singhai Jain (1):
>   i40e: Patch to enable Ethtool/netdev feature flag for NTUPLE control
> 
> Carolyn Wyborny (1):
>   igb: Add register defines needed for time sync functions
> 
> Catherine Sullivan (1):
>   i40e/i40evf: Bump build versions
> 
> Dan Carpenter (1):
>   i40e: potential array underflow in i40e_vc_process_vf_msg()
> 
> David Ertman (2):
>   e1000e: Fix Hardware Unit Hang
>   e1000e: Fix Explicitly set Transmit Control Register
> 
> Elizabeth Kappler (1):
>   i40e: Refactor and cleanup i40e_open(), adding i40e_vsi_open()
> 
> Jacob Keller (1):
>   ixgbe: add ixgbe_write_pci_cfg_word with ixgbe_removed check
> 
> Jesse Brandeburg (2):
>   i40e/i40evf: enable hardware feature head write back
>   i40e/i40evf: reduce context descriptors
> 
> Mark Rustad (2):
>   ixgbevf: Indicate removal state explicitly
>   ixgbevf: Protect ixgbevf_down with __IXGBEVF_DOWN bit
> 
> Mitch Williams (3):
>   i40e: support VF link state ndo
>   i40evf: correctly program RSS HLUT table
>   i40evf: use min_t
> 
>  drivers/net/ethernet/intel/e1000e/netdev.c         | 42 +++++++---
>  drivers/net/ethernet/intel/i40e/i40e.h             |  4 +-
>  drivers/net/ethernet/intel/i40e/i40e_main.c        | 89 ++++++++++++++++++---
>  drivers/net/ethernet/intel/i40e/i40e_txrx.c        | 48 ++++++++++--
>  drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 91 ++++++++++++++++++++--
>  drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h |  4 +
>  drivers/net/ethernet/intel/i40evf/i40e_txrx.c      | 48 ++++++++++--
>  drivers/net/ethernet/intel/i40evf/i40evf_main.c    | 23 +++---
>  drivers/net/ethernet/intel/igb/e1000_defines.h     | 70 ++++++++++++++++-
>  drivers/net/ethernet/intel/igb/e1000_regs.h        |  9 +++
>  drivers/net/ethernet/intel/igb/igb_ptp.c           |  4 +-
>  drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c     |  6 +-
>  drivers/net/ethernet/intel/ixgbe/ixgbe_common.h    |  1 +
>  drivers/net/ethernet/intel/ixgbe/ixgbe_main.c      |  9 +++
>  drivers/net/ethernet/intel/ixgbevf/ixgbevf.h       |  5 +-
>  drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c  | 18 +++--
>  16 files changed, 400 insertions(+), 71 deletions(-)
> 



[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [net-next v2 01/15] i40e: support VF link state ndo

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 18 Mar 2014 20:41:59 -0700

> +	vf = &(pf->vf[vf_id]);

These parenthesis are unnecessary, and in fact confusing.  Someone reading
this is going to look again and again trying to figure out why they are
needed.

Re: [net-next v2 02/15] i40evf: correctly program RSS HLUT table

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 18 Mar 2014 20:42:00 -0700

> +#define NEXT_QUEUE(_j) (++_j >= adapter->vsi_res->num_queue_pairs ? _j = 0 : _j)

I would strong recommend breaking the increment out of the test
and using a function rather than a macro:

static int next_queue(struct i40evf_adapter *adapter, int j)
{
	j += 1;

	return j >= adapter->vsi_res->num_queue_pairs ? 0 : j;
}

 ..

	j = next_queue(adapter, j);

Re: [net-next v2 06/15] i40e/i40evf: enable hardware feature head write back

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 18 Mar 2014 20:42:04 -0700

> +static inline u32 i40e_get_head(struct i40e_ring *tx_ring)
> +{
> +	void *head = (struct i40e_tx_desc *)tx_ring->desc + tx_ring->count;
> +	return le32_to_cpu(*(volatile __le32 *)head);

Please put an empty line between local variable declarations and code.

> +static inline u32 i40e_get_head(struct i40e_ring *tx_ring)
> +{
> +	void *head = (struct i40e_tx_desc *)tx_ring->desc + tx_ring->count;
> +	return le32_to_cpu(*(volatile __le32 *)head);

Likewise.

Re: [net-next v2 02/15] i40evf: correctly program RSS HLUT table

[Jeff Kirsher]

[-- Attachment #1: Type: text/plain, Size: 765 bytes --]

On Wed, 2014-03-19 at 10:08 +0200, Or Gerlitz wrote:
> On Wed, Mar 19, 2014 at 5:42 AM, Jeff Kirsher
> <jeffrey.t.kirsher@intel.com> wrote:
> > From: Mitch Williams <mitch.a.williams@intel.com>
> >
> > The HLUT programming loop in in i40evf_configure_rss was a) overly-
> > complicated, and b) just plain broken. Most of the entries ended up being
> > not written at all, so most of the flows ended up at queue zero.
> 
> In the V2 cover letter you wrote
> 
> dropped patch Mitch's patch "i40evf: correctly program RSS HLUT table"
> 
> so what is this patch?

Sorry, that was may bad.  In v2 I dropped patch 3 which was "i40evf:
Support RSS option in ethtool" patch.  I referenced the wrong patch in
the cover letter, thanks for noticing that Or.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [net-next v2 02/15] i40evf: correctly program RSS HLUT table

[Or Gerlitz]

On Wed, Mar 19, 2014 at 5:42 AM, Jeff Kirsher
<jeffrey.t.kirsher@intel.com> wrote:
> From: Mitch Williams <mitch.a.williams@intel.com>
>
> The HLUT programming loop in in i40evf_configure_rss was a) overly-
> complicated, and b) just plain broken. Most of the entries ended up being
> not written at all, so most of the flows ended up at queue zero.

In the V2 cover letter you wrote

dropped patch Mitch's patch "i40evf: correctly program RSS HLUT table"

so what is this patch?