[net-next 00/12][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-08

[net-next 00/12][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-08

[Jeff Kirsher]

This series contains updates to i40e and i40evf only.

Mitch fixes an issue where the client driver (i40iw) was attempting to
load on x710 devices (which do not support iWARP), so only register with
the client if iWARP is supported.

Jake fixes up error messages to better clarify to the user when adding a
invalid flow type.  Updates the driver to look up the MAC address from
eth_get_platform_mac_address() first before checking what the firmware
provides.  Cleans up code so we are not repeating a duplicate loop, by
checking both transmit and receive queues in a single loop.  Also cleans
up flags never used, so remove the definitions.

Alex does cleanup so that we are always updating pf->flags when a change
is made to the private flags.  Adds support for 3K buffers to the receive
path so that we can provide the additional padding needed in the event
of NET_IP_ALIGN being non-zero or a cache line being greater than 64.
Adds support for build_skb() to i40e/i40evf.

Maciej adjusts the scope of the rtnl lock held during reset because it
was stopping other PFs from running their reset procedures.

Alan reduces code complexity in i40e_detect_recover_hung_queue().

The following are changes since commit 3a9024f52c2e92a143195db43d2abbd5d0792c06:
  net: thunderx: Enable TSO and checksum offloads for ipv6
and are available in the git repository at:
  git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue 40GbE

Alan Brady (1):
  i40e: Simplify i40e_detect_recover_hung_queue logic

Alexander Duyck (4):
  i40e: Swap use of pf->flags and pf->hw_disabled_flags for ATR Eviction
  i40e/i40evf: Add support for using order 1 pages with a 3K buffer
  i40e/i40evf: Add support for padding start of frames
  i40e/i40evf: Use build_skb to build frames

Alice Michael (1):
  i40e: remove I40E_FLAG_NEED_LINK_UPDATE

Jacob Keller (4):
  i40e: update error message when trying to add invalid filters
  i40e: allow look-up of MAC address from Open Firmware or IDPROM
  i40e: remove extraneous loop in i40e_vsi_wait_queues_disabled
  i40e: clean up historic deprecated flag definitions

Maciej Sosin (1):
  i40e: Decrease the scope of rtnl lock

Mitch Williams (1):
  i40e: only register client on iWarp-capable devices

 drivers/net/ethernet/intel/i40e/i40e.h          |  10 +-
 drivers/net/ethernet/intel/i40e/i40e_ethtool.c  |   7 +-
 drivers/net/ethernet/intel/i40e/i40e_main.c     | 268 +++++++++++++-----------
 drivers/net/ethernet/intel/i40e/i40e_txrx.c     | 116 +++++++---
 drivers/net/ethernet/intel/i40e/i40e_txrx.h     |  85 +++++++-
 drivers/net/ethernet/intel/i40evf/i40e_txrx.c   |  89 ++++++--
 drivers/net/ethernet/intel/i40evf/i40e_txrx.h   |  82 +++++++-
 drivers/net/ethernet/intel/i40evf/i40evf.h      |   1 -
 drivers/net/ethernet/intel/i40evf/i40evf_main.c |  14 +-
 9 files changed, 486 insertions(+), 186 deletions(-)

-- 
2.12.2

[net-next 01/12] i40e: only register client on iWarp-capable devices

[Jeff Kirsher]

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

The client interface is only intended for use on devices that support
iWarp. Only register with the client if this is the case.

This fixes a panic when loading i40iw on X710 devices.

Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Reported-by: Stefan Assmann <sassmann@kpanic.de>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index d83430faaa41..1ee2759c38f7 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -11245,10 +11245,12 @@ static int i40e_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		  round_jiffies(jiffies + pf->service_timer_period));
 
 	/* add this PF to client device list and launch a client service task */
-	err = i40e_lan_add_device(pf);
-	if (err)
-		dev_info(&pdev->dev, "Failed to add PF to client API service list: %d\n",
-			 err);
+	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
+		err = i40e_lan_add_device(pf);
+		if (err)
+			dev_info(&pdev->dev, "Failed to add PF to client API service list: %d\n",
+				 err);
+	}
 
 #define PCI_SPEED_SIZE 8
 #define PCI_WIDTH_SIZE 8
@@ -11426,10 +11428,11 @@ static void i40e_remove(struct pci_dev *pdev)
 		i40e_vsi_release(pf->vsi[pf->lan_vsi]);
 
 	/* remove attached clients */
-	ret_code = i40e_lan_del_device(pf);
-	if (ret_code) {
-		dev_warn(&pdev->dev, "Failed to delete client device: %d\n",
-			 ret_code);
+	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
+		ret_code = i40e_lan_del_device(pf);
+		if (ret_code)
+			dev_warn(&pdev->dev, "Failed to delete client device: %d\n",
+				 ret_code);
 	}
 
 	/* shutdown and destroy the HMC */
-- 
2.12.2

[net-next 02/12] i40e: update error message when trying to add invalid filters

[Jeff Kirsher]

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

Re-word the error message displayed when adding a filter with an
invalid flow type. Additionally, report a distinct error message when
the IPv4 protocol is at fault.

Change-ID: Iba3d85b87f8d383c97c8bdd180df34a6adf3ee67
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_txrx.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index ebffca0cefac..d45f48c84ff7 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -533,14 +533,15 @@ int i40e_add_del_fdir(struct i40e_vsi *vsi,
 			break;
 		default:
 			/* We cannot support masking based on protocol */
-			goto unsupported_flow;
+			dev_info(&pf->pdev->dev, "Unsupported IPv4 protocol 0x%02x\n",
+				 input->ip4_proto);
+			return -EINVAL;
 		}
 		break;
 	default:
-unsupported_flow:
-		dev_info(&pf->pdev->dev, "Could not specify spec type %d\n",
+		dev_info(&pf->pdev->dev, "Unsupported flow type 0x%02x\n",
 			 input->flow_type);
-		ret = -EINVAL;
+		return -EINVAL;
 	}
 
 	/* The buffer allocated here will be normally be freed by
-- 
2.12.2

[net-next 03/12] i40e: Swap use of pf->flags and pf->hw_disabled_flags for ATR Eviction

[Jeff Kirsher]

From: Alexander Duyck <alexander.h.duyck@intel.com>

This is a minor cleanup so that we are always updating pf->flags when we
make a change to the private flags instead of updating a mix of either
pf->flags and/or pf->hw_disabled_flags.

In addition I went through and cleaned out all the spots where we were
using the X722 define in regards to this flag.

Lastly since we changed the logic I went through and flushed out any
redundancy and cleaned up the handling of the flags in the Tx path.

Change-ID: I79ff95a7272bb2533251ff11ef91e89ccb80b610
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_txrx.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index d45f48c84ff7..a9a97dd2c0a5 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -2263,8 +2263,7 @@ static void i40e_atr(struct i40e_ring *tx_ring, struct sk_buff *skb,
 	/* Due to lack of space, no more new filters can be programmed */
 	if (th->syn && (pf->hw_disabled_flags & I40E_FLAG_FD_ATR_ENABLED))
 		return;
-	if ((pf->flags & I40E_FLAG_HW_ATR_EVICT_CAPABLE) &&
-	    (!(pf->hw_disabled_flags & I40E_FLAG_HW_ATR_EVICT_CAPABLE))) {
+	if (pf->flags & I40E_FLAG_HW_ATR_EVICT_CAPABLE) {
 		/* HW ATR eviction will take care of removing filters on FIN
 		 * and RST packets.
 		 */
@@ -2326,8 +2325,7 @@ static void i40e_atr(struct i40e_ring *tx_ring, struct sk_buff *skb,
 			I40E_TXD_FLTR_QW1_CNTINDEX_SHIFT) &
 			I40E_TXD_FLTR_QW1_CNTINDEX_MASK;
 
-	if ((pf->flags & I40E_FLAG_HW_ATR_EVICT_CAPABLE) &&
-	    (!(pf->hw_disabled_flags & I40E_FLAG_HW_ATR_EVICT_CAPABLE)))
+	if (pf->flags & I40E_FLAG_HW_ATR_EVICT_CAPABLE)
 		dtype_cmd |= I40E_TXD_FLTR_QW1_ATR_MASK;
 
 	fdir_desc->qindex_flex_ptype_vsi = cpu_to_le32(flex_ptype);
-- 
2.12.2

[net-next 04/12] i40e: Decrease the scope of rtnl lock

[Jeff Kirsher]

From: Maciej Sosin <maciej.sosin@intel.com>

Previously rtnl lock was held during whole reset procedure that
was stopping other PFs running their reset procedures. In the result
reset was not handled properly and host reset was the only way
to recover.

Change-ID: I23c0771c0303caaa7bd64badbf0c667e25142954
Signed-off-by: Maciej Sosin <maciej.sosin@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e.h         |   2 +-
 drivers/net/ethernet/intel/i40e/i40e_ethtool.c |   6 +-
 drivers/net/ethernet/intel/i40e/i40e_main.c    | 138 +++++++++++++++++--------
 3 files changed, 101 insertions(+), 45 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index 421ea57128d3..686327c031fa 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -837,7 +837,7 @@ 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_do_reset(struct i40e_pf *pf, u32 reset_flags, bool lock_acquired);
 int i40e_config_rss(struct i40e_vsi *vsi, u8 *seed, u8 *lut, u16 lut_size);
 int i40e_get_rss(struct i40e_vsi *vsi, u8 *seed, u8 *lut, u16 lut_size);
 void i40e_fill_rss_lut(struct i40e_pf *pf, u8 *lut,
diff --git a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
index c0c1a0cdaa5b..68c0f204f93e 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
@@ -1852,7 +1852,7 @@ static void i40e_diag_test(struct net_device *netdev,
 			 * link then the following link test would have
 			 * to be moved to before the reset
 			 */
-			i40e_do_reset(pf, BIT(__I40E_PF_RESET_REQUESTED));
+			i40e_do_reset(pf, BIT(__I40E_PF_RESET_REQUESTED), true);
 
 		if (i40e_link_test(netdev, &data[I40E_ETH_TEST_LINK]))
 			eth_test->flags |= ETH_TEST_FL_FAILED;
@@ -1868,7 +1868,7 @@ static void i40e_diag_test(struct net_device *netdev,
 			eth_test->flags |= ETH_TEST_FL_FAILED;
 
 		clear_bit(__I40E_TESTING, &pf->state);
-		i40e_do_reset(pf, BIT(__I40E_PF_RESET_REQUESTED));
+		i40e_do_reset(pf, BIT(__I40E_PF_RESET_REQUESTED), true);
 
 		if (if_running)
 			i40e_open(netdev);
@@ -4099,7 +4099,7 @@ static int i40e_set_priv_flags(struct net_device *dev, u32 flags)
 	 */
 	if ((changed_flags & I40E_FLAG_VEB_STATS_ENABLED) ||
 	    ((changed_flags & I40E_FLAG_LEGACY_RX) && netif_running(dev)))
-		i40e_do_reset(pf, BIT(__I40E_PF_RESET_REQUESTED));
+		i40e_do_reset(pf, BIT(__I40E_PF_RESET_REQUESTED), true);
 
 	return 0;
 }
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 1ee2759c38f7..8181647f512e 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -50,13 +50,16 @@ static const char i40e_copyright[] = "Copyright (c) 2013 - 2014 Intel Corporatio
 
 /* a bit of forward declarations */
 static void i40e_vsi_reinit_locked(struct i40e_vsi *vsi);
-static void i40e_handle_reset_warning(struct i40e_pf *pf);
+static void i40e_handle_reset_warning(struct i40e_pf *pf, bool lock_acquired);
 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, 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);
+static void i40e_prep_for_reset(struct i40e_pf *pf, bool lock_acquired);
+static int i40e_reset(struct i40e_pf *pf);
+static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired);
 static void i40e_fdir_sb_setup(struct i40e_pf *pf);
 static int i40e_veb_get_bw_info(struct i40e_veb *veb);
 
@@ -5537,6 +5540,8 @@ int i40e_open(struct net_device *netdev)
  * Finish initialization of the VSI.
  *
  * Returns 0 on success, negative value on failure
+ *
+ * Note: expects to be called while under rtnl_lock()
  **/
 int i40e_vsi_open(struct i40e_vsi *vsi)
 {
@@ -5600,7 +5605,7 @@ int i40e_vsi_open(struct i40e_vsi *vsi)
 err_setup_tx:
 	i40e_vsi_free_tx_resources(vsi);
 	if (vsi == pf->vsi[pf->lan_vsi])
-		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED));
+		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED), true);
 
 	return err;
 }
@@ -5686,12 +5691,14 @@ int i40e_close(struct net_device *netdev)
  * i40e_do_reset - Start a PF or Core Reset sequence
  * @pf: board private structure
  * @reset_flags: which reset is requested
+ * @lock_acquired: indicates whether or not the lock has been acquired
+ * before this function was called.
  *
  * The essential difference in resets is that the PF Reset
  * doesn't clear the packet buffers, doesn't reset the PE
  * firmware, and doesn't bother the other PFs on the chip.
  **/
-void i40e_do_reset(struct i40e_pf *pf, u32 reset_flags)
+void i40e_do_reset(struct i40e_pf *pf, u32 reset_flags, bool lock_acquired)
 {
 	u32 val;
 
@@ -5737,7 +5744,7 @@ void i40e_do_reset(struct i40e_pf *pf, u32 reset_flags)
 		 * for the Core Reset.
 		 */
 		dev_dbg(&pf->pdev->dev, "PFR requested\n");
-		i40e_handle_reset_warning(pf);
+		i40e_handle_reset_warning(pf, lock_acquired);
 
 	} else if (reset_flags & BIT_ULL(__I40E_REINIT_REQUESTED)) {
 		int v;
@@ -5946,7 +5953,7 @@ static int i40e_handle_lldp_event(struct i40e_pf *pf,
 void i40e_do_reset_safe(struct i40e_pf *pf, u32 reset_flags)
 {
 	rtnl_lock();
-	i40e_do_reset(pf, reset_flags);
+	i40e_do_reset(pf, reset_flags, true);
 	rtnl_unlock();
 }
 
@@ -6348,7 +6355,6 @@ 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 |= BIT(__I40E_REINIT_REQUESTED);
 		clear_bit(__I40E_REINIT_REQUESTED, &pf->state);
@@ -6374,18 +6380,19 @@ static void i40e_reset_subtask(struct i40e_pf *pf)
 	 * precedence before starting a new reset sequence.
 	 */
 	if (test_bit(__I40E_RESET_INTR_RECEIVED, &pf->state)) {
-		i40e_handle_reset_warning(pf);
-		goto unlock;
+		i40e_prep_for_reset(pf, false);
+		i40e_reset(pf);
+		i40e_rebuild(pf, false, false);
 	}
 
 	/* If we're already down or resetting, just bail */
 	if (reset_flags &&
 	    !test_bit(__I40E_DOWN, &pf->state) &&
-	    !test_bit(__I40E_CONFIG_BUSY, &pf->state))
-		i40e_do_reset(pf, reset_flags);
-
-unlock:
-	rtnl_unlock();
+	    !test_bit(__I40E_CONFIG_BUSY, &pf->state)) {
+		rtnl_lock();
+		i40e_do_reset(pf, reset_flags, true);
+		rtnl_unlock();
+	}
 }
 
 /**
@@ -6873,10 +6880,12 @@ static void i40e_fdir_teardown(struct i40e_pf *pf)
 /**
  * i40e_prep_for_reset - prep for the core to reset
  * @pf: board private structure
+ * @lock_acquired: indicates whether or not the lock has been acquired
+ * before this function was called.
  *
  * Close up the VFs and other things in prep for PF Reset.
   **/
-static void i40e_prep_for_reset(struct i40e_pf *pf)
+static void i40e_prep_for_reset(struct i40e_pf *pf, bool lock_acquired)
 {
 	struct i40e_hw *hw = &pf->hw;
 	i40e_status ret = 0;
@@ -6891,7 +6900,12 @@ static void i40e_prep_for_reset(struct i40e_pf *pf)
 	dev_dbg(&pf->pdev->dev, "Tearing down internal switch for reset\n");
 
 	/* quiesce the VSIs and their queues that are not already DOWN */
+	/* pf_quiesce_all_vsi modifies netdev structures -rtnl_lock needed */
+	if (!lock_acquired)
+		rtnl_lock();
 	i40e_pf_quiesce_all_vsi(pf);
+	if (!lock_acquired)
+		rtnl_unlock();
 
 	for (v = 0; v < pf->num_alloc_vsi; v++) {
 		if (pf->vsi[v])
@@ -6926,29 +6940,39 @@ static void i40e_send_version(struct i40e_pf *pf)
 }
 
 /**
- * i40e_reset_and_rebuild - reset and rebuild using a saved config
+ * i40e_reset - wait for core reset to finish reset, reset pf if corer not seen
  * @pf: board private structure
- * @reinit: if the Main VSI needs to re-initialized.
  **/
-static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit)
+static int i40e_reset(struct i40e_pf *pf)
 {
 	struct i40e_hw *hw = &pf->hw;
-	u8 set_fc_aq_fail = 0;
 	i40e_status ret;
-	u32 val;
-	u32 v;
 
-	/* Now we wait for GRST to settle out.
-	 * We don't have to delete the VEBs or VSIs from the hw switch
-	 * because the reset will make them disappear.
-	 */
 	ret = i40e_pf_reset(hw);
 	if (ret) {
 		dev_info(&pf->pdev->dev, "PF reset failed, %d\n", ret);
 		set_bit(__I40E_RESET_FAILED, &pf->state);
-		goto clear_recovery;
+		clear_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state);
+	} else {
+		pf->pfr_count++;
 	}
-	pf->pfr_count++;
+	return ret;
+}
+
+/**
+ * i40e_rebuild - rebuild using a saved config
+ * @pf: board private structure
+ * @reinit: if the Main VSI needs to re-initialized.
+ * @lock_acquired: indicates whether or not the lock has been acquired
+ * before this function was called.
+ **/
+static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired)
+{
+	struct i40e_hw *hw = &pf->hw;
+	u8 set_fc_aq_fail = 0;
+	i40e_status ret;
+	u32 val;
+	int v;
 
 	if (test_bit(__I40E_DOWN, &pf->state))
 		goto clear_recovery;
@@ -6993,9 +7017,11 @@ static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit)
 	}
 #endif /* CONFIG_I40E_DCB */
 	/* do basic switch setup */
+	if (!lock_acquired)
+		rtnl_lock();
 	ret = i40e_setup_pf_switch(pf, reinit);
 	if (ret)
-		goto end_core_reset;
+		goto end_unlock;
 
 	/* The driver only wants link up/down and module qualification
 	 * reports from firmware.  Note the negative logic.
@@ -7066,7 +7092,7 @@ static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit)
 		if (ret) {
 			dev_info(&pf->pdev->dev,
 				 "rebuild of Main VSI failed: %d\n", ret);
-			goto end_core_reset;
+			goto end_unlock;
 		}
 	}
 
@@ -7117,6 +7143,9 @@ static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit)
 	/* tell the firmware that we're starting */
 	i40e_send_version(pf);
 
+end_unlock:
+if (!lock_acquired)
+	rtnl_unlock();
 end_core_reset:
 	clear_bit(__I40E_RESET_FAILED, &pf->state);
 clear_recovery:
@@ -7124,16 +7153,38 @@ static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit)
 }
 
 /**
+ * i40e_reset_and_rebuild - reset and rebuild using a saved config
+ * @pf: board private structure
+ * @reinit: if the Main VSI needs to re-initialized.
+ * @lock_acquired: indicates whether or not the lock has been acquired
+ * before this function was called.
+ **/
+static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit,
+				   bool lock_acquired)
+{
+	int ret;
+	/* Now we wait for GRST to settle out.
+	 * We don't have to delete the VEBs or VSIs from the hw switch
+	 * because the reset will make them disappear.
+	 */
+	ret = i40e_reset(pf);
+	if (!ret)
+		i40e_rebuild(pf, reinit, lock_acquired);
+}
+
+/**
  * i40e_handle_reset_warning - prep for the PF to reset, reset and rebuild
  * @pf: board private structure
  *
  * Close up the VFs and other things in prep for a Core Reset,
  * then get ready to rebuild the world.
+ * @lock_acquired: indicates whether or not the lock has been acquired
+ * before this function was called.
  **/
-static void i40e_handle_reset_warning(struct i40e_pf *pf)
+static void i40e_handle_reset_warning(struct i40e_pf *pf, bool lock_acquired)
 {
-	i40e_prep_for_reset(pf);
-	i40e_reset_and_rebuild(pf, false);
+	i40e_prep_for_reset(pf, lock_acquired);
+	i40e_reset_and_rebuild(pf, false, lock_acquired);
 }
 
 /**
@@ -8430,6 +8481,7 @@ static int i40e_pf_config_rss(struct i40e_pf *pf)
  *
  * returns 0 if rss is not enabled, if enabled returns the final rss queue
  * count which may be different from the requested queue count.
+ * Note: expects to be called while under rtnl_lock()
  **/
 int i40e_reconfig_rss_queues(struct i40e_pf *pf, int queue_count)
 {
@@ -8445,11 +8497,11 @@ int i40e_reconfig_rss_queues(struct i40e_pf *pf, int queue_count)
 		u16 qcount;
 
 		vsi->req_queue_pairs = queue_count;
-		i40e_prep_for_reset(pf);
+		i40e_prep_for_reset(pf, true);
 
 		pf->alloc_rss_size = new_rss_size;
 
-		i40e_reset_and_rebuild(pf, true);
+		i40e_reset_and_rebuild(pf, true, true);
 
 		/* Discard the user configured hash keys and lut, if less
 		 * queues are enabled.
@@ -8825,6 +8877,7 @@ static void i40e_clear_rss_lut(struct i40e_vsi *vsi)
  * i40e_set_features - set the netdev feature flags
  * @netdev: ptr to the netdev being adjusted
  * @features: the feature set that the stack is suggesting
+ * Note: expects to be called while under rtnl_lock()
  **/
 static int i40e_set_features(struct net_device *netdev,
 			     netdev_features_t features)
@@ -8848,7 +8901,7 @@ static int i40e_set_features(struct net_device *netdev,
 	need_reset = i40e_set_ntuple(pf, features);
 
 	if (need_reset)
-		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED));
+		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED), true);
 
 	return 0;
 }
@@ -9043,6 +9096,8 @@ static int i40e_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
  * is to change the mode then that requires a PF reset to
  * allow rebuild of the components with required hardware
  * bridge mode enabled.
+ *
+ * Note: expects to be called while under rtnl_lock()
  **/
 static int i40e_ndo_bridge_setlink(struct net_device *dev,
 				   struct nlmsghdr *nlh,
@@ -9098,7 +9153,8 @@ static int i40e_ndo_bridge_setlink(struct net_device *dev,
 				pf->flags |= I40E_FLAG_VEB_MODE_ENABLED;
 			else
 				pf->flags &= ~I40E_FLAG_VEB_MODE_ENABLED;
-			i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED));
+			i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED),
+				      true);
 			break;
 		}
 	}
@@ -11501,7 +11557,7 @@ static pci_ers_result_t i40e_pci_error_detected(struct pci_dev *pdev,
 	/* shutdown all operations */
 	if (!test_bit(__I40E_SUSPENDED, &pf->state)) {
 		rtnl_lock();
-		i40e_prep_for_reset(pf);
+		i40e_prep_for_reset(pf, true);
 		rtnl_unlock();
 	}
 
@@ -11570,7 +11626,7 @@ static void i40e_pci_error_resume(struct pci_dev *pdev)
 		return;
 
 	rtnl_lock();
-	i40e_handle_reset_warning(pf);
+	i40e_handle_reset_warning(pf, true);
 	rtnl_unlock();
 }
 
@@ -11633,7 +11689,7 @@ static void i40e_shutdown(struct pci_dev *pdev)
 	set_bit(__I40E_SUSPENDED, &pf->state);
 	set_bit(__I40E_DOWN, &pf->state);
 	rtnl_lock();
-	i40e_prep_for_reset(pf);
+	i40e_prep_for_reset(pf, true);
 	rtnl_unlock();
 
 	wr32(hw, I40E_PFPM_APM, (pf->wol_en ? I40E_PFPM_APM_APME_MASK : 0));
@@ -11652,7 +11708,7 @@ static void i40e_shutdown(struct pci_dev *pdev)
 		i40e_enable_mc_magic_wake(pf);
 
 	rtnl_lock();
-	i40e_prep_for_reset(pf);
+	i40e_prep_for_reset(pf, true);
 	rtnl_unlock();
 
 	wr32(hw, I40E_PFPM_APM,
@@ -11686,7 +11742,7 @@ static int i40e_suspend(struct pci_dev *pdev, pm_message_t state)
 		i40e_enable_mc_magic_wake(pf);
 
 	rtnl_lock();
-	i40e_prep_for_reset(pf);
+	i40e_prep_for_reset(pf, true);
 	rtnl_unlock();
 
 	wr32(hw, I40E_PFPM_APM, (pf->wol_en ? I40E_PFPM_APM_APME_MASK : 0));
@@ -11734,7 +11790,7 @@ static int i40e_resume(struct pci_dev *pdev)
 	if (test_and_clear_bit(__I40E_SUSPENDED, &pf->state)) {
 		clear_bit(__I40E_DOWN, &pf->state);
 		rtnl_lock();
-		i40e_reset_and_rebuild(pf, false);
+		i40e_reset_and_rebuild(pf, false, true);
 		rtnl_unlock();
 	}
 
-- 
2.12.2

[net-next 05/12] i40e: Simplify i40e_detect_recover_hung_queue logic

[Jeff Kirsher]

From: Alan Brady <alan.brady@intel.com>

This patch greatly reduces the unneeded complexity in the
i40e_detect_recover_hung_queue code path.  The previous implementation
set a 'hung bit' which would then get cleared while polling.  If the
detection routine was called a second time with the bit already set, we
would issue a software interrupt.  This patch makes it such that if
interrupts are disabled and we have pending TX descriptors, we trigger a
software interrupt since in, the worst case, queues are already clean
and we have an extra interrupt.

Additionally this patch removes the workaround for lost interrupts as
calling napi_reschedule in this context can cause software interrupts to
fire on the wrong CPU.

Change-ID: Iae108582a3ceb6229ed1d22e4ed6e69cf97aad8d
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e.h         |  4 --
 drivers/net/ethernet/intel/i40e/i40e_ethtool.c |  1 -
 drivers/net/ethernet/intel/i40e/i40e_main.c    | 59 +++++---------------------
 drivers/net/ethernet/intel/i40e/i40e_txrx.c    | 12 ++----
 drivers/net/ethernet/intel/i40e/i40e_txrx.h    |  3 +-
 5 files changed, 15 insertions(+), 64 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index 686327c031fa..110ef4204306 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -617,7 +617,6 @@ struct i40e_vsi {
 	u32 tx_busy;
 	u64 tx_linearize;
 	u64 tx_force_wb;
-	u64 tx_lost_interrupt;
 	u32 rx_buf_failed;
 	u32 rx_page_failed;
 
@@ -703,9 +702,6 @@ struct i40e_q_vector {
 
 	u8 num_ringpairs;	/* total number of ring pairs in vector */
 
-#define I40E_Q_VECTOR_HUNG_DETECT 0 /* Bit Index for hung detection logic */
-	unsigned long hung_detected; /* Set/Reset for hung_detection logic */
-
 	cpumask_t affinity_mask;
 	struct irq_affinity_notify affinity_notify;
 
diff --git a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
index 68c0f204f93e..10325b5a9805 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c
@@ -89,7 +89,6 @@ static const struct i40e_stats i40e_gstrings_misc_stats[] = {
 	I40E_VSI_STAT("rx_unknown_protocol", eth_stats.rx_unknown_protocol),
 	I40E_VSI_STAT("tx_linearize", tx_linearize),
 	I40E_VSI_STAT("tx_force_wb", tx_force_wb),
-	I40E_VSI_STAT("tx_lost_interrupt", tx_lost_interrupt),
 	I40E_VSI_STAT("rx_alloc_fail", rx_buf_failed),
 	I40E_VSI_STAT("rx_pg_alloc_fail", rx_page_failed),
 };
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 8181647f512e..22831a4a9099 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -737,7 +737,6 @@ static void i40e_update_vsi_stats(struct i40e_vsi *vsi)
 	struct i40e_eth_stats *oes;
 	struct i40e_eth_stats *es;     /* device's eth stats */
 	u32 tx_restart, tx_busy;
-	u64 tx_lost_interrupt;
 	struct i40e_ring *p;
 	u32 rx_page, rx_buf;
 	u64 bytes, packets;
@@ -763,7 +762,6 @@ static void i40e_update_vsi_stats(struct i40e_vsi *vsi)
 	rx_b = rx_p = 0;
 	tx_b = tx_p = 0;
 	tx_restart = tx_busy = tx_linearize = tx_force_wb = 0;
-	tx_lost_interrupt = 0;
 	rx_page = 0;
 	rx_buf = 0;
 	rcu_read_lock();
@@ -782,7 +780,6 @@ static void i40e_update_vsi_stats(struct i40e_vsi *vsi)
 		tx_busy += p->tx_stats.tx_busy;
 		tx_linearize += p->tx_stats.tx_linearize;
 		tx_force_wb += p->tx_stats.tx_force_wb;
-		tx_lost_interrupt += p->tx_stats.tx_lost_interrupt;
 
 		/* Rx queue is part of the same block as Tx queue */
 		p = &p[1];
@@ -801,7 +798,6 @@ static void i40e_update_vsi_stats(struct i40e_vsi *vsi)
 	vsi->tx_busy = tx_busy;
 	vsi->tx_linearize = tx_linearize;
 	vsi->tx_force_wb = tx_force_wb;
-	vsi->tx_lost_interrupt = tx_lost_interrupt;
 	vsi->rx_page_failed = rx_page;
 	vsi->rx_buf_failed = rx_buf;
 
@@ -4508,16 +4504,15 @@ static int i40e_pf_wait_queues_disabled(struct i40e_pf *pf)
  * @vsi: Pointer to VSI struct
  *
  * This function checks specified queue for given VSI. Detects hung condition.
- * Sets hung bit since it is two step process. Before next run of service task
- * if napi_poll runs, it reset 'hung' bit for respective q_vector. If not,
- * hung condition remain unchanged and during subsequent run, this function
- * issues SW interrupt to recover from hung condition.
+ * We proactively detect hung TX queues by checking if interrupts are disabled
+ * but there are pending descriptors.  If it appears hung, attempt to recover
+ * by triggering a SW interrupt.
  **/
 static void i40e_detect_recover_hung_queue(int q_idx, struct i40e_vsi *vsi)
 {
 	struct i40e_ring *tx_ring = NULL;
 	struct i40e_pf	*pf;
-	u32 head, val, tx_pending_hw;
+	u32 val, tx_pending;
 	int i;
 
 	pf = vsi->back;
@@ -4543,47 +4538,15 @@ static void i40e_detect_recover_hung_queue(int q_idx, struct i40e_vsi *vsi)
 	else
 		val = rd32(&pf->hw, I40E_PFINT_DYN_CTL0);
 
-	head = i40e_get_head(tx_ring);
+	tx_pending = i40e_get_tx_pending(tx_ring);
 
-	tx_pending_hw = i40e_get_tx_pending(tx_ring, false);
-
-	/* HW is done executing descriptors, updated HEAD write back,
-	 * but SW hasn't processed those descriptors. If interrupt is
-	 * not generated from this point ON, it could result into
-	 * dev_watchdog detecting timeout on those netdev_queue,
-	 * hence proactively trigger SW interrupt.
+	/* Interrupts are disabled and TX pending is non-zero,
+	 * trigger the SW interrupt (don't wait). Worst case
+	 * there will be one extra interrupt which may result
+	 * into not cleaning any queues because queues are cleaned.
 	 */
-	if (tx_pending_hw && (!(val & I40E_PFINT_DYN_CTLN_INTENA_MASK))) {
-		/* NAPI Poll didn't run and clear since it was set */
-		if (test_and_clear_bit(I40E_Q_VECTOR_HUNG_DETECT,
-				       &tx_ring->q_vector->hung_detected)) {
-			netdev_info(vsi->netdev, "VSI_seid %d, Hung TX queue %d, tx_pending_hw: %d, NTC:0x%x, HWB: 0x%x, NTU: 0x%x, TAIL: 0x%x\n",
-				    vsi->seid, q_idx, tx_pending_hw,
-				    tx_ring->next_to_clean, head,
-				    tx_ring->next_to_use,
-				    readl(tx_ring->tail));
-			netdev_info(vsi->netdev, "VSI_seid %d, Issuing force_wb for TX queue %d, Interrupt Reg: 0x%x\n",
-				    vsi->seid, q_idx, val);
-			i40e_force_wb(vsi, tx_ring->q_vector);
-		} else {
-			/* First Chance - detected possible hung */
-			set_bit(I40E_Q_VECTOR_HUNG_DETECT,
-				&tx_ring->q_vector->hung_detected);
-		}
-	}
-
-	/* This is the case where we have interrupts missing,
-	 * so the tx_pending in HW will most likely be 0, but we
-	 * will have tx_pending in SW since the WB happened but the
-	 * interrupt got lost.
-	 */
-	if ((!tx_pending_hw) && i40e_get_tx_pending(tx_ring, true) &&
-	    (!(val & I40E_PFINT_DYN_CTLN_INTENA_MASK))) {
-		local_bh_disable();
-		if (napi_reschedule(&tx_ring->q_vector->napi))
-			tx_ring->tx_stats.tx_lost_interrupt++;
-		local_bh_enable();
-	}
+	if (tx_pending && (!(val & I40E_PFINT_DYN_CTLN_INTENA_MASK)))
+		i40e_force_wb(vsi, tx_ring->q_vector);
 }
 
 /**
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index a9a97dd2c0a5..e95428c7aba0 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -711,19 +711,15 @@ void i40e_free_tx_resources(struct i40e_ring *tx_ring)
 /**
  * i40e_get_tx_pending - how many tx descriptors not processed
  * @tx_ring: the ring of descriptors
- * @in_sw: is tx_pending being checked in SW or HW
  *
  * Since there is no access to the ring head register
  * in XL710, we need to use our local copies
  **/
-u32 i40e_get_tx_pending(struct i40e_ring *ring, bool in_sw)
+u32 i40e_get_tx_pending(struct i40e_ring *ring)
 {
 	u32 head, tail;
 
-	if (!in_sw)
-		head = i40e_get_head(ring);
-	else
-		head = ring->next_to_clean;
+	head = i40e_get_head(ring);
 	tail = readl(ring->tail);
 
 	if (head != tail)
@@ -846,7 +842,7 @@ static bool i40e_clean_tx_irq(struct i40e_vsi *vsi,
 		 * them to be written back in case we stay in NAPI.
 		 * In this mode on X722 we do not enable Interrupt.
 		 */
-		unsigned int j = i40e_get_tx_pending(tx_ring, false);
+		unsigned int j = i40e_get_tx_pending(tx_ring);
 
 		if (budget &&
 		    ((j / WB_STRIDE) == 0) && (j > 0) &&
@@ -2126,8 +2122,6 @@ int i40e_napi_poll(struct napi_struct *napi, int budget)
 		return 0;
 	}
 
-	/* Clear hung_detected bit */
-	clear_bit(I40E_Q_VECTOR_HUNG_DETECT, &q_vector->hung_detected);
 	/* Since the actual Tx work is minimal, we can give the Tx a larger
 	 * budget and be more aggressive about cleaning up the Tx descriptors.
 	 */
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.h b/drivers/net/ethernet/intel/i40e/i40e_txrx.h
index d6609deace57..bc66ec4954d9 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.h
@@ -275,7 +275,6 @@ struct i40e_tx_queue_stats {
 	u64 tx_done_old;
 	u64 tx_linearize;
 	u64 tx_force_wb;
-	u64 tx_lost_interrupt;
 };
 
 struct i40e_rx_queue_stats {
@@ -400,7 +399,7 @@ void i40e_free_tx_resources(struct i40e_ring *tx_ring);
 void i40e_free_rx_resources(struct i40e_ring *rx_ring);
 int i40e_napi_poll(struct napi_struct *napi, int budget);
 void i40e_force_wb(struct i40e_vsi *vsi, struct i40e_q_vector *q_vector);
-u32 i40e_get_tx_pending(struct i40e_ring *ring, bool in_sw);
+u32 i40e_get_tx_pending(struct i40e_ring *ring);
 int __i40e_maybe_stop_tx(struct i40e_ring *tx_ring, int size);
 bool __i40e_chk_linearize(struct sk_buff *skb);
 
-- 
2.12.2

[net-next 06/12] i40e: allow look-up of MAC address from Open Firmware or IDPROM

[Jeff Kirsher]

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

Look up the MAC address from the eth_get_platform_mac_address() function
first before checking what the firmware provides. We already handle the
case of re-writing the MAC-VLAN filter, so there is no need to add extra
code for this. However, update the comment where we do this to indicate
that it does impact the Open Firmware MAC address case.

Change-ID: I73e59fbe0b0e7e6f3ee9f5170d0bd3a4d5faf4db
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c | 29 ++++++++++++++++-------------
 1 file changed, 16 insertions(+), 13 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 22831a4a9099..2111f120865a 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -9326,10 +9326,15 @@ static int i40e_config_netdev(struct i40e_vsi *vsi)
 	if (vsi->type == I40E_VSI_MAIN) {
 		SET_NETDEV_DEV(netdev, &pf->pdev->dev);
 		ether_addr_copy(mac_addr, hw->mac.perm_addr);
-		/* The following steps are necessary to properly keep track of
-		 * MAC-VLAN filters loaded into firmware - first we remove
-		 * filter that is automatically generated by firmware and then
-		 * add new filter both to the driver hash table and firmware.
+		/* The following steps are necessary for two reasons. First,
+		 * some older NVM configurations load a default MAC-VLAN
+		 * filter that will accept any tagged packet, and we want to
+		 * replace this with a normal filter. Additionally, it is
+		 * possible our MAC address was provided by the platform using
+		 * Open Firmware or similar.
+		 *
+		 * Thus, we need to remove the default filter and install one
+		 * specific to the MAC address.
 		 */
 		i40e_rm_default_mac_filter(vsi, mac_addr);
 		spin_lock_bh(&vsi->mac_filter_hash_lock);
@@ -10834,20 +10839,18 @@ static void i40e_print_features(struct i40e_pf *pf)
 
 /**
  * i40e_get_platform_mac_addr - get platform-specific MAC address
- *
  * @pdev: PCI device information struct
  * @pf: board private structure
  *
- * Look up the MAC address in Open Firmware  on systems that support it,
- * and use IDPROM on SPARC if no OF address is found. On return, the
- * I40E_FLAG_PF_MAC will be wset in pf->flags if a platform-specific value
- * has been selected.
+ * Look up the MAC address for the device. First we'll try
+ * eth_platform_get_mac_address, which will check Open Firmware, or arch
+ * specific fallback. Otherwise, we'll default to the stored value in
+ * firmware.
  **/
 static void i40e_get_platform_mac_addr(struct pci_dev *pdev, struct i40e_pf *pf)
 {
-	pf->flags &= ~I40E_FLAG_PF_MAC;
-	if (!eth_platform_get_mac_address(&pdev->dev, pf->hw.mac.addr))
-		pf->flags |= I40E_FLAG_PF_MAC;
+	if (eth_platform_get_mac_address(&pdev->dev, pf->hw.mac.addr))
+		i40e_get_mac_addr(&pf->hw, pf->hw.mac.addr);
 }
 
 /**
@@ -11061,9 +11064,9 @@ static int i40e_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		i40e_aq_stop_lldp(hw, true, NULL);
 	}
 
-	i40e_get_mac_addr(hw, hw->mac.addr);
 	/* allow a platform config to override the HW addr */
 	i40e_get_platform_mac_addr(pdev, pf);
+
 	if (!is_valid_ether_addr(hw->mac.addr)) {
 		dev_info(&pdev->dev, "invalid MAC address %pM\n", hw->mac.addr);
 		err = -EIO;
-- 
2.12.2

[net-next 07/12] i40e: remove extraneous loop in i40e_vsi_wait_queues_disabled

[Jeff Kirsher]

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

We can simply check both Tx and Rx queues in a single loop, rather than
repeating the loop twice.

Change-ID: Ic06f26b0e3c2620e0e33c1a2999edda488e647ad
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 2111f120865a..0e1240b704ef 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -4440,7 +4440,7 @@ static void i40e_pf_unquiesce_all_vsi(struct i40e_pf *pf)
  * i40e_vsi_wait_queues_disabled - Wait for VSI's queues to be disabled
  * @vsi: the VSI being configured
  *
- * This function waits for the given VSI's queues to be disabled.
+ * Wait until all queues on a given VSI have been disabled.
  **/
 static int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
 {
@@ -4449,7 +4449,7 @@ static int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
 
 	pf_q = vsi->base_queue;
 	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
-		/* Check and wait for the disable status of the queue */
+		/* Check and wait for the Tx queue */
 		ret = i40e_pf_txq_wait(pf, pf_q, false);
 		if (ret) {
 			dev_info(&pf->pdev->dev,
@@ -4457,11 +4457,7 @@ static int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
 				 vsi->seid, pf_q);
 			return ret;
 		}
-	}
-
-	pf_q = vsi->base_queue;
-	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
-		/* Check and wait for the disable status of the queue */
+		/* Check and wait for the Tx queue */
 		ret = i40e_pf_rxq_wait(pf, pf_q, false);
 		if (ret) {
 			dev_info(&pf->pdev->dev,
-- 
2.12.2

[net-next 08/12] i40e: remove I40E_FLAG_NEED_LINK_UPDATE

[Jeff Kirsher]

From: Alice Michael <alice.michael@intel.com>

The I40E_FLAG_NEED_LINK_UPDATE was never used. Remove the flag
definitions.

Change-ID: If59d0c6b4af85ca27281f3183c54b055adb439a4
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@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 | 1 -
 drivers/net/ethernet/intel/i40evf/i40evf.h  | 1 -
 3 files changed, 3 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index 110ef4204306..ef64dc60f857 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -389,7 +389,6 @@ struct i40e_pf {
 #define I40E_FLAG_MSIX_ENABLED			BIT_ULL(3)
 #define I40E_FLAG_RSS_ENABLED			BIT_ULL(6)
 #define I40E_FLAG_VMDQ_ENABLED			BIT_ULL(7)
-#define I40E_FLAG_NEED_LINK_UPDATE		BIT_ULL(9)
 #define I40E_FLAG_IWARP_ENABLED			BIT_ULL(10)
 #define I40E_FLAG_CLEAN_ADMINQ			BIT_ULL(14)
 #define I40E_FLAG_FILTER_SYNC			BIT_ULL(15)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 0e1240b704ef..3b1dab7a7cc9 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -11091,7 +11091,6 @@ static int i40e_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	INIT_WORK(&pf->service_task, i40e_service_task);
 	clear_bit(__I40E_SERVICE_SCHED, &pf->state);
-	pf->flags |= I40E_FLAG_NEED_LINK_UPDATE;
 
 	/* NVM bit on means WoL disabled for the port */
 	i40e_read_nvm_word(hw, I40E_SR_NVM_WAKE_ON_LAN, &wol_nvm_bits);
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf.h b/drivers/net/ethernet/intel/i40evf/i40evf.h
index d61ecf655091..35ded19e9cc2 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf.h
+++ b/drivers/net/ethernet/intel/i40evf/i40evf.h
@@ -205,7 +205,6 @@ struct i40evf_adapter {
 #define I40EVF_FLAG_IN_NETPOLL			BIT(4)
 #define I40EVF_FLAG_IMIR_ENABLED		BIT(5)
 #define I40EVF_FLAG_MQ_CAPABLE			BIT(6)
-#define I40EVF_FLAG_NEED_LINK_UPDATE		BIT(7)
 #define I40EVF_FLAG_PF_COMMS_FAILED		BIT(8)
 #define I40EVF_FLAG_RESET_PENDING		BIT(9)
 #define I40EVF_FLAG_RESET_NEEDED		BIT(10)
-- 
2.12.2

[net-next 09/12] i40e: clean up historic deprecated flag definitions

[Jeff Kirsher]

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

Since an early commit a few flags have no longer
been used. Remove these definitions to reduce code clutter.

Change-ID: I3589be4622574e747013cd4dc403e18b039f4965
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e.h | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h
index ef64dc60f857..e987503f8517 100644
--- a/drivers/net/ethernet/intel/i40e/i40e.h
+++ b/drivers/net/ethernet/intel/i40e/i40e.h
@@ -390,11 +390,8 @@ struct i40e_pf {
 #define I40E_FLAG_RSS_ENABLED			BIT_ULL(6)
 #define I40E_FLAG_VMDQ_ENABLED			BIT_ULL(7)
 #define I40E_FLAG_IWARP_ENABLED			BIT_ULL(10)
-#define I40E_FLAG_CLEAN_ADMINQ			BIT_ULL(14)
 #define I40E_FLAG_FILTER_SYNC			BIT_ULL(15)
 #define I40E_FLAG_SERVICE_CLIENT_REQUESTED	BIT_ULL(16)
-#define I40E_FLAG_PROCESS_MDD_EVENT		BIT_ULL(17)
-#define I40E_FLAG_PROCESS_VFLR_EVENT		BIT_ULL(18)
 #define I40E_FLAG_SRIOV_ENABLED			BIT_ULL(19)
 #define I40E_FLAG_DCB_ENABLED			BIT_ULL(20)
 #define I40E_FLAG_FD_SB_ENABLED			BIT_ULL(21)
-- 
2.12.2

[net-next 10/12] i40e/i40evf: Add support for using order 1 pages with a 3K buffer

[Jeff Kirsher]

From: Alexander Duyck <alexander.h.duyck@intel.com>

There are situations where adding padding to the front and back of an Rx
buffer will require that we add additional padding.  Specifically if
NET_IP_ALIGN is non-zero, or the MTU size is larger than 7.5K we would need
to use 2K buffers which leaves us with no room for the padding.

To preemptively address these cases I am adding support for 3K buffers to
the Rx path so that we can provide the additional padding needed in the
event of NET_IP_ALIGN being non-zero or a cache line being greater than 64.

Change-ID: I938bc1ba611285428df39a613cd66f98e60b55c7
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c     |  3 ++-
 drivers/net/ethernet/intel/i40e/i40e_txrx.c     | 27 +++++++++++++------------
 drivers/net/ethernet/intel/i40e/i40e_txrx.h     | 12 +++++++++++
 drivers/net/ethernet/intel/i40evf/i40e_txrx.c   | 27 +++++++++++++------------
 drivers/net/ethernet/intel/i40evf/i40e_txrx.h   | 12 +++++++++++
 drivers/net/ethernet/intel/i40evf/i40evf_main.c |  6 ++++++
 6 files changed, 60 insertions(+), 27 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 3b1dab7a7cc9..97489d69029a 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -3085,7 +3085,8 @@ static int i40e_vsi_configure_rx(struct i40e_vsi *vsi)
 #endif
 	} else {
 		vsi->max_frame = I40E_MAX_RXBUFFER;
-		vsi->rx_buf_len = I40E_RXBUFFER_2048;
+		vsi->rx_buf_len = (PAGE_SIZE < 8192) ? I40E_RXBUFFER_3072 :
+						       I40E_RXBUFFER_2048;
 	}
 
 	/* set up individual rings */
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index e95428c7aba0..bee16726c104 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -1138,14 +1138,15 @@ void i40e_clean_rx_ring(struct i40e_ring *rx_ring)
 		dma_sync_single_range_for_cpu(rx_ring->dev,
 					      rx_bi->dma,
 					      rx_bi->page_offset,
-					      I40E_RXBUFFER_2048,
+					      rx_ring->rx_buf_len,
 					      DMA_FROM_DEVICE);
 
 		/* free resources associated with mapping */
 		dma_unmap_page_attrs(rx_ring->dev, rx_bi->dma,
-				     PAGE_SIZE,
+				     i40e_rx_pg_size(rx_ring),
 				     DMA_FROM_DEVICE,
 				     I40E_RX_DMA_ATTR);
+
 		__page_frag_cache_drain(rx_bi->page, rx_bi->pagecnt_bias);
 
 		rx_bi->page = NULL;
@@ -1267,7 +1268,7 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
 	}
 
 	/* alloc new page for storage */
-	page = dev_alloc_page();
+	page = dev_alloc_pages(i40e_rx_pg_order(rx_ring));
 	if (unlikely(!page)) {
 		rx_ring->rx_stats.alloc_page_failed++;
 		return false;
@@ -1275,7 +1276,7 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
 
 	/* map page for use */
 	dma = dma_map_page_attrs(rx_ring->dev, page, 0,
-				 PAGE_SIZE,
+				 i40e_rx_pg_size(rx_ring),
 				 DMA_FROM_DEVICE,
 				 I40E_RX_DMA_ATTR);
 
@@ -1283,7 +1284,7 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
 	 * there isn't much point in holding memory we can't use
 	 */
 	if (dma_mapping_error(rx_ring->dev, dma)) {
-		__free_pages(page, 0);
+		__free_pages(page, i40e_rx_pg_order(rx_ring));
 		rx_ring->rx_stats.alloc_page_failed++;
 		return false;
 	}
@@ -1343,7 +1344,7 @@ bool i40e_alloc_rx_buffers(struct i40e_ring *rx_ring, u16 cleaned_count)
 		/* sync the buffer for use by the device */
 		dma_sync_single_range_for_device(rx_ring->dev, bi->dma,
 						 bi->page_offset,
-						 I40E_RXBUFFER_2048,
+						 rx_ring->rx_buf_len,
 						 DMA_FROM_DEVICE);
 
 		/* Refresh the desc even if buffer_addrs didn't change
@@ -1645,9 +1646,6 @@ static inline bool i40e_page_is_reusable(struct page *page)
  **/
 static bool i40e_can_reuse_rx_page(struct i40e_rx_buffer *rx_buffer)
 {
-#if (PAGE_SIZE >= 8192)
-	unsigned int last_offset = PAGE_SIZE - I40E_RXBUFFER_2048;
-#endif
 	unsigned int pagecnt_bias = rx_buffer->pagecnt_bias;
 	struct page *page = rx_buffer->page;
 
@@ -1660,7 +1658,9 @@ static bool i40e_can_reuse_rx_page(struct i40e_rx_buffer *rx_buffer)
 	if (unlikely((page_count(page) - pagecnt_bias) > 1))
 		return false;
 #else
-	if (rx_buffer->page_offset > last_offset)
+#define I40E_LAST_OFFSET \
+	(SKB_WITH_OVERHEAD(PAGE_SIZE) - I40E_RXBUFFER_2048)
+	if (rx_buffer->page_offset > I40E_LAST_OFFSET)
 		return false;
 #endif
 
@@ -1694,7 +1694,7 @@ static void i40e_add_rx_frag(struct i40e_ring *rx_ring,
 			     unsigned int size)
 {
 #if (PAGE_SIZE < 8192)
-	unsigned int truesize = I40E_RXBUFFER_2048;
+	unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
 #else
 	unsigned int truesize = SKB_DATA_ALIGN(size);
 #endif
@@ -1755,7 +1755,7 @@ static struct sk_buff *i40e_construct_skb(struct i40e_ring *rx_ring,
 {
 	void *va = page_address(rx_buffer->page) + rx_buffer->page_offset;
 #if (PAGE_SIZE < 8192)
-	unsigned int truesize = I40E_RXBUFFER_2048;
+	unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
 #else
 	unsigned int truesize = SKB_DATA_ALIGN(size);
 #endif
@@ -1821,7 +1821,8 @@ static void i40e_put_rx_buffer(struct i40e_ring *rx_ring,
 		rx_ring->rx_stats.page_reuse_count++;
 	} else {
 		/* we are not reusing the buffer so unmap it */
-		dma_unmap_page_attrs(rx_ring->dev, rx_buffer->dma, PAGE_SIZE,
+		dma_unmap_page_attrs(rx_ring->dev, rx_buffer->dma,
+				     i40e_rx_pg_size(rx_ring),
 				     DMA_FROM_DEVICE, I40E_RX_DMA_ATTR);
 		__page_frag_cache_drain(rx_buffer->page,
 					rx_buffer->pagecnt_bias);
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.h b/drivers/net/ethernet/intel/i40e/i40e_txrx.h
index bc66ec4954d9..2f618539a436 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.h
@@ -119,6 +119,7 @@ enum i40e_dyn_idx_t {
 #define I40E_RXBUFFER_256   256
 #define I40E_RXBUFFER_1536  1536  /* 128B aligned standard Ethernet frame */
 #define I40E_RXBUFFER_2048  2048
+#define I40E_RXBUFFER_3072  3072  /* Used for large frames w/ padding */
 #define I40E_MAX_RXBUFFER   9728  /* largest size for single descriptor */
 
 /* NOTE: netdev_alloc_skb reserves up to 64 bytes, NET_IP_ALIGN means we
@@ -389,6 +390,17 @@ struct i40e_ring_container {
 #define i40e_for_each_ring(pos, head) \
 	for (pos = (head).ring; pos != NULL; pos = pos->next)
 
+static inline unsigned int i40e_rx_pg_order(struct i40e_ring *ring)
+{
+#if (PAGE_SIZE < 8192)
+	if (ring->rx_buf_len > (PAGE_SIZE / 2))
+		return 1;
+#endif
+	return 0;
+}
+
+#define i40e_rx_pg_size(_ring) (PAGE_SIZE << i40e_rx_pg_order(_ring))
+
 bool i40e_alloc_rx_buffers(struct i40e_ring *rxr, u16 cleaned_count);
 netdev_tx_t i40e_lan_xmit_frame(struct sk_buff *skb, struct net_device *netdev);
 void i40e_clean_tx_ring(struct i40e_ring *tx_ring);
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index 95e383af41c4..6b60c19a794b 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -509,14 +509,15 @@ void i40evf_clean_rx_ring(struct i40e_ring *rx_ring)
 		dma_sync_single_range_for_cpu(rx_ring->dev,
 					      rx_bi->dma,
 					      rx_bi->page_offset,
-					      I40E_RXBUFFER_2048,
+					      rx_ring->rx_buf_len,
 					      DMA_FROM_DEVICE);
 
 		/* free resources associated with mapping */
 		dma_unmap_page_attrs(rx_ring->dev, rx_bi->dma,
-				     PAGE_SIZE,
+				     i40e_rx_pg_size(rx_ring),
 				     DMA_FROM_DEVICE,
 				     I40E_RX_DMA_ATTR);
+
 		__page_frag_cache_drain(rx_bi->page, rx_bi->pagecnt_bias);
 
 		rx_bi->page = NULL;
@@ -638,7 +639,7 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
 	}
 
 	/* alloc new page for storage */
-	page = dev_alloc_page();
+	page = dev_alloc_pages(i40e_rx_pg_order(rx_ring));
 	if (unlikely(!page)) {
 		rx_ring->rx_stats.alloc_page_failed++;
 		return false;
@@ -646,7 +647,7 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
 
 	/* map page for use */
 	dma = dma_map_page_attrs(rx_ring->dev, page, 0,
-				 PAGE_SIZE,
+				 i40e_rx_pg_size(rx_ring),
 				 DMA_FROM_DEVICE,
 				 I40E_RX_DMA_ATTR);
 
@@ -654,7 +655,7 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
 	 * there isn't much point in holding memory we can't use
 	 */
 	if (dma_mapping_error(rx_ring->dev, dma)) {
-		__free_pages(page, 0);
+		__free_pages(page, i40e_rx_pg_order(rx_ring));
 		rx_ring->rx_stats.alloc_page_failed++;
 		return false;
 	}
@@ -714,7 +715,7 @@ bool i40evf_alloc_rx_buffers(struct i40e_ring *rx_ring, u16 cleaned_count)
 		/* sync the buffer for use by the device */
 		dma_sync_single_range_for_device(rx_ring->dev, bi->dma,
 						 bi->page_offset,
-						 I40E_RXBUFFER_2048,
+						 rx_ring->rx_buf_len,
 						 DMA_FROM_DEVICE);
 
 		/* Refresh the desc even if buffer_addrs didn't change
@@ -1006,9 +1007,6 @@ static inline bool i40e_page_is_reusable(struct page *page)
  **/
 static bool i40e_can_reuse_rx_page(struct i40e_rx_buffer *rx_buffer)
 {
-#if (PAGE_SIZE >= 8192)
-	unsigned int last_offset = PAGE_SIZE - I40E_RXBUFFER_2048;
-#endif
 	unsigned int pagecnt_bias = rx_buffer->pagecnt_bias;
 	struct page *page = rx_buffer->page;
 
@@ -1021,7 +1019,9 @@ static bool i40e_can_reuse_rx_page(struct i40e_rx_buffer *rx_buffer)
 	if (unlikely((page_count(page) - pagecnt_bias) > 1))
 		return false;
 #else
-	if (rx_buffer->page_offset > last_offset)
+#define I40E_LAST_OFFSET \
+	(SKB_WITH_OVERHEAD(PAGE_SIZE) - I40E_RXBUFFER_2048)
+	if (rx_buffer->page_offset > I40E_LAST_OFFSET)
 		return false;
 #endif
 
@@ -1055,7 +1055,7 @@ static void i40e_add_rx_frag(struct i40e_ring *rx_ring,
 			     unsigned int size)
 {
 #if (PAGE_SIZE < 8192)
-	unsigned int truesize = I40E_RXBUFFER_2048;
+	unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
 #else
 	unsigned int truesize = SKB_DATA_ALIGN(size);
 #endif
@@ -1116,7 +1116,7 @@ static struct sk_buff *i40e_construct_skb(struct i40e_ring *rx_ring,
 {
 	void *va = page_address(rx_buffer->page) + rx_buffer->page_offset;
 #if (PAGE_SIZE < 8192)
-	unsigned int truesize = I40E_RXBUFFER_2048;
+	unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
 #else
 	unsigned int truesize = SKB_DATA_ALIGN(size);
 #endif
@@ -1182,7 +1182,8 @@ static void i40e_put_rx_buffer(struct i40e_ring *rx_ring,
 		rx_ring->rx_stats.page_reuse_count++;
 	} else {
 		/* we are not reusing the buffer so unmap it */
-		dma_unmap_page_attrs(rx_ring->dev, rx_buffer->dma, PAGE_SIZE,
+		dma_unmap_page_attrs(rx_ring->dev, rx_buffer->dma,
+				     i40e_rx_pg_size(rx_ring),
 				     DMA_FROM_DEVICE, I40E_RX_DMA_ATTR);
 		__page_frag_cache_drain(rx_buffer->page,
 					rx_buffer->pagecnt_bias);
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.h b/drivers/net/ethernet/intel/i40evf/i40e_txrx.h
index 3bb4d732e467..dc82f65267ec 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.h
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.h
@@ -106,6 +106,7 @@ enum i40e_dyn_idx_t {
 #define I40E_RXBUFFER_256   256
 #define I40E_RXBUFFER_1536  1536  /* 128B aligned standard Ethernet frame */
 #define I40E_RXBUFFER_2048  2048
+#define I40E_RXBUFFER_3072  3072  /* Used for large frames w/ padding */
 #define I40E_MAX_RXBUFFER   9728  /* largest size for single descriptor */
 
 /* NOTE: netdev_alloc_skb reserves up to 64 bytes, NET_IP_ALIGN means we
@@ -376,6 +377,17 @@ struct i40e_ring_container {
 #define i40e_for_each_ring(pos, head) \
 	for (pos = (head).ring; pos != NULL; pos = pos->next)
 
+static inline unsigned int i40e_rx_pg_order(struct i40e_ring *ring)
+{
+#if (PAGE_SIZE < 8192)
+	if (ring->rx_buf_len > (PAGE_SIZE / 2))
+		return 1;
+#endif
+	return 0;
+}
+
+#define i40e_rx_pg_size(_ring) (PAGE_SIZE << i40e_rx_pg_order(_ring))
+
 bool i40evf_alloc_rx_buffers(struct i40e_ring *rxr, u16 cleaned_count);
 netdev_tx_t i40evf_xmit_frame(struct sk_buff *skb, struct net_device *netdev);
 void i40evf_clean_tx_ring(struct i40e_ring *tx_ring);
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
index c690aba8e8d1..7d00abae6104 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
@@ -694,6 +694,12 @@ static void i40evf_configure_rx(struct i40evf_adapter *adapter)
 	/* Legacy Rx will always default to a 2048 buffer size. */
 #if (PAGE_SIZE < 8192)
 	if (!(adapter->flags & I40EVF_FLAG_LEGACY_RX)) {
+		/* For jumbo frames on systems with 4K pages we have to use
+		 * an order 1 page, so we might as well increase the size
+		 * of our Rx buffer to make better use of the available space
+		 */
+		rx_buf_len = I40E_RXBUFFER_3072;
+
 		/* We use a 1536 buffer size for configurations with
 		 * standard Ethernet mtu.  On x86 this gives us enough room
 		 * for shared info and 192 bytes of padding.
-- 
2.12.2

[net-next 11/12] i40e/i40evf: Add support for padding start of frames

[Jeff Kirsher]

From: Alexander Duyck <alexander.h.duyck@intel.com>

This patch adds padding to the start of frames to make room for headroom
for us to eventually start using build_skb.  Right now we guarantee at
least NET_SKB_PAD + NET_IP_ALIGN, however we allocate more space if more is
available.  For example on x86 the headroom should be 192 bytes.

On systems that have too large of a cache line size to support storing 1.5K
padding and shared info we default to using 3K buffers and reserve
everything that isn't used for skb_shared_info or the data buffer for
headroom.

Change-ID: I33c641c9a1ea10cf7cc484c2d20985368d2d709a
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_main.c     |  9 +++-
 drivers/net/ethernet/intel/i40e/i40e_txrx.c     | 15 +++++-
 drivers/net/ethernet/intel/i40e/i40e_txrx.h     | 70 ++++++++++++++++++++++++-
 drivers/net/ethernet/intel/i40evf/i40e_txrx.c   | 15 +++++-
 drivers/net/ethernet/intel/i40evf/i40e_txrx.h   | 70 ++++++++++++++++++++++++-
 drivers/net/ethernet/intel/i40evf/i40evf_main.c |  8 ++-
 6 files changed, 179 insertions(+), 8 deletions(-)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 97489d69029a..b6ec9beeebff 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -3038,6 +3038,12 @@ static int i40e_configure_rx_ring(struct i40e_ring *ring)
 		return -ENOMEM;
 	}
 
+	/* configure Rx buffer alignment */
+	if (!vsi->netdev || (vsi->back->flags & I40E_FLAG_LEGACY_RX))
+		clear_ring_build_skb_enabled(ring);
+	else
+		set_ring_build_skb_enabled(ring);
+
 	/* cache tail for quicker writes, and clear the reg before use */
 	ring->tail = hw->hw_addr + I40E_QRX_TAIL(pf_q);
 	writel(0, ring->tail);
@@ -3079,7 +3085,8 @@ static int i40e_vsi_configure_rx(struct i40e_vsi *vsi)
 		vsi->max_frame = I40E_MAX_RXBUFFER;
 		vsi->rx_buf_len = I40E_RXBUFFER_2048;
 #if (PAGE_SIZE < 8192)
-	} else if (vsi->netdev->mtu <= ETH_DATA_LEN) {
+	} else if (!I40E_2K_TOO_SMALL_WITH_PADDING &&
+		   (vsi->netdev->mtu <= ETH_DATA_LEN)) {
 		vsi->max_frame = I40E_RXBUFFER_1536 - NET_IP_ALIGN;
 		vsi->rx_buf_len = I40E_RXBUFFER_1536 - NET_IP_ALIGN;
 #endif
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index bee16726c104..f15e1bcf3555 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -1248,6 +1248,17 @@ static inline void i40e_release_rx_desc(struct i40e_ring *rx_ring, u32 val)
 }
 
 /**
+ * i40e_rx_offset - Return expected offset into page to access data
+ * @rx_ring: Ring we are requesting offset of
+ *
+ * Returns the offset value for ring into the data buffer.
+ */
+static inline unsigned int i40e_rx_offset(struct i40e_ring *rx_ring)
+{
+	return ring_uses_build_skb(rx_ring) ? I40E_SKB_PAD : 0;
+}
+
+/**
  * i40e_alloc_mapped_page - recycle or make a new page
  * @rx_ring: ring to use
  * @bi: rx_buffer struct to modify
@@ -1291,7 +1302,7 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
 
 	bi->dma = dma;
 	bi->page = page;
-	bi->page_offset = 0;
+	bi->page_offset = i40e_rx_offset(rx_ring);
 
 	/* initialize pagecnt_bias to 1 representing we fully own page */
 	bi->pagecnt_bias = 1;
@@ -1696,7 +1707,7 @@ static void i40e_add_rx_frag(struct i40e_ring *rx_ring,
 #if (PAGE_SIZE < 8192)
 	unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
 #else
-	unsigned int truesize = SKB_DATA_ALIGN(size);
+	unsigned int truesize = SKB_DATA_ALIGN(size + i40e_rx_offset(rx_ring));
 #endif
 
 	skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, rx_buffer->page,
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.h b/drivers/net/ethernet/intel/i40e/i40e_txrx.h
index 2f618539a436..f5de51124cae 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.h
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.h
@@ -135,6 +135,58 @@ enum i40e_dyn_idx_t {
 #define I40E_RX_DMA_ATTR \
 	(DMA_ATTR_SKIP_CPU_SYNC | DMA_ATTR_WEAK_ORDERING)
 
+/* Attempt to maximize the headroom available for incoming frames.  We
+ * use a 2K buffer for receives and need 1536/1534 to store the data for
+ * the frame.  This leaves us with 512 bytes of room.  From that we need
+ * to deduct the space needed for the shared info and the padding needed
+ * to IP align the frame.
+ *
+ * Note: For cache line sizes 256 or larger this value is going to end
+ *	 up negative.  In these cases we should fall back to the legacy
+ *	 receive path.
+ */
+#if (PAGE_SIZE < 8192)
+#define I40E_2K_TOO_SMALL_WITH_PADDING \
+((NET_SKB_PAD + I40E_RXBUFFER_1536) > SKB_WITH_OVERHEAD(I40E_RXBUFFER_2048))
+
+static inline int i40e_compute_pad(int rx_buf_len)
+{
+	int page_size, pad_size;
+
+	page_size = ALIGN(rx_buf_len, PAGE_SIZE / 2);
+	pad_size = SKB_WITH_OVERHEAD(page_size) - rx_buf_len;
+
+	return pad_size;
+}
+
+static inline int i40e_skb_pad(void)
+{
+	int rx_buf_len;
+
+	/* If a 2K buffer cannot handle a standard Ethernet frame then
+	 * optimize padding for a 3K buffer instead of a 1.5K buffer.
+	 *
+	 * For a 3K buffer we need to add enough padding to allow for
+	 * tailroom due to NET_IP_ALIGN possibly shifting us out of
+	 * cache-line alignment.
+	 */
+	if (I40E_2K_TOO_SMALL_WITH_PADDING)
+		rx_buf_len = I40E_RXBUFFER_3072 + SKB_DATA_ALIGN(NET_IP_ALIGN);
+	else
+		rx_buf_len = I40E_RXBUFFER_1536;
+
+	/* if needed make room for NET_IP_ALIGN */
+	rx_buf_len -= NET_IP_ALIGN;
+
+	return i40e_compute_pad(rx_buf_len);
+}
+
+#define I40E_SKB_PAD i40e_skb_pad()
+#else
+#define I40E_2K_TOO_SMALL_WITH_PADDING false
+#define I40E_SKB_PAD (NET_SKB_PAD + NET_IP_ALIGN)
+#endif
+
 /**
  * i40e_test_staterr - tests bits in Rx descriptor status and error fields
  * @rx_desc: pointer to receive descriptor (in le64 format)
@@ -341,7 +393,8 @@ struct i40e_ring {
 	u8 packet_stride;
 
 	u16 flags;
-#define I40E_TXR_FLAGS_WB_ON_ITR	BIT(0)
+#define I40E_TXR_FLAGS_WB_ON_ITR		BIT(0)
+#define I40E_RXR_FLAGS_BUILD_SKB_ENABLED	BIT(1)
 
 	/* stats structs */
 	struct i40e_queue_stats	stats;
@@ -369,6 +422,21 @@ struct i40e_ring {
 					 */
 } ____cacheline_internodealigned_in_smp;
 
+static inline bool ring_uses_build_skb(struct i40e_ring *ring)
+{
+	return !!(ring->flags & I40E_RXR_FLAGS_BUILD_SKB_ENABLED);
+}
+
+static inline void set_ring_build_skb_enabled(struct i40e_ring *ring)
+{
+	ring->flags |= I40E_RXR_FLAGS_BUILD_SKB_ENABLED;
+}
+
+static inline void clear_ring_build_skb_enabled(struct i40e_ring *ring)
+{
+	ring->flags &= ~I40E_RXR_FLAGS_BUILD_SKB_ENABLED;
+}
+
 enum i40e_latency_range {
 	I40E_LOWEST_LATENCY = 0,
 	I40E_LOW_LATENCY = 1,
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index 6b60c19a794b..a71f81a9291f 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -619,6 +619,17 @@ static inline void i40e_release_rx_desc(struct i40e_ring *rx_ring, u32 val)
 }
 
 /**
+ * i40e_rx_offset - Return expected offset into page to access data
+ * @rx_ring: Ring we are requesting offset of
+ *
+ * Returns the offset value for ring into the data buffer.
+ */
+static inline unsigned int i40e_rx_offset(struct i40e_ring *rx_ring)
+{
+	return ring_uses_build_skb(rx_ring) ? I40E_SKB_PAD : 0;
+}
+
+/**
  * i40e_alloc_mapped_page - recycle or make a new page
  * @rx_ring: ring to use
  * @bi: rx_buffer struct to modify
@@ -662,7 +673,7 @@ static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
 
 	bi->dma = dma;
 	bi->page = page;
-	bi->page_offset = 0;
+	bi->page_offset = i40e_rx_offset(rx_ring);
 
 	/* initialize pagecnt_bias to 1 representing we fully own page */
 	bi->pagecnt_bias = 1;
@@ -1057,7 +1068,7 @@ static void i40e_add_rx_frag(struct i40e_ring *rx_ring,
 #if (PAGE_SIZE < 8192)
 	unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
 #else
-	unsigned int truesize = SKB_DATA_ALIGN(size);
+	unsigned int truesize = SKB_DATA_ALIGN(size + i40e_rx_offset(rx_ring));
 #endif
 
 	skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, rx_buffer->page,
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.h b/drivers/net/ethernet/intel/i40evf/i40e_txrx.h
index dc82f65267ec..901282c87cf6 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.h
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.h
@@ -122,6 +122,58 @@ enum i40e_dyn_idx_t {
 #define I40E_RX_DMA_ATTR \
 	(DMA_ATTR_SKIP_CPU_SYNC | DMA_ATTR_WEAK_ORDERING)
 
+/* Attempt to maximize the headroom available for incoming frames.  We
+ * use a 2K buffer for receives and need 1536/1534 to store the data for
+ * the frame.  This leaves us with 512 bytes of room.  From that we need
+ * to deduct the space needed for the shared info and the padding needed
+ * to IP align the frame.
+ *
+ * Note: For cache line sizes 256 or larger this value is going to end
+ *	 up negative.  In these cases we should fall back to the legacy
+ *	 receive path.
+ */
+#if (PAGE_SIZE < 8192)
+#define I40E_2K_TOO_SMALL_WITH_PADDING \
+((NET_SKB_PAD + I40E_RXBUFFER_1536) > SKB_WITH_OVERHEAD(I40E_RXBUFFER_2048))
+
+static inline int i40e_compute_pad(int rx_buf_len)
+{
+	int page_size, pad_size;
+
+	page_size = ALIGN(rx_buf_len, PAGE_SIZE / 2);
+	pad_size = SKB_WITH_OVERHEAD(page_size) - rx_buf_len;
+
+	return pad_size;
+}
+
+static inline int i40e_skb_pad(void)
+{
+	int rx_buf_len;
+
+	/* If a 2K buffer cannot handle a standard Ethernet frame then
+	 * optimize padding for a 3K buffer instead of a 1.5K buffer.
+	 *
+	 * For a 3K buffer we need to add enough padding to allow for
+	 * tailroom due to NET_IP_ALIGN possibly shifting us out of
+	 * cache-line alignment.
+	 */
+	if (I40E_2K_TOO_SMALL_WITH_PADDING)
+		rx_buf_len = I40E_RXBUFFER_3072 + SKB_DATA_ALIGN(NET_IP_ALIGN);
+	else
+		rx_buf_len = I40E_RXBUFFER_1536;
+
+	/* if needed make room for NET_IP_ALIGN */
+	rx_buf_len -= NET_IP_ALIGN;
+
+	return i40e_compute_pad(rx_buf_len);
+}
+
+#define I40E_SKB_PAD i40e_skb_pad()
+#else
+#define I40E_2K_TOO_SMALL_WITH_PADDING false
+#define I40E_SKB_PAD (NET_SKB_PAD + NET_IP_ALIGN)
+#endif
+
 /**
  * i40e_test_staterr - tests bits in Rx descriptor status and error fields
  * @rx_desc: pointer to receive descriptor (in le64 format)
@@ -328,7 +380,8 @@ struct i40e_ring {
 	u8 packet_stride;
 
 	u16 flags;
-#define I40E_TXR_FLAGS_WB_ON_ITR	BIT(0)
+#define I40E_TXR_FLAGS_WB_ON_ITR		BIT(0)
+#define I40E_RXR_FLAGS_BUILD_SKB_ENABLED	BIT(1)
 
 	/* stats structs */
 	struct i40e_queue_stats	stats;
@@ -356,6 +409,21 @@ struct i40e_ring {
 					 */
 } ____cacheline_internodealigned_in_smp;
 
+static inline bool ring_uses_build_skb(struct i40e_ring *ring)
+{
+	return !!(ring->flags & I40E_RXR_FLAGS_BUILD_SKB_ENABLED);
+}
+
+static inline void set_ring_build_skb_enabled(struct i40e_ring *ring)
+{
+	ring->flags |= I40E_RXR_FLAGS_BUILD_SKB_ENABLED;
+}
+
+static inline void clear_ring_build_skb_enabled(struct i40e_ring *ring)
+{
+	ring->flags &= ~I40E_RXR_FLAGS_BUILD_SKB_ENABLED;
+}
+
 enum i40e_latency_range {
 	I40E_LOWEST_LATENCY = 0,
 	I40E_LOW_LATENCY = 1,
diff --git a/drivers/net/ethernet/intel/i40evf/i40evf_main.c b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
index 7d00abae6104..12a930e879af 100644
--- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c
+++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c
@@ -704,7 +704,8 @@ static void i40evf_configure_rx(struct i40evf_adapter *adapter)
 		 * standard Ethernet mtu.  On x86 this gives us enough room
 		 * for shared info and 192 bytes of padding.
 		 */
-		if (netdev->mtu <= ETH_DATA_LEN)
+		if (!I40E_2K_TOO_SMALL_WITH_PADDING &&
+		    (netdev->mtu <= ETH_DATA_LEN))
 			rx_buf_len = I40E_RXBUFFER_1536 - NET_IP_ALIGN;
 	}
 #endif
@@ -712,6 +713,11 @@ static void i40evf_configure_rx(struct i40evf_adapter *adapter)
 	for (i = 0; i < adapter->num_active_queues; i++) {
 		adapter->rx_rings[i].tail = hw->hw_addr + I40E_QRX_TAIL1(i);
 		adapter->rx_rings[i].rx_buf_len = rx_buf_len;
+
+		if (adapter->flags & I40EVF_FLAG_LEGACY_RX)
+			clear_ring_build_skb_enabled(&adapter->rx_rings[i]);
+		else
+			set_ring_build_skb_enabled(&adapter->rx_rings[i]);
 	}
 }
 
-- 
2.12.2

[net-next 12/12] i40e/i40evf: Use build_skb to build frames

[Jeff Kirsher]

From: Alexander Duyck <alexander.h.duyck@intel.com>

This patch is meant to improve the performance of the Rx path.
Specifically by using build_skb we have several distinct advantages.

In the case of small frames we were previously using a copy-break approach.
This means that we were allocating a page fragment to use for skb->head,
and were having to copy the packet into that region.  Both of those calls
are now avoided since we just build the skb around the data.

In the case of large frames the gains are much more significant.
Specifically we were having to allocate skb->head, and copy the headers as
before.  However in addition we were having to parse the header using
eth_get_headlen which could be quite expensive.  All of this is avoided by
building the frame around the data.  I have seen gains as high as 30% when
using VXLAN for instance due to just header pulling overhead.

Finally with all this in place it also sets us up to start looking at
enabling XDP.  Specifically we now have a path in which the data is in the
page and the frame is built around it.  So if we parse it with XDP before
we call build_skb we can take care of any necessary processing there.

Change-ID: Id4bdd618e94473d41f892417e5d8019639e421e3
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/i40e/i40e_txrx.c   | 47 +++++++++++++++++++++++++++
 drivers/net/ethernet/intel/i40evf/i40e_txrx.c | 47 +++++++++++++++++++++++++++
 2 files changed, 94 insertions(+)

diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index f15e1bcf3555..20691d2bf113 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -1816,6 +1816,51 @@ static struct sk_buff *i40e_construct_skb(struct i40e_ring *rx_ring,
 }
 
 /**
+ * i40e_build_skb - Build skb around an existing buffer
+ * @rx_ring: Rx descriptor ring to transact packets on
+ * @rx_buffer: Rx buffer to pull data from
+ * @size: size of buffer to add to skb
+ *
+ * This function builds an skb around an existing Rx buffer, taking care
+ * to set up the skb correctly and avoid any memcpy overhead.
+ */
+static struct sk_buff *i40e_build_skb(struct i40e_ring *rx_ring,
+				      struct i40e_rx_buffer *rx_buffer,
+				      unsigned int size)
+{
+	void *va = page_address(rx_buffer->page) + rx_buffer->page_offset;
+#if (PAGE_SIZE < 8192)
+	unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
+#else
+	unsigned int truesize = SKB_DATA_ALIGN(size);
+#endif
+	struct sk_buff *skb;
+
+	/* prefetch first cache line of first page */
+	prefetch(va);
+#if L1_CACHE_BYTES < 128
+	prefetch(va + L1_CACHE_BYTES);
+#endif
+	/* build an skb around the page buffer */
+	skb = build_skb(va - I40E_SKB_PAD, truesize);
+	if (unlikely(!skb))
+		return NULL;
+
+	/* update pointers within the skb to store the data */
+	skb_reserve(skb, I40E_SKB_PAD);
+	__skb_put(skb, size);
+
+	/* buffer is used by skb, update page_offset */
+#if (PAGE_SIZE < 8192)
+	rx_buffer->page_offset ^= truesize;
+#else
+	rx_buffer->page_offset += truesize;
+#endif
+
+	return skb;
+}
+
+/**
  * i40e_put_rx_buffer - Clean up used buffer and either recycle or free
  * @rx_ring: rx descriptor ring to transact packets on
  * @rx_buffer: rx buffer to pull data from
@@ -1939,6 +1984,8 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
 		/* retrieve a buffer from the ring */
 		if (skb)
 			i40e_add_rx_frag(rx_ring, rx_buffer, skb, size);
+		else if (ring_uses_build_skb(rx_ring))
+			skb = i40e_build_skb(rx_ring, rx_buffer, size);
 		else
 			skb = i40e_construct_skb(rx_ring, rx_buffer, size);
 
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index a71f81a9291f..460171edc412 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -1177,6 +1177,51 @@ static struct sk_buff *i40e_construct_skb(struct i40e_ring *rx_ring,
 }
 
 /**
+ * i40e_build_skb - Build skb around an existing buffer
+ * @rx_ring: Rx descriptor ring to transact packets on
+ * @rx_buffer: Rx buffer to pull data from
+ * @size: size of buffer to add to skb
+ *
+ * This function builds an skb around an existing Rx buffer, taking care
+ * to set up the skb correctly and avoid any memcpy overhead.
+ */
+static struct sk_buff *i40e_build_skb(struct i40e_ring *rx_ring,
+				      struct i40e_rx_buffer *rx_buffer,
+				      unsigned int size)
+{
+	void *va = page_address(rx_buffer->page) + rx_buffer->page_offset;
+#if (PAGE_SIZE < 8192)
+	unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
+#else
+	unsigned int truesize = SKB_DATA_ALIGN(size);
+#endif
+	struct sk_buff *skb;
+
+	/* prefetch first cache line of first page */
+	prefetch(va);
+#if L1_CACHE_BYTES < 128
+	prefetch(va + L1_CACHE_BYTES);
+#endif
+	/* build an skb around the page buffer */
+	skb = build_skb(va - I40E_SKB_PAD, truesize);
+	if (unlikely(!skb))
+		return NULL;
+
+	/* update pointers within the skb to store the data */
+	skb_reserve(skb, I40E_SKB_PAD);
+	__skb_put(skb, size);
+
+	/* buffer is used by skb, update page_offset */
+#if (PAGE_SIZE < 8192)
+	rx_buffer->page_offset ^= truesize;
+#else
+	rx_buffer->page_offset += truesize;
+#endif
+
+	return skb;
+}
+
+/**
  * i40e_put_rx_buffer - Clean up used buffer and either recycle or free
  * @rx_ring: rx descriptor ring to transact packets on
  * @rx_buffer: rx buffer to pull data from
@@ -1295,6 +1340,8 @@ static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
 		/* retrieve a buffer from the ring */
 		if (skb)
 			i40e_add_rx_frag(rx_ring, rx_buffer, skb, size);
+		else if (ring_uses_build_skb(rx_ring))
+			skb = i40e_build_skb(rx_ring, rx_buffer, size);
 		else
 			skb = i40e_construct_skb(rx_ring, rx_buffer, size);
 
-- 
2.12.2

Re: [net-next 00/12][pull request] 40GbE Intel Wired LAN Driver Updates 2017-04-08

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Sat,  8 Apr 2017 02:55:41 -0700

> This series contains updates to i40e and i40evf only.

Pulled, thanks Jeff.