[net-next 00/11][pull request] Intel Wired LAN Driver Updates

[net-next 00/11][pull request] Intel Wired LAN Driver Updates

[Jeff Kirsher]

The following series contains updates to e1000, e1000e, igb and
netdev/ixgbe.  There are 2 fixes and the remaining patches are
either add support or cleanup.

Here is a list of the new support added:
 - igb adds support for byte queue limits and basic runtime PM
 - e1000e adds Receive Packet Steering (RPS)
 - FCoE adds ndo_get_fcoe_hbainfo() call

The following are changes since commit fa0f5aa74316c636427ac92dad0bc5714c34ca17:
  net_sched: qdisc_alloc_handle() can be too slow
and are available in the git repository at:
  git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next master

Bruce Allan (5):
  e1000e: cleanup Rx checksum offload code
  e1000e: add Receive Packet Steering (RPS) support
  e1000e: re-enable alternate MAC address for all devices which support
    it
  e1000e: convert head, tail and itr_register offsets to __iomem
    pointers
  e1000e: pass pointer to ring struct instead of adapter struct

Eric Dumazet (1):
  igb: Add support for byte queue limits.

Jesse Brandeburg (1):
  e1000: fix lockdep splat in shutdown handler

Koki Sanagi (1):
  igb: reset PHY after recovering from PHY power down

Neerav Parikh (2):
  netdev: FCoE: Add new ndo_get_fcoe_hbainfo() call
  ixgbe: FCoE: Add support for ndo_get_fcoe_hbainfo() call

Yan, Zheng (1):
  igb: add basic runtime PM support

 drivers/net/ethernet/intel/e1000/e1000_main.c |    8 +-
 drivers/net/ethernet/intel/e1000e/defines.h   |    7 +
 drivers/net/ethernet/intel/e1000e/e1000.h     |   27 +-
 drivers/net/ethernet/intel/e1000e/ethtool.c   |   10 +-
 drivers/net/ethernet/intel/e1000e/hw.h        |    9 +-
 drivers/net/ethernet/intel/e1000e/lib.c       |    7 +-
 drivers/net/ethernet/intel/e1000e/netdev.c    |  397 +++++++++++++++----------
 drivers/net/ethernet/intel/igb/igb.h          |    5 +
 drivers/net/ethernet/intel/igb/igb_ethtool.c  |   16 +
 drivers/net/ethernet/intel/igb/igb_main.c     |  142 ++++++++--
 drivers/net/ethernet/intel/ixgbe/ixgbe.h      |    3 +
 drivers/net/ethernet/intel/ixgbe/ixgbe_fcoe.c |   83 +++++
 drivers/net/ethernet/intel/ixgbe/ixgbe_main.c |    5 +-
 include/linux/netdevice.h                     |   26 ++
 14 files changed, 541 insertions(+), 204 deletions(-)

-- 
1.7.7.4

[net-next 01/11] e1000: fix lockdep splat in shutdown handler

[Jeff Kirsher]

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

As reported by Steven Rostedt, e1000 has a lockdep splat added
during the recent merge window.  The issue is that
cancel_delayed_work is called while holding our private mutex.

There is no reason that I can see to hold the mutex during pci
shutdown, it was more just paranoia that I put the mutex_lock
around the call to e1000_down.

In a quick survey lots of drivers handle locking differently when
being called by the pci layer.  The assumption here is that we
don't need the mutexes' protection in this function because
the driver could not be unloaded while in the shutdown handler
which is only called at reboot or poweroff.

Reported-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Steven Rostedt <rostedt@goodmis.org>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/e1000/e1000_main.c |    8 +-------
 1 files changed, 1 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/intel/e1000/e1000_main.c b/drivers/net/ethernet/intel/e1000/e1000_main.c
index 985d589..934d5aa 100644
--- a/drivers/net/ethernet/intel/e1000/e1000_main.c
+++ b/drivers/net/ethernet/intel/e1000/e1000_main.c
@@ -4724,8 +4724,6 @@ static int __e1000_shutdown(struct pci_dev *pdev, bool *enable_wake)
 
 	netif_device_detach(netdev);
 
-	mutex_lock(&adapter->mutex);
-
 	if (netif_running(netdev)) {
 		WARN_ON(test_bit(__E1000_RESETTING, &adapter->flags));
 		e1000_down(adapter);
@@ -4733,10 +4731,8 @@ static int __e1000_shutdown(struct pci_dev *pdev, bool *enable_wake)
 
 #ifdef CONFIG_PM
 	retval = pci_save_state(pdev);
-	if (retval) {
-		mutex_unlock(&adapter->mutex);
+	if (retval)
 		return retval;
-	}
 #endif
 
 	status = er32(STATUS);
@@ -4791,8 +4787,6 @@ static int __e1000_shutdown(struct pci_dev *pdev, bool *enable_wake)
 	if (netif_running(netdev))
 		e1000_free_irq(adapter);
 
-	mutex_unlock(&adapter->mutex);
-
 	pci_disable_device(pdev);
 
 	return 0;
-- 
1.7.7.4

[net-next 02/11] igb: Add support for byte queue limits.

[Jeff Kirsher]

From: Eric Dumazet <eric.dumazet@gmail.com>

This adds support for byte queue limits (BQL)

Since this driver collects bytes count in 'bytecount' field, use it also
in igb_tx_map()

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Alexander Duyck <alexander.h.duyck@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/igb.h      |    5 +++++
 drivers/net/ethernet/intel/igb/igb_main.c |    5 +++++
 2 files changed, 10 insertions(+), 0 deletions(-)

diff --git a/drivers/net/ethernet/intel/igb/igb.h b/drivers/net/ethernet/intel/igb/igb.h
index c69feeb..3d12e67 100644
--- a/drivers/net/ethernet/intel/igb/igb.h
+++ b/drivers/net/ethernet/intel/igb/igb.h
@@ -447,4 +447,9 @@ static inline s32 igb_get_phy_info(struct e1000_hw *hw)
 	return 0;
 }
 
+static inline struct netdev_queue *txring_txq(const struct igb_ring *tx_ring)
+{
+	return netdev_get_tx_queue(tx_ring->netdev, tx_ring->queue_index);
+}
+
 #endif /* _IGB_H_ */
diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c
index 89d576c..dcc68cc 100644
--- a/drivers/net/ethernet/intel/igb/igb_main.c
+++ b/drivers/net/ethernet/intel/igb/igb_main.c
@@ -3201,6 +3201,7 @@ static void igb_clean_tx_ring(struct igb_ring *tx_ring)
 		buffer_info = &tx_ring->tx_buffer_info[i];
 		igb_unmap_and_free_tx_resource(tx_ring, buffer_info);
 	}
+	netdev_tx_reset_queue(txring_txq(tx_ring));
 
 	size = sizeof(struct igb_tx_buffer) * tx_ring->count;
 	memset(tx_ring->tx_buffer_info, 0, size);
@@ -4238,6 +4239,8 @@ static void igb_tx_map(struct igb_ring *tx_ring,
 		frag++;
 	}
 
+	netdev_tx_sent_queue(txring_txq(tx_ring), first->bytecount);
+
 	/* write last descriptor with RS and EOP bits */
 	cmd_type |= cpu_to_le32(size) | cpu_to_le32(IGB_TXD_DCMD);
 	tx_desc->read.cmd_type_len = cmd_type;
@@ -5777,6 +5780,8 @@ static bool igb_clean_tx_irq(struct igb_q_vector *q_vector)
 		}
 	}
 
+	netdev_tx_completed_queue(txring_txq(tx_ring),
+				  total_packets, total_bytes);
 	i += tx_ring->count;
 	tx_ring->next_to_clean = i;
 	u64_stats_update_begin(&tx_ring->tx_syncp);
-- 
1.7.7.4

[net-next 03/11] igb: add basic runtime PM support

[Jeff Kirsher]

From: "Yan, Zheng" <zheng.z.yan@intel.com>

Use the runtime power management framework to add basic runtime PM support
to the igb driver. Namely, make the driver suspend the device when the link
is off and set it up for generating a wakeup event after the link has been
detected again. This feature is disabled by default.

Based on e1000e's runtime PM code.

Changes since v1:
Don't suspend the device when shutting down the interface.
Avoid race between runtime suspending and ethtool operations.

Signed-off-by: Zheng Yan <zheng.z.yan@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/igb_ethtool.c |   16 +++
 drivers/net/ethernet/intel/igb/igb_main.c    |  136 ++++++++++++++++++++++----
 2 files changed, 133 insertions(+), 19 deletions(-)

diff --git a/drivers/net/ethernet/intel/igb/igb_ethtool.c b/drivers/net/ethernet/intel/igb/igb_ethtool.c
index f1206be..7998bf4 100644
--- a/drivers/net/ethernet/intel/igb/igb_ethtool.c
+++ b/drivers/net/ethernet/intel/igb/igb_ethtool.c
@@ -36,6 +36,7 @@
 #include <linux/ethtool.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
+#include <linux/pm_runtime.h>
 
 #include "igb.h"
 
@@ -2161,6 +2162,19 @@ static void igb_get_strings(struct net_device *netdev, u32 stringset, u8 *data)
 	}
 }
 
+static int igb_ethtool_begin(struct net_device *netdev)
+{
+	struct igb_adapter *adapter = netdev_priv(netdev);
+	pm_runtime_get_sync(&adapter->pdev->dev);
+	return 0;
+}
+
+static void igb_ethtool_complete(struct net_device *netdev)
+{
+	struct igb_adapter *adapter = netdev_priv(netdev);
+	pm_runtime_put(&adapter->pdev->dev);
+}
+
 static const struct ethtool_ops igb_ethtool_ops = {
 	.get_settings           = igb_get_settings,
 	.set_settings           = igb_set_settings,
@@ -2187,6 +2201,8 @@ static const struct ethtool_ops igb_ethtool_ops = {
 	.get_ethtool_stats      = igb_get_ethtool_stats,
 	.get_coalesce           = igb_get_coalesce,
 	.set_coalesce           = igb_set_coalesce,
+	.begin			= igb_ethtool_begin,
+	.complete		= igb_ethtool_complete,
 };
 
 void igb_set_ethtool_ops(struct net_device *netdev)
diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c
index dcc68cc..fac71e2 100644
--- a/drivers/net/ethernet/intel/igb/igb_main.c
+++ b/drivers/net/ethernet/intel/igb/igb_main.c
@@ -53,6 +53,7 @@
 #include <linux/if_ether.h>
 #include <linux/aer.h>
 #include <linux/prefetch.h>
+#include <linux/pm_runtime.h>
 #ifdef CONFIG_IGB_DCA
 #include <linux/dca.h>
 #endif
@@ -172,8 +173,18 @@ static int igb_check_vf_assignment(struct igb_adapter *adapter);
 #endif
 
 #ifdef CONFIG_PM
-static int igb_suspend(struct pci_dev *, pm_message_t);
-static int igb_resume(struct pci_dev *);
+static int igb_suspend(struct device *);
+static int igb_resume(struct device *);
+#ifdef CONFIG_PM_RUNTIME
+static int igb_runtime_suspend(struct device *dev);
+static int igb_runtime_resume(struct device *dev);
+static int igb_runtime_idle(struct device *dev);
+#endif
+static const struct dev_pm_ops igb_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(igb_suspend, igb_resume)
+	SET_RUNTIME_PM_OPS(igb_runtime_suspend, igb_runtime_resume,
+			igb_runtime_idle)
+};
 #endif
 static void igb_shutdown(struct pci_dev *);
 #ifdef CONFIG_IGB_DCA
@@ -214,9 +225,7 @@ static struct pci_driver igb_driver = {
 	.probe    = igb_probe,
 	.remove   = __devexit_p(igb_remove),
 #ifdef CONFIG_PM
-	/* Power Management Hooks */
-	.suspend  = igb_suspend,
-	.resume   = igb_resume,
+	.driver.pm = &igb_pm_ops,
 #endif
 	.shutdown = igb_shutdown,
 	.err_handler = &igb_err_handler
@@ -2111,6 +2120,8 @@ static int __devinit igb_probe(struct pci_dev *pdev,
 	default:
 		break;
 	}
+
+	pm_runtime_put_noidle(&pdev->dev);
 	return 0;
 
 err_register:
@@ -2150,6 +2161,8 @@ static void __devexit igb_remove(struct pci_dev *pdev)
 	struct igb_adapter *adapter = netdev_priv(netdev);
 	struct e1000_hw *hw = &adapter->hw;
 
+	pm_runtime_get_noresume(&pdev->dev);
+
 	/*
 	 * The watchdog timer may be rescheduled, so explicitly
 	 * disable watchdog from being rescheduled.
@@ -2472,16 +2485,22 @@ static int __devinit igb_sw_init(struct igb_adapter *adapter)
  * handler is registered with the OS, the watchdog timer is started,
  * and the stack is notified that the interface is ready.
  **/
-static int igb_open(struct net_device *netdev)
+static int __igb_open(struct net_device *netdev, bool resuming)
 {
 	struct igb_adapter *adapter = netdev_priv(netdev);
 	struct e1000_hw *hw = &adapter->hw;
+	struct pci_dev *pdev = adapter->pdev;
 	int err;
 	int i;
 
 	/* disallow open during test */
-	if (test_bit(__IGB_TESTING, &adapter->state))
+	if (test_bit(__IGB_TESTING, &adapter->state)) {
+		WARN_ON(resuming);
 		return -EBUSY;
+	}
+
+	if (!resuming)
+		pm_runtime_get_sync(&pdev->dev);
 
 	netif_carrier_off(netdev);
 
@@ -2527,6 +2546,9 @@ static int igb_open(struct net_device *netdev)
 
 	netif_tx_start_all_queues(netdev);
 
+	if (!resuming)
+		pm_runtime_put(&pdev->dev);
+
 	/* start the watchdog. */
 	hw->mac.get_link_status = 1;
 	schedule_work(&adapter->watchdog_task);
@@ -2541,10 +2563,17 @@ err_setup_rx:
 	igb_free_all_tx_resources(adapter);
 err_setup_tx:
 	igb_reset(adapter);
+	if (!resuming)
+		pm_runtime_put(&pdev->dev);
 
 	return err;
 }
 
+static int igb_open(struct net_device *netdev)
+{
+	return __igb_open(netdev, false);
+}
+
 /**
  * igb_close - Disables a network interface
  * @netdev: network interface device structure
@@ -2556,21 +2585,32 @@ err_setup_tx:
  * needs to be disabled.  A global MAC reset is issued to stop the
  * hardware, and all transmit and receive resources are freed.
  **/
-static int igb_close(struct net_device *netdev)
+static int __igb_close(struct net_device *netdev, bool suspending)
 {
 	struct igb_adapter *adapter = netdev_priv(netdev);
+	struct pci_dev *pdev = adapter->pdev;
 
 	WARN_ON(test_bit(__IGB_RESETTING, &adapter->state));
-	igb_down(adapter);
 
+	if (!suspending)
+		pm_runtime_get_sync(&pdev->dev);
+
+	igb_down(adapter);
 	igb_free_irq(adapter);
 
 	igb_free_all_tx_resources(adapter);
 	igb_free_all_rx_resources(adapter);
 
+	if (!suspending)
+		pm_runtime_put_sync(&pdev->dev);
 	return 0;
 }
 
+static int igb_close(struct net_device *netdev)
+{
+	return __igb_close(netdev, false);
+}
+
 /**
  * igb_setup_tx_resources - allocate Tx resources (Descriptors)
  * @tx_ring: tx descriptor ring (for a specific queue) to setup
@@ -3631,6 +3671,9 @@ static void igb_watchdog_task(struct work_struct *work)
 
 	link = igb_has_link(adapter);
 	if (link) {
+		/* Cancel scheduled suspend requests. */
+		pm_runtime_resume(netdev->dev.parent);
+
 		if (!netif_carrier_ok(netdev)) {
 			u32 ctrl;
 			hw->mac.ops.get_speed_and_duplex(hw,
@@ -3702,6 +3745,9 @@ static void igb_watchdog_task(struct work_struct *work)
 			if (!test_bit(__IGB_DOWN, &adapter->state))
 				mod_timer(&adapter->phy_info_timer,
 					  round_jiffies(jiffies + 2 * HZ));
+
+			pm_schedule_suspend(netdev->dev.parent,
+					    MSEC_PER_SEC * 5);
 		}
 	}
 
@@ -6588,13 +6634,14 @@ err_inval:
 	return -EINVAL;
 }
 
-static int __igb_shutdown(struct pci_dev *pdev, bool *enable_wake)
+static int __igb_shutdown(struct pci_dev *pdev, bool *enable_wake,
+			  bool runtime)
 {
 	struct net_device *netdev = pci_get_drvdata(pdev);
 	struct igb_adapter *adapter = netdev_priv(netdev);
 	struct e1000_hw *hw = &adapter->hw;
 	u32 ctrl, rctl, status;
-	u32 wufc = adapter->wol;
+	u32 wufc = runtime ? E1000_WUFC_LNKC : adapter->wol;
 #ifdef CONFIG_PM
 	int retval = 0;
 #endif
@@ -6602,7 +6649,7 @@ static int __igb_shutdown(struct pci_dev *pdev, bool *enable_wake)
 	netif_device_detach(netdev);
 
 	if (netif_running(netdev))
-		igb_close(netdev);
+		__igb_close(netdev, true);
 
 	igb_clear_interrupt_scheme(adapter);
 
@@ -6661,12 +6708,13 @@ static int __igb_shutdown(struct pci_dev *pdev, bool *enable_wake)
 }
 
 #ifdef CONFIG_PM
-static int igb_suspend(struct pci_dev *pdev, pm_message_t state)
+static int igb_suspend(struct device *dev)
 {
 	int retval;
 	bool wake;
+	struct pci_dev *pdev = to_pci_dev(dev);
 
-	retval = __igb_shutdown(pdev, &wake);
+	retval = __igb_shutdown(pdev, &wake, 0);
 	if (retval)
 		return retval;
 
@@ -6680,8 +6728,9 @@ static int igb_suspend(struct pci_dev *pdev, pm_message_t state)
 	return 0;
 }
 
-static int igb_resume(struct pci_dev *pdev)
+static int igb_resume(struct device *dev)
 {
+	struct pci_dev *pdev = to_pci_dev(dev);
 	struct net_device *netdev = pci_get_drvdata(pdev);
 	struct igb_adapter *adapter = netdev_priv(netdev);
 	struct e1000_hw *hw = &adapter->hw;
@@ -6702,7 +6751,18 @@ static int igb_resume(struct pci_dev *pdev)
 	pci_enable_wake(pdev, PCI_D3hot, 0);
 	pci_enable_wake(pdev, PCI_D3cold, 0);
 
-	if (igb_init_interrupt_scheme(adapter)) {
+	if (!rtnl_is_locked()) {
+		/*
+		 * shut up ASSERT_RTNL() warning in
+		 * netif_set_real_num_tx/rx_queues.
+		 */
+		rtnl_lock();
+		err = igb_init_interrupt_scheme(adapter);
+		rtnl_unlock();
+	} else {
+		err = igb_init_interrupt_scheme(adapter);
+	}
+	if (err) {
 		dev_err(&pdev->dev, "Unable to allocate memory for queues\n");
 		return -ENOMEM;
 	}
@@ -6715,23 +6775,61 @@ static int igb_resume(struct pci_dev *pdev)
 
 	wr32(E1000_WUS, ~0);
 
-	if (netif_running(netdev)) {
-		err = igb_open(netdev);
+	if (netdev->flags & IFF_UP) {
+		err = __igb_open(netdev, true);
 		if (err)
 			return err;
 	}
 
 	netif_device_attach(netdev);
+	return 0;
+}
+
+#ifdef CONFIG_PM_RUNTIME
+static int igb_runtime_idle(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct net_device *netdev = pci_get_drvdata(pdev);
+	struct igb_adapter *adapter = netdev_priv(netdev);
+
+	if (!igb_has_link(adapter))
+		pm_schedule_suspend(dev, MSEC_PER_SEC * 5);
+
+	return -EBUSY;
+}
+
+static int igb_runtime_suspend(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	int retval;
+	bool wake;
+
+	retval = __igb_shutdown(pdev, &wake, 1);
+	if (retval)
+		return retval;
+
+	if (wake) {
+		pci_prepare_to_sleep(pdev);
+	} else {
+		pci_wake_from_d3(pdev, false);
+		pci_set_power_state(pdev, PCI_D3hot);
+	}
 
 	return 0;
 }
+
+static int igb_runtime_resume(struct device *dev)
+{
+	return igb_resume(dev);
+}
+#endif /* CONFIG_PM_RUNTIME */
 #endif
 
 static void igb_shutdown(struct pci_dev *pdev)
 {
 	bool wake;
 
-	__igb_shutdown(pdev, &wake);
+	__igb_shutdown(pdev, &wake, 0);
 
 	if (system_state == SYSTEM_POWER_OFF) {
 		pci_wake_from_d3(pdev, wake);
-- 
1.7.7.4

[net-next 04/11] igb: reset PHY after recovering from PHY power down

[Jeff Kirsher]

From: Koki Sanagi <sanagi.koki@jp.fujitsu.com>

According to 82576_Datasheet.pdf, PHY setting is lost after PHY power down.
So resetting PHY is needed when recovering from PHY power down to set a default
setting to PHY register.
Owing to this lack, NIC doesn't link up in some rare situation.

Signed-off-by: Koki Sanagi <sanagi.koki@jp.fujitsu.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/igb_main.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c
index fac71e2..01e5e89 100644
--- a/drivers/net/ethernet/intel/igb/igb_main.c
+++ b/drivers/net/ethernet/intel/igb/igb_main.c
@@ -1507,6 +1507,7 @@ void igb_power_up_link(struct igb_adapter *adapter)
 		igb_power_up_phy_copper(&adapter->hw);
 	else
 		igb_power_up_serdes_link_82575(&adapter->hw);
+	igb_reset_phy(&adapter->hw);
 }
 
 /**
-- 
1.7.7.4

[net-next 05/11] netdev: FCoE: Add new ndo_get_fcoe_hbainfo() call

[Jeff Kirsher]

From: Neerav Parikh <Neerav.Parikh@intel.com>

This adds a new ndo_get_fcoe_hbainfo() call in
net_device_ops for FCoE protocol stack.

If supported by the underlying device, the FCoE protocol
stack will call this to get device specific information
from the underlying device.
This information will then be utilized by the FCoE protocol
stack to register Fiber Channel HBA attributes with the
Fiber Channel Management Service via Fabric Device
Management Interface (FDMI) as per the T11 FC-GS
specification.

Signed-off-by: Neerav Parikh <Neerav.Parikh@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 include/linux/netdevice.h |   26 ++++++++++++++++++++++++++
 1 files changed, 26 insertions(+), 0 deletions(-)

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index a776a67..4009ce6 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -707,6 +707,23 @@ struct netdev_tc_txq {
 	u16 offset;
 };
 
+#if defined(CONFIG_FCOE) || defined(CONFIG_FCOE_MODULE)
+/*
+ * This structure is to hold information about the device
+ * configured to run FCoE protocol stack.
+ */
+struct netdev_fcoe_hbainfo {
+	char	manufacturer[64];
+	char	serial_number[64];
+	char	hardware_version[64];
+	char	driver_version[64];
+	char	optionrom_version[64];
+	char	firmware_version[64];
+	char	model[256];
+	char	model_description[256];
+};
+#endif
+
 /*
  * This structure defines the management hooks for network devices.
  * The following hooks can be defined; unless noted otherwise, they are
@@ -847,6 +864,13 @@ struct netdev_tc_txq {
  *	perform necessary setup and returns 1 to indicate the device is set up
  *	successfully to perform DDP on this I/O, otherwise this returns 0.
  *
+ * int (*ndo_fcoe_get_hbainfo)(struct net_device *dev,
+ *			       struct netdev_fcoe_hbainfo *hbainfo);
+ *	Called when the FCoE Protocol stack wants information on the underlying
+ *	device. This information is utilized by the FCoE protocol stack to
+ *	register attributes with Fiber Channel management service as per the
+ *	FC-GS Fabric Device Management Information(FDMI) specification.
+ *
  * int (*ndo_fcoe_get_wwn)(struct net_device *dev, u64 *wwn, int type);
  *	Called when the underlying device wants to override default World Wide
  *	Name (WWN) generation mechanism in FCoE protocol stack to pass its own
@@ -950,6 +974,8 @@ struct net_device_ops {
 						       u16 xid,
 						       struct scatterlist *sgl,
 						       unsigned int sgc);
+	int			(*ndo_fcoe_get_hbainfo)(struct net_device *dev,
+					struct netdev_fcoe_hbainfo *hbainfo);
 #endif
 
 #if IS_ENABLED(CONFIG_LIBFCOE)
-- 
1.7.7.4

[net-next 06/11] ixgbe: FCoE: Add support for ndo_get_fcoe_hbainfo() call

[Jeff Kirsher]

From: Neerav Parikh <Neerav.Parikh@intel.com>

This patch implements support for ndo_get_fcoe_hbainfo()
call in the ixgbe driver.

This function will be called by the FCoE protocol stack to
obtain device specific information from the underlying
device configured to do FCoE.

Signed-off-by: Neerav Parikh <Neerav.Parikh@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
 drivers/net/ethernet/intel/ixgbe/ixgbe.h      |    3 +
 drivers/net/ethernet/intel/ixgbe/ixgbe_fcoe.c |   83 +++++++++++++++++++++++++
 drivers/net/ethernet/intel/ixgbe/ixgbe_main.c |    5 +-
 3 files changed, 90 insertions(+), 1 deletions(-)

diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe.h b/drivers/net/ethernet/intel/ixgbe/ixgbe.h
index a8368d5..258164d 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe.h
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe.h
@@ -560,6 +560,7 @@ extern int ixgbe_copy_dcb_cfg(struct ixgbe_dcb_config *src_dcb_cfg,
 
 extern char ixgbe_driver_name[];
 extern const char ixgbe_driver_version[];
+extern char ixgbe_default_device_descr[];
 
 extern void ixgbe_up(struct ixgbe_adapter *adapter);
 extern void ixgbe_down(struct ixgbe_adapter *adapter);
@@ -627,6 +628,8 @@ extern u8 ixgbe_fcoe_getapp(struct ixgbe_adapter *adapter);
 extern u8 ixgbe_fcoe_setapp(struct ixgbe_adapter *adapter, u8 up);
 #endif /* CONFIG_IXGBE_DCB */
 extern int ixgbe_fcoe_get_wwn(struct net_device *netdev, u64 *wwn, int type);
+extern int ixgbe_fcoe_get_hbainfo(struct net_device *netdev,
+				  struct netdev_fcoe_hbainfo *info);
 #endif /* IXGBE_FCOE */
 
 #endif /* _IXGBE_H_ */
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_fcoe.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_fcoe.c
index df3b1be..d18d615 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_fcoe.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_fcoe.c
@@ -855,3 +855,86 @@ int ixgbe_fcoe_get_wwn(struct net_device *netdev, u64 *wwn, int type)
 	}
 	return rc;
 }
+
+/**
+ * ixgbe_fcoe_get_hbainfo - get FCoE HBA information
+ * @netdev : ixgbe adapter
+ * @info : HBA information
+ *
+ * Returns ixgbe HBA information
+ *
+ * Returns : 0 on success
+ */
+int ixgbe_fcoe_get_hbainfo(struct net_device *netdev,
+			   struct netdev_fcoe_hbainfo *info)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+	struct ixgbe_hw *hw = &adapter->hw;
+	int i, pos;
+	u8 buf[8];
+
+	if (!info)
+		return -EINVAL;
+
+	/* Don't return information on unsupported devices */
+	if (hw->mac.type != ixgbe_mac_82599EB &&
+	    hw->mac.type != ixgbe_mac_X540)
+		return -EINVAL;
+
+	/* Manufacturer */
+	snprintf(info->manufacturer, sizeof(info->manufacturer),
+		 "Intel Corporation");
+
+	/* Serial Number */
+
+	/* Get the PCI-e Device Serial Number Capability */
+	pos = pci_find_ext_capability(adapter->pdev, PCI_EXT_CAP_ID_DSN);
+	if (pos) {
+		pos += 4;
+		for (i = 0; i < 8; i++)
+			pci_read_config_byte(adapter->pdev, pos + i, &buf[i]);
+
+		snprintf(info->serial_number, sizeof(info->serial_number),
+			 "%02X%02X%02X%02X%02X%02X%02X%02X",
+			 buf[7], buf[6], buf[5], buf[4],
+			 buf[3], buf[2], buf[1], buf[0]);
+	} else
+		snprintf(info->serial_number, sizeof(info->serial_number),
+			 "Unknown");
+
+	/* Hardware Version */
+	snprintf(info->hardware_version,
+		 sizeof(info->hardware_version),
+		 "Rev %d", hw->revision_id);
+	/* Driver Name/Version */
+	snprintf(info->driver_version,
+		 sizeof(info->driver_version),
+		 "%s v%s",
+		 ixgbe_driver_name,
+		 ixgbe_driver_version);
+	/* Firmware Version */
+	snprintf(info->firmware_version,
+		 sizeof(info->firmware_version),
+		 "0x%08x",
+		 (adapter->eeprom_verh << 16) |
+		  adapter->eeprom_verl);
+
+	/* Model */
+	if (hw->mac.type == ixgbe_mac_82599EB) {
+		snprintf(info->model,
+			 sizeof(info->model),
+			 "Intel 82599");
+	} else {
+		snprintf(info->model,
+			 sizeof(info->model),
+			 "Intel X540");
+	}
+
+	/* Model Description */
+	snprintf(info->model_description,
+		 sizeof(info->model_description),
+		 "%s",
+		 ixgbe_default_device_descr);
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
index 74669a8..1ee5d0f 100644
--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
@@ -55,6 +55,8 @@
 char ixgbe_driver_name[] = "ixgbe";
 static const char ixgbe_driver_string[] =
 			      "Intel(R) 10 Gigabit PCI Express Network Driver";
+char ixgbe_default_device_descr[] =
+			      "Intel(R) 10 Gigabit Network Connection";
 #define MAJ 3
 #define MIN 6
 #define BUILD 7
@@ -7293,6 +7295,7 @@ static const struct net_device_ops ixgbe_netdev_ops = {
 	.ndo_fcoe_enable = ixgbe_fcoe_enable,
 	.ndo_fcoe_disable = ixgbe_fcoe_disable,
 	.ndo_fcoe_get_wwn = ixgbe_fcoe_get_wwn,
+	.ndo_fcoe_get_hbainfo = ixgbe_fcoe_get_hbainfo,
 #endif /* IXGBE_FCOE */
 	.ndo_set_features = ixgbe_set_features,
 	.ndo_fix_features = ixgbe_fix_features,
@@ -7722,7 +7725,7 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
 	/* add san mac addr to netdev */
 	ixgbe_add_sanmac_netdev(netdev);
 
-	e_dev_info("Intel(R) 10 Gigabit Network Connection\n");
+	e_dev_info("%s\n", ixgbe_default_device_descr);
 	cards_found++;
 	return 0;
 
-- 
1.7.7.4

[net-next 07/11] e1000e: cleanup Rx checksum offload code

[Jeff Kirsher]

From: Bruce Allan <bruce.w.allan@intel.com>

1) cleanup whitespace in e1000_rx_checksum() function header comment
2) do not check hardware checksum when Rx checksum is disabled
3) reduce duplicated calls to le16_to_cpu() by just using it within
   e1000_rx_checksum() instead of in each call to the function

Signed-off-by: Bruce Allan <bruce.w.allan@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/e1000e/netdev.c |   25 ++++++++++++++-----------
 1 files changed, 14 insertions(+), 11 deletions(-)

diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index 3911401..e01ffce 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -487,10 +487,10 @@ static void e1000_receive_skb(struct e1000_adapter *adapter,
 
 /**
  * e1000_rx_checksum - Receive Checksum Offload
- * @adapter:     board private structure
- * @status_err:  receive descriptor status and error fields
- * @csum:	receive descriptor csum field
- * @sk_buff:     socket buffer with received data
+ * @adapter: board private structure
+ * @status_err: receive descriptor status and error fields
+ * @csum: receive descriptor csum field
+ * @sk_buff: socket buffer with received data
  **/
 static void e1000_rx_checksum(struct e1000_adapter *adapter, u32 status_err,
 			      u32 csum, struct sk_buff *skb)
@@ -500,9 +500,14 @@ static void e1000_rx_checksum(struct e1000_adapter *adapter, u32 status_err,
 
 	skb_checksum_none_assert(skb);
 
+	/* Rx checksum disabled */
+	if (!(adapter->netdev->features & NETIF_F_RXCSUM))
+		return;
+
 	/* Ignore Checksum bit is set */
 	if (status & E1000_RXD_STAT_IXSM)
 		return;
+
 	/* TCP/UDP checksum error bit is set */
 	if (errors & E1000_RXD_ERR_TCPE) {
 		/* let the stack verify checksum errors */
@@ -524,7 +529,7 @@ static void e1000_rx_checksum(struct e1000_adapter *adapter, u32 status_err,
 		 * Hardware complements the payload checksum, so we undo it
 		 * and then put the value in host order for further stack use.
 		 */
-		__sum16 sum = (__force __sum16)htons(csum);
+		__sum16 sum = (__force __sum16)htons(le16_to_cpu(csum));
 		skb->csum = csum_unfold(~sum);
 		skb->ip_summed = CHECKSUM_COMPLETE;
 	}
@@ -957,8 +962,7 @@ static bool e1000_clean_rx_irq(struct e1000_adapter *adapter,
 
 		/* Receive Checksum Offload */
 		e1000_rx_checksum(adapter, staterr,
-				  le16_to_cpu(rx_desc->wb.lower.hi_dword.
-					      csum_ip.csum), skb);
+				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
 
 		e1000_receive_skb(adapter, netdev, skb, staterr,
 				  rx_desc->wb.upper.vlan);
@@ -1318,8 +1322,8 @@ copydone:
 		total_rx_bytes += skb->len;
 		total_rx_packets++;
 
-		e1000_rx_checksum(adapter, staterr, le16_to_cpu(
-			rx_desc->wb.lower.hi_dword.csum_ip.csum), skb);
+		e1000_rx_checksum(adapter, staterr,
+				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
 
 		if (rx_desc->wb.upper.header_status &
 			   cpu_to_le16(E1000_RXDPS_HDRSTAT_HDRSP))
@@ -1491,8 +1495,7 @@ static bool e1000_clean_jumbo_rx_irq(struct e1000_adapter *adapter,
 
 		/* Receive Checksum Offload XXX recompute due to CRC strip? */
 		e1000_rx_checksum(adapter, staterr,
-				  le16_to_cpu(rx_desc->wb.lower.hi_dword.
-					      csum_ip.csum), skb);
+				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
 
 		/* probably a little skewed due to removing CRC */
 		total_rx_bytes += skb->len;
-- 
1.7.7.4

[net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[Jeff Kirsher]

From: Bruce Allan <bruce.w.allan@intel.com>

Enable RPS by default.  Disallow jumbo frames when both receive checksum
and receive hashing are enabled because the hardware cannot do both IP
payload checksum (enabled when receive checksum is enabled when using
packet split which is used for jumbo frames) and provide RSS hash at the
same time.

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/defines.h |    7 ++
 drivers/net/ethernet/intel/e1000e/hw.h      |    5 ++
 drivers/net/ethernet/intel/e1000e/netdev.c  |   93 +++++++++++++++++++++++++--
 3 files changed, 100 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/intel/e1000e/defines.h b/drivers/net/ethernet/intel/e1000e/defines.h
index c516a74..efa0310 100644
--- a/drivers/net/ethernet/intel/e1000e/defines.h
+++ b/drivers/net/ethernet/intel/e1000e/defines.h
@@ -126,6 +126,12 @@
     E1000_RXDEXT_STATERR_CXE |            \
     E1000_RXDEXT_STATERR_RXE)
 
+#define E1000_MRQC_RSS_FIELD_IPV4_TCP          0x00010000
+#define E1000_MRQC_RSS_FIELD_IPV4              0x00020000
+#define E1000_MRQC_RSS_FIELD_IPV6_TCP_EX       0x00040000
+#define E1000_MRQC_RSS_FIELD_IPV6              0x00100000
+#define E1000_MRQC_RSS_FIELD_IPV6_TCP          0x00200000
+
 #define E1000_RXDPS_HDRSTAT_HDRSP              0x00008000
 
 /* Management Control */
@@ -326,6 +332,7 @@
 /* Receive Checksum Control */
 #define E1000_RXCSUM_TUOFL     0x00000200   /* TCP / UDP checksum offload */
 #define E1000_RXCSUM_IPPCSE    0x00001000   /* IP payload checksum enable */
+#define E1000_RXCSUM_PCSD      0x00002000   /* packet checksum disabled */
 
 /* Header split receive */
 #define E1000_RFCTL_NFSW_DIS            0x00000040
diff --git a/drivers/net/ethernet/intel/e1000e/hw.h b/drivers/net/ethernet/intel/e1000e/hw.h
index 2967039..93c349e 100644
--- a/drivers/net/ethernet/intel/e1000e/hw.h
+++ b/drivers/net/ethernet/intel/e1000e/hw.h
@@ -204,6 +204,7 @@ enum e1e_registers {
 	E1000_WUC      = 0x05800, /* Wakeup Control - RW */
 	E1000_WUFC     = 0x05808, /* Wakeup Filter Control - RW */
 	E1000_WUS      = 0x05810, /* Wakeup Status - RO */
+	E1000_MRQC     = 0x05818, /* Multiple Receive Control - RW */
 	E1000_MANC     = 0x05820, /* Management Control - RW */
 	E1000_FFLT     = 0x05F00, /* Flexible Filter Length Table - RW Array */
 	E1000_HOST_IF  = 0x08800, /* Host Interface */
@@ -219,6 +220,10 @@ enum e1e_registers {
 	E1000_SWSM      = 0x05B50, /* SW Semaphore */
 	E1000_FWSM      = 0x05B54, /* FW Semaphore */
 	E1000_SWSM2     = 0x05B58, /* Driver-only SW semaphore */
+	E1000_RETA_BASE = 0x05C00, /* Redirection Table - RW */
+#define E1000_RETA(_n)	(E1000_RETA_BASE + ((_n) * 4))
+	E1000_RSSRK_BASE = 0x05C80, /* RSS Random Key - RW */
+#define E1000_RSSRK(_n)	(E1000_RSSRK_BASE + ((_n) * 4))
 	E1000_FFLT_DBG  = 0x05F04, /* Debug Register */
 	E1000_PCH_RAICC_BASE = 0x05F50, /* Receive Address Initial CRC */
 #define E1000_PCH_RAICC(_n)	(E1000_PCH_RAICC_BASE + ((_n) * 4))
diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index e01ffce..fa7bf63 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -845,6 +845,13 @@ check_page:
 	}
 }
 
+static inline void e1000_rx_hash(struct net_device *netdev, __le32 rss,
+				 struct sk_buff *skb)
+{
+	if (netdev->features & NETIF_F_RXHASH)
+		skb->rxhash = le32_to_cpu(rss);
+}
+
 /**
  * e1000_clean_rx_irq - Send received data up the network stack; legacy
  * @adapter: board private structure
@@ -964,6 +971,8 @@ static bool e1000_clean_rx_irq(struct e1000_adapter *adapter,
 		e1000_rx_checksum(adapter, staterr,
 				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
 
+		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);
+
 		e1000_receive_skb(adapter, netdev, skb, staterr,
 				  rx_desc->wb.upper.vlan);
 
@@ -1325,6 +1334,8 @@ copydone:
 		e1000_rx_checksum(adapter, staterr,
 				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
 
+		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);
+
 		if (rx_desc->wb.upper.header_status &
 			   cpu_to_le16(E1000_RXDPS_HDRSTAT_HDRSP))
 			adapter->rx_hdr_split++;
@@ -1497,6 +1508,8 @@ static bool e1000_clean_jumbo_rx_irq(struct e1000_adapter *adapter,
 		e1000_rx_checksum(adapter, staterr,
 				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
 
+		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);
+
 		/* probably a little skewed due to removing CRC */
 		total_rx_bytes += skb->len;
 		total_rx_packets++;
@@ -3271,6 +3284,49 @@ static void e1000e_set_rx_mode(struct net_device *netdev)
 		e1000e_vlan_strip_disable(adapter);
 }
 
+static void e1000e_setup_rss_hash(struct e1000_adapter *adapter)
+{
+	struct e1000_hw *hw = &adapter->hw;
+	u32 mrqc, rxcsum;
+	int j;
+	static const u8 rsshash[40] = {
+		0x6d, 0x5a, 0x56, 0xda, 0x25, 0x5b, 0x0e, 0xc2, 0x41, 0x67,
+		0x25, 0x3d, 0x43, 0xa3, 0x8f, 0xb0, 0xd0, 0xca, 0x2b, 0xcb,
+		0xae, 0x7b, 0x30, 0xb4, 0x77, 0xcb, 0x2d, 0xa3, 0x80, 0x30,
+		0xf2, 0x0c, 0x6a, 0x42, 0xb7, 0x3b, 0xbe, 0xac, 0x01, 0xfa
+	};
+
+	/* Fill out hash function seeds */
+	for (j = 0; j < 10; j++) {
+		u32 rsskey = rsshash[(j * 4)];
+		rsskey |= rsshash[(j * 4) + 1] << 8;
+		rsskey |= rsshash[(j * 4) + 2] << 16;
+		rsskey |= rsshash[(j * 4) + 3] << 24;
+		E1000_WRITE_REG_ARRAY(hw, E1000_RSSRK(0), j, rsskey);
+	}
+
+	/* Direct all traffic to queue 0 */
+	for (j = 0; j < 32; j++)
+		ew32(RETA(j), 0);
+
+	/*
+	 * Disable raw packet checksumming so that RSS hash is placed in
+	 * descriptor on writeback.
+	 */
+	rxcsum = er32(RXCSUM);
+	rxcsum |= E1000_RXCSUM_PCSD;
+
+	ew32(RXCSUM, rxcsum);
+
+	mrqc = (E1000_MRQC_RSS_FIELD_IPV4 |
+		E1000_MRQC_RSS_FIELD_IPV4_TCP |
+		E1000_MRQC_RSS_FIELD_IPV6 |
+		E1000_MRQC_RSS_FIELD_IPV6_TCP |
+		E1000_MRQC_RSS_FIELD_IPV6_TCP_EX);
+
+	ew32(MRQC, mrqc);
+}
+
 /**
  * e1000_configure - configure the hardware for Rx and Tx
  * @adapter: private board structure
@@ -3283,6 +3339,9 @@ static void e1000_configure(struct e1000_adapter *adapter)
 	e1000_init_manageability_pt(adapter);
 
 	e1000_configure_tx(adapter);
+
+	if (adapter->netdev->features & NETIF_F_RXHASH)
+		e1000e_setup_rss_hash(adapter);
 	e1000_setup_rctl(adapter);
 	e1000_configure_rx(adapter);
 	adapter->alloc_rx_buf(adapter, e1000_desc_unused(adapter->rx_ring),
@@ -5168,10 +5227,22 @@ static int e1000_change_mtu(struct net_device *netdev, int new_mtu)
 	int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;
 
 	/* Jumbo frame support */
-	if ((max_frame > ETH_FRAME_LEN + ETH_FCS_LEN) &&
-	    !(adapter->flags & FLAG_HAS_JUMBO_FRAMES)) {
-		e_err("Jumbo Frames not supported.\n");
-		return -EINVAL;
+	if (max_frame > ETH_FRAME_LEN + ETH_FCS_LEN) {
+		if (!(adapter->flags & FLAG_HAS_JUMBO_FRAMES)) {
+			e_err("Jumbo Frames not supported.\n");
+			return -EINVAL;
+		}
+
+		/*
+		 * IP payload checksum (enabled with jumbos/packet-split when
+		 * Rx checksum is enabled) and generation of RSS hash is
+		 * mutually exclusive in the hardware.
+		 */
+		if ((netdev->features & NETIF_F_RXCSUM) &&
+		    (netdev->features & NETIF_F_RXHASH)) {
+			e_err("Jumbo frames cannot be enabled when both receive checksum offload and receive hashing are enabled.  Disable one of the receive offload features before enabling jumbos.\n");
+			return -EINVAL;
+		}
 	}
 
 	/* Supported frame sizes */
@@ -5943,9 +6014,20 @@ static int e1000_set_features(struct net_device *netdev,
 		adapter->flags |= FLAG_TSO_FORCE;
 
 	if (!(changed & (NETIF_F_HW_VLAN_RX | NETIF_F_HW_VLAN_TX |
-			 NETIF_F_RXCSUM)))
+			 NETIF_F_RXCSUM | NETIF_F_RXHASH)))
 		return 0;
 
+	/*
+	 * IP payload checksum (enabled with jumbos/packet-split when Rx
+	 * checksum is enabled) and generation of RSS hash is mutually
+	 * exclusive in the hardware.
+	 */
+	if (adapter->rx_ps_pages &&
+	    (features & NETIF_F_RXCSUM) && (features & NETIF_F_RXHASH)) {
+		e_err("Enabling both receive checksum offload and receive hashing is not possible with jumbo frames.  Disable jumbos or enable only one of the receive offload features.\n");
+		return -EINVAL;
+	}
+
 	if (netif_running(netdev))
 		e1000e_reinit_locked(adapter);
 	else
@@ -6136,6 +6218,7 @@ static int __devinit e1000_probe(struct pci_dev *pdev,
 			    NETIF_F_HW_VLAN_TX |
 			    NETIF_F_TSO |
 			    NETIF_F_TSO6 |
+			    NETIF_F_RXHASH |
 			    NETIF_F_RXCSUM |
 			    NETIF_F_HW_CSUM);
 
-- 
1.7.7.4

[net-next 09/11] e1000e: re-enable alternate MAC address for all devices which support it

[Jeff Kirsher]

From: Bruce Allan <bruce.w.allan@intel.com>

Signed-off-by: Bruce Allan <bruce.w.allan@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/e1000e/lib.c |    7 ++-----
 1 files changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/intel/e1000e/lib.c b/drivers/net/ethernet/intel/e1000e/lib.c
index 0893ab1..78e3f4c 100644
--- a/drivers/net/ethernet/intel/e1000e/lib.c
+++ b/drivers/net/ethernet/intel/e1000e/lib.c
@@ -187,11 +187,8 @@ s32 e1000_check_alt_mac_addr_generic(struct e1000_hw *hw)
 	if (ret_val)
 		goto out;
 
-	/* Check for LOM (vs. NIC) or one of two valid mezzanine cards */
-	if (!((nvm_data & NVM_COMPAT_LOM) ||
-	      (hw->adapter->pdev->device == E1000_DEV_ID_82571EB_SERDES_DUAL) ||
-	      (hw->adapter->pdev->device == E1000_DEV_ID_82571EB_SERDES_QUAD) ||
-	      (hw->adapter->pdev->device == E1000_DEV_ID_82571EB_SERDES)))
+	/* not supported on older hardware or 82573 */
+	if ((hw->mac.type < e1000_82571) || (hw->mac.type == e1000_82573))
 		goto out;
 
 	ret_val = e1000_read_nvm(hw, NVM_ALT_MAC_ADDR_PTR, 1,
-- 
1.7.7.4

[net-next 10/11] e1000e: convert head, tail and itr_register offsets to __iomem pointers

[Jeff Kirsher]

From: Bruce Allan <bruce.w.allan@intel.com>

The Tx/Rx head and tail registers and itr_register are always at known
addresses based on the __iomem address at which the PCI region (from BAR 0)
is mapped and known offsets within the region for each of these registers.
Store and use the full address rather than just the region offset to reduce
unnecessary address calculations.  Also, change current u8 __iomem pointers
to void __iomem pointers.

Signed-off-by: Bruce Allan <bruce.w.allan@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/e1000e/e1000.h  |    6 ++--
 drivers/net/ethernet/intel/e1000e/hw.h     |    4 +-
 drivers/net/ethernet/intel/e1000e/netdev.c |   53 +++++++++++++--------------
 3 files changed, 31 insertions(+), 32 deletions(-)

diff --git a/drivers/net/ethernet/intel/e1000e/e1000.h b/drivers/net/ethernet/intel/e1000e/e1000.h
index f478a22..f6ac197 100644
--- a/drivers/net/ethernet/intel/e1000e/e1000.h
+++ b/drivers/net/ethernet/intel/e1000e/e1000.h
@@ -242,8 +242,8 @@ struct e1000_ring {
 	u16 next_to_use;
 	u16 next_to_clean;
 
-	u16 head;
-	u16 tail;
+	void __iomem *head;
+	void __iomem *tail;
 
 	/* array of buffer information structs */
 	struct e1000_buffer *buffer_info;
@@ -251,7 +251,7 @@ struct e1000_ring {
 	char name[IFNAMSIZ + 5];
 	u32 ims_val;
 	u32 itr_val;
-	u16 itr_register;
+	void __iomem *itr_register;
 	int set_itr;
 
 	struct sk_buff *rx_skb_top;
diff --git a/drivers/net/ethernet/intel/e1000e/hw.h b/drivers/net/ethernet/intel/e1000e/hw.h
index 93c349e..a15da47 100644
--- a/drivers/net/ethernet/intel/e1000e/hw.h
+++ b/drivers/net/ethernet/intel/e1000e/hw.h
@@ -969,8 +969,8 @@ struct e1000_dev_spec_ich8lan {
 struct e1000_hw {
 	struct e1000_adapter *adapter;
 
-	u8 __iomem *hw_addr;
-	u8 __iomem *flash_address;
+	void __iomem *hw_addr;
+	void __iomem *flash_address;
 
 	struct e1000_mac_info  mac;
 	struct e1000_fc_info   fc;
diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index fa7bf63..bbbc296 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -550,7 +550,7 @@ static void e1000_rx_checksum(struct e1000_adapter *adapter, u32 status_err,
  * which has bit 24 set while ME is accessing Host CSR registers, wait
  * if it is set and try again a number of times.
  **/
-static inline s32 e1000e_update_tail_wa(struct e1000_hw *hw, u8 __iomem * tail,
+static inline s32 e1000e_update_tail_wa(struct e1000_hw *hw, void __iomem *tail,
 					unsigned int i)
 {
 	unsigned int j = 0;
@@ -569,10 +569,9 @@ static inline s32 e1000e_update_tail_wa(struct e1000_hw *hw, u8 __iomem * tail,
 
 static void e1000e_update_rdt_wa(struct e1000_adapter *adapter, unsigned int i)
 {
-	u8 __iomem *tail = (adapter->hw.hw_addr + adapter->rx_ring->tail);
 	struct e1000_hw *hw = &adapter->hw;
 
-	if (e1000e_update_tail_wa(hw, tail, i)) {
+	if (e1000e_update_tail_wa(hw, adapter->rx_ring->tail, i)) {
 		u32 rctl = er32(RCTL);
 		ew32(RCTL, rctl & ~E1000_RCTL_EN);
 		e_err("ME firmware caused invalid RDT - resetting\n");
@@ -582,10 +581,9 @@ static void e1000e_update_rdt_wa(struct e1000_adapter *adapter, unsigned int i)
 
 static void e1000e_update_tdt_wa(struct e1000_adapter *adapter, unsigned int i)
 {
-	u8 __iomem *tail = (adapter->hw.hw_addr + adapter->tx_ring->tail);
 	struct e1000_hw *hw = &adapter->hw;
 
-	if (e1000e_update_tail_wa(hw, tail, i)) {
+	if (e1000e_update_tail_wa(hw, adapter->tx_ring->tail, i)) {
 		u32 tctl = er32(TCTL);
 		ew32(TCTL, tctl & ~E1000_TCTL_EN);
 		e_err("ME firmware caused invalid TDT - resetting\n");
@@ -651,7 +649,7 @@ map_skb:
 			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
 				e1000e_update_rdt_wa(adapter, i);
 			else
-				writel(i, adapter->hw.hw_addr + rx_ring->tail);
+				writel(i, rx_ring->tail);
 		}
 		i++;
 		if (i == rx_ring->count)
@@ -754,8 +752,7 @@ static void e1000_alloc_rx_buffers_ps(struct e1000_adapter *adapter,
 			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
 				e1000e_update_rdt_wa(adapter, i << 1);
 			else
-				writel(i << 1,
-				       adapter->hw.hw_addr + rx_ring->tail);
+				writel(i << 1, rx_ring->tail);
 		}
 
 		i++;
@@ -841,7 +838,7 @@ check_page:
 		if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
 			e1000e_update_rdt_wa(adapter, i);
 		else
-			writel(i, adapter->hw.hw_addr + rx_ring->tail);
+			writel(i, rx_ring->tail);
 	}
 }
 
@@ -1076,8 +1073,8 @@ static void e1000_print_hw_hang(struct work_struct *work)
 	      "PHY 1000BASE-T Status  <%x>\n"
 	      "PHY Extended Status    <%x>\n"
 	      "PCI Status             <%x>\n",
-	      readl(adapter->hw.hw_addr + tx_ring->head),
-	      readl(adapter->hw.hw_addr + tx_ring->tail),
+	      readl(tx_ring->head),
+	      readl(tx_ring->tail),
 	      tx_ring->next_to_use,
 	      tx_ring->next_to_clean,
 	      tx_ring->buffer_info[eop].time_stamp,
@@ -1617,8 +1614,8 @@ static void e1000_clean_rx_ring(struct e1000_adapter *adapter)
 	rx_ring->next_to_use = 0;
 	adapter->flags2 &= ~FLAG2_IS_DISCARDING;
 
-	writel(0, adapter->hw.hw_addr + rx_ring->head);
-	writel(0, adapter->hw.hw_addr + rx_ring->tail);
+	writel(0, rx_ring->head);
+	writel(0, rx_ring->tail);
 }
 
 static void e1000e_downshift_workaround(struct work_struct *work)
@@ -1814,7 +1811,7 @@ static irqreturn_t e1000_intr_msix_rx(int irq, void *data)
 	 */
 	if (adapter->rx_ring->set_itr) {
 		writel(1000000000 / (adapter->rx_ring->itr_val * 256),
-		       adapter->hw.hw_addr + adapter->rx_ring->itr_register);
+		       adapter->rx_ring->itr_register);
 		adapter->rx_ring->set_itr = 0;
 	}
 
@@ -1855,9 +1852,9 @@ static void e1000_configure_msix(struct e1000_adapter *adapter)
 	adapter->eiac_mask |= rx_ring->ims_val;
 	if (rx_ring->itr_val)
 		writel(1000000000 / (rx_ring->itr_val * 256),
-		       hw->hw_addr + rx_ring->itr_register);
+		       rx_ring->itr_register);
 	else
-		writel(1, hw->hw_addr + rx_ring->itr_register);
+		writel(1, rx_ring->itr_register);
 	ivar = E1000_IVAR_INT_ALLOC_VALID | vector;
 
 	/* Configure Tx vector */
@@ -1865,9 +1862,9 @@ static void e1000_configure_msix(struct e1000_adapter *adapter)
 	vector++;
 	if (tx_ring->itr_val)
 		writel(1000000000 / (tx_ring->itr_val * 256),
-		       hw->hw_addr + tx_ring->itr_register);
+		       tx_ring->itr_register);
 	else
-		writel(1, hw->hw_addr + tx_ring->itr_register);
+		writel(1, tx_ring->itr_register);
 	adapter->eiac_mask |= tx_ring->ims_val;
 	ivar |= ((E1000_IVAR_INT_ALLOC_VALID | vector) << 8);
 
@@ -1982,7 +1979,8 @@ static int e1000_request_msix(struct e1000_adapter *adapter)
 			  netdev);
 	if (err)
 		goto out;
-	adapter->rx_ring->itr_register = E1000_EITR_82574(vector);
+	adapter->rx_ring->itr_register = adapter->hw.hw_addr +
+	    E1000_EITR_82574(vector);
 	adapter->rx_ring->itr_val = adapter->itr;
 	vector++;
 
@@ -1997,7 +1995,8 @@ static int e1000_request_msix(struct e1000_adapter *adapter)
 			  netdev);
 	if (err)
 		goto out;
-	adapter->tx_ring->itr_register = E1000_EITR_82574(vector);
+	adapter->tx_ring->itr_register = adapter->hw.hw_addr +
+	    E1000_EITR_82574(vector);
 	adapter->tx_ring->itr_val = adapter->itr;
 	vector++;
 
@@ -2288,8 +2287,8 @@ static void e1000_clean_tx_ring(struct e1000_adapter *adapter)
 	tx_ring->next_to_use = 0;
 	tx_ring->next_to_clean = 0;
 
-	writel(0, adapter->hw.hw_addr + tx_ring->head);
-	writel(0, adapter->hw.hw_addr + tx_ring->tail);
+	writel(0, tx_ring->head);
+	writel(0, tx_ring->tail);
 }
 
 /**
@@ -2773,8 +2772,8 @@ static void e1000_configure_tx(struct e1000_adapter *adapter)
 	ew32(TDLEN, tdlen);
 	ew32(TDH, 0);
 	ew32(TDT, 0);
-	tx_ring->head = E1000_TDH;
-	tx_ring->tail = E1000_TDT;
+	tx_ring->head = adapter->hw.hw_addr + E1000_TDH;
+	tx_ring->tail = adapter->hw.hw_addr + E1000_TDT;
 
 	/* Set the default values for the Tx Inter Packet Gap timer */
 	tipg = DEFAULT_82543_TIPG_IPGT_COPPER;          /*  8  */
@@ -3088,8 +3087,8 @@ static void e1000_configure_rx(struct e1000_adapter *adapter)
 	ew32(RDLEN, rdlen);
 	ew32(RDH, 0);
 	ew32(RDT, 0);
-	rx_ring->head = E1000_RDH;
-	rx_ring->tail = E1000_RDT;
+	rx_ring->head = adapter->hw.hw_addr + E1000_RDH;
+	rx_ring->tail = adapter->hw.hw_addr + E1000_RDT;
 
 	/* Enable Receive Checksum Offload for TCP and UDP */
 	rxcsum = er32(RXCSUM);
@@ -4921,7 +4920,7 @@ static void e1000_tx_queue(struct e1000_adapter *adapter,
 	if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
 		e1000e_update_tdt_wa(adapter, i);
 	else
-		writel(i, adapter->hw.hw_addr + tx_ring->tail);
+		writel(i, tx_ring->tail);
 
 	/*
 	 * we need this if more than one processor can write to our tail
-- 
1.7.7.4

[net-next 11/11] e1000e: pass pointer to ring struct instead of adapter struct

[Jeff Kirsher]

From: Bruce Allan <bruce.w.allan@intel.com>

For ring-specific functions, pass a pointer to the ring struct instead of a
pointer to the adapter struct.

Signed-off-by: Bruce Allan <bruce.w.allan@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/e1000e/e1000.h   |   21 ++-
 drivers/net/ethernet/intel/e1000e/ethtool.c |   10 +-
 drivers/net/ethernet/intel/e1000e/netdev.c  |  232 ++++++++++++++-------------
 3 files changed, 136 insertions(+), 127 deletions(-)

diff --git a/drivers/net/ethernet/intel/e1000e/e1000.h b/drivers/net/ethernet/intel/e1000e/e1000.h
index f6ac197..db604cd 100644
--- a/drivers/net/ethernet/intel/e1000e/e1000.h
+++ b/drivers/net/ethernet/intel/e1000e/e1000.h
@@ -234,6 +234,7 @@ struct e1000_buffer {
 };
 
 struct e1000_ring {
+	struct e1000_adapter *adapter;	/* back pointer to adapter */
 	void *desc;			/* pointer to ring memory  */
 	dma_addr_t dma;			/* phys address of ring    */
 	unsigned int size;		/* length of ring in bytes */
@@ -334,11 +335,10 @@ struct e1000_adapter {
 	/*
 	 * Rx
 	 */
-	bool (*clean_rx) (struct e1000_adapter *adapter,
-			  int *work_done, int work_to_do)
-						____cacheline_aligned_in_smp;
-	void (*alloc_rx_buf) (struct e1000_adapter *adapter,
-			      int cleaned_count, gfp_t gfp);
+	bool (*clean_rx) (struct e1000_ring *ring, int *work_done,
+			  int work_to_do) ____cacheline_aligned_in_smp;
+	void (*alloc_rx_buf) (struct e1000_ring *ring, int cleaned_count,
+			      gfp_t gfp);
 	struct e1000_ring *rx_ring;
 
 	u32 rx_int_delay;
@@ -398,6 +398,9 @@ struct e1000_adapter {
 
 	bool idle_check;
 	int phy_hang_count;
+
+	u16 tx_ring_count;
+	u16 rx_ring_count;
 };
 
 struct e1000_info {
@@ -492,10 +495,10 @@ extern void e1000e_down(struct e1000_adapter *adapter);
 extern void e1000e_reinit_locked(struct e1000_adapter *adapter);
 extern void e1000e_reset(struct e1000_adapter *adapter);
 extern void e1000e_power_up_phy(struct e1000_adapter *adapter);
-extern int e1000e_setup_rx_resources(struct e1000_adapter *adapter);
-extern int e1000e_setup_tx_resources(struct e1000_adapter *adapter);
-extern void e1000e_free_rx_resources(struct e1000_adapter *adapter);
-extern void e1000e_free_tx_resources(struct e1000_adapter *adapter);
+extern int e1000e_setup_rx_resources(struct e1000_ring *ring);
+extern int e1000e_setup_tx_resources(struct e1000_ring *ring);
+extern void e1000e_free_rx_resources(struct e1000_ring *ring);
+extern void e1000e_free_tx_resources(struct e1000_ring *ring);
 extern struct rtnl_link_stats64 *e1000e_get_stats64(struct net_device *netdev,
                                                     struct rtnl_link_stats64
                                                     *stats);
diff --git a/drivers/net/ethernet/intel/e1000e/ethtool.c b/drivers/net/ethernet/intel/e1000e/ethtool.c
index fb2c28e..f0bf726 100644
--- a/drivers/net/ethernet/intel/e1000e/ethtool.c
+++ b/drivers/net/ethernet/intel/e1000e/ethtool.c
@@ -656,10 +656,10 @@ static int e1000_set_ringparam(struct net_device *netdev,
 
 	if (netif_running(adapter->netdev)) {
 		/* Try to get new resources before deleting old */
-		err = e1000e_setup_rx_resources(adapter);
+		err = e1000e_setup_rx_resources(rx_ring);
 		if (err)
 			goto err_setup_rx;
-		err = e1000e_setup_tx_resources(adapter);
+		err = e1000e_setup_tx_resources(tx_ring);
 		if (err)
 			goto err_setup_tx;
 
@@ -669,8 +669,8 @@ static int e1000_set_ringparam(struct net_device *netdev,
 		 */
 		adapter->rx_ring = rx_old;
 		adapter->tx_ring = tx_old;
-		e1000e_free_rx_resources(adapter);
-		e1000e_free_tx_resources(adapter);
+		e1000e_free_rx_resources(adapter->rx_ring);
+		e1000e_free_tx_resources(adapter->tx_ring);
 		kfree(tx_old);
 		kfree(rx_old);
 		adapter->rx_ring = rx_ring;
@@ -683,7 +683,7 @@ static int e1000_set_ringparam(struct net_device *netdev,
 	clear_bit(__E1000_RESETTING, &adapter->state);
 	return 0;
 err_setup_tx:
-	e1000e_free_rx_resources(adapter);
+	e1000e_free_rx_resources(rx_ring);
 err_setup_rx:
 	adapter->rx_ring = rx_old;
 	adapter->tx_ring = tx_old;
diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
index bbbc296..9b69a6b 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -567,11 +567,12 @@ static inline s32 e1000e_update_tail_wa(struct e1000_hw *hw, void __iomem *tail,
 	return 0;
 }
 
-static void e1000e_update_rdt_wa(struct e1000_adapter *adapter, unsigned int i)
+static void e1000e_update_rdt_wa(struct e1000_ring *rx_ring, unsigned int i)
 {
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct e1000_hw *hw = &adapter->hw;
 
-	if (e1000e_update_tail_wa(hw, adapter->rx_ring->tail, i)) {
+	if (e1000e_update_tail_wa(hw, rx_ring->tail, i)) {
 		u32 rctl = er32(RCTL);
 		ew32(RCTL, rctl & ~E1000_RCTL_EN);
 		e_err("ME firmware caused invalid RDT - resetting\n");
@@ -579,11 +580,12 @@ static void e1000e_update_rdt_wa(struct e1000_adapter *adapter, unsigned int i)
 	}
 }
 
-static void e1000e_update_tdt_wa(struct e1000_adapter *adapter, unsigned int i)
+static void e1000e_update_tdt_wa(struct e1000_ring *tx_ring, unsigned int i)
 {
+	struct e1000_adapter *adapter = tx_ring->adapter;
 	struct e1000_hw *hw = &adapter->hw;
 
-	if (e1000e_update_tail_wa(hw, adapter->tx_ring->tail, i)) {
+	if (e1000e_update_tail_wa(hw, tx_ring->tail, i)) {
 		u32 tctl = er32(TCTL);
 		ew32(TCTL, tctl & ~E1000_TCTL_EN);
 		e_err("ME firmware caused invalid TDT - resetting\n");
@@ -593,14 +595,14 @@ static void e1000e_update_tdt_wa(struct e1000_adapter *adapter, unsigned int i)
 
 /**
  * e1000_alloc_rx_buffers - Replace used receive buffers
- * @adapter: address of board private structure
+ * @rx_ring: Rx descriptor ring
  **/
-static void e1000_alloc_rx_buffers(struct e1000_adapter *adapter,
+static void e1000_alloc_rx_buffers(struct e1000_ring *rx_ring,
 				   int cleaned_count, gfp_t gfp)
 {
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct net_device *netdev = adapter->netdev;
 	struct pci_dev *pdev = adapter->pdev;
-	struct e1000_ring *rx_ring = adapter->rx_ring;
 	union e1000_rx_desc_extended *rx_desc;
 	struct e1000_buffer *buffer_info;
 	struct sk_buff *skb;
@@ -647,7 +649,7 @@ map_skb:
 			 */
 			wmb();
 			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
-				e1000e_update_rdt_wa(adapter, i);
+				e1000e_update_rdt_wa(rx_ring, i);
 			else
 				writel(i, rx_ring->tail);
 		}
@@ -662,15 +664,15 @@ map_skb:
 
 /**
  * e1000_alloc_rx_buffers_ps - Replace used receive buffers; packet split
- * @adapter: address of board private structure
+ * @rx_ring: Rx descriptor ring
  **/
-static void e1000_alloc_rx_buffers_ps(struct e1000_adapter *adapter,
+static void e1000_alloc_rx_buffers_ps(struct e1000_ring *rx_ring,
 				      int cleaned_count, gfp_t gfp)
 {
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct net_device *netdev = adapter->netdev;
 	struct pci_dev *pdev = adapter->pdev;
 	union e1000_rx_desc_packet_split *rx_desc;
-	struct e1000_ring *rx_ring = adapter->rx_ring;
 	struct e1000_buffer *buffer_info;
 	struct e1000_ps_page *ps_page;
 	struct sk_buff *skb;
@@ -750,7 +752,7 @@ static void e1000_alloc_rx_buffers_ps(struct e1000_adapter *adapter,
 			 */
 			wmb();
 			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
-				e1000e_update_rdt_wa(adapter, i << 1);
+				e1000e_update_rdt_wa(rx_ring, i << 1);
 			else
 				writel(i << 1, rx_ring->tail);
 		}
@@ -767,17 +769,17 @@ no_buffers:
 
 /**
  * e1000_alloc_jumbo_rx_buffers - Replace used jumbo receive buffers
- * @adapter: address of board private structure
+ * @rx_ring: Rx descriptor ring
  * @cleaned_count: number of buffers to allocate this pass
  **/
 
-static void e1000_alloc_jumbo_rx_buffers(struct e1000_adapter *adapter,
+static void e1000_alloc_jumbo_rx_buffers(struct e1000_ring *rx_ring,
 					 int cleaned_count, gfp_t gfp)
 {
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct net_device *netdev = adapter->netdev;
 	struct pci_dev *pdev = adapter->pdev;
 	union e1000_rx_desc_extended *rx_desc;
-	struct e1000_ring *rx_ring = adapter->rx_ring;
 	struct e1000_buffer *buffer_info;
 	struct sk_buff *skb;
 	unsigned int i;
@@ -836,7 +838,7 @@ check_page:
 		 * such as IA-64). */
 		wmb();
 		if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
-			e1000e_update_rdt_wa(adapter, i);
+			e1000e_update_rdt_wa(rx_ring, i);
 		else
 			writel(i, rx_ring->tail);
 	}
@@ -850,19 +852,19 @@ static inline void e1000_rx_hash(struct net_device *netdev, __le32 rss,
 }
 
 /**
- * e1000_clean_rx_irq - Send received data up the network stack; legacy
- * @adapter: board private structure
+ * e1000_clean_rx_irq - Send received data up the network stack
+ * @rx_ring: Rx descriptor ring
  *
  * the return value indicates whether actual cleaning was done, there
  * is no guarantee that everything was cleaned
  **/
-static bool e1000_clean_rx_irq(struct e1000_adapter *adapter,
-			       int *work_done, int work_to_do)
+static bool e1000_clean_rx_irq(struct e1000_ring *rx_ring, int *work_done,
+			       int work_to_do)
 {
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct net_device *netdev = adapter->netdev;
 	struct pci_dev *pdev = adapter->pdev;
 	struct e1000_hw *hw = &adapter->hw;
-	struct e1000_ring *rx_ring = adapter->rx_ring;
 	union e1000_rx_desc_extended *rx_desc, *next_rxd;
 	struct e1000_buffer *buffer_info, *next_buffer;
 	u32 length, staterr;
@@ -978,7 +980,7 @@ next_desc:
 
 		/* return some buffers to hardware, one at a time is too slow */
 		if (cleaned_count >= E1000_RX_BUFFER_WRITE) {
-			adapter->alloc_rx_buf(adapter, cleaned_count,
+			adapter->alloc_rx_buf(rx_ring, cleaned_count,
 					      GFP_ATOMIC);
 			cleaned_count = 0;
 		}
@@ -993,16 +995,18 @@ next_desc:
 
 	cleaned_count = e1000_desc_unused(rx_ring);
 	if (cleaned_count)
-		adapter->alloc_rx_buf(adapter, cleaned_count, GFP_ATOMIC);
+		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
 
 	adapter->total_rx_bytes += total_rx_bytes;
 	adapter->total_rx_packets += total_rx_packets;
 	return cleaned;
 }
 
-static void e1000_put_txbuf(struct e1000_adapter *adapter,
-			     struct e1000_buffer *buffer_info)
+static void e1000_put_txbuf(struct e1000_ring *tx_ring,
+			    struct e1000_buffer *buffer_info)
 {
+	struct e1000_adapter *adapter = tx_ring->adapter;
+
 	if (buffer_info->dma) {
 		if (buffer_info->mapped_as_page)
 			dma_unmap_page(&adapter->pdev->dev, buffer_info->dma,
@@ -1090,16 +1094,16 @@ static void e1000_print_hw_hang(struct work_struct *work)
 
 /**
  * e1000_clean_tx_irq - Reclaim resources after transmit completes
- * @adapter: board private structure
+ * @tx_ring: Tx descriptor ring
  *
  * the return value indicates whether actual cleaning was done, there
  * is no guarantee that everything was cleaned
  **/
-static bool e1000_clean_tx_irq(struct e1000_adapter *adapter)
+static bool e1000_clean_tx_irq(struct e1000_ring *tx_ring)
 {
+	struct e1000_adapter *adapter = tx_ring->adapter;
 	struct net_device *netdev = adapter->netdev;
 	struct e1000_hw *hw = &adapter->hw;
-	struct e1000_ring *tx_ring = adapter->tx_ring;
 	struct e1000_tx_desc *tx_desc, *eop_desc;
 	struct e1000_buffer *buffer_info;
 	unsigned int i, eop;
@@ -1129,7 +1133,7 @@ static bool e1000_clean_tx_irq(struct e1000_adapter *adapter)
 				}
 			}
 
-			e1000_put_txbuf(adapter, buffer_info);
+			e1000_put_txbuf(tx_ring, buffer_info);
 			tx_desc->upper.data = 0;
 
 			i++;
@@ -1183,19 +1187,19 @@ static bool e1000_clean_tx_irq(struct e1000_adapter *adapter)
 
 /**
  * e1000_clean_rx_irq_ps - Send received data up the network stack; packet split
- * @adapter: board private structure
+ * @rx_ring: Rx descriptor ring
  *
  * the return value indicates whether actual cleaning was done, there
  * is no guarantee that everything was cleaned
  **/
-static bool e1000_clean_rx_irq_ps(struct e1000_adapter *adapter,
-				  int *work_done, int work_to_do)
+static bool e1000_clean_rx_irq_ps(struct e1000_ring *rx_ring, int *work_done,
+				  int work_to_do)
 {
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct e1000_hw *hw = &adapter->hw;
 	union e1000_rx_desc_packet_split *rx_desc, *next_rxd;
 	struct net_device *netdev = adapter->netdev;
 	struct pci_dev *pdev = adapter->pdev;
-	struct e1000_ring *rx_ring = adapter->rx_ring;
 	struct e1000_buffer *buffer_info, *next_buffer;
 	struct e1000_ps_page *ps_page;
 	struct sk_buff *skb;
@@ -1346,7 +1350,7 @@ next_desc:
 
 		/* return some buffers to hardware, one at a time is too slow */
 		if (cleaned_count >= E1000_RX_BUFFER_WRITE) {
-			adapter->alloc_rx_buf(adapter, cleaned_count,
+			adapter->alloc_rx_buf(rx_ring, cleaned_count,
 					      GFP_ATOMIC);
 			cleaned_count = 0;
 		}
@@ -1361,7 +1365,7 @@ next_desc:
 
 	cleaned_count = e1000_desc_unused(rx_ring);
 	if (cleaned_count)
-		adapter->alloc_rx_buf(adapter, cleaned_count, GFP_ATOMIC);
+		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
 
 	adapter->total_rx_bytes += total_rx_bytes;
 	adapter->total_rx_packets += total_rx_packets;
@@ -1387,13 +1391,12 @@ static void e1000_consume_page(struct e1000_buffer *bi, struct sk_buff *skb,
  * the return value indicates whether actual cleaning was done, there
  * is no guarantee that everything was cleaned
  **/
-
-static bool e1000_clean_jumbo_rx_irq(struct e1000_adapter *adapter,
-                                     int *work_done, int work_to_do)
+static bool e1000_clean_jumbo_rx_irq(struct e1000_ring *rx_ring, int *work_done,
+				     int work_to_do)
 {
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct net_device *netdev = adapter->netdev;
 	struct pci_dev *pdev = adapter->pdev;
-	struct e1000_ring *rx_ring = adapter->rx_ring;
 	union e1000_rx_desc_extended *rx_desc, *next_rxd;
 	struct e1000_buffer *buffer_info, *next_buffer;
 	u32 length, staterr;
@@ -1526,7 +1529,7 @@ next_desc:
 
 		/* return some buffers to hardware, one at a time is too slow */
 		if (unlikely(cleaned_count >= E1000_RX_BUFFER_WRITE)) {
-			adapter->alloc_rx_buf(adapter, cleaned_count,
+			adapter->alloc_rx_buf(rx_ring, cleaned_count,
 					      GFP_ATOMIC);
 			cleaned_count = 0;
 		}
@@ -1541,7 +1544,7 @@ next_desc:
 
 	cleaned_count = e1000_desc_unused(rx_ring);
 	if (cleaned_count)
-		adapter->alloc_rx_buf(adapter, cleaned_count, GFP_ATOMIC);
+		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
 
 	adapter->total_rx_bytes += total_rx_bytes;
 	adapter->total_rx_packets += total_rx_packets;
@@ -1550,11 +1553,11 @@ next_desc:
 
 /**
  * e1000_clean_rx_ring - Free Rx Buffers per Queue
- * @adapter: board private structure
+ * @rx_ring: Rx descriptor ring
  **/
-static void e1000_clean_rx_ring(struct e1000_adapter *adapter)
+static void e1000_clean_rx_ring(struct e1000_ring *rx_ring)
 {
-	struct e1000_ring *rx_ring = adapter->rx_ring;
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct e1000_buffer *buffer_info;
 	struct e1000_ps_page *ps_page;
 	struct pci_dev *pdev = adapter->pdev;
@@ -1794,7 +1797,7 @@ static irqreturn_t e1000_intr_msix_tx(int irq, void *data)
 	adapter->total_tx_bytes = 0;
 	adapter->total_tx_packets = 0;
 
-	if (!e1000_clean_tx_irq(adapter))
+	if (!e1000_clean_tx_irq(tx_ring))
 		/* Ring was not completely cleaned, so fire another interrupt */
 		ew32(ICS, tx_ring->ims_val);
 
@@ -1805,14 +1808,15 @@ static irqreturn_t e1000_intr_msix_rx(int irq, void *data)
 {
 	struct net_device *netdev = data;
 	struct e1000_adapter *adapter = netdev_priv(netdev);
+	struct e1000_ring *rx_ring = adapter->rx_ring;
 
 	/* Write the ITR value calculated at the end of the
 	 * previous interrupt.
 	 */
-	if (adapter->rx_ring->set_itr) {
-		writel(1000000000 / (adapter->rx_ring->itr_val * 256),
-		       adapter->rx_ring->itr_register);
-		adapter->rx_ring->set_itr = 0;
+	if (rx_ring->set_itr) {
+		writel(1000000000 / (rx_ring->itr_val * 256),
+		       rx_ring->itr_register);
+		rx_ring->set_itr = 0;
 	}
 
 	if (napi_schedule_prep(&adapter->napi)) {
@@ -2177,13 +2181,13 @@ static int e1000_alloc_ring_dma(struct e1000_adapter *adapter,
 
 /**
  * e1000e_setup_tx_resources - allocate Tx resources (Descriptors)
- * @adapter: board private structure
+ * @tx_ring: Tx descriptor ring
  *
  * Return 0 on success, negative on failure
  **/
-int e1000e_setup_tx_resources(struct e1000_adapter *adapter)
+int e1000e_setup_tx_resources(struct e1000_ring *tx_ring)
 {
-	struct e1000_ring *tx_ring = adapter->tx_ring;
+	struct e1000_adapter *adapter = tx_ring->adapter;
 	int err = -ENOMEM, size;
 
 	size = sizeof(struct e1000_buffer) * tx_ring->count;
@@ -2211,13 +2215,13 @@ err:
 
 /**
  * e1000e_setup_rx_resources - allocate Rx resources (Descriptors)
- * @adapter: board private structure
+ * @rx_ring: Rx descriptor ring
  *
  * Returns 0 on success, negative on failure
  **/
-int e1000e_setup_rx_resources(struct e1000_adapter *adapter)
+int e1000e_setup_rx_resources(struct e1000_ring *rx_ring)
 {
-	struct e1000_ring *rx_ring = adapter->rx_ring;
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct e1000_buffer *buffer_info;
 	int i, size, desc_len, err = -ENOMEM;
 
@@ -2264,18 +2268,18 @@ err:
 
 /**
  * e1000_clean_tx_ring - Free Tx Buffers
- * @adapter: board private structure
+ * @tx_ring: Tx descriptor ring
  **/
-static void e1000_clean_tx_ring(struct e1000_adapter *adapter)
+static void e1000_clean_tx_ring(struct e1000_ring *tx_ring)
 {
-	struct e1000_ring *tx_ring = adapter->tx_ring;
+	struct e1000_adapter *adapter = tx_ring->adapter;
 	struct e1000_buffer *buffer_info;
 	unsigned long size;
 	unsigned int i;
 
 	for (i = 0; i < tx_ring->count; i++) {
 		buffer_info = &tx_ring->buffer_info[i];
-		e1000_put_txbuf(adapter, buffer_info);
+		e1000_put_txbuf(tx_ring, buffer_info);
 	}
 
 	netdev_reset_queue(adapter->netdev);
@@ -2293,16 +2297,16 @@ static void e1000_clean_tx_ring(struct e1000_adapter *adapter)
 
 /**
  * e1000e_free_tx_resources - Free Tx Resources per Queue
- * @adapter: board private structure
+ * @tx_ring: Tx descriptor ring
  *
  * Free all transmit software resources
  **/
-void e1000e_free_tx_resources(struct e1000_adapter *adapter)
+void e1000e_free_tx_resources(struct e1000_ring *tx_ring)
 {
+	struct e1000_adapter *adapter = tx_ring->adapter;
 	struct pci_dev *pdev = adapter->pdev;
-	struct e1000_ring *tx_ring = adapter->tx_ring;
 
-	e1000_clean_tx_ring(adapter);
+	e1000_clean_tx_ring(tx_ring);
 
 	vfree(tx_ring->buffer_info);
 	tx_ring->buffer_info = NULL;
@@ -2314,18 +2318,17 @@ void e1000e_free_tx_resources(struct e1000_adapter *adapter)
 
 /**
  * e1000e_free_rx_resources - Free Rx Resources
- * @adapter: board private structure
+ * @rx_ring: Rx descriptor ring
  *
  * Free all receive software resources
  **/
-
-void e1000e_free_rx_resources(struct e1000_adapter *adapter)
+void e1000e_free_rx_resources(struct e1000_ring *rx_ring)
 {
+	struct e1000_adapter *adapter = rx_ring->adapter;
 	struct pci_dev *pdev = adapter->pdev;
-	struct e1000_ring *rx_ring = adapter->rx_ring;
 	int i;
 
-	e1000_clean_rx_ring(adapter);
+	e1000_clean_rx_ring(rx_ring);
 
 	for (i = 0; i < rx_ring->count; i++)
 		kfree(rx_ring->buffer_info[i].ps_pages);
@@ -2479,13 +2482,19 @@ set_itr_now:
  **/
 static int __devinit e1000_alloc_queues(struct e1000_adapter *adapter)
 {
-	adapter->tx_ring = kzalloc(sizeof(struct e1000_ring), GFP_KERNEL);
+	int size = sizeof(struct e1000_ring);
+
+	adapter->tx_ring = kzalloc(size, GFP_KERNEL);
 	if (!adapter->tx_ring)
 		goto err;
+	adapter->tx_ring->count = adapter->tx_ring_count;
+	adapter->tx_ring->adapter = adapter;
 
-	adapter->rx_ring = kzalloc(sizeof(struct e1000_ring), GFP_KERNEL);
+	adapter->rx_ring = kzalloc(size, GFP_KERNEL);
 	if (!adapter->rx_ring)
 		goto err;
+	adapter->rx_ring->count = adapter->rx_ring_count;
+	adapter->rx_ring->adapter = adapter;
 
 	return 0;
 err:
@@ -2513,10 +2522,10 @@ static int e1000_clean(struct napi_struct *napi, int budget)
 	    !(adapter->rx_ring->ims_val & adapter->tx_ring->ims_val))
 		goto clean_rx;
 
-	tx_cleaned = e1000_clean_tx_irq(adapter);
+	tx_cleaned = e1000_clean_tx_irq(adapter->tx_ring);
 
 clean_rx:
-	adapter->clean_rx(adapter, &work_done, budget);
+	adapter->clean_rx(adapter->rx_ring, &work_done, budget);
 
 	if (!tx_cleaned)
 		work_done = budget;
@@ -3332,6 +3341,8 @@ static void e1000e_setup_rss_hash(struct e1000_adapter *adapter)
  **/
 static void e1000_configure(struct e1000_adapter *adapter)
 {
+	struct e1000_ring *rx_ring = adapter->rx_ring;
+
 	e1000e_set_rx_mode(adapter->netdev);
 
 	e1000_restore_vlan(adapter);
@@ -3343,8 +3354,7 @@ static void e1000_configure(struct e1000_adapter *adapter)
 		e1000e_setup_rss_hash(adapter);
 	e1000_setup_rctl(adapter);
 	e1000_configure_rx(adapter);
-	adapter->alloc_rx_buf(adapter, e1000_desc_unused(adapter->rx_ring),
-			      GFP_KERNEL);
+	adapter->alloc_rx_buf(rx_ring, e1000_desc_unused(rx_ring), GFP_KERNEL);
 }
 
 /**
@@ -3654,8 +3664,8 @@ void e1000e_down(struct e1000_adapter *adapter)
 	spin_unlock(&adapter->stats64_lock);
 
 	e1000e_flush_descriptors(adapter);
-	e1000_clean_tx_ring(adapter);
-	e1000_clean_rx_ring(adapter);
+	e1000_clean_tx_ring(adapter->tx_ring);
+	e1000_clean_rx_ring(adapter->rx_ring);
 
 	adapter->link_speed = 0;
 	adapter->link_duplex = 0;
@@ -3695,6 +3705,8 @@ static int __devinit e1000_sw_init(struct e1000_adapter *adapter)
 	adapter->rx_ps_bsize0 = 128;
 	adapter->max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
 	adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN;
+	adapter->tx_ring_count = E1000_DEFAULT_TXD;
+	adapter->rx_ring_count = E1000_DEFAULT_RXD;
 
 	spin_lock_init(&adapter->stats64_lock);
 
@@ -3853,12 +3865,12 @@ static int e1000_open(struct net_device *netdev)
 	netif_carrier_off(netdev);
 
 	/* allocate transmit descriptors */
-	err = e1000e_setup_tx_resources(adapter);
+	err = e1000e_setup_tx_resources(adapter->tx_ring);
 	if (err)
 		goto err_setup_tx;
 
 	/* allocate receive descriptors */
-	err = e1000e_setup_rx_resources(adapter);
+	err = e1000e_setup_rx_resources(adapter->rx_ring);
 	if (err)
 		goto err_setup_rx;
 
@@ -3934,9 +3946,9 @@ static int e1000_open(struct net_device *netdev)
 err_req_irq:
 	e1000e_release_hw_control(adapter);
 	e1000_power_down_phy(adapter);
-	e1000e_free_rx_resources(adapter);
+	e1000e_free_rx_resources(adapter->rx_ring);
 err_setup_rx:
-	e1000e_free_tx_resources(adapter);
+	e1000e_free_tx_resources(adapter->tx_ring);
 err_setup_tx:
 	e1000e_reset(adapter);
 	pm_runtime_put_sync(&pdev->dev);
@@ -3972,8 +3984,8 @@ static int e1000_close(struct net_device *netdev)
 	}
 	e1000_power_down_phy(adapter);
 
-	e1000e_free_tx_resources(adapter);
-	e1000e_free_rx_resources(adapter);
+	e1000e_free_tx_resources(adapter->tx_ring);
+	e1000e_free_rx_resources(adapter->rx_ring);
 
 	/*
 	 * kill manageability vlan ID if supported, but not if a vlan with
@@ -4630,10 +4642,8 @@ link_up:
 #define E1000_TX_FLAGS_VLAN_MASK	0xffff0000
 #define E1000_TX_FLAGS_VLAN_SHIFT	16
 
-static int e1000_tso(struct e1000_adapter *adapter,
-		     struct sk_buff *skb)
+static int e1000_tso(struct e1000_ring *tx_ring, struct sk_buff *skb)
 {
-	struct e1000_ring *tx_ring = adapter->tx_ring;
 	struct e1000_context_desc *context_desc;
 	struct e1000_buffer *buffer_info;
 	unsigned int i;
@@ -4702,9 +4712,9 @@ static int e1000_tso(struct e1000_adapter *adapter,
 	return 1;
 }
 
-static bool e1000_tx_csum(struct e1000_adapter *adapter, struct sk_buff *skb)
+static bool e1000_tx_csum(struct e1000_ring *tx_ring, struct sk_buff *skb)
 {
-	struct e1000_ring *tx_ring = adapter->tx_ring;
+	struct e1000_adapter *adapter = tx_ring->adapter;
 	struct e1000_context_desc *context_desc;
 	struct e1000_buffer *buffer_info;
 	unsigned int i;
@@ -4765,12 +4775,11 @@ static bool e1000_tx_csum(struct e1000_adapter *adapter, struct sk_buff *skb)
 #define E1000_MAX_PER_TXD	8192
 #define E1000_MAX_TXD_PWR	12
 
-static int e1000_tx_map(struct e1000_adapter *adapter,
-			struct sk_buff *skb, unsigned int first,
-			unsigned int max_per_txd, unsigned int nr_frags,
-			unsigned int mss)
+static int e1000_tx_map(struct e1000_ring *tx_ring, struct sk_buff *skb,
+			unsigned int first, unsigned int max_per_txd,
+			unsigned int nr_frags, unsigned int mss)
 {
-	struct e1000_ring *tx_ring = adapter->tx_ring;
+	struct e1000_adapter *adapter = tx_ring->adapter;
 	struct pci_dev *pdev = adapter->pdev;
 	struct e1000_buffer *buffer_info;
 	unsigned int len = skb_headlen(skb);
@@ -4856,16 +4865,15 @@ dma_error:
 			i += tx_ring->count;
 		i--;
 		buffer_info = &tx_ring->buffer_info[i];
-		e1000_put_txbuf(adapter, buffer_info);
+		e1000_put_txbuf(tx_ring, buffer_info);
 	}
 
 	return 0;
 }
 
-static void e1000_tx_queue(struct e1000_adapter *adapter,
-			   int tx_flags, int count)
+static void e1000_tx_queue(struct e1000_ring *tx_ring, int tx_flags, int count)
 {
-	struct e1000_ring *tx_ring = adapter->tx_ring;
+	struct e1000_adapter *adapter = tx_ring->adapter;
 	struct e1000_tx_desc *tx_desc = NULL;
 	struct e1000_buffer *buffer_info;
 	u32 txd_upper = 0, txd_lower = E1000_TXD_CMD_IFCS;
@@ -4918,7 +4926,7 @@ static void e1000_tx_queue(struct e1000_adapter *adapter,
 	tx_ring->next_to_use = i;
 
 	if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
-		e1000e_update_tdt_wa(adapter, i);
+		e1000e_update_tdt_wa(tx_ring, i);
 	else
 		writel(i, tx_ring->tail);
 
@@ -4968,11 +4976,11 @@ static int e1000_transfer_dhcp_info(struct e1000_adapter *adapter,
 	return 0;
 }
 
-static int __e1000_maybe_stop_tx(struct net_device *netdev, int size)
+static int __e1000_maybe_stop_tx(struct e1000_ring *tx_ring, int size)
 {
-	struct e1000_adapter *adapter = netdev_priv(netdev);
+	struct e1000_adapter *adapter = tx_ring->adapter;
 
-	netif_stop_queue(netdev);
+	netif_stop_queue(adapter->netdev);
 	/*
 	 * Herbert's original patch had:
 	 *  smp_mb__after_netif_stop_queue();
@@ -4984,22 +4992,20 @@ static int __e1000_maybe_stop_tx(struct net_device *netdev, int size)
 	 * We need to check again in a case another CPU has just
 	 * made room available.
 	 */
-	if (e1000_desc_unused(adapter->tx_ring) < size)
+	if (e1000_desc_unused(tx_ring) < size)
 		return -EBUSY;
 
 	/* A reprieve! */
-	netif_start_queue(netdev);
+	netif_start_queue(adapter->netdev);
 	++adapter->restart_queue;
 	return 0;
 }
 
-static int e1000_maybe_stop_tx(struct net_device *netdev, int size)
+static int e1000_maybe_stop_tx(struct e1000_ring *tx_ring, int size)
 {
-	struct e1000_adapter *adapter = netdev_priv(netdev);
-
-	if (e1000_desc_unused(adapter->tx_ring) >= size)
+	if (e1000_desc_unused(tx_ring) >= size)
 		return 0;
-	return __e1000_maybe_stop_tx(netdev, size);
+	return __e1000_maybe_stop_tx(tx_ring, size);
 }
 
 #define TXD_USE_COUNT(S, X) (((S) >> (X)) + 1 )
@@ -5085,7 +5091,7 @@ static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
 	 * need: count + 2 desc gap to keep tail from touching
 	 * head, otherwise try next time
 	 */
-	if (e1000_maybe_stop_tx(netdev, count + 2))
+	if (e1000_maybe_stop_tx(tx_ring, count + 2))
 		return NETDEV_TX_BUSY;
 
 	if (vlan_tx_tag_present(skb)) {
@@ -5095,7 +5101,7 @@ static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
 
 	first = tx_ring->next_to_use;
 
-	tso = e1000_tso(adapter, skb);
+	tso = e1000_tso(tx_ring, skb);
 	if (tso < 0) {
 		dev_kfree_skb_any(skb);
 		return NETDEV_TX_OK;
@@ -5103,7 +5109,7 @@ static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
 
 	if (tso)
 		tx_flags |= E1000_TX_FLAGS_TSO;
-	else if (e1000_tx_csum(adapter, skb))
+	else if (e1000_tx_csum(tx_ring, skb))
 		tx_flags |= E1000_TX_FLAGS_CSUM;
 
 	/*
@@ -5115,12 +5121,12 @@ static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
 		tx_flags |= E1000_TX_FLAGS_IPV4;
 
 	/* if count is 0 then mapping error has occurred */
-	count = e1000_tx_map(adapter, skb, first, max_per_txd, nr_frags, mss);
+	count = e1000_tx_map(tx_ring, skb, first, max_per_txd, nr_frags, mss);
 	if (count) {
 		netdev_sent_queue(netdev, skb->len);
-		e1000_tx_queue(adapter, tx_flags, count);
+		e1000_tx_queue(tx_ring, tx_flags, count);
 		/* Make sure there is space in the ring for the next send. */
-		e1000_maybe_stop_tx(netdev, MAX_SKB_FRAGS + 2);
+		e1000_maybe_stop_tx(tx_ring, MAX_SKB_FRAGS + 2);
 
 	} else {
 		dev_kfree_skb_any(skb);
-- 
1.7.7.4

Re: [net-next 05/11] netdev: FCoE: Add new ndo_get_fcoe_hbainfo() call

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue,  3 Jan 2012 11:19:10 -0800

> +	int			(*ndo_fcoe_get_hbainfo)(struct net_device *dev,
> +					struct netdev_fcoe_hbainfo *hbainfo);

Please line up the second line's argument correctly, it should be:

> +	int			(*ndo_fcoe_get_hbainfo)(struct net_device *dev,
> +							struct netdev_fcoe_hbainfo *hbainfo);

Thanks.

Re: [net-next 07/11] e1000e: cleanup Rx checksum offload code

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue,  3 Jan 2012 11:19:12 -0800

> -		__sum16 sum = (__force __sum16)htons(csum);
> +		__sum16 sum = (__force __sum16)htons(le16_to_cpu(csum));

Looks like a NOP.  It's essentially "cpu_to_le16(le16_to_cpu(csum)" as
far as I can tell.

Re: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[David Miller]

From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue,  3 Jan 2012 11:19:13 -0800

> +		e_err("Enabling both receive checksum offload and receive hashing is not possible with jumbo frames.  Disable jumbos or enable only one of the receive offload features.\n");

Please split this up into multiple lines like this:

	e_err("..."
	      "..."
	      "...");

Thanks.

RE: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[Allan, Bruce W]

>-----Original Message-----
>From: David Miller [mailto:davem@davemloft.net]
>Sent: Tuesday, January 03, 2012 12:10 PM
>To: Kirsher, Jeffrey T
>Cc: Allan, Bruce W; netdev@vger.kernel.org; gospo@redhat.com;
>sassmann@redhat.com
>Subject: Re: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support
>
>From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
>Date: Tue,  3 Jan 2012 11:19:13 -0800
>
>> +		e_err("Enabling both receive checksum offload and receive hashing is
>not possible with jumbo frames.  Disable jumbos or enable only one of the
>receive offload features.\n");
>
>Please split this up into multiple lines like this:
>
>	e_err("..."
>	      "..."
>	      "...");
>
>Thanks.

I was under the impression from recently accepted patches that the preferred
method for log/error message strings was to keep them as a single string
(rather than broken across multiple lines) for easy searching.  Is that not
the case here?

>From ./Documentation/CodingStyle..."never break user-visible strings such as
printk messages, because that breaks the ability to grep for them."

Do you still want this change?

Thanks,
Bruce.

RE: [net-next 07/11] e1000e: cleanup Rx checksum offload code

[Allan, Bruce W]

>-----Original Message-----
>From: David Miller [mailto:davem@davemloft.net]
>Sent: Tuesday, January 03, 2012 12:09 PM
>To: Kirsher, Jeffrey T
>Cc: Allan, Bruce W; netdev@vger.kernel.org; gospo@redhat.com;
>sassmann@redhat.com
>Subject: Re: [net-next 07/11] e1000e: cleanup Rx checksum offload code
>
>From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
>Date: Tue,  3 Jan 2012 11:19:12 -0800
>
>> -		__sum16 sum = (__force __sum16)htons(csum);
>> +		__sum16 sum = (__force __sum16)htons(le16_to_cpu(csum));
>
>Looks like a NOP.  It's essentially "cpu_to_le16(le16_to_cpu(csum)" as
>far as I can tell.

OK, I'll look into fixing this.

Thanks,
Bruce.

Re: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[David Miller]

From: "Allan, Bruce W" <bruce.w.allan@intel.com>
Date: Tue, 3 Jan 2012 20:16:44 +0000

> I was under the impression from recently accepted patches that the preferred
> method for log/error message strings was to keep them as a single string
> (rather than broken across multiple lines) for easy searching.  Is that not
> the case here?
> 
>>From ./Documentation/CodingStyle..."never break user-visible strings such as
> printk messages, because that breaks the ability to grep for them."
> 
> Do you still want this change?

Fair enough, I rescind my request and this patch is fine as-is.

RE: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[Allan, Bruce W]

>-----Original Message-----
>From: David Miller [mailto:davem@davemloft.net]
>Sent: Tuesday, January 03, 2012 12:28 PM
>To: Allan, Bruce W
>Cc: Kirsher, Jeffrey T; netdev@vger.kernel.org; gospo@redhat.com;
>sassmann@redhat.com
>Subject: Re: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support
>
>From: "Allan, Bruce W" <bruce.w.allan@intel.com>
>Date: Tue, 3 Jan 2012 20:16:44 +0000
>
>> I was under the impression from recently accepted patches that the preferred
>> method for log/error message strings was to keep them as a single string
>> (rather than broken across multiple lines) for easy searching.  Is that not
>> the case here?
>>
>>>From ./Documentation/CodingStyle..."never break user-visible strings such as
>> printk messages, because that breaks the ability to grep for them."
>>
>> Do you still want this change?
>
>Fair enough, I rescind my request and this patch is fine as-is.

Okay, thanks Dave.

Bruce.

RE: [net-next 05/11] netdev: FCoE: Add new ndo_get_fcoe_hbainfo() call

[Parikh, Neerav]

> -----Original Message-----
> From: David Miller [mailto:davem@davemloft.net]
> Sent: Tuesday, January 03, 2012 12:08 PM
> To: Kirsher, Jeffrey T
> Cc: Parikh, Neerav; netdev@vger.kernel.org; gospo@redhat.com;
> sassmann@redhat.com
> Subject: Re: [net-next 05/11] netdev: FCoE: Add new
> ndo_get_fcoe_hbainfo() call
> 
> From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
> Date: Tue,  3 Jan 2012 11:19:10 -0800
> 
> > +	int			(*ndo_fcoe_get_hbainfo)(struct net_device *dev,
> > +					struct netdev_fcoe_hbainfo *hbainfo);
> 
> Please line up the second line's argument correctly, it should be:
> 
I was just trying to make checkpatch.pl happy as the line went beyond
80 character limit.

> > +	int			(*ndo_fcoe_get_hbainfo)(struct net_device *dev,
> > +							struct netdev_fcoe_hbainfo
> *hbainfo);
> 
> Thanks.

Re: [net-next 07/11] e1000e: cleanup Rx checksum offload code

[Ben Hutchings]

On Tue, 2012-01-03 at 15:08 -0500, David Miller wrote:
> From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
> Date: Tue,  3 Jan 2012 11:19:12 -0800
> 
> > -		__sum16 sum = (__force __sum16)htons(csum);
> > +		__sum16 sum = (__force __sum16)htons(le16_to_cpu(csum));
> 
> Looks like a NOP.  It's essentially "cpu_to_le16(le16_to_cpu(csum)" as
> far as I can tell.

Looks like a swab() to me...

Ben.

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.

Re: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[Ben Hutchings]

On Tue, 2012-01-03 at 11:19 -0800, Jeff Kirsher wrote:
> From: Bruce Allan <bruce.w.allan@intel.com>
> 
> Enable RPS by default.  Disallow jumbo frames when both receive checksum
> and receive hashing are enabled because the hardware cannot do both IP
> payload checksum (enabled when receive checksum is enabled when using
> packet split which is used for jumbo frames) and provide RSS hash at the
> same time.
[...]

This doesn't seem to have much to do with RPS.  RPS can use a hardware
hash but doesn't require it.

You should also implement the ethtool command to query flow hashing
behaviour (ETHTOOL_GRXFH command, get_rxnfc operation).

Ben.

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.

Re: [net-next 07/11] e1000e: cleanup Rx checksum offload code

[David Miller]

From: Ben Hutchings <bhutchings@solarflare.com>
Date: Tue, 3 Jan 2012 21:00:12 +0000

> On Tue, 2012-01-03 at 15:08 -0500, David Miller wrote:
>> From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
>> Date: Tue,  3 Jan 2012 11:19:12 -0800
>> 
>> > -		__sum16 sum = (__force __sum16)htons(csum);
>> > +		__sum16 sum = (__force __sum16)htons(le16_to_cpu(csum));
>> 
>> Looks like a NOP.  It's essentially "cpu_to_le16(le16_to_cpu(csum)" as
>> far as I can tell.
> 
> Looks like a swab() to me...

I don't see how it can be.  It's effectively doing a 16-bit swap and
then a 16-bit swap again, which is a NOP.

It's the same as "(__force __sum16) csum" and I bet if the code above
works, then this expression I'm suggesting will work too.

Re: [net-next 07/11] e1000e: cleanup Rx checksum offload code

[Ben Hutchings]

On Tue, 2012-01-03 at 16:15 -0500, David Miller wrote:
> From: Ben Hutchings <bhutchings@solarflare.com>
> Date: Tue, 3 Jan 2012 21:00:12 +0000
> 
> > On Tue, 2012-01-03 at 15:08 -0500, David Miller wrote:
> >> From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
> >> Date: Tue,  3 Jan 2012 11:19:12 -0800
> >> 
> >> > -		__sum16 sum = (__force __sum16)htons(csum);
> >> > +		__sum16 sum = (__force __sum16)htons(le16_to_cpu(csum));
> >> 
> >> Looks like a NOP.  It's essentially "cpu_to_le16(le16_to_cpu(csum)" as
> >> far as I can tell.
> > 
> > Looks like a swab() to me...
> 
> I don't see how it can be.  It's effectively doing a 16-bit swap and
> then a 16-bit swap again, which is a NOP.
> 
> It's the same as "(__force __sum16) csum" and I bet if the code above
> works, then this expression I'm suggesting will work too.

In this part of the universe, network order is big-endian.

Ben.

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.

Re: [net-next 07/11] e1000e: cleanup Rx checksum offload code

[David Miller]

From: Ben Hutchings <bhutchings@solarflare.com>
Date: Tue, 3 Jan 2012 21:24:02 +0000

> On Tue, 2012-01-03 at 16:15 -0500, David Miller wrote:
>> From: Ben Hutchings <bhutchings@solarflare.com>
>> Date: Tue, 3 Jan 2012 21:00:12 +0000
>> 
>> > On Tue, 2012-01-03 at 15:08 -0500, David Miller wrote:
>> >> From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
>> >> Date: Tue,  3 Jan 2012 11:19:12 -0800
>> >> 
>> >> > -		__sum16 sum = (__force __sum16)htons(csum);
>> >> > +		__sum16 sum = (__force __sum16)htons(le16_to_cpu(csum));
>> >> 
>> >> Looks like a NOP.  It's essentially "cpu_to_le16(le16_to_cpu(csum)" as
>> >> far as I can tell.
>> > 
>> > Looks like a swab() to me...
>> 
>> I don't see how it can be.  It's effectively doing a 16-bit swap and
>> then a 16-bit swap again, which is a NOP.
>> 
>> It's the same as "(__force __sum16) csum" and I bet if the code above
>> works, then this expression I'm suggesting will work too.
> 
> In this part of the universe, network order is big-endian.

Indeed, you're right, therefore this is something like swab().
:-)

RE: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[Joe Perches]

On Tue, 2012-01-03 at 20:28 +0000, Allan, Bruce W wrote:
> >From: David Miller [mailto:davem@davemloft.net]
> >Fair enough, I rescind my request and this patch is fine as-is.
> Okay, thanks Dave.

A couple of other comments though.

On Tue, 2012-01-03 at 11:19 -0800, Jeff Kirsher wrote:
> From: Bruce Allan <bruce.w.allan@intel.com>
> 
> Enable RPS by default.  Disallow jumbo frames when both receive checksum
> and receive hashing are enabled because the hardware cannot do both IP
> payload checksum (enabled when receive checksum is enabled when using
> packet split which is used for jumbo frames) and provide RSS hash at the
> same time.
[]
> diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c
[]
> +static void e1000e_setup_rss_hash(struct e1000_adapter *adapter)
[]
> +	static const u8 rsshash[40] = {
[]
+	};
> +
> +	/* Fill out hash function seeds */
> +	for (j = 0; j < 10; j++) {
> +		u32 rsskey = rsshash[(j * 4)];
> +		rsskey |= rsshash[(j * 4) + 1] << 8;
> +		rsskey |= rsshash[(j * 4) + 2] << 16;
> +		rsskey |= rsshash[(j * 4) + 3] << 24;

Strictly, don't these shifts first need a cast to u32?

	u32 rsskey = rsshash[j * 4];
	rsskey |= ((u32)rsshash[j * 4 + 1]) << 8;
	rsskey |= ((u32)rsshash[j * 4 + 2]) << 16;
	rsskey |= ((u32)rsshash[j * 4 + 3]) << 24;

[]

> +	if (adapter->rx_ps_pages &&
> +	    (features & NETIF_F_RXCSUM) && (features & NETIF_F_RXHASH)) {
> +		e_err("Enabling both receive checksum offload and receive hashing is not possible with jumbo frames.  Disable jumbos or enable only one of the receive offload features.\n");
> +		return -EINVAL;
> +	}
 
This is a very long output string.
I think keeping individual dmesg lines shorter is preferred.
It might be better to use 2 e_err lines like:

		e_err("Enabling both receive checksum offload and receive hashing is not possible with jumbo frames\n");
		e_err("Disable jumbos or enable only one of the receive offload features\n");

or

		e_err("Enabling both receive checksum offload and receive hashing is not possible with jumbo frames\n"
		      "Disable jumbos or enable only one of the receive offload features\n");

Re: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[Harvey Harrison]

On Tue, Jan 3, 2012 at 4:12 PM, Joe Perches <joe@perches.com> wrote:
> On Tue, 2012-01-03 at 20:28 +0000, Allan, Bruce W wrote:
>> +
>> +     /* Fill out hash function seeds */
>> +     for (j = 0; j < 10; j++) {
>> +             u32 rsskey = rsshash[(j * 4)];
>> +             rsskey |= rsshash[(j * 4) + 1] << 8;
>> +             rsskey |= rsshash[(j * 4) + 2] << 16;
>> +             rsskey |= rsshash[(j * 4) + 3] << 24;
>

Perhaps consider get_unaligned_le32()

Cheers,

Harvey

RE: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[Allan, Bruce W]

>-----Original Message-----
>From: Ben Hutchings [mailto:bhutchings@solarflare.com]
>Sent: Tuesday, January 03, 2012 1:05 PM
>To: Kirsher, Jeffrey T
>Cc: davem@davemloft.net; Allan, Bruce W; netdev@vger.kernel.org;
>gospo@redhat.com; sassmann@redhat.com
>Subject: Re: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support
>
>On Tue, 2012-01-03 at 11:19 -0800, Jeff Kirsher wrote:
>> From: Bruce Allan <bruce.w.allan@intel.com>
>>
>> Enable RPS by default.  Disallow jumbo frames when both receive checksum
>> and receive hashing are enabled because the hardware cannot do both IP
>> payload checksum (enabled when receive checksum is enabled when using
>> packet split which is used for jumbo frames) and provide RSS hash at the
>> same time.
>[...]
>
>This doesn't seem to have much to do with RPS.  RPS can use a hardware
>hash but doesn't require it.
>
>You should also implement the ethtool command to query flow hashing
>behaviour (ETHTOOL_GRXFH command, get_rxnfc operation).
>
>Ben.
>
>--
>Ben Hutchings, Staff Engineer, Solarflare
>Not speaking for my employer; that's the marketing department's job.
>They asked us to note that Solarflare product names are trademarked.

Thanks for the feedback Ben.  One question: is the data returned from
an ETHTOOL_GRXFH command with the get_rxnfc operation supposed to be
the default for the device or the current setting?

Thanks,
Bruce.

RE: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support

[Ben Hutchings]

On Fri, 2012-01-06 at 17:22 +0000, Allan, Bruce W wrote:
> >-----Original Message-----
> >From: Ben Hutchings [mailto:bhutchings@solarflare.com]
> >Sent: Tuesday, January 03, 2012 1:05 PM
> >To: Kirsher, Jeffrey T
> >Cc: davem@davemloft.net; Allan, Bruce W; netdev@vger.kernel.org;
> >gospo@redhat.com; sassmann@redhat.com
> >Subject: Re: [net-next 08/11] e1000e: add Receive Packet Steering (RPS) support
> >
> >On Tue, 2012-01-03 at 11:19 -0800, Jeff Kirsher wrote:
> >> From: Bruce Allan <bruce.w.allan@intel.com>
> >>
> >> Enable RPS by default.  Disallow jumbo frames when both receive checksum
> >> and receive hashing are enabled because the hardware cannot do both IP
> >> payload checksum (enabled when receive checksum is enabled when using
> >> packet split which is used for jumbo frames) and provide RSS hash at the
> >> same time.
> >[...]
> >
> >This doesn't seem to have much to do with RPS.  RPS can use a hardware
> >hash but doesn't require it.
> >
> >You should also implement the ethtool command to query flow hashing
> >behaviour (ETHTOOL_GRXFH command, get_rxnfc operation).
> >
> >Ben.
> >
> >--
> >Ben Hutchings, Staff Engineer, Solarflare
> >Not speaking for my employer; that's the marketing department's job.
> >They asked us to note that Solarflare product names are trademarked.
> 
> Thanks for the feedback Ben.  One question: is the data returned from
> an ETHTOOL_GRXFH command with the get_rxnfc operation supposed to be
> the default for the device or the current setting?

It's supposed to return the current setting.  There's no way to find out
the default once the behaviour has been changed.

Ben.

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.