Merge branches 'cma', 'cxgb3', 'ehca', 'ipoib', 'mad', 'mlx4' and 'nes' into for-next

MAINTAINERS

@@ -2899,9 +2899,9 @@ S:	Maintained
 
 NETEFFECT IWARP RNIC DRIVER (IW_NES)
 P:	Faisal Latif
-M:	flatif@neteffect.com
+M:	faisal.latif@intel.com
 P:	Chien Tung
-M:	ctung@neteffect.com
+M:	chien.tin.tung@intel.com
 L:	general@lists.openfabrics.org
 W:	http://www.neteffect.com
 S:	Supported

drivers/infiniband/core/mad.c

@@ -406,19 +406,15 @@ static int register_snoop_agent(struct ib_mad_qp_info *qp_info,
 
 	if (i == qp_info->snoop_table_size) {
 		/* Grow table. */
-		new_snoop_table = kmalloc(sizeof mad_snoop_priv *
-					  qp_info->snoop_table_size + 1,
-					  GFP_ATOMIC);
+		new_snoop_table = krealloc(qp_info->snoop_table,
+					   sizeof mad_snoop_priv *
+					   (qp_info->snoop_table_size + 1),
+					   GFP_ATOMIC);
 		if (!new_snoop_table) {
 			i = -ENOMEM;
 			goto out;
 		}
-		if (qp_info->snoop_table) {
-			memcpy(new_snoop_table, qp_info->snoop_table,
-			       sizeof mad_snoop_priv *
-			       qp_info->snoop_table_size);
-			kfree(qp_info->snoop_table);
-		}
+
 		qp_info->snoop_table = new_snoop_table;
 		qp_info->snoop_table_size++;
 	}

drivers/infiniband/hw/cxgb3/iwch_cm.c

@@ -1942,6 +1942,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 fail3:
 	cxgb3_free_atid(ep->com.tdev, ep->atid);
 fail2:
+	cm_id->rem_ref(cm_id);
 	put_ep(&ep->com);
 out:
 	return err;

drivers/infiniband/hw/ehca/ehca_classes.h

@@ -128,6 +128,8 @@ struct ehca_shca {
 	/* MR pgsize: bit 0-3 means 4K, 64K, 1M, 16M respectively */
 	u32 hca_cap_mr_pgsize;
 	int max_mtu;
+	int max_num_qps;
+	int max_num_cqs;
 	atomic_t num_cqs;
 	atomic_t num_qps;
 };

drivers/infiniband/hw/ehca/ehca_cq.c

@@ -132,9 +132,9 @@ struct ib_cq *ehca_create_cq(struct ib_device *device, int cqe, int comp_vector,
 	if (cqe >= 0xFFFFFFFF - 64 - additional_cqe)
 		return ERR_PTR(-EINVAL);
 
-	if (!atomic_add_unless(&shca->num_cqs, 1, ehca_max_cq)) {
+	if (!atomic_add_unless(&shca->num_cqs, 1, shca->max_num_cqs)) {
 		ehca_err(device, "Unable to create CQ, max number of %i "
-			"CQs reached.", ehca_max_cq);
+			"CQs reached.", shca->max_num_cqs);
 		ehca_err(device, "To increase the maximum number of CQs "
 			"use the number_of_cqs module parameter.\n");
 		return ERR_PTR(-ENOSPC);

drivers/infiniband/hw/ehca/ehca_main.c

@@ -44,6 +44,8 @@
 #include <linux/slab.h>
 #endif
 
+#include <linux/notifier.h>
+#include <linux/memory.h>
 #include "ehca_classes.h"
 #include "ehca_iverbs.h"
 #include "ehca_mrmw.h"
@@ -366,22 +368,23 @@ static int ehca_sense_attributes(struct ehca_shca *shca)
 			shca->hca_cap_mr_pgsize |= pgsize_map[i + 1];
 
 	/* Set maximum number of CQs and QPs to calculate EQ size */
-	if (ehca_max_qp == -1)
-		ehca_max_qp = min_t(int, rblock->max_qp, EHCA_MAX_NUM_QUEUES);
-	else if (ehca_max_qp < 1 || ehca_max_qp > rblock->max_qp) {
-		ehca_gen_err("Requested number of QPs is out of range (1 - %i) "
-			"specified by HW", rblock->max_qp);
-		ret = -EINVAL;
-		goto sense_attributes1;
+	if (shca->max_num_qps == -1)
+		shca->max_num_qps = min_t(int, rblock->max_qp,
+					  EHCA_MAX_NUM_QUEUES);
+	else if (shca->max_num_qps < 1 || shca->max_num_qps > rblock->max_qp) {
+		ehca_gen_warn("The requested number of QPs is out of range "
+			      "(1 - %i) specified by HW. Value is set to %i",
+			      rblock->max_qp, rblock->max_qp);
+		shca->max_num_qps = rblock->max_qp;
 	}
 
-	if (ehca_max_cq == -1)
-		ehca_max_cq = min_t(int, rblock->max_cq, EHCA_MAX_NUM_QUEUES);
-	else if (ehca_max_cq < 1 || ehca_max_cq > rblock->max_cq) {
-		ehca_gen_err("Requested number of CQs is out of range (1 - %i) "
-			"specified by HW", rblock->max_cq);
-		ret = -EINVAL;
-		goto sense_attributes1;
+	if (shca->max_num_cqs == -1)
+		shca->max_num_cqs = min_t(int, rblock->max_cq,
+					  EHCA_MAX_NUM_QUEUES);
+	else if (shca->max_num_cqs < 1 || shca->max_num_cqs > rblock->max_cq) {
+		ehca_gen_warn("The requested number of CQs is out of range "
+			      "(1 - %i) specified by HW. Value is set to %i",
+			      rblock->max_cq, rblock->max_cq);
 	}
 
 	/* query max MTU from first port -- it's the same for all ports */
@@ -733,9 +736,13 @@ static int __devinit ehca_probe(struct of_device *dev,
 		ehca_gen_err("Cannot allocate shca memory.");
 		return -ENOMEM;
 	}
+
 	mutex_init(&shca->modify_mutex);
 	atomic_set(&shca->num_cqs, 0);
 	atomic_set(&shca->num_qps, 0);
+	shca->max_num_qps = ehca_max_qp;
+	shca->max_num_cqs = ehca_max_cq;
+
 	for (i = 0; i < ARRAY_SIZE(shca->sport); i++)
 		spin_lock_init(&shca->sport[i].mod_sqp_lock);
 
@@ -755,7 +762,7 @@ static int __devinit ehca_probe(struct of_device *dev,
 		goto probe1;
 	}
 
-	eq_size = 2 * ehca_max_cq + 4 * ehca_max_qp;
+	eq_size = 2 * shca->max_num_cqs + 4 * shca->max_num_qps;
 	/* create event queues */
 	ret = ehca_create_eq(shca, &shca->eq, EHCA_EQ, eq_size);
 	if (ret) {
@@ -964,6 +971,41 @@ void ehca_poll_eqs(unsigned long data)
 	spin_unlock(&shca_list_lock);
 }
 
+static int ehca_mem_notifier(struct notifier_block *nb,
+			     unsigned long action, void *data)
+{
+	static unsigned long ehca_dmem_warn_time;
+
+	switch (action) {
+	case MEM_CANCEL_OFFLINE:
+	case MEM_CANCEL_ONLINE:
+	case MEM_ONLINE:
+	case MEM_OFFLINE:
+		return NOTIFY_OK;
+	case MEM_GOING_ONLINE:
+	case MEM_GOING_OFFLINE:
+		/* only ok if no hca is attached to the lpar */
+		spin_lock(&shca_list_lock);
+		if (list_empty(&shca_list)) {
+			spin_unlock(&shca_list_lock);
+			return NOTIFY_OK;
+		} else {
+			spin_unlock(&shca_list_lock);
+			if (printk_timed_ratelimit(&ehca_dmem_warn_time,
+						   30 * 1000))
+				ehca_gen_err("DMEM operations are not allowed"
+					     "as long as an ehca adapter is"
+					     "attached to the LPAR");
+			return NOTIFY_BAD;
+		}
+	}
+	return NOTIFY_OK;
+}
+
+static struct notifier_block ehca_mem_nb = {
+	.notifier_call = ehca_mem_notifier,
+};
+
 static int __init ehca_module_init(void)
 {
 	int ret;
@@ -991,6 +1033,12 @@ static int __init ehca_module_init(void)
 		goto module_init2;
 	}
 
+	ret = register_memory_notifier(&ehca_mem_nb);
+	if (ret) {
+		ehca_gen_err("Failed registering memory add/remove notifier");
+		goto module_init3;
+	}
+
 	if (ehca_poll_all_eqs != 1) {
 		ehca_gen_err("WARNING!!!");
 		ehca_gen_err("It is possible to lose interrupts.");
@@ -1003,6 +1051,9 @@ static int __init ehca_module_init(void)
 
 	return 0;
 
+module_init3:
+	ibmebus_unregister_driver(&ehca_driver);
+
 module_init2:
 	ehca_destroy_slab_caches();
 
@@ -1018,6 +1069,8 @@ static void __exit ehca_module_exit(void)
 
 	ibmebus_unregister_driver(&ehca_driver);
 
+	unregister_memory_notifier(&ehca_mem_nb);
+
 	ehca_destroy_slab_caches();
 
 	ehca_destroy_comp_pool();

drivers/infiniband/hw/ehca/ehca_qp.c

@@ -465,9 +465,9 @@ static struct ehca_qp *internal_create_qp(
 	u32 swqe_size = 0, rwqe_size = 0, ib_qp_num;
 	unsigned long flags;
 
-	if (!atomic_add_unless(&shca->num_qps, 1, ehca_max_qp)) {
+	if (!atomic_add_unless(&shca->num_qps, 1, shca->max_num_qps)) {
 		ehca_err(pd->device, "Unable to create QP, max number of %i "
-			 "QPs reached.", ehca_max_qp);
+			 "QPs reached.", shca->max_num_qps);
 		ehca_err(pd->device, "To increase the maximum number of QPs "
 			 "use the number_of_qps module parameter.\n");
 		return ERR_PTR(-ENOSPC);
@@ -502,6 +502,12 @@ static struct ehca_qp *internal_create_qp(
 	if (init_attr->srq) {
 		my_srq = container_of(init_attr->srq, struct ehca_qp, ib_srq);
 
+		if (qp_type == IB_QPT_UC) {
+			ehca_err(pd->device, "UC with SRQ not supported");
+			atomic_dec(&shca->num_qps);
+			return ERR_PTR(-EINVAL);
+		}
+
 		has_srq = 1;
 		parms.ext_type = EQPT_SRQBASE;
 		parms.srq_qpn = my_srq->real_qp_num;

drivers/infiniband/hw/mlx4/mad.c

@@ -298,7 +298,7 @@ int mlx4_ib_mad_init(struct mlx4_ib_dev *dev)
 	int p, q;
 	int ret;
 
-	for (p = 0; p < dev->dev->caps.num_ports; ++p)
+	for (p = 0; p < dev->num_ports; ++p)
 		for (q = 0; q <= 1; ++q) {
 			agent = ib_register_mad_agent(&dev->ib_dev, p + 1,
 						      q ? IB_QPT_GSI : IB_QPT_SMI,
@@ -314,7 +314,7 @@ int mlx4_ib_mad_init(struct mlx4_ib_dev *dev)
 	return 0;
 
 err:
-	for (p = 0; p < dev->dev->caps.num_ports; ++p)
+	for (p = 0; p < dev->num_ports; ++p)
 		for (q = 0; q <= 1; ++q)
 			if (dev->send_agent[p][q])
 				ib_unregister_mad_agent(dev->send_agent[p][q]);
@@ -327,7 +327,7 @@ void mlx4_ib_mad_cleanup(struct mlx4_ib_dev *dev)
 	struct ib_mad_agent *agent;
 	int p, q;
 
-	for (p = 0; p < dev->dev->caps.num_ports; ++p) {
+	for (p = 0; p < dev->num_ports; ++p) {
 		for (q = 0; q <= 1; ++q) {
 			agent = dev->send_agent[p][q];
 			dev->send_agent[p][q] = NULL;

drivers/infiniband/hw/mlx4/main.c

@@ -574,7 +574,10 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 	ibdev->ib_dev.owner		= THIS_MODULE;
 	ibdev->ib_dev.node_type		= RDMA_NODE_IB_CA;
 	ibdev->ib_dev.local_dma_lkey	= dev->caps.reserved_lkey;
-	ibdev->ib_dev.phys_port_cnt	= dev->caps.num_ports;
+	ibdev->num_ports = 0;
+	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB)
+		ibdev->num_ports++;
+	ibdev->ib_dev.phys_port_cnt     = ibdev->num_ports;
 	ibdev->ib_dev.num_comp_vectors	= 1;
 	ibdev->ib_dev.dma_device	= &dev->pdev->dev;
 
@@ -691,7 +694,7 @@ static void mlx4_ib_remove(struct mlx4_dev *dev, void *ibdev_ptr)
 	struct mlx4_ib_dev *ibdev = ibdev_ptr;
 	int p;
 
-	for (p = 1; p <= dev->caps.num_ports; ++p)
+	for (p = 1; p <= ibdev->num_ports; ++p)
 		mlx4_CLOSE_PORT(dev, p);
 
 	mlx4_ib_mad_cleanup(ibdev);
@@ -706,6 +709,10 @@ static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr,
 			  enum mlx4_dev_event event, int port)
 {
 	struct ib_event ibev;
+	struct mlx4_ib_dev *ibdev = to_mdev((struct ib_device *) ibdev_ptr);
+
+	if (port > ibdev->num_ports)
+		return;
 
 	switch (event) {
 	case MLX4_DEV_EVENT_PORT_UP:

drivers/infiniband/hw/mlx4/mlx4_ib.h

@@ -162,6 +162,7 @@ struct mlx4_ib_ah {
 struct mlx4_ib_dev {
 	struct ib_device	ib_dev;
 	struct mlx4_dev	       *dev;
+	int			num_ports;
 	void __iomem	       *uar_map;
 
 	struct mlx4_uar		priv_uar;

drivers/infiniband/hw/mlx4/qp.c

@@ -451,6 +451,7 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
 			    struct ib_qp_init_attr *init_attr,
 			    struct ib_udata *udata, int sqpn, struct mlx4_ib_qp *qp)
 {
+	int qpn;
 	int err;
 
 	mutex_init(&qp->mutex);
@@ -545,9 +546,17 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
 		}
 	}
 
-	err = mlx4_qp_alloc(dev->dev, sqpn, &qp->mqp);
+	if (sqpn) {
+		qpn = sqpn;
+	} else {
+		err = mlx4_qp_reserve_range(dev->dev, 1, 1, &qpn);
+		if (err)
+			goto err_wrid;
+	}
+
+	err = mlx4_qp_alloc(dev->dev, qpn, &qp->mqp);
 	if (err)
-		goto err_wrid;
+		goto err_qpn;
 
 	/*
 	 * Hardware wants QPN written in big-endian order (after
@@ -560,6 +569,10 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
 
 	return 0;
 
+err_qpn:
+	if (!sqpn)
+		mlx4_qp_release_range(dev->dev, qpn, 1);
+
 err_wrid:
 	if (pd->uobject) {
 		if (!init_attr->srq)
@@ -655,6 +668,10 @@ static void destroy_qp_common(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp,
 	mlx4_ib_unlock_cqs(send_cq, recv_cq);
 
 	mlx4_qp_free(dev->dev, &qp->mqp);
+
+	if (!is_sqp(dev, qp))
+		mlx4_qp_release_range(dev->dev, qp->mqp.qpn, 1);
+
 	mlx4_mtt_cleanup(dev->dev, &qp->mtt);
 
 	if (is_user) {

drivers/infiniband/ulp/ipoib/ipoib.h

@@ -507,6 +507,7 @@ int ipoib_pkey_dev_delay_open(struct net_device *dev);
 void ipoib_drain_cq(struct net_device *dev);
 
 void ipoib_set_ethtool_ops(struct net_device *dev);
+int ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca);
 
 #ifdef CONFIG_INFINIBAND_IPOIB_CM
 

drivers/infiniband/ulp/ipoib/ipoib_ethtool.c

@@ -42,6 +42,13 @@ static void ipoib_get_drvinfo(struct net_device *netdev,
 	strncpy(drvinfo->driver, "ipoib", sizeof(drvinfo->driver) - 1);
 }
 
+static u32 ipoib_get_rx_csum(struct net_device *dev)
+{
+	struct ipoib_dev_priv *priv = netdev_priv(dev);
+	return test_bit(IPOIB_FLAG_CSUM, &priv->flags) &&
+		!test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags);
+}
+
 static int ipoib_get_coalesce(struct net_device *dev,
 			      struct ethtool_coalesce *coal)
 {
@@ -129,7 +136,7 @@ static void ipoib_get_ethtool_stats(struct net_device *dev,
 
 static const struct ethtool_ops ipoib_ethtool_ops = {
 	.get_drvinfo		= ipoib_get_drvinfo,
-	.get_tso		= ethtool_op_get_tso,
+	.get_rx_csum		= ipoib_get_rx_csum,
 	.get_coalesce		= ipoib_get_coalesce,
 	.set_coalesce		= ipoib_set_coalesce,
 	.get_flags		= ethtool_op_get_flags,

drivers/infiniband/ulp/ipoib/ipoib_ib.c

@@ -685,10 +685,6 @@ int ipoib_ib_dev_open(struct net_device *dev)
 	queue_delayed_work(ipoib_workqueue, &priv->ah_reap_task,
 			   round_jiffies_relative(HZ));
 
-	init_timer(&priv->poll_timer);
-	priv->poll_timer.function = ipoib_ib_tx_timer_func;
-	priv->poll_timer.data = (unsigned long)dev;
-
 	set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags);
 
 	return 0;
@@ -906,6 +902,9 @@ int ipoib_ib_dev_init(struct net_device *dev, struct ib_device *ca, int port)
 		return -ENODEV;
 	}
 
+	setup_timer(&priv->poll_timer, ipoib_ib_tx_timer_func,
+		    (unsigned long) dev);
+
 	if (dev->flags & IFF_UP) {
 		if (ipoib_ib_dev_open(dev)) {
 			ipoib_transport_dev_cleanup(dev);

drivers/infiniband/ulp/ipoib/ipoib_main.c

@@ -1173,11 +1173,48 @@ int ipoib_add_pkey_attr(struct net_device *dev)
 	return device_create_file(&dev->dev, &dev_attr_pkey);
 }
 
+int ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca)
+{
+	struct ib_device_attr *device_attr;
+	int result = -ENOMEM;
+
+	device_attr = kmalloc(sizeof *device_attr, GFP_KERNEL);
+	if (!device_attr) {
+		printk(KERN_WARNING "%s: allocation of %zu bytes failed\n",
+		       hca->name, sizeof *device_attr);
+		return result;
+	}
+
+	result = ib_query_device(hca, device_attr);
+	if (result) {
+		printk(KERN_WARNING "%s: ib_query_device failed (ret = %d)\n",
+		       hca->name, result);
+		kfree(device_attr);
+		return result;
+	}
+	priv->hca_caps = device_attr->device_cap_flags;
+
+	kfree(device_attr);
+
+	if (priv->hca_caps & IB_DEVICE_UD_IP_CSUM) {
+		set_bit(IPOIB_FLAG_CSUM, &priv->flags);
+		priv->dev->features |= NETIF_F_SG | NETIF_F_IP_CSUM;
+	}
+
+	if (lro)
+		priv->dev->features |= NETIF_F_LRO;
+
+	if (priv->dev->features & NETIF_F_SG && priv->hca_caps & IB_DEVICE_UD_TSO)
+		priv->dev->features |= NETIF_F_TSO;
+
+	return 0;
+}
+
+
 static struct net_device *ipoib_add_port(const char *format,
 					 struct ib_device *hca, u8 port)
 {
 	struct ipoib_dev_priv *priv;
-	struct ib_device_attr *device_attr;
 	struct ib_port_attr attr;
 	int result = -ENOMEM;
 
@@ -1206,31 +1243,8 @@ static struct net_device *ipoib_add_port(const char *format,
 		goto device_init_failed;
 	}
 
-	device_attr = kmalloc(sizeof *device_attr, GFP_KERNEL);
-	if (!device_attr) {
-		printk(KERN_WARNING "%s: allocation of %zu bytes failed\n",
-		       hca->name, sizeof *device_attr);
+	if (ipoib_set_dev_features(priv, hca))
 		goto device_init_failed;
-	}
-
-	result = ib_query_device(hca, device_attr);
-	if (result) {
-		printk(KERN_WARNING "%s: ib_query_device failed (ret = %d)\n",
-		       hca->name, result);
-		kfree(device_attr);
-		goto device_init_failed;
-	}
-	priv->hca_caps = device_attr->device_cap_flags;
-
-	kfree(device_attr);
-
-	if (priv->hca_caps & IB_DEVICE_UD_IP_CSUM) {
-		set_bit(IPOIB_FLAG_CSUM, &priv->flags);
-		priv->dev->features |= NETIF_F_SG | NETIF_F_IP_CSUM;
-	}
-
-	if (lro)
-		priv->dev->features |= NETIF_F_LRO;
 
 	/*
 	 * Set the full membership bit, so that we join the right
@@ -1266,9 +1280,6 @@ static struct net_device *ipoib_add_port(const char *format,
 		goto event_failed;
 	}
 
-	if (priv->dev->features & NETIF_F_SG && priv->hca_caps & IB_DEVICE_UD_TSO)
-		priv->dev->features |= NETIF_F_TSO;
-
 	result = register_netdev(priv->dev);
 	if (result) {
 		printk(KERN_WARNING "%s: couldn't register ipoib port %d; error %d\n",

drivers/infiniband/ulp/ipoib/ipoib_vlan.c

@@ -93,6 +93,10 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey)
 	priv->mcast_mtu  = priv->admin_mtu = priv->dev->mtu;
 	set_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags);
 
+	result = ipoib_set_dev_features(priv, ppriv->ca);
+	if (result)
+		goto device_init_failed;
+
 	priv->pkey = pkey;
 
 	memcpy(priv->dev->dev_addr, ppriv->dev->dev_addr, INFINIBAND_ALEN);

drivers/net/Kconfig

@@ -2465,6 +2465,15 @@ config PASEMI_MAC
 	  This driver supports the on-chip 1/10Gbit Ethernet controller on
 	  PA Semi's PWRficient line of chips.
 
+config MLX4_EN
+	tristate "Mellanox Technologies 10Gbit Ethernet support"
+	depends on PCI && INET
+	select MLX4_CORE
+	select INET_LRO
+	help
+	  This driver supports Mellanox Technologies ConnectX Ethernet
+	  devices.
+
 config MLX4_CORE
 	tristate
 	depends on PCI

drivers/net/mlx4/Makefile

 obj-$(CONFIG_MLX4_CORE)		+= mlx4_core.o
 
 mlx4_core-y :=	alloc.o catas.o cmd.o cq.o eq.o fw.o icm.o intf.o main.o mcg.o \
-		mr.o pd.o profile.o qp.o reset.o srq.o
+		mr.o pd.o port.o profile.o qp.o reset.o srq.o
+
+obj-$(CONFIG_MLX4_EN)               += mlx4_en.o
+
+mlx4_en-y := 	en_main.o en_tx.o en_rx.o en_params.o en_port.o en_cq.o \
+		en_resources.o en_netdev.o

drivers/net/mlx4/alloc.c

@@ -47,13 +47,16 @@ u32 mlx4_bitmap_alloc(struct mlx4_bitmap *bitmap)
 
 	obj = find_next_zero_bit(bitmap->table, bitmap->max, bitmap->last);
 	if (obj >= bitmap->max) {
-		bitmap->top = (bitmap->top + bitmap->max) & bitmap->mask;
+		bitmap->top = (bitmap->top + bitmap->max + bitmap->reserved_top)
+				& bitmap->mask;
 		obj = find_first_zero_bit(bitmap->table, bitmap->max);
 	}
 
 	if (obj < bitmap->max) {
 		set_bit(obj, bitmap->table);
-		bitmap->last = (obj + 1) & (bitmap->max - 1);
+		bitmap->last = (obj + 1);
+		if (bitmap->last == bitmap->max)
+			bitmap->last = 0;
 		obj |= bitmap->top;
 	} else
 		obj = -1;
@@ -65,16 +68,90 @@ u32 mlx4_bitmap_alloc(struct mlx4_bitmap *bitmap)
 
 void mlx4_bitmap_free(struct mlx4_bitmap *bitmap, u32 obj)
 {
-	obj &= bitmap->max - 1;
+	mlx4_bitmap_free_range(bitmap, obj, 1);
+}
+
+static unsigned long find_aligned_range(unsigned long *bitmap,
+					u32 start, u32 nbits,
+					int len, int align)
+{
+	unsigned long end, i;
+
+again:
+	start = ALIGN(start, align);
+
+	while ((start < nbits) && test_bit(start, bitmap))
+		start += align;
+
+	if (start >= nbits)
+		return -1;
+
+	end = start+len;
+	if (end > nbits)
+		return -1;
+
+	for (i = start + 1; i < end; i++) {
+		if (test_bit(i, bitmap)) {
+			start = i + 1;
+			goto again;
+		}
+	}
+
+	return start;
+}
+
+u32 mlx4_bitmap_alloc_range(struct mlx4_bitmap *bitmap, int cnt, int align)
+{
+	u32 obj, i;
+
+	if (likely(cnt == 1 && align == 1))
+		return mlx4_bitmap_alloc(bitmap);
+
+	spin_lock(&bitmap->lock);
+
+	obj = find_aligned_range(bitmap->table, bitmap->last,
+				 bitmap->max, cnt, align);
+	if (obj >= bitmap->max) {
+		bitmap->top = (bitmap->top + bitmap->max + bitmap->reserved_top)
+				& bitmap->mask;
+		obj = find_aligned_range(bitmap->table, 0, bitmap->max,
+					 cnt, align);
+	}
+
+	if (obj < bitmap->max) {
+		for (i = 0; i < cnt; i++)
+			set_bit(obj + i, bitmap->table);
+		if (obj == bitmap->last) {
+			bitmap->last = (obj + cnt);
+			if (bitmap->last >= bitmap->max)
+				bitmap->last = 0;
+		}
+		obj |= bitmap->top;
+	} else
+		obj = -1;
+
+	spin_unlock(&bitmap->lock);
+
+	return obj;
+}
+
+void mlx4_bitmap_free_range(struct mlx4_bitmap *bitmap, u32 obj, int cnt)
+{
+	u32 i;
+
+	obj &= bitmap->max + bitmap->reserved_top - 1;
 
 	spin_lock(&bitmap->lock);
-	clear_bit(obj, bitmap->table);
+	for (i = 0; i < cnt; i++)
+		clear_bit(obj + i, bitmap->table);
 	bitmap->last = min(bitmap->last, obj);
-	bitmap->top = (bitmap->top + bitmap->max) & bitmap->mask;
+	bitmap->top = (bitmap->top + bitmap->max + bitmap->reserved_top)
+			& bitmap->mask;
 	spin_unlock(&bitmap->lock);
 }
 
-int mlx4_bitmap_init(struct mlx4_bitmap *bitmap, u32 num, u32 mask, u32 reserved)
+int mlx4_bitmap_init(struct mlx4_bitmap *bitmap, u32 num, u32 mask,
+		     u32 reserved_bot, u32 reserved_top)
 {
 	int i;
 
@@ -84,14 +161,16 @@ int mlx4_bitmap_init(struct mlx4_bitmap *bitmap, u32 num, u32 mask, u32 reserved
 
 	bitmap->last = 0;
 	bitmap->top  = 0;
-	bitmap->max  = num;
+	bitmap->max  = num - reserved_top;
 	bitmap->mask = mask;
+	bitmap->reserved_top = reserved_top;
 	spin_lock_init(&bitmap->lock);
-	bitmap->table = kzalloc(BITS_TO_LONGS(num) * sizeof (long), GFP_KERNEL);
+	bitmap->table = kzalloc(BITS_TO_LONGS(bitmap->max) *
+				sizeof (long), GFP_KERNEL);
 	if (!bitmap->table)
 		return -ENOMEM;
 
-	for (i = 0; i < reserved; ++i)
+	for (i = 0; i < reserved_bot; ++i)
 		set_bit(i, bitmap->table);
 
 	return 0;

drivers/net/mlx4/cq.c

@@ -300,7 +300,7 @@ int mlx4_init_cq_table(struct mlx4_dev *dev)
 	INIT_RADIX_TREE(&cq_table->tree, GFP_ATOMIC);
 
 	err = mlx4_bitmap_init(&cq_table->bitmap, dev->caps.num_cqs,
-			       dev->caps.num_cqs - 1, dev->caps.reserved_cqs);
+			       dev->caps.num_cqs - 1, dev->caps.reserved_cqs, 0);
 	if (err)
 		return err;
 

drivers/net/mlx4/en_cq.c

+/*
+ * Copyright (c) 2007 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/mlx4/cq.h>
+#include <linux/mlx4/qp.h>
+#include <linux/mlx4/cmd.h>
+
+#include "mlx4_en.h"
+
+static void mlx4_en_cq_event(struct mlx4_cq *cq, enum mlx4_event event)
+{
+	return;
+}
+
+
+int mlx4_en_create_cq(struct mlx4_en_priv *priv,
+		      struct mlx4_en_cq *cq,
+		      int entries, int ring, enum cq_type mode)
+{
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err;
+
+	cq->size = entries;
+	if (mode == RX)
+		cq->buf_size = cq->size * sizeof(struct mlx4_cqe);
+	else
+		cq->buf_size = sizeof(struct mlx4_cqe);
+
+	cq->ring = ring;
+	cq->is_tx = mode;
+	spin_lock_init(&cq->lock);
+
+	err = mlx4_alloc_hwq_res(mdev->dev, &cq->wqres,
+				cq->buf_size, 2 * PAGE_SIZE);
+	if (err)
+		return err;
+
+	err = mlx4_en_map_buffer(&cq->wqres.buf);
+	if (err)
+		mlx4_free_hwq_res(mdev->dev, &cq->wqres, cq->buf_size);
+
+	return err;
+}
+
+int mlx4_en_activate_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq)
+{
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err;
+
+	cq->dev = mdev->pndev[priv->port];
+	cq->mcq.set_ci_db  = cq->wqres.db.db;
+	cq->mcq.arm_db     = cq->wqres.db.db + 1;
+	*cq->mcq.set_ci_db = 0;
+	*cq->mcq.arm_db    = 0;
+	cq->buf = (struct mlx4_cqe *) cq->wqres.buf.direct.buf;
+	memset(cq->buf, 0, cq->buf_size);
+
+	err = mlx4_cq_alloc(mdev->dev, cq->size, &cq->wqres.mtt, &mdev->priv_uar,
+			    cq->wqres.db.dma, &cq->mcq, cq->is_tx);
+	if (err)
+		return err;
+
+	cq->mcq.comp  = cq->is_tx ? mlx4_en_tx_irq : mlx4_en_rx_irq;
+	cq->mcq.event = mlx4_en_cq_event;
+
+	if (cq->is_tx) {
+		init_timer(&cq->timer);
+		cq->timer.function = mlx4_en_poll_tx_cq;
+		cq->timer.data = (unsigned long) cq;
+	} else {
+		netif_napi_add(cq->dev, &cq->napi, mlx4_en_poll_rx_cq, 64);
+		napi_enable(&cq->napi);
+	}
+
+	return 0;
+}
+
+void mlx4_en_destroy_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq)
+{
+	struct mlx4_en_dev *mdev = priv->mdev;
+
+	mlx4_en_unmap_buffer(&cq->wqres.buf);
+	mlx4_free_hwq_res(mdev->dev, &cq->wqres, cq->buf_size);
+	cq->buf_size = 0;
+	cq->buf = NULL;
+}
+
+void mlx4_en_deactivate_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq)
+{
+	struct mlx4_en_dev *mdev = priv->mdev;
+
+	if (cq->is_tx)
+		del_timer(&cq->timer);
+	else
+		napi_disable(&cq->napi);
+
+	mlx4_cq_free(mdev->dev, &cq->mcq);
+}
+
+/* Set rx cq moderation parameters */
+int mlx4_en_set_cq_moder(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq)
+{
+	return mlx4_cq_modify(priv->mdev->dev, &cq->mcq,
+			      cq->moder_cnt, cq->moder_time);
+}
+
+int mlx4_en_arm_cq(struct mlx4_en_priv *priv, struct mlx4_en_cq *cq)
+{
+	cq->armed = 1;
+	mlx4_cq_arm(&cq->mcq, MLX4_CQ_DB_REQ_NOT, priv->mdev->uar_map,
+		    &priv->mdev->uar_lock);
+
+	return 0;
+}
+
+

drivers/net/mlx4/en_main.c

+/*
+ * Copyright (c) 2007 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/cpumask.h>
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <linux/netdevice.h>
+#include <linux/cpumask.h>
+
+#include <linux/mlx4/driver.h>
+#include <linux/mlx4/device.h>
+#include <linux/mlx4/cmd.h>
+
+#include "mlx4_en.h"
+
+MODULE_AUTHOR("Liran Liss, Yevgeny Petrilin");
+MODULE_DESCRIPTION("Mellanox ConnectX HCA Ethernet driver");
+MODULE_LICENSE("Dual BSD/GPL");
+MODULE_VERSION(DRV_VERSION " ("DRV_RELDATE")");
+
+static const char mlx4_en_version[] =
+	DRV_NAME ": Mellanox ConnectX HCA Ethernet driver v"
+	DRV_VERSION " (" DRV_RELDATE ")\n";
+
+static void mlx4_en_event(struct mlx4_dev *dev, void *endev_ptr,
+			  enum mlx4_dev_event event, int port)
+{
+	struct mlx4_en_dev *mdev = (struct mlx4_en_dev *) endev_ptr;
+	struct mlx4_en_priv *priv;
+
+	if (!mdev->pndev[port])
+		return;
+
+	priv = netdev_priv(mdev->pndev[port]);
+	switch (event) {
+	case MLX4_DEV_EVENT_PORT_UP:
+	case MLX4_DEV_EVENT_PORT_DOWN:
+		/* To prevent races, we poll the link state in a separate
+		  task rather than changing it here */
+		priv->link_state = event;
+		queue_work(mdev->workqueue, &priv->linkstate_task);
+		break;
+
+	case MLX4_DEV_EVENT_CATASTROPHIC_ERROR:
+		mlx4_err(mdev, "Internal error detected, restarting device\n");
+		break;
+
+	default:
+		mlx4_warn(mdev, "Unhandled event: %d\n", event);
+	}
+}
+
+static void mlx4_en_remove(struct mlx4_dev *dev, void *endev_ptr)
+{
+	struct mlx4_en_dev *mdev = endev_ptr;
+	int i;
+
+	mutex_lock(&mdev->state_lock);
+	mdev->device_up = false;
+	mutex_unlock(&mdev->state_lock);
+
+	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_ETH)
+		if (mdev->pndev[i])
+			mlx4_en_destroy_netdev(mdev->pndev[i]);
+
+	flush_workqueue(mdev->workqueue);
+	destroy_workqueue(mdev->workqueue);
+	mlx4_mr_free(dev, &mdev->mr);
+	mlx4_uar_free(dev, &mdev->priv_uar);
+	mlx4_pd_free(dev, mdev->priv_pdn);
+	kfree(mdev);
+}
+
+static void *mlx4_en_add(struct mlx4_dev *dev)
+{
+	static int mlx4_en_version_printed;
+	struct mlx4_en_dev *mdev;
+	int i;
+	int err;
+
+	if (!mlx4_en_version_printed) {
+		printk(KERN_INFO "%s", mlx4_en_version);
+		mlx4_en_version_printed++;
+	}
+
+	mdev = kzalloc(sizeof *mdev, GFP_KERNEL);
+	if (!mdev) {
+		dev_err(&dev->pdev->dev, "Device struct alloc failed, "
+			"aborting.\n");
+		err = -ENOMEM;
+		goto err_free_res;
+	}
+
+	if (mlx4_pd_alloc(dev, &mdev->priv_pdn))
+		goto err_free_dev;
+
+	if (mlx4_uar_alloc(dev, &mdev->priv_uar))
+		goto err_pd;
+
+	mdev->uar_map = ioremap(mdev->priv_uar.pfn << PAGE_SHIFT, PAGE_SIZE);
+	if (!mdev->uar_map)
+		goto err_uar;
+	spin_lock_init(&mdev->uar_lock);
+
+	mdev->dev = dev;
+	mdev->dma_device = &(dev->pdev->dev);
+	mdev->pdev = dev->pdev;
+	mdev->device_up = false;
+
+	mdev->LSO_support = !!(dev->caps.flags & (1 << 15));
+	if (!mdev->LSO_support)
+		mlx4_warn(mdev, "LSO not supported, please upgrade to later "
+				"FW version to enable LSO\n");
+
+	if (mlx4_mr_alloc(mdev->dev, mdev->priv_pdn, 0, ~0ull,
+			 MLX4_PERM_LOCAL_WRITE |  MLX4_PERM_LOCAL_READ,
+			 0, 0, &mdev->mr)) {
+		mlx4_err(mdev, "Failed allocating memory region\n");
+		goto err_uar;
+	}
+	if (mlx4_mr_enable(mdev->dev, &mdev->mr)) {
+		mlx4_err(mdev, "Failed enabling memory region\n");
+		goto err_mr;
+	}
+
+	/* Build device profile according to supplied module parameters */
+	err = mlx4_en_get_profile(mdev);
+	if (err) {
+		mlx4_err(mdev, "Bad module parameters, aborting.\n");
+		goto err_mr;
+	}
+
+	/* Configure wich ports to start according to module parameters */
+	mdev->port_cnt = 0;
+	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_ETH)
+		mdev->port_cnt++;
+
+	/* If we did not receive an explicit number of Rx rings, default to
+	 * the number of completion vectors populated by the mlx4_core */
+	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_ETH) {
+		mlx4_info(mdev, "Using %d tx rings for port:%d\n",
+			  mdev->profile.prof[i].tx_ring_num, i);
+		if (!mdev->profile.prof[i].rx_ring_num) {
+			mdev->profile.prof[i].rx_ring_num = 1;
+			mlx4_info(mdev, "Defaulting to %d rx rings for port:%d\n",
+				  1, i);
+		} else
+			mlx4_info(mdev, "Using %d rx rings for port:%d\n",
+				  mdev->profile.prof[i].rx_ring_num, i);
+	}
+
+	/* Create our own workqueue for reset/multicast tasks
+	 * Note: we cannot use the shared workqueue because of deadlocks caused
+	 *       by the rtnl lock */
+	mdev->workqueue = create_singlethread_workqueue("mlx4_en");
+	if (!mdev->workqueue) {
+		err = -ENOMEM;
+		goto err_close_nic;
+	}
+
+	/* At this stage all non-port specific tasks are complete:
+	 * mark the card state as up */
+	mutex_init(&mdev->state_lock);
+	mdev->device_up = true;
+
+	/* Setup ports */
+
+	/* Create a netdev for each port */
+	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_ETH) {
+		mlx4_info(mdev, "Activating port:%d\n", i);
+		if (mlx4_en_init_netdev(mdev, i, &mdev->profile.prof[i])) {
+			mdev->pndev[i] = NULL;
+			goto err_free_netdev;
+		}
+	}
+	return mdev;
+
+
+err_free_netdev:
+	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_ETH) {
+		if (mdev->pndev[i])
+			mlx4_en_destroy_netdev(mdev->pndev[i]);
+	}
+
+	mutex_lock(&mdev->state_lock);
+	mdev->device_up = false;
+	mutex_unlock(&mdev->state_lock);
+	flush_workqueue(mdev->workqueue);
+
+	/* Stop event queue before we drop down to release shared SW state */
+
+err_close_nic:
+	destroy_workqueue(mdev->workqueue);
+err_mr:
+	mlx4_mr_free(dev, &mdev->mr);
+err_uar:
+	mlx4_uar_free(dev, &mdev->priv_uar);
+err_pd:
+	mlx4_pd_free(dev, mdev->priv_pdn);
+err_free_dev:
+	kfree(mdev);
+err_free_res:
+	return NULL;
+}
+
+static struct mlx4_interface mlx4_en_interface = {
+	.add	= mlx4_en_add,
+	.remove	= mlx4_en_remove,
+	.event	= mlx4_en_event,
+};
+
+static int __init mlx4_en_init(void)
+{
+	return mlx4_register_interface(&mlx4_en_interface);
+}
+
+static void __exit mlx4_en_cleanup(void)
+{
+	mlx4_unregister_interface(&mlx4_en_interface);
+}
+
+module_init(mlx4_en_init);
+module_exit(mlx4_en_cleanup);
+

drivers/net/mlx4/en_params.c

+/*
+ * Copyright (c) 2007 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/ethtool.h>
+#include <linux/netdevice.h>
+
+#include "mlx4_en.h"
+#include "en_port.h"
+
+#define MLX4_EN_PARM_INT(X, def_val, desc) \
+	static unsigned int X = def_val;\
+	module_param(X , uint, 0444); \
+	MODULE_PARM_DESC(X, desc);
+
+
+/*
+ * Device scope module parameters
+ */
+
+
+/* Use a XOR rathern than Toeplitz hash function for RSS */
+MLX4_EN_PARM_INT(rss_xor, 0, "Use XOR hash function for RSS");
+
+/* RSS hash type mask - default to <saddr, daddr, sport, dport> */
+MLX4_EN_PARM_INT(rss_mask, 0xf, "RSS hash type bitmask");
+
+/* Number of LRO sessions per Rx ring (rounded up to a power of two) */
+MLX4_EN_PARM_INT(num_lro, MLX4_EN_MAX_LRO_DESCRIPTORS,
+		 "Number of LRO sessions per ring or disabled (0)");
+
+/* Priority pausing */
+MLX4_EN_PARM_INT(pptx, MLX4_EN_DEF_TX_PAUSE,
+		 "Pause policy on TX: 0 never generate pause frames "
+		 "1 generate pause frames according to RX buffer threshold");
+MLX4_EN_PARM_INT(pprx, MLX4_EN_DEF_RX_PAUSE,
+		 "Pause policy on RX: 0 ignore received pause frames "
+		 "1 respect received pause frames");
+MLX4_EN_PARM_INT(pfctx, 0, "Priority based Flow Control policy on TX[7:0]."
+			   " Per priority bit mask");
+MLX4_EN_PARM_INT(pfcrx, 0, "Priority based Flow Control policy on RX[7:0]."
+			   " Per priority bit mask");
+
+/* Interrupt moderation tunning */
+MLX4_EN_PARM_INT(rx_moder_cnt, MLX4_EN_AUTO_CONF,
+	       "Max coalesced descriptors for Rx interrupt moderation");
+MLX4_EN_PARM_INT(rx_moder_time, MLX4_EN_AUTO_CONF,
+	       "Timeout following last packet for Rx interrupt moderation");
+MLX4_EN_PARM_INT(auto_moder, 1, "Enable dynamic interrupt moderation");
+
+MLX4_EN_PARM_INT(rx_ring_num1, 0, "Number or Rx rings for port 1 (0 = #cores)");
+MLX4_EN_PARM_INT(rx_ring_num2, 0, "Number or Rx rings for port 2 (0 = #cores)");
+
+MLX4_EN_PARM_INT(tx_ring_size1, MLX4_EN_AUTO_CONF, "Tx ring size for port 1");
+MLX4_EN_PARM_INT(tx_ring_size2, MLX4_EN_AUTO_CONF, "Tx ring size for port 2");
+MLX4_EN_PARM_INT(rx_ring_size1, MLX4_EN_AUTO_CONF, "Rx ring size for port 1");
+MLX4_EN_PARM_INT(rx_ring_size2, MLX4_EN_AUTO_CONF, "Rx ring size for port 2");
+
+
+int mlx4_en_get_profile(struct mlx4_en_dev *mdev)
+{
+	struct mlx4_en_profile *params = &mdev->profile;
+
+	params->rx_moder_cnt = min_t(int, rx_moder_cnt, MLX4_EN_AUTO_CONF);
+	params->rx_moder_time = min_t(int, rx_moder_time, MLX4_EN_AUTO_CONF);
+	params->auto_moder = auto_moder;
+	params->rss_xor = (rss_xor != 0);
+	params->rss_mask = rss_mask & 0x1f;
+	params->num_lro = min_t(int, num_lro , MLX4_EN_MAX_LRO_DESCRIPTORS);
+	params->rx_pause = pprx;
+	params->rx_ppp = pfcrx;
+	params->tx_pause = pptx;
+	params->tx_ppp = pfctx;
+	if (params->rx_ppp || params->tx_ppp) {
+		params->prof[1].tx_ring_num = MLX4_EN_TX_RING_NUM;
+		params->prof[2].tx_ring_num = MLX4_EN_TX_RING_NUM;
+	} else {
+		params->prof[1].tx_ring_num = 1;
+		params->prof[2].tx_ring_num = 1;
+	}
+	params->prof[1].rx_ring_num = min_t(int, rx_ring_num1, MAX_RX_RINGS);
+	params->prof[2].rx_ring_num = min_t(int, rx_ring_num2, MAX_RX_RINGS);
+
+	if (tx_ring_size1 == MLX4_EN_AUTO_CONF)
+		tx_ring_size1 = MLX4_EN_DEF_TX_RING_SIZE;
+	params->prof[1].tx_ring_size =
+		(tx_ring_size1 < MLX4_EN_MIN_TX_SIZE) ?
+		 MLX4_EN_MIN_TX_SIZE : roundup_pow_of_two(tx_ring_size1);
+
+	if (tx_ring_size2 == MLX4_EN_AUTO_CONF)
+		tx_ring_size2 = MLX4_EN_DEF_TX_RING_SIZE;
+	params->prof[2].tx_ring_size =
+		(tx_ring_size2 < MLX4_EN_MIN_TX_SIZE) ?
+		 MLX4_EN_MIN_TX_SIZE : roundup_pow_of_two(tx_ring_size2);
+
+	if (rx_ring_size1 == MLX4_EN_AUTO_CONF)
+		rx_ring_size1 = MLX4_EN_DEF_RX_RING_SIZE;
+	params->prof[1].rx_ring_size =
+		(rx_ring_size1 < MLX4_EN_MIN_RX_SIZE) ?
+		 MLX4_EN_MIN_RX_SIZE : roundup_pow_of_two(rx_ring_size1);
+
+	if (rx_ring_size2 == MLX4_EN_AUTO_CONF)
+		rx_ring_size2 = MLX4_EN_DEF_RX_RING_SIZE;
+	params->prof[2].rx_ring_size =
+		(rx_ring_size2 < MLX4_EN_MIN_RX_SIZE) ?
+		 MLX4_EN_MIN_RX_SIZE : roundup_pow_of_two(rx_ring_size2);
+	return 0;
+}
+
+
+/*
+ * Ethtool support
+ */
+
+static void mlx4_en_update_lro_stats(struct mlx4_en_priv *priv)
+{
+	int i;
+
+	priv->port_stats.lro_aggregated = 0;
+	priv->port_stats.lro_flushed = 0;
+	priv->port_stats.lro_no_desc = 0;
+
+	for (i = 0; i < priv->rx_ring_num; i++) {
+		priv->port_stats.lro_aggregated += priv->rx_ring[i].lro.stats.aggregated;
+		priv->port_stats.lro_flushed += priv->rx_ring[i].lro.stats.flushed;
+		priv->port_stats.lro_no_desc += priv->rx_ring[i].lro.stats.no_desc;
+	}
+}
+
+static void
+mlx4_en_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *drvinfo)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+
+	sprintf(drvinfo->driver, DRV_NAME " (%s)", mdev->dev->board_id);
+	strncpy(drvinfo->version, DRV_VERSION " (" DRV_RELDATE ")", 32);
+	sprintf(drvinfo->fw_version, "%d.%d.%d",
+		(u16) (mdev->dev->caps.fw_ver >> 32),
+		(u16) ((mdev->dev->caps.fw_ver >> 16) & 0xffff),
+		(u16) (mdev->dev->caps.fw_ver & 0xffff));
+	strncpy(drvinfo->bus_info, pci_name(mdev->dev->pdev), 32);
+	drvinfo->n_stats = 0;
+	drvinfo->regdump_len = 0;
+	drvinfo->eedump_len = 0;
+}
+
+static u32 mlx4_en_get_tso(struct net_device *dev)
+{
+	return (dev->features & NETIF_F_TSO) != 0;
+}
+
+static int mlx4_en_set_tso(struct net_device *dev, u32 data)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+
+	if (data) {
+		if (!priv->mdev->LSO_support)
+			return -EPERM;
+		dev->features |= (NETIF_F_TSO | NETIF_F_TSO6);
+	} else
+		dev->features &= ~(NETIF_F_TSO | NETIF_F_TSO6);
+	return 0;
+}
+
+static u32 mlx4_en_get_rx_csum(struct net_device *dev)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	return priv->rx_csum;
+}
+
+static int mlx4_en_set_rx_csum(struct net_device *dev, u32 data)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	priv->rx_csum = (data != 0);
+	return 0;
+}
+
+static const char main_strings[][ETH_GSTRING_LEN] = {
+	"rx_packets", "tx_packets", "rx_bytes", "tx_bytes", "rx_errors",
+	"tx_errors", "rx_dropped", "tx_dropped", "multicast", "collisions",
+	"rx_length_errors", "rx_over_errors", "rx_crc_errors",
+	"rx_frame_errors", "rx_fifo_errors", "rx_missed_errors",
+	"tx_aborted_errors", "tx_carrier_errors", "tx_fifo_errors",
+	"tx_heartbeat_errors", "tx_window_errors",
+
+	/* port statistics */
+	"lro_aggregated", "lro_flushed", "lro_no_desc", "tso_packets",
+	"queue_stopped", "wake_queue", "tx_timeout", "rx_alloc_failed",
+	"rx_csum_good", "rx_csum_none", "tx_chksum_offload",
+
+	/* packet statistics */
+	"broadcast", "rx_prio_0", "rx_prio_1", "rx_prio_2", "rx_prio_3",
+	"rx_prio_4", "rx_prio_5", "rx_prio_6", "rx_prio_7", "tx_prio_0",
+	"tx_prio_1", "tx_prio_2", "tx_prio_3", "tx_prio_4", "tx_prio_5",
+	"tx_prio_6", "tx_prio_7",
+};
+#define NUM_MAIN_STATS	21
+#define NUM_ALL_STATS	(NUM_MAIN_STATS + NUM_PORT_STATS + NUM_PKT_STATS + NUM_PERF_STATS)
+
+static u32 mlx4_en_get_msglevel(struct net_device *dev)
+{
+	return ((struct mlx4_en_priv *) netdev_priv(dev))->msg_enable;
+}
+
+static void mlx4_en_set_msglevel(struct net_device *dev, u32 val)
+{
+	((struct mlx4_en_priv *) netdev_priv(dev))->msg_enable = val;
+}
+
+static void mlx4_en_get_wol(struct net_device *netdev,
+			    struct ethtool_wolinfo *wol)
+{
+	wol->supported = 0;
+	wol->wolopts = 0;
+
+	return;
+}
+
+static int mlx4_en_get_sset_count(struct net_device *dev, int sset)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+
+	if (sset != ETH_SS_STATS)
+		return -EOPNOTSUPP;
+
+	return NUM_ALL_STATS + (priv->tx_ring_num + priv->rx_ring_num) * 2;
+}
+
+static void mlx4_en_get_ethtool_stats(struct net_device *dev,
+		struct ethtool_stats *stats, uint64_t *data)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	int index = 0;
+	int i;
+
+	spin_lock_bh(&priv->stats_lock);
+
+	mlx4_en_update_lro_stats(priv);
+
+	for (i = 0; i < NUM_MAIN_STATS; i++)
+		data[index++] = ((unsigned long *) &priv->stats)[i];
+	for (i = 0; i < NUM_PORT_STATS; i++)
+		data[index++] = ((unsigned long *) &priv->port_stats)[i];
+	for (i = 0; i < priv->tx_ring_num; i++) {
+		data[index++] = priv->tx_ring[i].packets;
+		data[index++] = priv->tx_ring[i].bytes;
+	}
+	for (i = 0; i < priv->rx_ring_num; i++) {
+		data[index++] = priv->rx_ring[i].packets;
+		data[index++] = priv->rx_ring[i].bytes;
+	}
+	for (i = 0; i < NUM_PKT_STATS; i++)
+		data[index++] = ((unsigned long *) &priv->pkstats)[i];
+	spin_unlock_bh(&priv->stats_lock);
+
+}
+
+static void mlx4_en_get_strings(struct net_device *dev,
+				uint32_t stringset, uint8_t *data)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	int index = 0;
+	int i;
+
+	if (stringset != ETH_SS_STATS)
+		return;
+
+	/* Add main counters */
+	for (i = 0; i < NUM_MAIN_STATS; i++)
+		strcpy(data + (index++) * ETH_GSTRING_LEN, main_strings[i]);
+	for (i = 0; i < NUM_PORT_STATS; i++)
+		strcpy(data + (index++) * ETH_GSTRING_LEN,
+			main_strings[i + NUM_MAIN_STATS]);
+	for (i = 0; i < priv->tx_ring_num; i++) {
+		sprintf(data + (index++) * ETH_GSTRING_LEN,
+			"tx%d_packets", i);
+		sprintf(data + (index++) * ETH_GSTRING_LEN,
+			"tx%d_bytes", i);
+	}
+	for (i = 0; i < priv->rx_ring_num; i++) {
+		sprintf(data + (index++) * ETH_GSTRING_LEN,
+			"rx%d_packets", i);
+		sprintf(data + (index++) * ETH_GSTRING_LEN,
+			"rx%d_bytes", i);
+	}
+	for (i = 0; i < NUM_PKT_STATS; i++)
+		strcpy(data + (index++) * ETH_GSTRING_LEN,
+			main_strings[i + NUM_MAIN_STATS + NUM_PORT_STATS]);
+}
+
+static int mlx4_en_get_settings(struct net_device *dev, struct ethtool_cmd *cmd)
+{
+	cmd->autoneg = AUTONEG_DISABLE;
+	cmd->supported = SUPPORTED_10000baseT_Full;
+	cmd->advertising = SUPPORTED_10000baseT_Full;
+	if (netif_carrier_ok(dev)) {
+		cmd->speed = SPEED_10000;
+		cmd->duplex = DUPLEX_FULL;
+	} else {
+		cmd->speed = -1;
+		cmd->duplex = -1;
+	}
+	return 0;
+}
+
+static int mlx4_en_set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
+{
+	if ((cmd->autoneg == AUTONEG_ENABLE) ||
+	    (cmd->speed != SPEED_10000) || (cmd->duplex != DUPLEX_FULL))
+		return -EINVAL;
+
+	/* Nothing to change */
+	return 0;
+}
+
+static int mlx4_en_get_coalesce(struct net_device *dev,
+			      struct ethtool_coalesce *coal)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+
+	coal->tx_coalesce_usecs = 0;
+	coal->tx_max_coalesced_frames = 0;
+	coal->rx_coalesce_usecs = priv->rx_usecs;
+	coal->rx_max_coalesced_frames = priv->rx_frames;
+
+	coal->pkt_rate_low = priv->pkt_rate_low;
+	coal->rx_coalesce_usecs_low = priv->rx_usecs_low;
+	coal->pkt_rate_high = priv->pkt_rate_high;
+	coal->rx_coalesce_usecs_high = priv->rx_usecs_high;
+	coal->rate_sample_interval = priv->sample_interval;
+	coal->use_adaptive_rx_coalesce = priv->adaptive_rx_coal;
+	return 0;
+}
+
+static int mlx4_en_set_coalesce(struct net_device *dev,
+			      struct ethtool_coalesce *coal)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	int err, i;
+
+	priv->rx_frames = (coal->rx_max_coalesced_frames ==
+			   MLX4_EN_AUTO_CONF) ?
+				MLX4_EN_RX_COAL_TARGET /
+				priv->dev->mtu + 1 :
+				coal->rx_max_coalesced_frames;
+	priv->rx_usecs = (coal->rx_coalesce_usecs ==
+			  MLX4_EN_AUTO_CONF) ?
+				MLX4_EN_RX_COAL_TIME :
+				coal->rx_coalesce_usecs;
+
+	/* Set adaptive coalescing params */
+	priv->pkt_rate_low = coal->pkt_rate_low;
+	priv->rx_usecs_low = coal->rx_coalesce_usecs_low;
+	priv->pkt_rate_high = coal->pkt_rate_high;
+	priv->rx_usecs_high = coal->rx_coalesce_usecs_high;
+	priv->sample_interval = coal->rate_sample_interval;
+	priv->adaptive_rx_coal = coal->use_adaptive_rx_coalesce;
+	priv->last_moder_time = MLX4_EN_AUTO_CONF;
+	if (priv->adaptive_rx_coal)
+		return 0;
+
+	for (i = 0; i < priv->rx_ring_num; i++) {
+		priv->rx_cq[i].moder_cnt = priv->rx_frames;
+		priv->rx_cq[i].moder_time = priv->rx_usecs;
+		err = mlx4_en_set_cq_moder(priv, &priv->rx_cq[i]);
+		if (err)
+			return err;
+	}
+	return 0;
+}
+
+static int mlx4_en_set_pauseparam(struct net_device *dev,
+				struct ethtool_pauseparam *pause)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+	int err;
+
+	mdev->profile.tx_pause = pause->tx_pause != 0;
+	mdev->profile.rx_pause = pause->rx_pause != 0;
+	err = mlx4_SET_PORT_general(mdev->dev, priv->port,
+				    priv->rx_skb_size + ETH_FCS_LEN,
+				    mdev->profile.tx_pause,
+				    mdev->profile.tx_ppp,
+				    mdev->profile.rx_pause,
+				    mdev->profile.rx_ppp);
+	if (err)
+		mlx4_err(mdev, "Failed setting pause params to\n");
+
+	return err;
+}
+
+static void mlx4_en_get_pauseparam(struct net_device *dev,
+				 struct ethtool_pauseparam *pause)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+
+	pause->tx_pause = mdev->profile.tx_pause;
+	pause->rx_pause = mdev->profile.rx_pause;
+}
+
+static void mlx4_en_get_ringparam(struct net_device *dev,
+				  struct ethtool_ringparam *param)
+{
+	struct mlx4_en_priv *priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = priv->mdev;
+
+	memset(param, 0, sizeof(*param));
+	param->rx_max_pending = mdev->dev->caps.max_rq_sg;
+	param->tx_max_pending = mdev->dev->caps.max_sq_sg;
+	param->rx_pending = mdev->profile.prof[priv->port].rx_ring_size;
+	param->tx_pending = mdev->profile.prof[priv->port].tx_ring_size;
+}
+
+const struct ethtool_ops mlx4_en_ethtool_ops = {
+	.get_drvinfo = mlx4_en_get_drvinfo,
+	.get_settings = mlx4_en_get_settings,
+	.set_settings = mlx4_en_set_settings,
+#ifdef NETIF_F_TSO
+	.get_tso = mlx4_en_get_tso,
+	.set_tso = mlx4_en_set_tso,
+#endif
+	.get_sg = ethtool_op_get_sg,
+	.set_sg = ethtool_op_set_sg,
+	.get_link = ethtool_op_get_link,
+	.get_rx_csum = mlx4_en_get_rx_csum,
+	.set_rx_csum = mlx4_en_set_rx_csum,
+	.get_tx_csum = ethtool_op_get_tx_csum,
+	.set_tx_csum = ethtool_op_set_tx_ipv6_csum,
+	.get_strings = mlx4_en_get_strings,
+	.get_sset_count = mlx4_en_get_sset_count,
+	.get_ethtool_stats = mlx4_en_get_ethtool_stats,
+	.get_wol = mlx4_en_get_wol,
+	.get_msglevel = mlx4_en_get_msglevel,
+	.set_msglevel = mlx4_en_set_msglevel,
+	.get_coalesce = mlx4_en_get_coalesce,
+	.set_coalesce = mlx4_en_set_coalesce,
+	.get_pauseparam = mlx4_en_get_pauseparam,
+	.set_pauseparam = mlx4_en_set_pauseparam,
+	.get_ringparam = mlx4_en_get_ringparam,
+	.get_flags = ethtool_op_get_flags,
+	.set_flags = ethtool_op_set_flags,
+};
+
+
+
+
+

drivers/net/mlx4/en_port.c

+/*
+ * Copyright (c) 2007 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+
+#include <linux/if_vlan.h>
+
+#include <linux/mlx4/device.h>
+#include <linux/mlx4/cmd.h>
+
+#include "en_port.h"
+#include "mlx4_en.h"
+
+
+int mlx4_SET_MCAST_FLTR(struct mlx4_dev *dev, u8 port,
+			u64 mac, u64 clear, u8 mode)
+{
+	return mlx4_cmd(dev, (mac | (clear << 63)), port, mode,
+			MLX4_CMD_SET_MCAST_FLTR, MLX4_CMD_TIME_CLASS_B);
+}
+
+int mlx4_SET_VLAN_FLTR(struct mlx4_dev *dev, u8 port, struct vlan_group *grp)
+{
+	struct mlx4_cmd_mailbox *mailbox;
+	struct mlx4_set_vlan_fltr_mbox *filter;
+	int i;
+	int j;
+	int index = 0;
+	u32 entry;
+	int err = 0;
+
+	mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(mailbox))
+		return PTR_ERR(mailbox);
+
+	filter = mailbox->buf;
+	if (grp) {
+		memset(filter, 0, sizeof *filter);
+		for (i = VLAN_FLTR_SIZE - 1; i >= 0; i--) {
+			entry = 0;
+			for (j = 0; j < 32; j++)
+				if (vlan_group_get_device(grp, index++))
+					entry |= 1 << j;
+			filter->entry[i] = cpu_to_be32(entry);
+		}
+	} else {
+		/* When no vlans are configured we block all vlans */
+		memset(filter, 0, sizeof(*filter));
+	}
+	err = mlx4_cmd(dev, mailbox->dma, port, 0, MLX4_CMD_SET_VLAN_FLTR,
+		       MLX4_CMD_TIME_CLASS_B);
+	mlx4_free_cmd_mailbox(dev, mailbox);
+	return err;
+}
+
+
+int mlx4_SET_PORT_general(struct mlx4_dev *dev, u8 port, int mtu,
+			  u8 pptx, u8 pfctx, u8 pprx, u8 pfcrx)
+{
+	struct mlx4_cmd_mailbox *mailbox;
+	struct mlx4_set_port_general_context *context;
+	int err;
+	u32 in_mod;
+
+	mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(mailbox))
+		return PTR_ERR(mailbox);
+	context = mailbox->buf;
+	memset(context, 0, sizeof *context);
+
+	context->flags = SET_PORT_GEN_ALL_VALID;
+	context->mtu = cpu_to_be16(mtu);
+	context->pptx = (pptx * (!pfctx)) << 7;
+	context->pfctx = pfctx;
+	context->pprx = (pprx * (!pfcrx)) << 7;
+	context->pfcrx = pfcrx;
+
+	in_mod = MLX4_SET_PORT_GENERAL << 8 | port;
+	err = mlx4_cmd(dev, mailbox->dma, in_mod, 1, MLX4_CMD_SET_PORT,
+		       MLX4_CMD_TIME_CLASS_B);
+
+	mlx4_free_cmd_mailbox(dev, mailbox);
+	return err;
+}
+
+int mlx4_SET_PORT_qpn_calc(struct mlx4_dev *dev, u8 port, u32 base_qpn,
+			   u8 promisc)
+{
+	struct mlx4_cmd_mailbox *mailbox;
+	struct mlx4_set_port_rqp_calc_context *context;
+	int err;
+	u32 in_mod;
+
+	mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(mailbox))
+		return PTR_ERR(mailbox);
+	context = mailbox->buf;
+	memset(context, 0, sizeof *context);
+
+	context->base_qpn = cpu_to_be32(base_qpn);
+	context->promisc = cpu_to_be32(promisc << SET_PORT_PROMISC_SHIFT | base_qpn);
+	context->mcast = cpu_to_be32(1 << SET_PORT_PROMISC_SHIFT | base_qpn);
+	context->intra_no_vlan = 0;
+	context->no_vlan = MLX4_NO_VLAN_IDX;
+	context->intra_vlan_miss = 0;
+	context->vlan_miss = MLX4_VLAN_MISS_IDX;
+
+	in_mod = MLX4_SET_PORT_RQP_CALC << 8 | port;
+	err = mlx4_cmd(dev, mailbox->dma, in_mod, 1, MLX4_CMD_SET_PORT,
+		       MLX4_CMD_TIME_CLASS_B);
+
+	mlx4_free_cmd_mailbox(dev, mailbox);
+	return err;
+}
+
+
+int mlx4_en_DUMP_ETH_STATS(struct mlx4_en_dev *mdev, u8 port, u8 reset)
+{
+	struct mlx4_en_stat_out_mbox *mlx4_en_stats;
+	struct mlx4_en_priv *priv = netdev_priv(mdev->pndev[port]);
+	struct net_device_stats *stats = &priv->stats;
+	struct mlx4_cmd_mailbox *mailbox;
+	u64 in_mod = reset << 8 | port;
+	int err;
+
+	mailbox = mlx4_alloc_cmd_mailbox(mdev->dev);
+	if (IS_ERR(mailbox))
+		return PTR_ERR(mailbox);
+	memset(mailbox->buf, 0, sizeof(*mlx4_en_stats));
+	err = mlx4_cmd_box(mdev->dev, 0, mailbox->dma, in_mod, 0,
+			   MLX4_CMD_DUMP_ETH_STATS, MLX4_CMD_TIME_CLASS_B);
+	if (err)
+		goto out;
+
+	mlx4_en_stats = mailbox->buf;
+
+	spin_lock_bh(&priv->stats_lock);
+
+	stats->rx_packets = be32_to_cpu(mlx4_en_stats->RTOTFRMS) -
+			    be32_to_cpu(mlx4_en_stats->RDROP);
+	stats->tx_packets = be64_to_cpu(mlx4_en_stats->TTOT_prio_0) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_prio_1) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_prio_2) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_prio_3) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_prio_4) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_prio_5) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_prio_6) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_prio_7) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_novlan) +
+			    be64_to_cpu(mlx4_en_stats->TTOT_loopbk);
+	stats->rx_bytes = be64_to_cpu(mlx4_en_stats->ROCT_prio_0) +
+			  be64_to_cpu(mlx4_en_stats->ROCT_prio_1) +
+			  be64_to_cpu(mlx4_en_stats->ROCT_prio_2) +
+			  be64_to_cpu(mlx4_en_stats->ROCT_prio_3) +
+			  be64_to_cpu(mlx4_en_stats->ROCT_prio_4) +
+			  be64_to_cpu(mlx4_en_stats->ROCT_prio_5) +
+			  be64_to_cpu(mlx4_en_stats->ROCT_prio_6) +
+			  be64_to_cpu(mlx4_en_stats->ROCT_prio_7) +
+			  be64_to_cpu(mlx4_en_stats->ROCT_novlan);
+
+	stats->tx_bytes = be64_to_cpu(mlx4_en_stats->TTTLOCT_prio_0) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_prio_1) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_prio_2) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_prio_3) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_prio_4) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_prio_5) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_prio_6) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_prio_7) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_novlan) +
+			  be64_to_cpu(mlx4_en_stats->TTTLOCT_loopbk);
+
+	stats->rx_errors = be64_to_cpu(mlx4_en_stats->PCS) +
+			   be32_to_cpu(mlx4_en_stats->RdropLength) +
+			   be32_to_cpu(mlx4_en_stats->RJBBR) +
+			   be32_to_cpu(mlx4_en_stats->RCRC) +
+			   be32_to_cpu(mlx4_en_stats->RRUNT);
+	stats->tx_errors = be32_to_cpu(mlx4_en_stats->TDROP);
+	stats->multicast = be64_to_cpu(mlx4_en_stats->MCAST_prio_0) +
+			   be64_to_cpu(mlx4_en_stats->MCAST_prio_1) +
+			   be64_to_cpu(mlx4_en_stats->MCAST_prio_2) +
+			   be64_to_cpu(mlx4_en_stats->MCAST_prio_3) +
+			   be64_to_cpu(mlx4_en_stats->MCAST_prio_4) +
+			   be64_to_cpu(mlx4_en_stats->MCAST_prio_5) +
+			   be64_to_cpu(mlx4_en_stats->MCAST_prio_6) +
+			   be64_to_cpu(mlx4_en_stats->MCAST_prio_7) +
+			   be64_to_cpu(mlx4_en_stats->MCAST_novlan);
+	stats->collisions = 0;
+	stats->rx_length_errors = be32_to_cpu(mlx4_en_stats->RdropLength);
+	stats->rx_over_errors = be32_to_cpu(mlx4_en_stats->RdropOvflw);
+	stats->rx_crc_errors = be32_to_cpu(mlx4_en_stats->RCRC);
+	stats->rx_frame_errors = 0;
+	stats->rx_fifo_errors = be32_to_cpu(mlx4_en_stats->RdropOvflw);
+	stats->rx_missed_errors = be32_to_cpu(mlx4_en_stats->RdropOvflw);
+	stats->tx_aborted_errors = 0;
+	stats->tx_carrier_errors = 0;
+	stats->tx_fifo_errors = 0;
+	stats->tx_heartbeat_errors = 0;
+	stats->tx_window_errors = 0;
+
+	priv->pkstats.broadcast =
+				be64_to_cpu(mlx4_en_stats->RBCAST_prio_0) +
+				be64_to_cpu(mlx4_en_stats->RBCAST_prio_1) +
+				be64_to_cpu(mlx4_en_stats->RBCAST_prio_2) +
+				be64_to_cpu(mlx4_en_stats->RBCAST_prio_3) +
+				be64_to_cpu(mlx4_en_stats->RBCAST_prio_4) +
+				be64_to_cpu(mlx4_en_stats->RBCAST_prio_5) +
+				be64_to_cpu(mlx4_en_stats->RBCAST_prio_6) +
+				be64_to_cpu(mlx4_en_stats->RBCAST_prio_7) +
+				be64_to_cpu(mlx4_en_stats->RBCAST_novlan);
+	priv->pkstats.rx_prio[0] = be64_to_cpu(mlx4_en_stats->RTOT_prio_0);
+	priv->pkstats.rx_prio[1] = be64_to_cpu(mlx4_en_stats->RTOT_prio_1);
+	priv->pkstats.rx_prio[2] = be64_to_cpu(mlx4_en_stats->RTOT_prio_2);
+	priv->pkstats.rx_prio[3] = be64_to_cpu(mlx4_en_stats->RTOT_prio_3);
+	priv->pkstats.rx_prio[4] = be64_to_cpu(mlx4_en_stats->RTOT_prio_4);
+	priv->pkstats.rx_prio[5] = be64_to_cpu(mlx4_en_stats->RTOT_prio_5);
+	priv->pkstats.rx_prio[6] = be64_to_cpu(mlx4_en_stats->RTOT_prio_6);
+	priv->pkstats.rx_prio[7] = be64_to_cpu(mlx4_en_stats->RTOT_prio_7);
+	priv->pkstats.tx_prio[0] = be64_to_cpu(mlx4_en_stats->TTOT_prio_0);
+	priv->pkstats.tx_prio[1] = be64_to_cpu(mlx4_en_stats->TTOT_prio_1);
+	priv->pkstats.tx_prio[2] = be64_to_cpu(mlx4_en_stats->TTOT_prio_2);
+	priv->pkstats.tx_prio[3] = be64_to_cpu(mlx4_en_stats->TTOT_prio_3);
+	priv->pkstats.tx_prio[4] = be64_to_cpu(mlx4_en_stats->TTOT_prio_4);
+	priv->pkstats.tx_prio[5] = be64_to_cpu(mlx4_en_stats->TTOT_prio_5);
+	priv->pkstats.tx_prio[6] = be64_to_cpu(mlx4_en_stats->TTOT_prio_6);
+	priv->pkstats.tx_prio[7] = be64_to_cpu(mlx4_en_stats->TTOT_prio_7);
+	spin_unlock_bh(&priv->stats_lock);
+
+out:
+	mlx4_free_cmd_mailbox(mdev->dev, mailbox);
+	return err;
+}
+

drivers/net/mlx4/en_port.h

+/*
+ * Copyright (c) 2007 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#ifndef _MLX4_EN_PORT_H_
+#define _MLX4_EN_PORT_H_
+
+
+#define SET_PORT_GEN_ALL_VALID	0x7
+#define SET_PORT_PROMISC_SHIFT	31
+
+enum {
+	MLX4_CMD_SET_VLAN_FLTR  = 0x47,
+	MLX4_CMD_SET_MCAST_FLTR = 0x48,
+	MLX4_CMD_DUMP_ETH_STATS = 0x49,
+};
+
+struct mlx4_set_port_general_context {
+	u8 reserved[3];
+	u8 flags;
+	u16 reserved2;
+	__be16 mtu;
+	u8 pptx;
+	u8 pfctx;
+	u16 reserved3;
+	u8 pprx;
+	u8 pfcrx;
+	u16 reserved4;
+};
+
+struct mlx4_set_port_rqp_calc_context {
+	__be32 base_qpn;
+	__be32 flags;
+	u8 reserved[3];
+	u8 mac_miss;
+	u8 intra_no_vlan;
+	u8 no_vlan;
+	u8 intra_vlan_miss;
+	u8 vlan_miss;
+	u8 reserved2[3];
+	u8 no_vlan_prio;
+	__be32 promisc;
+	__be32 mcast;
+};
+
+#define VLAN_FLTR_SIZE	128
+struct mlx4_set_vlan_fltr_mbox {
+	__be32 entry[VLAN_FLTR_SIZE];
+};
+
+
+enum {
+	MLX4_MCAST_CONFIG       = 0,
+	MLX4_MCAST_DISABLE      = 1,
+	MLX4_MCAST_ENABLE       = 2,
+};
+
+
+struct mlx4_en_stat_out_mbox {
+	/* Received frames with a length of 64 octets */
+	__be64 R64_prio_0;
+	__be64 R64_prio_1;
+	__be64 R64_prio_2;
+	__be64 R64_prio_3;
+	__be64 R64_prio_4;
+	__be64 R64_prio_5;
+	__be64 R64_prio_6;
+	__be64 R64_prio_7;
+	__be64 R64_novlan;
+	/* Received frames with a length of 127 octets */
+	__be64 R127_prio_0;
+	__be64 R127_prio_1;
+	__be64 R127_prio_2;
+	__be64 R127_prio_3;
+	__be64 R127_prio_4;
+	__be64 R127_prio_5;
+	__be64 R127_prio_6;
+	__be64 R127_prio_7;
+	__be64 R127_novlan;
+	/* Received frames with a length of 255 octets */
+	__be64 R255_prio_0;
+	__be64 R255_prio_1;
+	__be64 R255_prio_2;
+	__be64 R255_prio_3;
+	__be64 R255_prio_4;
+	__be64 R255_prio_5;
+	__be64 R255_prio_6;
+	__be64 R255_prio_7;
+	__be64 R255_novlan;
+	/* Received frames with a length of 511 octets */
+	__be64 R511_prio_0;
+	__be64 R511_prio_1;
+	__be64 R511_prio_2;
+	__be64 R511_prio_3;
+	__be64 R511_prio_4;
+	__be64 R511_prio_5;
+	__be64 R511_prio_6;
+	__be64 R511_prio_7;
+	__be64 R511_novlan;
+	/* Received frames with a length of 1023 octets */
+	__be64 R1023_prio_0;
+	__be64 R1023_prio_1;
+	__be64 R1023_prio_2;
+	__be64 R1023_prio_3;
+	__be64 R1023_prio_4;
+	__be64 R1023_prio_5;
+	__be64 R1023_prio_6;
+	__be64 R1023_prio_7;
+	__be64 R1023_novlan;
+	/* Received frames with a length of 1518 octets */
+	__be64 R1518_prio_0;
+	__be64 R1518_prio_1;
+	__be64 R1518_prio_2;
+	__be64 R1518_prio_3;
+	__be64 R1518_prio_4;
+	__be64 R1518_prio_5;
+	__be64 R1518_prio_6;
+	__be64 R1518_prio_7;
+	__be64 R1518_novlan;
+	/* Received frames with a length of 1522 octets */
+	__be64 R1522_prio_0;
+	__be64 R1522_prio_1;
+	__be64 R1522_prio_2;
+	__be64 R1522_prio_3;
+	__be64 R1522_prio_4;
+	__be64 R1522_prio_5;
+	__be64 R1522_prio_6;
+	__be64 R1522_prio_7;
+	__be64 R1522_novlan;
+	/* Received frames with a length of 1548 octets */
+	__be64 R1548_prio_0;
+	__be64 R1548_prio_1;
+	__be64 R1548_prio_2;
+	__be64 R1548_prio_3;
+	__be64 R1548_prio_4;
+	__be64 R1548_prio_5;
+	__be64 R1548_prio_6;
+	__be64 R1548_prio_7;
+	__be64 R1548_novlan;
+	/* Received frames with a length of 1548 < octets < MTU */
+	__be64 R2MTU_prio_0;
+	__be64 R2MTU_prio_1;
+	__be64 R2MTU_prio_2;
+	__be64 R2MTU_prio_3;
+	__be64 R2MTU_prio_4;
+	__be64 R2MTU_prio_5;
+	__be64 R2MTU_prio_6;
+	__be64 R2MTU_prio_7;
+	__be64 R2MTU_novlan;
+	/* Received frames with a length of MTU< octets and good CRC */
+	__be64 RGIANT_prio_0;
+	__be64 RGIANT_prio_1;
+	__be64 RGIANT_prio_2;
+	__be64 RGIANT_prio_3;
+	__be64 RGIANT_prio_4;
+	__be64 RGIANT_prio_5;
+	__be64 RGIANT_prio_6;
+	__be64 RGIANT_prio_7;
+	__be64 RGIANT_novlan;
+	/* Received broadcast frames with good CRC */
+	__be64 RBCAST_prio_0;
+	__be64 RBCAST_prio_1;
+	__be64 RBCAST_prio_2;
+	__be64 RBCAST_prio_3;
+	__be64 RBCAST_prio_4;
+	__be64 RBCAST_prio_5;
+	__be64 RBCAST_prio_6;
+	__be64 RBCAST_prio_7;
+	__be64 RBCAST_novlan;
+	/* Received multicast frames with good CRC */
+	__be64 MCAST_prio_0;
+	__be64 MCAST_prio_1;
+	__be64 MCAST_prio_2;
+	__be64 MCAST_prio_3;
+	__be64 MCAST_prio_4;
+	__be64 MCAST_prio_5;
+	__be64 MCAST_prio_6;
+	__be64 MCAST_prio_7;
+	__be64 MCAST_novlan;
+	/* Received unicast not short or GIANT frames with good CRC */
+	__be64 RTOTG_prio_0;
+	__be64 RTOTG_prio_1;
+	__be64 RTOTG_prio_2;
+	__be64 RTOTG_prio_3;
+	__be64 RTOTG_prio_4;
+	__be64 RTOTG_prio_5;
+	__be64 RTOTG_prio_6;
+	__be64 RTOTG_prio_7;
+	__be64 RTOTG_novlan;
+
+	/* Count of total octets of received frames, includes framing characters */
+	__be64 RTTLOCT_prio_0;
+	/* Count of total octets of received frames, not including framing
+	   characters */
+	__be64 RTTLOCT_NOFRM_prio_0;
+	/* Count of Total number of octets received
+	   (only for frames without errors) */
+	__be64 ROCT_prio_0;
+
+	__be64 RTTLOCT_prio_1;
+	__be64 RTTLOCT_NOFRM_prio_1;
+	__be64 ROCT_prio_1;
+
+	__be64 RTTLOCT_prio_2;
+	__be64 RTTLOCT_NOFRM_prio_2;
+	__be64 ROCT_prio_2;
+
+	__be64 RTTLOCT_prio_3;
+	__be64 RTTLOCT_NOFRM_prio_3;
+	__be64 ROCT_prio_3;
+
+	__be64 RTTLOCT_prio_4;
+	__be64 RTTLOCT_NOFRM_prio_4;
+	__be64 ROCT_prio_4;
+
+	__be64 RTTLOCT_prio_5;
+	__be64 RTTLOCT_NOFRM_prio_5;
+	__be64 ROCT_prio_5;
+
+	__be64 RTTLOCT_prio_6;
+	__be64 RTTLOCT_NOFRM_prio_6;
+	__be64 ROCT_prio_6;
+
+	__be64 RTTLOCT_prio_7;
+	__be64 RTTLOCT_NOFRM_prio_7;
+	__be64 ROCT_prio_7;
+
+	__be64 RTTLOCT_novlan;
+	__be64 RTTLOCT_NOFRM_novlan;
+	__be64 ROCT_novlan;
+
+	/* Count of Total received frames including bad frames */
+	__be64 RTOT_prio_0;
+	/* Count of  Total number of received frames with 802.1Q encapsulation */
+	__be64 R1Q_prio_0;
+	__be64 reserved1;
+
+	__be64 RTOT_prio_1;
+	__be64 R1Q_prio_1;
+	__be64 reserved2;
+
+	__be64 RTOT_prio_2;
+	__be64 R1Q_prio_2;
+	__be64 reserved3;
+
+	__be64 RTOT_prio_3;
+	__be64 R1Q_prio_3;
+	__be64 reserved4;
+
+	__be64 RTOT_prio_4;
+	__be64 R1Q_prio_4;
+	__be64 reserved5;
+
+	__be64 RTOT_prio_5;
+	__be64 R1Q_prio_5;
+	__be64 reserved6;
+
+	__be64 RTOT_prio_6;
+	__be64 R1Q_prio_6;
+	__be64 reserved7;
+
+	__be64 RTOT_prio_7;
+	__be64 R1Q_prio_7;
+	__be64 reserved8;
+
+	__be64 RTOT_novlan;
+	__be64 R1Q_novlan;
+	__be64 reserved9;
+
+	/* Total number of Successfully Received Control Frames */
+	__be64 RCNTL;
+	__be64 reserved10;
+	__be64 reserved11;
+	__be64 reserved12;
+	/* Count of received frames with a length/type field  value between 46
+	   (42 for VLANtagged frames) and 1500 (also 1500 for VLAN-tagged frames),
+	   inclusive */
+	__be64 RInRangeLengthErr;
+	/* Count of received frames with length/type field between 1501 and 1535
+	   decimal, inclusive */
+	__be64 ROutRangeLengthErr;
+	/* Count of received frames that are longer than max allowed size for
+	   802.3 frames (1518/1522) */
+	__be64 RFrmTooLong;
+	/* Count frames received with PCS error */
+	__be64 PCS;
+
+	/* Transmit frames with a length of 64 octets */
+	__be64 T64_prio_0;
+	__be64 T64_prio_1;
+	__be64 T64_prio_2;
+	__be64 T64_prio_3;
+	__be64 T64_prio_4;
+	__be64 T64_prio_5;
+	__be64 T64_prio_6;
+	__be64 T64_prio_7;
+	__be64 T64_novlan;
+	__be64 T64_loopbk;
+	/* Transmit frames with a length of 65 to 127 octets. */
+	__be64 T127_prio_0;
+	__be64 T127_prio_1;
+	__be64 T127_prio_2;
+	__be64 T127_prio_3;
+	__be64 T127_prio_4;
+	__be64 T127_prio_5;
+	__be64 T127_prio_6;
+	__be64 T127_prio_7;
+	__be64 T127_novlan;
+	__be64 T127_loopbk;
+	/* Transmit frames with a length of 128 to 255 octets */
+	__be64 T255_prio_0;
+	__be64 T255_prio_1;
+	__be64 T255_prio_2;
+	__be64 T255_prio_3;
+	__be64 T255_prio_4;
+	__be64 T255_prio_5;
+	__be64 T255_prio_6;
+	__be64 T255_prio_7;
+	__be64 T255_novlan;
+	__be64 T255_loopbk;
+	/* Transmit frames with a length of 256 to 511 octets */
+	__be64 T511_prio_0;
+	__be64 T511_prio_1;
+	__be64 T511_prio_2;
+	__be64 T511_prio_3;
+	__be64 T511_prio_4;
+	__be64 T511_prio_5;
+	__be64 T511_prio_6;
+	__be64 T511_prio_7;
+	__be64 T511_novlan;
+	__be64 T511_loopbk;
+	/* Transmit frames with a length of 512 to 1023 octets */
+	__be64 T1023_prio_0;
+	__be64 T1023_prio_1;
+	__be64 T1023_prio_2;
+	__be64 T1023_prio_3;
+	__be64 T1023_prio_4;
+	__be64 T1023_prio_5;
+	__be64 T1023_prio_6;
+	__be64 T1023_prio_7;
+	__be64 T1023_novlan;
+	__be64 T1023_loopbk;
+	/* Transmit frames with a length of 1024 to 1518 octets */
+	__be64 T1518_prio_0;
+	__be64 T1518_prio_1;
+	__be64 T1518_prio_2;
+	__be64 T1518_prio_3;
+	__be64 T1518_prio_4;
+	__be64 T1518_prio_5;
+	__be64 T1518_prio_6;
+	__be64 T1518_prio_7;
+	__be64 T1518_novlan;
+	__be64 T1518_loopbk;
+	/* Counts transmit frames with a length of 1519 to 1522 bytes */
+	__be64 T1522_prio_0;
+	__be64 T1522_prio_1;
+	__be64 T1522_prio_2;
+	__be64 T1522_prio_3;
+	__be64 T1522_prio_4;
+	__be64 T1522_prio_5;
+	__be64 T1522_prio_6;
+	__be64 T1522_prio_7;
+	__be64 T1522_novlan;
+	__be64 T1522_loopbk;
+	/* Transmit frames with a length of 1523 to 1548 octets */
+	__be64 T1548_prio_0;
+	__be64 T1548_prio_1;
+	__be64 T1548_prio_2;
+	__be64 T1548_prio_3;
+	__be64 T1548_prio_4;
+	__be64 T1548_prio_5;
+	__be64 T1548_prio_6;
+	__be64 T1548_prio_7;
+	__be64 T1548_novlan;
+	__be64 T1548_loopbk;
+	/* Counts transmit frames with a length of 1549 to MTU bytes */
+	__be64 T2MTU_prio_0;
+	__be64 T2MTU_prio_1;
+	__be64 T2MTU_prio_2;
+	__be64 T2MTU_prio_3;
+	__be64 T2MTU_prio_4;
+	__be64 T2MTU_prio_5;
+	__be64 T2MTU_prio_6;
+	__be64 T2MTU_prio_7;
+	__be64 T2MTU_novlan;
+	__be64 T2MTU_loopbk;
+	/* Transmit frames with a length greater than MTU octets and a good CRC. */
+	__be64 TGIANT_prio_0;
+	__be64 TGIANT_prio_1;
+	__be64 TGIANT_prio_2;
+	__be64 TGIANT_prio_3;
+	__be64 TGIANT_prio_4;
+	__be64 TGIANT_prio_5;
+	__be64 TGIANT_prio_6;
+	__be64 TGIANT_prio_7;
+	__be64 TGIANT_novlan;
+	__be64 TGIANT_loopbk;
+	/* Transmit broadcast frames with a good CRC */
+	__be64 TBCAST_prio_0;
+	__be64 TBCAST_prio_1;
+	__be64 TBCAST_prio_2;
+	__be64 TBCAST_prio_3;
+	__be64 TBCAST_prio_4;
+	__be64 TBCAST_prio_5;
+	__be64 TBCAST_prio_6;
+	__be64 TBCAST_prio_7;
+	__be64 TBCAST_novlan;
+	__be64 TBCAST_loopbk;
+	/* Transmit multicast frames with a good CRC */
+	__be64 TMCAST_prio_0;
+	__be64 TMCAST_prio_1;
+	__be64 TMCAST_prio_2;
+	__be64 TMCAST_prio_3;
+	__be64 TMCAST_prio_4;
+	__be64 TMCAST_prio_5;
+	__be64 TMCAST_prio_6;
+	__be64 TMCAST_prio_7;
+	__be64 TMCAST_novlan;
+	__be64 TMCAST_loopbk;
+	/* Transmit good frames that are neither broadcast nor multicast */
+	__be64 TTOTG_prio_0;
+	__be64 TTOTG_prio_1;
+	__be64 TTOTG_prio_2;
+	__be64 TTOTG_prio_3;
+	__be64 TTOTG_prio_4;
+	__be64 TTOTG_prio_5;
+	__be64 TTOTG_prio_6;
+	__be64 TTOTG_prio_7;
+	__be64 TTOTG_novlan;
+	__be64 TTOTG_loopbk;
+
+	/* total octets of transmitted frames, including framing characters */
+	__be64 TTTLOCT_prio_0;
+	/* total octets of transmitted frames, not including framing characters */
+	__be64 TTTLOCT_NOFRM_prio_0;
+	/* ifOutOctets */
+	__be64 TOCT_prio_0;
+
+	__be64 TTTLOCT_prio_1;
+	__be64 TTTLOCT_NOFRM_prio_1;
+	__be64 TOCT_prio_1;
+
+	__be64 TTTLOCT_prio_2;
+	__be64 TTTLOCT_NOFRM_prio_2;
+	__be64 TOCT_prio_2;
+
+	__be64 TTTLOCT_prio_3;
+	__be64 TTTLOCT_NOFRM_prio_3;
+	__be64 TOCT_prio_3;
+
+	__be64 TTTLOCT_prio_4;
+	__be64 TTTLOCT_NOFRM_prio_4;
+	__be64 TOCT_prio_4;
+
+	__be64 TTTLOCT_prio_5;
+	__be64 TTTLOCT_NOFRM_prio_5;
+	__be64 TOCT_prio_5;
+
+	__be64 TTTLOCT_prio_6;
+	__be64 TTTLOCT_NOFRM_prio_6;
+	__be64 TOCT_prio_6;
+
+	__be64 TTTLOCT_prio_7;
+	__be64 TTTLOCT_NOFRM_prio_7;
+	__be64 TOCT_prio_7;
+
+	__be64 TTTLOCT_novlan;
+	__be64 TTTLOCT_NOFRM_novlan;
+	__be64 TOCT_novlan;
+
+	__be64 TTTLOCT_loopbk;
+	__be64 TTTLOCT_NOFRM_loopbk;
+	__be64 TOCT_loopbk;
+
+	/* Total frames transmitted with a good CRC that are not aborted  */
+	__be64 TTOT_prio_0;
+	/* Total number of frames transmitted with 802.1Q encapsulation */
+	__be64 T1Q_prio_0;
+	__be64 reserved13;
+
+	__be64 TTOT_prio_1;
+	__be64 T1Q_prio_1;
+	__be64 reserved14;
+
+	__be64 TTOT_prio_2;
+	__be64 T1Q_prio_2;
+	__be64 reserved15;
+
+	__be64 TTOT_prio_3;
+	__be64 T1Q_prio_3;
+	__be64 reserved16;
+
+	__be64 TTOT_prio_4;
+	__be64 T1Q_prio_4;
+	__be64 reserved17;
+
+	__be64 TTOT_prio_5;
+	__be64 T1Q_prio_5;
+	__be64 reserved18;
+
+	__be64 TTOT_prio_6;
+	__be64 T1Q_prio_6;
+	__be64 reserved19;
+
+	__be64 TTOT_prio_7;
+	__be64 T1Q_prio_7;
+	__be64 reserved20;
+
+	__be64 TTOT_novlan;
+	__be64 T1Q_novlan;
+	__be64 reserved21;
+
+	__be64 TTOT_loopbk;
+	__be64 T1Q_loopbk;
+	__be64 reserved22;
+
+	/* Received frames with a length greater than MTU octets and a bad CRC */
+	__be32 RJBBR;
+	/* Received frames with a bad CRC that are not runts, jabbers,
+	   or alignment errors */
+	__be32 RCRC;
+	/* Received frames with SFD with a length of less than 64 octets and a
+	   bad CRC */
+	__be32 RRUNT;
+	/* Received frames with a length less than 64 octets and a good CRC */
+	__be32 RSHORT;
+	/* Total Number of Received Packets Dropped */
+	__be32 RDROP;
+	/* Drop due to overflow  */
+	__be32 RdropOvflw;
+	/* Drop due to overflow */
+	__be32 RdropLength;
+	/* Total of good frames. Does not include frames received with
+	   frame-too-long, FCS, or length errors */
+	__be32 RTOTFRMS;
+	/* Total dropped Xmited packets */
+	__be32 TDROP;
+};
+
+
+#endif

drivers/net/mlx4/en_resources.c

+/*
+ * Copyright (c) 2007 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/vmalloc.h>
+#include <linux/mlx4/qp.h>
+
+#include "mlx4_en.h"
+
+void mlx4_en_fill_qp_context(struct mlx4_en_priv *priv, int size, int stride,
+			     int is_tx, int rss, int qpn, int cqn, int srqn,
+			     struct mlx4_qp_context *context)
+{
+	struct mlx4_en_dev *mdev = priv->mdev;
+
+	memset(context, 0, sizeof *context);
+	context->flags = cpu_to_be32(7 << 16 | rss << 13);
+	context->pd = cpu_to_be32(mdev->priv_pdn);
+	context->mtu_msgmax = 0xff;
+	context->rq_size_stride = 0;
+	if (is_tx)
+		context->sq_size_stride = ilog2(size) << 3 | (ilog2(stride) - 4);
+	else
+		context->sq_size_stride = 1;
+	context->usr_page = cpu_to_be32(mdev->priv_uar.index);
+	context->local_qpn = cpu_to_be32(qpn);
+	context->pri_path.ackto = 1 & 0x07;
+	context->pri_path.sched_queue = 0x83 | (priv->port - 1) << 6;
+	context->pri_path.counter_index = 0xff;
+	context->cqn_send = cpu_to_be32(cqn);
+	context->cqn_recv = cpu_to_be32(cqn);
+	context->db_rec_addr = cpu_to_be64(priv->res.db.dma << 2);
+	if (!rss)
+		context->srqn = cpu_to_be32(MLX4_EN_USE_SRQ | srqn);
+}
+
+
+int mlx4_en_map_buffer(struct mlx4_buf *buf)
+{
+	struct page **pages;
+	int i;
+
+	if (BITS_PER_LONG == 64 || buf->nbufs == 1)
+		return 0;
+
+	pages = kmalloc(sizeof *pages * buf->nbufs, GFP_KERNEL);
+	if (!pages)
+		return -ENOMEM;
+
+	for (i = 0; i < buf->nbufs; ++i)
+		pages[i] = virt_to_page(buf->page_list[i].buf);
+
+	buf->direct.buf = vmap(pages, buf->nbufs, VM_MAP, PAGE_KERNEL);
+	kfree(pages);
+	if (!buf->direct.buf)
+		return -ENOMEM;
+
+	return 0;
+}
+
+void mlx4_en_unmap_buffer(struct mlx4_buf *buf)
+{
+	if (BITS_PER_LONG == 64 || buf->nbufs == 1)
+		return;
+
+	vunmap(buf->direct.buf);
+}

drivers/net/mlx4/eq.c

@@ -558,7 +558,7 @@ int mlx4_init_eq_table(struct mlx4_dev *dev)
 	int i;
 
 	err = mlx4_bitmap_init(&priv->eq_table.bitmap, dev->caps.num_eqs,
-			       dev->caps.num_eqs - 1, dev->caps.reserved_eqs);
+			       dev->caps.num_eqs - 1, dev->caps.reserved_eqs, 0);
 	if (err)
 		return err;
 

drivers/net/mlx4/fw.c

@@ -88,6 +88,7 @@ static void dump_dev_cap_flags(struct mlx4_dev *dev, u32 flags)
 		[ 8] = "P_Key violation counter",
 		[ 9] = "Q_Key violation counter",
 		[10] = "VMM",
+		[12] = "DPDP",
 		[16] = "MW support",
 		[17] = "APM support",
 		[18] = "Atomic ops support",
@@ -346,7 +347,7 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 			MLX4_GET(field, outbox, QUERY_DEV_CAP_VL_PORT_OFFSET);
 			dev_cap->max_vl[i]	   = field >> 4;
 			MLX4_GET(field, outbox, QUERY_DEV_CAP_MTU_WIDTH_OFFSET);
-			dev_cap->max_mtu[i]	   = field >> 4;
+			dev_cap->ib_mtu[i]	   = field >> 4;
 			dev_cap->max_port_width[i] = field & 0xf;
 			MLX4_GET(field, outbox, QUERY_DEV_CAP_MAX_GID_OFFSET);
 			dev_cap->max_gids[i]	   = 1 << (field & 0xf);
@@ -354,9 +355,13 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 			dev_cap->max_pkeys[i]	   = 1 << (field & 0xf);
 		}
 	} else {
+#define QUERY_PORT_SUPPORTED_TYPE_OFFSET	0x00
 #define QUERY_PORT_MTU_OFFSET			0x01
+#define QUERY_PORT_ETH_MTU_OFFSET		0x02
 #define QUERY_PORT_WIDTH_OFFSET			0x06
 #define QUERY_PORT_MAX_GID_PKEY_OFFSET		0x07
+#define QUERY_PORT_MAC_OFFSET			0x08
+#define QUERY_PORT_MAX_MACVLAN_OFFSET		0x0a
 #define QUERY_PORT_MAX_VL_OFFSET		0x0b
 
 		for (i = 1; i <= dev_cap->num_ports; ++i) {
@@ -365,8 +370,10 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 			if (err)
 				goto out;
 
+			MLX4_GET(field, outbox, QUERY_PORT_SUPPORTED_TYPE_OFFSET);
+			dev_cap->supported_port_types[i] = field & 3;
 			MLX4_GET(field, outbox, QUERY_PORT_MTU_OFFSET);
-			dev_cap->max_mtu[i]	   = field & 0xf;
+			dev_cap->ib_mtu[i]	   = field & 0xf;
 			MLX4_GET(field, outbox, QUERY_PORT_WIDTH_OFFSET);
 			dev_cap->max_port_width[i] = field & 0xf;
 			MLX4_GET(field, outbox, QUERY_PORT_MAX_GID_PKEY_OFFSET);
@@ -374,6 +381,11 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 			dev_cap->max_pkeys[i]	   = 1 << (field & 0xf);
 			MLX4_GET(field, outbox, QUERY_PORT_MAX_VL_OFFSET);
 			dev_cap->max_vl[i]	   = field & 0xf;
+			MLX4_GET(field, outbox, QUERY_PORT_MAX_MACVLAN_OFFSET);
+			dev_cap->log_max_macs[i]  = field & 0xf;
+			dev_cap->log_max_vlans[i] = field >> 4;
+			MLX4_GET(dev_cap->eth_mtu[i], outbox, QUERY_PORT_ETH_MTU_OFFSET);
+			MLX4_GET(dev_cap->def_mac[i], outbox, QUERY_PORT_MAC_OFFSET);
 		}
 	}
 
@@ -407,7 +419,7 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 	mlx4_dbg(dev, "Max CQEs: %d, max WQEs: %d, max SRQ WQEs: %d\n",
 		 dev_cap->max_cq_sz, dev_cap->max_qp_sz, dev_cap->max_srq_sz);
 	mlx4_dbg(dev, "Local CA ACK delay: %d, max MTU: %d, port width cap: %d\n",
-		 dev_cap->local_ca_ack_delay, 128 << dev_cap->max_mtu[1],
+		 dev_cap->local_ca_ack_delay, 128 << dev_cap->ib_mtu[1],
 		 dev_cap->max_port_width[1]);
 	mlx4_dbg(dev, "Max SQ desc size: %d, max SQ S/G: %d\n",
 		 dev_cap->max_sq_desc_sz, dev_cap->max_sq_sg);
@@ -819,7 +831,7 @@ int mlx4_INIT_PORT(struct mlx4_dev *dev, int port)
 		flags |= (dev->caps.port_width_cap[port] & 0xf) << INIT_PORT_PORT_WIDTH_SHIFT;
 		MLX4_PUT(inbox, flags,		  INIT_PORT_FLAGS_OFFSET);
 
-		field = 128 << dev->caps.mtu_cap[port];
+		field = 128 << dev->caps.ib_mtu_cap[port];
 		MLX4_PUT(inbox, field, INIT_PORT_MTU_OFFSET);
 		field = dev->caps.gid_table_len[port];
 		MLX4_PUT(inbox, field, INIT_PORT_MAX_GID_OFFSET);

drivers/net/mlx4/fw.h

@@ -66,11 +66,13 @@ struct mlx4_dev_cap {
 	int local_ca_ack_delay;
 	int num_ports;
 	u32 max_msg_sz;
-	int max_mtu[MLX4_MAX_PORTS + 1];
+	int ib_mtu[MLX4_MAX_PORTS + 1];
 	int max_port_width[MLX4_MAX_PORTS + 1];
 	int max_vl[MLX4_MAX_PORTS + 1];
 	int max_gids[MLX4_MAX_PORTS + 1];
 	int max_pkeys[MLX4_MAX_PORTS + 1];
+	u64 def_mac[MLX4_MAX_PORTS + 1];
+	u16 eth_mtu[MLX4_MAX_PORTS + 1];
 	u16 stat_rate_support;
 	u32 flags;
 	int reserved_uars;
@@ -102,6 +104,9 @@ struct mlx4_dev_cap {
 	u32 reserved_lkey;
 	u64 max_icm_sz;
 	int max_gso_sz;
+	u8  supported_port_types[MLX4_MAX_PORTS + 1];
+	u8  log_max_macs[MLX4_MAX_PORTS + 1];
+	u8  log_max_vlans[MLX4_MAX_PORTS + 1];
 };
 
 struct mlx4_adapter {

drivers/net/mlx4/mcg.c

@@ -368,8 +368,8 @@ int mlx4_init_mcg_table(struct mlx4_dev *dev)
 	struct mlx4_priv *priv = mlx4_priv(dev);
 	int err;
 
-	err = mlx4_bitmap_init(&priv->mcg_table.bitmap,
-			       dev->caps.num_amgms, dev->caps.num_amgms - 1, 0);
+	err = mlx4_bitmap_init(&priv->mcg_table.bitmap, dev->caps.num_amgms,
+			       dev->caps.num_amgms - 1, 0, 0);
 	if (err)
 		return err;
 

drivers/net/mlx4/mlx4.h

@@ -111,6 +111,7 @@ struct mlx4_bitmap {
 	u32			last;
 	u32			top;
 	u32			max;
+	u32                     reserved_top;
 	u32			mask;
 	spinlock_t		lock;
 	unsigned long	       *table;
@@ -251,6 +252,38 @@ struct mlx4_catas_err {
 	struct list_head	list;
 };
 
+#define MLX4_MAX_MAC_NUM	128
+#define MLX4_MAC_TABLE_SIZE	(MLX4_MAX_MAC_NUM << 3)
+
+struct mlx4_mac_table {
+	__be64			entries[MLX4_MAX_MAC_NUM];
+	int			refs[MLX4_MAX_MAC_NUM];
+	struct mutex		mutex;
+	int			total;
+	int			max;
+};
+
+#define MLX4_MAX_VLAN_NUM	128
+#define MLX4_VLAN_TABLE_SIZE	(MLX4_MAX_VLAN_NUM << 2)
+
+struct mlx4_vlan_table {
+	__be32			entries[MLX4_MAX_VLAN_NUM];
+	int			refs[MLX4_MAX_VLAN_NUM];
+	struct mutex		mutex;
+	int			total;
+	int			max;
+};
+
+struct mlx4_port_info {
+	struct mlx4_dev	       *dev;
+	int			port;
+	char			dev_name[16];
+	struct device_attribute port_attr;
+	enum mlx4_port_type	tmp_type;
+	struct mlx4_mac_table	mac_table;
+	struct mlx4_vlan_table	vlan_table;
+};
+
 struct mlx4_priv {
 	struct mlx4_dev		dev;
 
@@ -279,6 +312,8 @@ struct mlx4_priv {
 
 	struct mlx4_uar		driver_uar;
 	void __iomem	       *kar;
+	struct mlx4_port_info	port[MLX4_MAX_PORTS + 1];
+	struct mutex		port_mutex;
 };
 
 static inline struct mlx4_priv *mlx4_priv(struct mlx4_dev *dev)
@@ -288,7 +323,10 @@ static inline struct mlx4_priv *mlx4_priv(struct mlx4_dev *dev)
 
 u32 mlx4_bitmap_alloc(struct mlx4_bitmap *bitmap);
 void mlx4_bitmap_free(struct mlx4_bitmap *bitmap, u32 obj);
-int mlx4_bitmap_init(struct mlx4_bitmap *bitmap, u32 num, u32 mask, u32 reserved);
+u32 mlx4_bitmap_alloc_range(struct mlx4_bitmap *bitmap, int cnt, int align);
+void mlx4_bitmap_free_range(struct mlx4_bitmap *bitmap, u32 obj, int cnt);
+int mlx4_bitmap_init(struct mlx4_bitmap *bitmap, u32 num, u32 mask,
+		     u32 reserved_bot, u32 resetrved_top);
 void mlx4_bitmap_cleanup(struct mlx4_bitmap *bitmap);
 
 int mlx4_reset(struct mlx4_dev *dev);
@@ -346,4 +384,9 @@ void mlx4_srq_event(struct mlx4_dev *dev, u32 srqn, int event_type);
 
 void mlx4_handle_catas_err(struct mlx4_dev *dev);
 
+void mlx4_init_mac_table(struct mlx4_dev *dev, struct mlx4_mac_table *table);
+void mlx4_init_vlan_table(struct mlx4_dev *dev, struct mlx4_vlan_table *table);
+
+int mlx4_SET_PORT(struct mlx4_dev *dev, u8 port);
+
 #endif /* MLX4_H */

drivers/net/mlx4/mr.c

@@ -461,7 +461,7 @@ int mlx4_init_mr_table(struct mlx4_dev *dev)
 	int err;
 
 	err = mlx4_bitmap_init(&mr_table->mpt_bitmap, dev->caps.num_mpts,
-			       ~0, dev->caps.reserved_mrws);
+			       ~0, dev->caps.reserved_mrws, 0);
 	if (err)
 		return err;
 

drivers/net/mlx4/pd.c

@@ -62,7 +62,7 @@ int mlx4_init_pd_table(struct mlx4_dev *dev)
 	struct mlx4_priv *priv = mlx4_priv(dev);
 
 	return mlx4_bitmap_init(&priv->pd_bitmap, dev->caps.num_pds,
-				(1 << 24) - 1, dev->caps.reserved_pds);
+				(1 << 24) - 1, dev->caps.reserved_pds, 0);
 }
 
 void mlx4_cleanup_pd_table(struct mlx4_dev *dev)
@@ -100,7 +100,7 @@ int mlx4_init_uar_table(struct mlx4_dev *dev)
 
 	return mlx4_bitmap_init(&mlx4_priv(dev)->uar_table.bitmap,
 				dev->caps.num_uars, dev->caps.num_uars - 1,
-				max(128, dev->caps.reserved_uars));
+				max(128, dev->caps.reserved_uars), 0);
 }
 
 void mlx4_cleanup_uar_table(struct mlx4_dev *dev)

drivers/net/mlx4/qp.c

@@ -147,19 +147,42 @@ int mlx4_qp_modify(struct mlx4_dev *dev, struct mlx4_mtt *mtt,
 }
 EXPORT_SYMBOL_GPL(mlx4_qp_modify);
 
-int mlx4_qp_alloc(struct mlx4_dev *dev, int sqpn, struct mlx4_qp *qp)
+int mlx4_qp_reserve_range(struct mlx4_dev *dev, int cnt, int align, int *base)
+{
+	struct mlx4_priv *priv = mlx4_priv(dev);
+	struct mlx4_qp_table *qp_table = &priv->qp_table;
+	int qpn;
+
+	qpn = mlx4_bitmap_alloc_range(&qp_table->bitmap, cnt, align);
+	if (qpn == -1)
+		return -ENOMEM;
+
+	*base = qpn;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(mlx4_qp_reserve_range);
+
+void mlx4_qp_release_range(struct mlx4_dev *dev, int base_qpn, int cnt)
+{
+	struct mlx4_priv *priv = mlx4_priv(dev);
+	struct mlx4_qp_table *qp_table = &priv->qp_table;
+	if (base_qpn < dev->caps.sqp_start + 8)
+		return;
+
+	mlx4_bitmap_free_range(&qp_table->bitmap, base_qpn, cnt);
+}
+EXPORT_SYMBOL_GPL(mlx4_qp_release_range);
+
+int mlx4_qp_alloc(struct mlx4_dev *dev, int qpn, struct mlx4_qp *qp)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
 	struct mlx4_qp_table *qp_table = &priv->qp_table;
 	int err;
 
-	if (sqpn)
-		qp->qpn = sqpn;
-	else {
-		qp->qpn = mlx4_bitmap_alloc(&qp_table->bitmap);
-		if (qp->qpn == -1)
-			return -ENOMEM;
-	}
+	if (!qpn)
+		return -EINVAL;
+
+	qp->qpn = qpn;
 
 	err = mlx4_table_get(dev, &qp_table->qp_table, qp->qpn);
 	if (err)
@@ -208,9 +231,6 @@ int mlx4_qp_alloc(struct mlx4_dev *dev, int sqpn, struct mlx4_qp *qp)
 	mlx4_table_put(dev, &qp_table->qp_table, qp->qpn);
 
 err_out:
-	if (!sqpn)
-		mlx4_bitmap_free(&qp_table->bitmap, qp->qpn);
-
 	return err;
 }
 EXPORT_SYMBOL_GPL(mlx4_qp_alloc);
@@ -239,9 +259,6 @@ void mlx4_qp_free(struct mlx4_dev *dev, struct mlx4_qp *qp)
 	mlx4_table_put(dev, &qp_table->altc_table, qp->qpn);
 	mlx4_table_put(dev, &qp_table->auxc_table, qp->qpn);
 	mlx4_table_put(dev, &qp_table->qp_table, qp->qpn);
-
-	if (qp->qpn >= dev->caps.sqp_start + 8)
-		mlx4_bitmap_free(&qp_table->bitmap, qp->qpn);
 }
 EXPORT_SYMBOL_GPL(mlx4_qp_free);
 
@@ -255,6 +272,7 @@ int mlx4_init_qp_table(struct mlx4_dev *dev)
 {
 	struct mlx4_qp_table *qp_table = &mlx4_priv(dev)->qp_table;
 	int err;
+	int reserved_from_top = 0;
 
 	spin_lock_init(&qp_table->lock);
 	INIT_RADIX_TREE(&dev->qp_table_tree, GFP_ATOMIC);
@@ -264,9 +282,40 @@ int mlx4_init_qp_table(struct mlx4_dev *dev)
 	 * block of special QPs must be aligned to a multiple of 8, so
 	 * round up.
 	 */
-	dev->caps.sqp_start = ALIGN(dev->caps.reserved_qps, 8);
+	dev->caps.sqp_start =
+		ALIGN(dev->caps.reserved_qps_cnt[MLX4_QP_REGION_FW], 8);
+
+	{
+		int sort[MLX4_NUM_QP_REGION];
+		int i, j, tmp;
+		int last_base = dev->caps.num_qps;
+
+		for (i = 1; i < MLX4_NUM_QP_REGION; ++i)
+			sort[i] = i;
+
+		for (i = MLX4_NUM_QP_REGION; i > 0; --i) {
+			for (j = 2; j < i; ++j) {
+				if (dev->caps.reserved_qps_cnt[sort[j]] >
+				    dev->caps.reserved_qps_cnt[sort[j - 1]]) {
+					tmp             = sort[j];
+					sort[j]         = sort[j - 1];
+					sort[j - 1]     = tmp;
+				}
+			}
+		}
+
+		for (i = 1; i < MLX4_NUM_QP_REGION; ++i) {
+			last_base -= dev->caps.reserved_qps_cnt[sort[i]];
+			dev->caps.reserved_qps_base[sort[i]] = last_base;
+			reserved_from_top +=
+				dev->caps.reserved_qps_cnt[sort[i]];
+		}
+
+	}
+
 	err = mlx4_bitmap_init(&qp_table->bitmap, dev->caps.num_qps,
-			       (1 << 24) - 1, dev->caps.sqp_start + 8);
+			       (1 << 23) - 1, dev->caps.sqp_start + 8,
+			       reserved_from_top);
 	if (err)
 		return err;
 

drivers/net/mlx4/srq.c

@@ -245,7 +245,7 @@ int mlx4_init_srq_table(struct mlx4_dev *dev)
 	INIT_RADIX_TREE(&srq_table->tree, GFP_ATOMIC);
 
 	err = mlx4_bitmap_init(&srq_table->bitmap, dev->caps.num_srqs,
-			       dev->caps.num_srqs - 1, dev->caps.reserved_srqs);
+			       dev->caps.num_srqs - 1, dev->caps.reserved_srqs, 0);
 	if (err)
 		return err;
 

include/linux/mlx4/cmd.h

@@ -132,6 +132,15 @@ enum {
 	MLX4_MAILBOX_SIZE	=  4096
 };
 
+enum {
+	/* set port opcode modifiers */
+	MLX4_SET_PORT_GENERAL   = 0x0,
+	MLX4_SET_PORT_RQP_CALC  = 0x1,
+	MLX4_SET_PORT_MAC_TABLE = 0x2,
+	MLX4_SET_PORT_VLAN_TABLE = 0x3,
+	MLX4_SET_PORT_PRIO_MAP  = 0x4,
+};
+
 struct mlx4_dev;
 
 struct mlx4_cmd_mailbox {