Merge branch 'mlx5-RDMA-netdevice'

Saeed Mahameed says:

====================
Mellanox, mlx5 RDMA net device support

This series provides the lower level mlx5 support of RDMA netdevice
creation API [1] suggested and introduced by Intel's HFI OPA VNIC
netdevice driver [2], to enable IPoIB mlx5 RDMA netdevice creation.

mlx5 IPoIB RDMA netdev will serve as an acceleration netdevice for the current
IPoIB ULP generic netdevice, providing:
	- mlx5 RSS support.
	- mlx5 HW RX,TX offloads (checksum, TSO, LRO, etc ..).
	- Full mlx5 HW features transparent to the ULP itself.

The idea here is to reuse and benefit from the already implemented mlx5e netdevice
management and channels API for both etherent and RDMA netdevices, since both IPoIB
and Ethernet netdevices share same common mlx5 HW resources (with some small
exceptions) and share most of the control/data path logic, it is more natural to
have them share the same code.

The differences between IPoIB and Ethernet netdevices can be summarized to:

Steering:
In mlx5, IPoIB traffic is sent and received from an underlay special QP, and in Ethernet
the traffic is handled by vports and vport steering is managed by e-switch or FW.

For IPoIB traffic to get steered correctly the only thing we need to do is to create RSS
HW contexts for RX and TX HW contexts for TX (similar to mlx5e) with the underlay QP attached to
them (underlay QP will be 0 in case of Ethernet).

RX,TX:
Since IPoIB traffic is different, slightly modified RX and TX handlers are required,
still we do some code reuse in data path via common helper functions.

All of the other generic netdevice and mlx5 aspects will be shared between mlx5 Ethernet
and IPoIB netdevices, e.g.
	- Channels creation and handling (RQs,SQs,CQs, NAPI, interrupt moderation, etc..)
	- Offloads, checksum, GRO, LRO, TSO, and more.
        - netdevice logic and non Ethernet specific ndos (open/close, etc..)

In order to achieve what we want:

In patchet 1 to 3, Erez added the supported for underlay QP in mlx5_ifc and refactored
the mlx5 steering code to accept the underlay QP as a parameter for creating steering
objects and enabled flow steering for IB link.

Then we are going to use the mlx5e netdevice profile, which is already used to separate between
NIC and VF representors netdevices, to create new type of IPoIB netdevice profile.

For that, one small refactoring is required to make mlx5e netdevice profile management
more genetic and agnostic to link type which is done in patch #4.

In patch #5, we introduce ipoib.c to host all of mlx5 IPoIB (mlx5i) specific logic and a
skeleton for the IPoIB mlx5 netdevice profile, and we will start filling it in next patches,
using mlx5e already existing APIs.

Patch #6 and #7, Implement init/cleanup RX mlx5i netdev profile handlers to create mlx5 RSS
resources, same as mlx5e but without vlan and L2 steering tables.

Patch #8, Implement init/cleanup TX mlx5i netdev profile handlers, to create TX resources
same as mlx5e but with one TC (tc = 0) support.

Patch #9, Implement mlx5i open/close ndos, where we reuese the mlx5e channels API, to start/stop TX/RX channels.

Patch #10, Create the underlay QP and attach it to mlx5i RSS and TX HW contexts.

Patch #11 and #12, Break down the mlx5e xmit flow into smaller helper function and implement the
mlx5i IPoIB xmit routine.

Patch #13 and #14, Have an RX handler per netdevice profile. We already do this before this series
in a non clean way to separate between NIC netdev and VF representor RX handlers, in patch 13 we make
the RX handler generic and bound to a profile and in patch 14 we implement the IPoIB RX handlers.

Patch #15, Small cleanup to avoid e-switch with IPoIB netdev.

In order to enable mlx5 IPoIB, a merge between the IPoIB RDMA netdev offolad support [3]
- which was alread submitted to the rdma mailing list - and this series is required
plus an extra small patch [4] which will connect between both sides and actually enables the offload.

Once both patch-sets are merged into linux we will have to submit the extra small patch [4], to enable
the feature.

Thanks,
Saeed.

[1] https://patchwork.kernel.org/patch/9676637/

[2] https://lwn.net/Articles/715453/
    https://patchwork.kernel.org/patch/9587815/

[3] https://patchwork.kernel.org/patch/9672069/
[4] https://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux.git/commit/?id=0141db6a686e32294dee015b7d07706162ba48d8
====================

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

17 files changed, 1158 insertions, 312 deletions

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig
index 117170014e88..a84b652f9b54 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig
+++ b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig
@@ -31,3 +31,10 @@ config MLX5_CORE_EN_DCB
 	  This flag is depended on the kernel's DCB support.
 
 	  If unsure, set to Y
+
+config MLX5_CORE_IPOIB
+	bool "Mellanox Technologies ConnectX-4 IPoIB offloads support"
+	depends on MLX5_CORE_EN
+	default y
+	---help---
+	  MLX5 IPoIB offloads & acceleration support.
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
index 9f43beb86250..9e644615f07a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile
+++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
@@ -11,3 +11,5 @@ mlx5_core-$(CONFIG_MLX5_CORE_EN) += wq.o eswitch.o eswitch_offloads.o \
 		en_tc.o en_arfs.o en_rep.o en_fs_ethtool.o en_selftest.o
 
 mlx5_core-$(CONFIG_MLX5_CORE_EN_DCB) +=  en_dcbnl.o
+
+mlx5_core-$(CONFIG_MLX5_CORE_IPOIB) += ipoib.o
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index b7feecfbb5a5..0881325fba04 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -37,6 +37,7 @@
 #include <linux/timecounter.h>
 #include <linux/net_tstamp.h>
 #include <linux/ptp_clock_kernel.h>
+#include <linux/crash_dump.h>
 #include <linux/mlx5/driver.h>
 #include <linux/mlx5/qp.h>
 #include <linux/mlx5/cq.h>
@@ -153,6 +154,14 @@ static inline int mlx5_max_log_rq_size(int wq_type)
 	}
 }
 
+static inline int mlx5e_get_max_num_channels(struct mlx5_core_dev *mdev)
+{
+	return is_kdump_kernel() ?
+		MLX5E_MIN_NUM_CHANNELS :
+		min_t(int, mdev->priv.eq_table.num_comp_vectors,
+		      MLX5E_MAX_NUM_CHANNELS);
+}
+
 struct mlx5e_tx_wqe {
 	struct mlx5_wqe_ctrl_seg ctrl;
 	struct mlx5_wqe_eth_seg  eth;
@@ -295,6 +304,7 @@ struct mlx5e_cq {
 } ____cacheline_aligned_in_smp;
 
 struct mlx5e_tx_wqe_info {
+	struct sk_buff *skb;
 	u32 num_bytes;
 	u8  num_wqebbs;
 	u8  num_dma;
@@ -336,7 +346,6 @@ struct mlx5e_txqsq {
 
 	/* write@xmit, read@completion */
 	struct {
-		struct sk_buff           **skb;
 		struct mlx5e_sq_dma       *dma_fifo;
 		struct mlx5e_tx_wqe_info  *wqe_info;
 	} db;
@@ -770,6 +779,10 @@ struct mlx5e_profile {
 	void	(*disable)(struct mlx5e_priv *priv);
 	void	(*update_stats)(struct mlx5e_priv *priv);
 	int	(*max_nch)(struct mlx5_core_dev *mdev);
+	struct {
+		mlx5e_fp_handle_rx_cqe handle_rx_cqe;
+		mlx5e_fp_handle_rx_cqe handle_rx_cqe_mpwqe;
+	} rx_handlers;
 	int	max_tc;
 };
 
@@ -874,6 +887,8 @@ typedef int (*mlx5e_fp_hw_modify)(struct mlx5e_priv *priv);
 void mlx5e_switch_priv_channels(struct mlx5e_priv *priv,
 				struct mlx5e_channels *new_chs,
 				mlx5e_fp_hw_modify hw_modify);
+void mlx5e_activate_priv_channels(struct mlx5e_priv *priv);
+void mlx5e_deactivate_priv_channels(struct mlx5e_priv *priv);
 
 void mlx5e_build_default_indir_rqt(struct mlx5_core_dev *mdev,
 				   u32 *indirection_rqt, int len,
@@ -990,21 +1005,30 @@ int mlx5e_attr_get(struct net_device *dev, struct switchdev_attr *attr);
 void mlx5e_handle_rx_cqe_rep(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe);
 void mlx5e_update_hw_rep_counters(struct mlx5e_priv *priv);
 
+/* common netdev helpers */
+int mlx5e_create_indirect_rqt(struct mlx5e_priv *priv);
+
+int mlx5e_create_indirect_tirs(struct mlx5e_priv *priv);
+void mlx5e_destroy_indirect_tirs(struct mlx5e_priv *priv);
+
 int mlx5e_create_direct_rqts(struct mlx5e_priv *priv);
-void mlx5e_destroy_rqt(struct mlx5e_priv *priv, struct mlx5e_rqt *rqt);
+void mlx5e_destroy_direct_rqts(struct mlx5e_priv *priv);
 int mlx5e_create_direct_tirs(struct mlx5e_priv *priv);
 void mlx5e_destroy_direct_tirs(struct mlx5e_priv *priv);
+void mlx5e_destroy_rqt(struct mlx5e_priv *priv, struct mlx5e_rqt *rqt);
+
+int mlx5e_create_ttc_table(struct mlx5e_priv *priv, u32 underlay_qpn);
+void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv);
+
+int mlx5e_create_tis(struct mlx5_core_dev *mdev, int tc,
+		     u32 underlay_qpn, u32 *tisn);
+void mlx5e_destroy_tis(struct mlx5_core_dev *mdev, u32 tisn);
+
 int mlx5e_create_tises(struct mlx5e_priv *priv);
 void mlx5e_cleanup_nic_tx(struct mlx5e_priv *priv);
 int mlx5e_close(struct net_device *netdev);
 int mlx5e_open(struct net_device *netdev);
 void mlx5e_update_stats_work(struct work_struct *work);
-struct net_device *mlx5e_create_netdev(struct mlx5_core_dev *mdev,
-				       const struct mlx5e_profile *profile,
-				       void *ppriv);
-void mlx5e_destroy_netdev(struct mlx5_core_dev *mdev, struct mlx5e_priv *priv);
-int mlx5e_attach_netdev(struct mlx5_core_dev *mdev, struct net_device *netdev);
-void mlx5e_detach_netdev(struct mlx5_core_dev *mdev, struct net_device *netdev);
 u32 mlx5e_choose_lro_timeout(struct mlx5_core_dev *mdev, u32 wanted_timeout);
 
 int mlx5e_get_offload_stats(int attr_id, const struct net_device *dev,
@@ -1012,5 +1036,16 @@ int mlx5e_get_offload_stats(int attr_id, const struct net_device *dev,
 bool mlx5e_has_offload_stats(const struct net_device *dev, int attr_id);
 
 bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv);
-bool mlx5e_is_vf_vport_rep(struct mlx5e_priv *priv);
+
+/* mlx5e generic netdev management API */
+struct net_device*
+mlx5e_create_netdev(struct mlx5_core_dev *mdev, const struct mlx5e_profile *profile,
+		    void *ppriv);
+int mlx5e_attach_netdev(struct mlx5e_priv *priv);
+void mlx5e_detach_netdev(struct mlx5e_priv *priv);
+void mlx5e_destroy_netdev(struct mlx5e_priv *priv);
+void mlx5e_build_nic_params(struct mlx5_core_dev *mdev,
+			    struct mlx5e_params *params,
+			    u16 max_channels);
+
 #endif /* __MLX5_EN_H__ */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
index c4e9cc79f5c7..c8a005326e30 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c
@@ -321,10 +321,16 @@ static int arfs_create_table(struct mlx5e_priv *priv,
 {
 	struct mlx5e_arfs_tables *arfs = &priv->fs.arfs;
 	struct mlx5e_flow_table *ft = &arfs->arfs_tables[type].ft;
+	struct mlx5_flow_table_attr ft_attr = {};
 	int err;
 
-	ft->t = mlx5_create_flow_table(priv->fs.ns, MLX5E_NIC_PRIO,
-				       MLX5E_ARFS_TABLE_SIZE, MLX5E_ARFS_FT_LEVEL, 0);
+	ft->num_groups = 0;
+
+	ft_attr.max_fte = MLX5E_ARFS_TABLE_SIZE;
+	ft_attr.level = MLX5E_ARFS_FT_LEVEL;
+	ft_attr.prio = MLX5E_NIC_PRIO;
+
+	ft->t = mlx5_create_flow_table(priv->fs.ns, &ft_attr);
 	if (IS_ERR(ft->t)) {
 		err = PTR_ERR(ft->t);
 		ft->t = NULL;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index 5376d69a6b1a..576d6787b484 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -792,7 +792,7 @@ err:
 	return err;
 }
 
-static void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv)
+void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv)
 {
 	struct mlx5e_ttc_table *ttc = &priv->fs.ttc;
 
@@ -800,14 +800,19 @@ static void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv)
 	mlx5e_destroy_flow_table(&ttc->ft);
 }
 
-static int mlx5e_create_ttc_table(struct mlx5e_priv *priv)
+int mlx5e_create_ttc_table(struct mlx5e_priv *priv, u32 underlay_qpn)
 {
 	struct mlx5e_ttc_table *ttc = &priv->fs.ttc;
+	struct mlx5_flow_table_attr ft_attr = {};
 	struct mlx5e_flow_table *ft = &ttc->ft;
 	int err;
 
-	ft->t = mlx5_create_flow_table(priv->fs.ns, MLX5E_NIC_PRIO,
-				       MLX5E_TTC_TABLE_SIZE, MLX5E_TTC_FT_LEVEL, 0);
+	ft_attr.max_fte = MLX5E_TTC_TABLE_SIZE;
+	ft_attr.level = MLX5E_TTC_FT_LEVEL;
+	ft_attr.prio = MLX5E_NIC_PRIO;
+	ft_attr.underlay_qpn = underlay_qpn;
+
+	ft->t = mlx5_create_flow_table(priv->fs.ns, &ft_attr);
 	if (IS_ERR(ft->t)) {
 		err = PTR_ERR(ft->t);
 		ft->t = NULL;
@@ -973,12 +978,16 @@ static int mlx5e_create_l2_table(struct mlx5e_priv *priv)
 {
 	struct mlx5e_l2_table *l2_table = &priv->fs.l2;
 	struct mlx5e_flow_table *ft = &l2_table->ft;
+	struct mlx5_flow_table_attr ft_attr = {};
 	int err;
 
 	ft->num_groups = 0;
-	ft->t = mlx5_create_flow_table(priv->fs.ns, MLX5E_NIC_PRIO,
-				       MLX5E_L2_TABLE_SIZE, MLX5E_L2_FT_LEVEL, 0);
 
+	ft_attr.max_fte = MLX5E_L2_TABLE_SIZE;
+	ft_attr.level = MLX5E_L2_FT_LEVEL;
+	ft_attr.prio = MLX5E_NIC_PRIO;
+
+	ft->t = mlx5_create_flow_table(priv->fs.ns, &ft_attr);
 	if (IS_ERR(ft->t)) {
 		err = PTR_ERR(ft->t);
 		ft->t = NULL;
@@ -1076,11 +1085,16 @@ static int mlx5e_create_vlan_table_groups(struct mlx5e_flow_table *ft)
 static int mlx5e_create_vlan_table(struct mlx5e_priv *priv)
 {
 	struct mlx5e_flow_table *ft = &priv->fs.vlan.ft;
+	struct mlx5_flow_table_attr ft_attr = {};
 	int err;
 
 	ft->num_groups = 0;
-	ft->t = mlx5_create_flow_table(priv->fs.ns, MLX5E_NIC_PRIO,
-				       MLX5E_VLAN_TABLE_SIZE, MLX5E_VLAN_FT_LEVEL, 0);
+
+	ft_attr.max_fte = MLX5E_VLAN_TABLE_SIZE;
+	ft_attr.level = MLX5E_VLAN_FT_LEVEL;
+	ft_attr.prio = MLX5E_NIC_PRIO;
+
+	ft->t = mlx5_create_flow_table(priv->fs.ns, &ft_attr);
 
 	if (IS_ERR(ft->t)) {
 		err = PTR_ERR(ft->t);
@@ -1133,7 +1147,7 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv)
 		priv->netdev->hw_features &= ~NETIF_F_NTUPLE;
 	}
 
-	err = mlx5e_create_ttc_table(priv);
+	err = mlx5e_create_ttc_table(priv, 0);
 	if (err) {
 		netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n",
 			   err);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 8b7b7e604ea0..061b20c73071 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -31,7 +31,6 @@
  */
 
 #include <net/tc_act/tc_gact.h>
-#include <linux/crash_dump.h>
 #include <net/pkt_cls.h>
 #include <linux/mlx5/fs.h>
 #include <net/vxlan.h>
@@ -586,15 +585,17 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c,
 
 	switch (rq->wq_type) {
 	case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ:
-		if (mlx5e_is_vf_vport_rep(c->priv)) {
-			err = -EINVAL;
-			goto err_rq_wq_destroy;
-		}
 
-		rq->handle_rx_cqe = mlx5e_handle_rx_cqe_mpwrq;
 		rq->alloc_wqe = mlx5e_alloc_rx_mpwqe;
 		rq->dealloc_wqe = mlx5e_dealloc_rx_mpwqe;
 
+		rq->handle_rx_cqe = c->priv->profile->rx_handlers.handle_rx_cqe_mpwqe;
+		if (!rq->handle_rx_cqe) {
+			err = -EINVAL;
+			netdev_err(c->netdev, "RX handler of MPWQE RQ is not set, err %d\n", err);
+			goto err_rq_wq_destroy;
+		}
+
 		rq->mpwqe_stride_sz = BIT(params->mpwqe_log_stride_sz);
 		rq->mpwqe_num_strides = BIT(params->mpwqe_log_num_strides);
 
@@ -617,15 +618,17 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c,
 			err = -ENOMEM;
 			goto err_rq_wq_destroy;
 		}
-
-		if (mlx5e_is_vf_vport_rep(c->priv))
-			rq->handle_rx_cqe = mlx5e_handle_rx_cqe_rep;
-		else
-			rq->handle_rx_cqe = mlx5e_handle_rx_cqe;
-
 		rq->alloc_wqe = mlx5e_alloc_rx_wqe;
 		rq->dealloc_wqe = mlx5e_dealloc_rx_wqe;
 
+		rq->handle_rx_cqe = c->priv->profile->rx_handlers.handle_rx_cqe;
+		if (!rq->handle_rx_cqe) {
+			kfree(rq->dma_info);
+			err = -EINVAL;
+			netdev_err(c->netdev, "RX handler of RQ is not set, err %d\n", err);
+			goto err_rq_wq_destroy;
+		}
+
 		rq->buff.wqe_sz = params->lro_en  ?
 				params->lro_wqe_sz :
 				MLX5E_SW2HW_MTU(c->netdev->mtu);
@@ -1043,7 +1046,6 @@ static void mlx5e_free_txqsq_db(struct mlx5e_txqsq *sq)
 {
 	kfree(sq->db.wqe_info);
 	kfree(sq->db.dma_fifo);
-	kfree(sq->db.skb);
 }
 
 static int mlx5e_alloc_txqsq_db(struct mlx5e_txqsq *sq, int numa)
@@ -1051,13 +1053,11 @@ static int mlx5e_alloc_txqsq_db(struct mlx5e_txqsq *sq, int numa)
 	int wq_sz = mlx5_wq_cyc_get_size(&sq->wq);
 	int df_sz = wq_sz * MLX5_SEND_WQEBB_NUM_DS;
 
-	sq->db.skb = kzalloc_node(wq_sz * sizeof(*sq->db.skb),
-				      GFP_KERNEL, numa);
 	sq->db.dma_fifo = kzalloc_node(df_sz * sizeof(*sq->db.dma_fifo),
 					   GFP_KERNEL, numa);
 	sq->db.wqe_info = kzalloc_node(wq_sz * sizeof(*sq->db.wqe_info),
 					   GFP_KERNEL, numa);
-	if (!sq->db.skb || !sq->db.dma_fifo || !sq->db.wqe_info) {
+	if (!sq->db.dma_fifo || !sq->db.wqe_info) {
 		mlx5e_free_txqsq_db(sq);
 		return -ENOMEM;
 	}
@@ -1296,7 +1296,7 @@ static void mlx5e_deactivate_txqsq(struct mlx5e_txqsq *sq)
 	if (mlx5e_wqc_has_room_for(&sq->wq, sq->cc, sq->pc, 1)) {
 		struct mlx5e_tx_wqe *nop;
 
-		sq->db.skb[(sq->pc & sq->wq.sz_m1)] = NULL;
+		sq->db.wqe_info[(sq->pc & sq->wq.sz_m1)].skb = NULL;
 		nop = mlx5e_post_nop(&sq->wq, sq->sqn, &sq->pc);
 		mlx5e_notify_hw(&sq->wq, sq->pc, sq->uar_map, &nop->ctrl);
 	}
@@ -1710,14 +1710,6 @@ static int mlx5e_set_tx_maxrate(struct net_device *dev, int index, u32 rate)
 	return err;
 }
 
-static inline int mlx5e_get_max_num_channels(struct mlx5_core_dev *mdev)
-{
-	return is_kdump_kernel() ?
-		MLX5E_MIN_NUM_CHANNELS :
-		min_t(int, mdev->priv.eq_table.num_comp_vectors,
-		      MLX5E_MAX_NUM_CHANNELS);
-}
-
 static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix,
 			      struct mlx5e_params *params,
 			      struct mlx5e_channel_param *cparam,
@@ -2124,11 +2116,15 @@ void mlx5e_destroy_rqt(struct mlx5e_priv *priv, struct mlx5e_rqt *rqt)
 	mlx5_core_destroy_rqt(priv->mdev, rqt->rqtn);
 }
 
-static int mlx5e_create_indirect_rqts(struct mlx5e_priv *priv)
+int mlx5e_create_indirect_rqt(struct mlx5e_priv *priv)
 {
 	struct mlx5e_rqt *rqt = &priv->indir_rqt;
+	int err;
 
-	return mlx5e_create_rqt(priv, MLX5E_INDIR_RQT_SIZE, rqt);
+	err = mlx5e_create_rqt(priv, MLX5E_INDIR_RQT_SIZE, rqt);
+	if (err)
+		mlx5_core_warn(priv->mdev, "create indirect rqts failed, %d\n", err);
+	return err;
 }
 
 int mlx5e_create_direct_rqts(struct mlx5e_priv *priv)
@@ -2147,12 +2143,21 @@ int mlx5e_create_direct_rqts(struct mlx5e_priv *priv)
 	return 0;
 
 err_destroy_rqts:
+	mlx5_core_warn(priv->mdev, "create direct rqts failed, %d\n", err);
 	for (ix--; ix >= 0; ix--)
 		mlx5e_destroy_rqt(priv, &priv->direct_tir[ix].rqt);
 
 	return err;
 }
 
+void mlx5e_destroy_direct_rqts(struct mlx5e_priv *priv)
+{
+	int i;
+
+	for (i = 0; i < priv->profile->max_nch(priv->mdev); i++)
+		mlx5e_destroy_rqt(priv, &priv->direct_tir[i].rqt);
+}
+
 static int mlx5e_rx_hash_fn(int hfunc)
 {
 	return (hfunc == ETH_RSS_HASH_TOP) ?
@@ -2543,7 +2548,13 @@ static void mlx5e_build_channels_tx_maps(struct mlx5e_priv *priv)
 	}
 }
 
-static void mlx5e_activate_priv_channels(struct mlx5e_priv *priv)
+static bool mlx5e_is_eswitch_vport_mngr(struct mlx5_core_dev *mdev)
+{
+	return (MLX5_CAP_GEN(mdev, vport_group_manager) &&
+		MLX5_CAP_GEN(mdev, port_type) == MLX5_CAP_PORT_TYPE_ETH);
+}
+
+void mlx5e_activate_priv_channels(struct mlx5e_priv *priv)
 {
 	int num_txqs = priv->channels.num * priv->channels.params.num_tc;
 	struct net_device *netdev = priv->netdev;
@@ -2556,18 +2567,18 @@ static void mlx5e_activate_priv_channels(struct mlx5e_priv *priv)
 	mlx5e_activate_channels(&priv->channels);
 	netif_tx_start_all_queues(priv->netdev);
 
-	if (MLX5_CAP_GEN(priv->mdev, vport_group_manager))
+	if (mlx5e_is_eswitch_vport_mngr(priv->mdev))
 		mlx5e_add_sqs_fwd_rules(priv);
 
 	mlx5e_wait_channels_min_rx_wqes(&priv->channels);
 	mlx5e_redirect_rqts_to_channels(priv, &priv->channels);
 }
 
-static void mlx5e_deactivate_priv_channels(struct mlx5e_priv *priv)
+void mlx5e_deactivate_priv_channels(struct mlx5e_priv *priv)
 {
 	mlx5e_redirect_rqts_to_drop(priv);
 
-	if (MLX5_CAP_GEN(priv->mdev, vport_group_manager))
+	if (mlx5e_is_eswitch_vport_mngr(priv->mdev))
 		mlx5e_remove_sqs_fwd_rules(priv);
 
 	/* FIXME: This is a W/A only for tx timeout watch dog false alarm when
@@ -2755,24 +2766,25 @@ static void mlx5e_close_drop_rq(struct mlx5e_rq *drop_rq)
 	mlx5e_free_cq(&drop_rq->cq);
 }
 
-static int mlx5e_create_tis(struct mlx5e_priv *priv, int tc)
+int mlx5e_create_tis(struct mlx5_core_dev *mdev, int tc,
+		     u32 underlay_qpn, u32 *tisn)
 {
-	struct mlx5_core_dev *mdev = priv->mdev;
 	u32 in[MLX5_ST_SZ_DW(create_tis_in)] = {0};
 	void *tisc = MLX5_ADDR_OF(create_tis_in, in, ctx);
 
 	MLX5_SET(tisc, tisc, prio, tc << 1);
+	MLX5_SET(tisc, tisc, underlay_qpn, underlay_qpn);
 	MLX5_SET(tisc, tisc, transport_domain, mdev->mlx5e_res.td.tdn);
 
 	if (mlx5_lag_is_lacp_owner(mdev))
 		MLX5_SET(tisc, tisc, strict_lag_tx_port_affinity, 1);
 
-	return mlx5_core_create_tis(mdev, in, sizeof(in), &priv->tisn[tc]);
+	return mlx5_core_create_tis(mdev, in, sizeof(in), tisn);
 }
 
-static void mlx5e_destroy_tis(struct mlx5e_priv *priv, int tc)
+void mlx5e_destroy_tis(struct mlx5_core_dev *mdev, u32 tisn)
 {
-	mlx5_core_destroy_tis(priv->mdev, priv->tisn[tc]);
+	mlx5_core_destroy_tis(mdev, tisn);
 }
 
 int mlx5e_create_tises(struct mlx5e_priv *priv)
@@ -2781,7 +2793,7 @@ int mlx5e_create_tises(struct mlx5e_priv *priv)
 	int tc;
 
 	for (tc = 0; tc < priv->profile->max_tc; tc++) {
-		err = mlx5e_create_tis(priv, tc);
+		err = mlx5e_create_tis(priv->mdev, tc, 0, &priv->tisn[tc]);
 		if (err)
 			goto err_close_tises;
 	}
@@ -2790,7 +2802,7 @@ int mlx5e_create_tises(struct mlx5e_priv *priv)
 
 err_close_tises:
 	for (tc--; tc >= 0; tc--)
-		mlx5e_destroy_tis(priv, tc);
+		mlx5e_destroy_tis(priv->mdev, priv->tisn[tc]);
 
 	return err;
 }
@@ -2800,7 +2812,7 @@ void mlx5e_cleanup_nic_tx(struct mlx5e_priv *priv)
 	int tc;
 
 	for (tc = 0; tc < priv->profile->max_tc; tc++)
-		mlx5e_destroy_tis(priv, tc);
+		mlx5e_destroy_tis(priv->mdev, priv->tisn[tc]);
 }
 
 static void mlx5e_build_indir_tir_ctx(struct mlx5e_priv *priv,
@@ -2827,7 +2839,7 @@ static void mlx5e_build_direct_tir_ctx(struct mlx5e_priv *priv, u32 rqtn, u32 *t
 	MLX5_SET(tirc, tirc, rx_hash_fn, MLX5_RX_HASH_FN_INVERTED_XOR8);
 }
 
-static int mlx5e_create_indirect_tirs(struct mlx5e_priv *priv)
+int mlx5e_create_indirect_tirs(struct mlx5e_priv *priv)
 {
 	struct mlx5e_tir *tir;
 	void *tirc;
@@ -2856,6 +2868,7 @@ static int mlx5e_create_indirect_tirs(struct mlx5e_priv *priv)
 	return 0;
 
 err_destroy_tirs:
+	mlx5_core_warn(priv->mdev, "create indirect tirs failed, %d\n", err);
 	for (tt--; tt >= 0; tt--)
 		mlx5e_destroy_tir(priv->mdev, &priv->indir_tir[tt]);
 
@@ -2894,6 +2907,7 @@ int mlx5e_create_direct_tirs(struct mlx5e_priv *priv)
 	return 0;
 
 err_destroy_ch_tirs:
+	mlx5_core_warn(priv->mdev, "create direct tirs failed, %d\n", err);
 	for (ix--; ix >= 0; ix--)
 		mlx5e_destroy_tir(priv->mdev, &priv->direct_tir[ix]);
 
@@ -2902,7 +2916,7 @@ err_destroy_ch_tirs:
 	return err;
 }
 
-static void mlx5e_destroy_indirect_tirs(struct mlx5e_priv *priv)
+void mlx5e_destroy_indirect_tirs(struct mlx5e_priv *priv)
 {
 	int i;
 
@@ -3803,9 +3817,9 @@ u32 mlx5e_choose_lro_timeout(struct mlx5_core_dev *mdev, u32 wanted_timeout)
 	return MLX5_CAP_ETH(mdev, lro_timer_supported_periods[i]);
 }
 
-static void mlx5e_build_nic_params(struct mlx5_core_dev *mdev,
-				   struct mlx5e_params *params,
-				   u16 max_channels)
+void mlx5e_build_nic_params(struct mlx5_core_dev *mdev,
+			    struct mlx5e_params *params,
+			    u16 max_channels)
 {
 	u8 cq_period_mode = 0;
 	u32 link_speed = 0;
@@ -3835,6 +3849,7 @@ static void mlx5e_build_nic_params(struct mlx5_core_dev *mdev,
 	mlx5e_set_rq_params(mdev, params);
 
 	/* HW LRO */
+	/* TODO: && MLX5_CAP_ETH(mdev, lro_cap) */
 	if (params->rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ)
 		params->lro_en = true;
 	params->lro_timeout = mlx5e_choose_lro_timeout(mdev, MLX5E_DEFAULT_LRO_TIMEOUT);
@@ -4040,31 +4055,22 @@ static int mlx5e_init_nic_rx(struct mlx5e_priv *priv)
 {
 	struct mlx5_core_dev *mdev = priv->mdev;
 	int err;
-	int i;
 
-	err = mlx5e_create_indirect_rqts(priv);
-	if (err) {
-		mlx5_core_warn(mdev, "create indirect rqts failed, %d\n", err);
+	err = mlx5e_create_indirect_rqt(priv);
+	if (err)
 		return err;
-	}
 
 	err = mlx5e_create_direct_rqts(priv);
-	if (err) {
-		mlx5_core_warn(mdev, "create direct rqts failed, %d\n", err);
+	if (err)
 		goto err_destroy_indirect_rqts;
-	}
 
 	err = mlx5e_create_indirect_tirs(priv);
-	if (err) {
-		mlx5_core_warn(mdev, "create indirect tirs failed, %d\n", err);
+	if (err)
 		goto err_destroy_direct_rqts;
-	}
 
 	err = mlx5e_create_direct_tirs(priv);
-	if (err) {
-		mlx5_core_warn(mdev, "create direct tirs failed, %d\n", err);
+	if (err)
 		goto err_destroy_indirect_tirs;
-	}
 
 	err = mlx5e_create_flow_steering(priv);
 	if (err) {
@@ -4085,8 +4091,7 @@ err_destroy_direct_tirs:
 err_destroy_indirect_tirs:
 	mlx5e_destroy_indirect_tirs(priv);
 err_destroy_direct_rqts:
-	for (i = 0; i < priv->profile->max_nch(mdev); i++)
-		mlx5e_destroy_rqt(priv, &priv->direct_tir[i].rqt);
+	mlx5e_destroy_direct_rqts(priv);
 err_destroy_indirect_rqts:
 	mlx5e_destroy_rqt(priv, &priv->indir_rqt);
 	return err;
@@ -4094,14 +4099,11 @@ err_destroy_indirect_rqts:
 
 static void mlx5e_cleanup_nic_rx(struct mlx5e_priv *priv)
 {
-	int i;
-
 	mlx5e_tc_cleanup(priv);
 	mlx5e_destroy_flow_steering(priv);
 	mlx5e_destroy_direct_tirs(priv);
 	mlx5e_destroy_indirect_tirs(priv);
-	for (i = 0; i < priv->profile->max_nch(priv->mdev); i++)
-		mlx5e_destroy_rqt(priv, &priv->direct_tir[i].rqt);
+	mlx5e_destroy_direct_rqts(priv);
 	mlx5e_destroy_rqt(priv, &priv->indir_rqt);
 }
 
@@ -4121,12 +4123,57 @@ static int mlx5e_init_nic_tx(struct mlx5e_priv *priv)
 	return 0;
 }
 
+static void mlx5e_register_vport_rep(struct mlx5_core_dev *mdev)
+{
+	struct mlx5_eswitch *esw = mdev->priv.eswitch;
+	int total_vfs = MLX5_TOTAL_VPORTS(mdev);
+	int vport;
+	u8 mac[ETH_ALEN];
+
+	if (!MLX5_CAP_GEN(mdev, vport_group_manager))
+		return;
+
+	mlx5_query_nic_vport_mac_address(mdev, 0, mac);
+
+	for (vport = 1; vport < total_vfs; vport++) {
+		struct mlx5_eswitch_rep rep;
+
+		rep.load = mlx5e_vport_rep_load;
+		rep.unload = mlx5e_vport_rep_unload;
+		rep.vport = vport;
+		ether_addr_copy(rep.hw_id, mac);
+		mlx5_eswitch_register_vport_rep(esw, vport, &rep);
+	}
+}
+
+static void mlx5e_unregister_vport_rep(struct mlx5_core_dev *mdev)
+{
+	struct mlx5_eswitch *esw = mdev->priv.eswitch;
+	int total_vfs = MLX5_TOTAL_VPORTS(mdev);
+	int vport;
+
+	if (!MLX5_CAP_GEN(mdev, vport_group_manager))
+		return;
+
+	for (vport = 1; vport < total_vfs; vport++)
+		mlx5_eswitch_unregister_vport_rep(esw, vport);
+}
+
 static void mlx5e_nic_enable(struct mlx5e_priv *priv)
 {
 	struct net_device *netdev = priv->netdev;
 	struct mlx5_core_dev *mdev = priv->mdev;
 	struct mlx5_eswitch *esw = mdev->priv.eswitch;
 	struct mlx5_eswitch_rep rep;
+	u16 max_mtu;
+
+	mlx5e_init_l2_addr(priv);
+
+	/* MTU range: 68 - hw-specific max */
+	netdev->min_mtu = ETH_MIN_MTU;
+	mlx5_query_port_max_mtu(priv->mdev, &max_mtu, 1);
+	netdev->max_mtu = MLX5E_HW2SW_MTU(max_mtu);
+	mlx5e_set_dev_port_mtu(priv);
 
 	mlx5_lag_add(mdev, netdev);
 
@@ -4141,6 +4188,8 @@ static void mlx5e_nic_enable(struct mlx5e_priv *priv)
 		mlx5_eswitch_register_vport_rep(esw, 0, &rep);
 	}
 
+	mlx5e_register_vport_rep(mdev);
+
 	if (netdev->reg_state != NETREG_REGISTERED)
 		return;
 
@@ -4152,6 +4201,12 @@ static void mlx5e_nic_enable(struct mlx5e_priv *priv)
 	}
 
 	queue_work(priv->wq, &priv->set_rx_mode_work);
+
+	rtnl_lock();
+	if (netif_running(netdev))
+		mlx5e_open(netdev);
+	netif_device_attach(netdev);
+	rtnl_unlock();
 }
 
 static void mlx5e_nic_disable(struct mlx5e_priv *priv)
@@ -4159,7 +4214,14 @@ static void mlx5e_nic_disable(struct mlx5e_priv *priv)
 	struct mlx5_core_dev *mdev = priv->mdev;
 	struct mlx5_eswitch *esw = mdev->priv.eswitch;
 
+	rtnl_lock();
+	if (netif_running(priv->netdev))
+		mlx5e_close(priv->netdev);
+	netif_device_detach(priv->netdev);
+	rtnl_unlock();
+
 	queue_work(priv->wq, &priv->set_rx_mode_work);
+	mlx5e_unregister_vport_rep(mdev);
 	if (MLX5_CAP_GEN(mdev, vport_group_manager))
 		mlx5_eswitch_unregister_vport_rep(esw, 0);
 	mlx5e_disable_async_events(priv);
@@ -4177,9 +4239,13 @@ static const struct mlx5e_profile mlx5e_nic_profile = {
 	.disable	   = mlx5e_nic_disable,
 	.update_stats	   = mlx5e_update_stats,
 	.max_nch	   = mlx5e_get_max_num_channels,
+	.rx_handlers.handle_rx_cqe       = mlx5e_handle_rx_cqe,
+	.rx_handlers.handle_rx_cqe_mpwqe = mlx5e_handle_rx_cqe_mpwrq,
 	.max_tc		   = MLX5E_MAX_NUM_TC,
 };
 
+/* mlx5e generic netdev management API (move to en_common.c) */
+
 struct net_device *mlx5e_create_netdev(struct mlx5_core_dev *mdev,
 				       const struct mlx5e_profile *profile,
 				       void *ppriv)
@@ -4219,14 +4285,12 @@ err_cleanup_nic:
 	return NULL;
 }
 
-int mlx5e_attach_netdev(struct mlx5_core_dev *mdev, struct net_device *netdev)
+int mlx5e_attach_netdev(struct mlx5e_priv *priv)
 {
+	struct mlx5_core_dev *mdev = priv->mdev;
 	const struct mlx5e_profile *profile;
-	struct mlx5e_priv *priv;
-	u16 max_mtu;
 	int err;
 
-	priv = netdev_priv(netdev);
 	profile = priv->profile;
 	clear_bit(MLX5E_STATE_DESTROYING, &priv->state);
 
@@ -4246,24 +4310,9 @@ int mlx5e_attach_netdev(struct mlx5_core_dev *mdev, struct net_device *netdev)
 
 	mlx5e_create_q_counter(priv);
 
-	mlx5e_init_l2_addr(priv);
-
-	/* MTU range: 68 - hw-specific max */
-	netdev->min_mtu = ETH_MIN_MTU;
-	mlx5_query_port_max_mtu(priv->mdev, &max_mtu, 1);
-	netdev->max_mtu = MLX5E_HW2SW_MTU(max_mtu);
-
-	mlx5e_set_dev_port_mtu(priv);
-
 	if (profile->enable)
 		profile->enable(priv);
 
-	rtnl_lock();
-	if (netif_running(netdev))
-		mlx5e_open(netdev);
-	netif_device_attach(netdev);
-	rtnl_unlock();
-
 	return 0;
 
 err_close_drop_rq:
@@ -4276,55 +4325,12 @@ out:
 	return err;
 }
 
-static void mlx5e_register_vport_rep(struct mlx5_core_dev *mdev)
+void mlx5e_detach_netdev(struct mlx5e_priv *priv)
 {
-	struct mlx5_eswitch *esw = mdev->priv.eswitch;
-	int total_vfs = MLX5_TOTAL_VPORTS(mdev);
-	int vport;
-	u8 mac[ETH_ALEN];