[PATCH v7 10/12] net/nfp: add flower representor framework

[Chaoyong He <chaoyong.he@corigine.com>]

Adds the framework to support flower representors. The number of VF
representors are parsed from the command line. For physical port
representors the current logic aims to create a representor for
each physical port present on the hardware.

An eth_dev is created for each physical port and VF, and flower
firmware requires a MAC repr cmsg to be transmitted to firmware
with info about the number of physical ports configured.

Reify messages are sent to hardware for each physical port representor.
An rte_ring is also created per representor so that traffic can be
pushed and pulled to this interface.

To up and down the real device represented by a flower representor port
a port mod message is used to convey that info to the firmware. This
message will be used in the dev_ops callbacks of flower representors.

Each cmsg generated by the driver is prepended with a cmsg header.
This commit also adds the logic to fill in the header of cmsgs.

Also add the Rx and Tx path for flower representors. For Rx packets are
dequeued from the representor ring and passed to the eth_dev. For Tx
the first queue of the PF vNIC is used. Metadata about the representor
is added before the packet is sent down to firmware.

Signed-off-by: Chaoyong He <chaoyong.he@corigine.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund@corigine.com>
---
 drivers/net/nfp/flower/nfp_flower.c             |  59 +++
 drivers/net/nfp/flower/nfp_flower.h             |  18 +
 drivers/net/nfp/flower/nfp_flower_cmsg.c        | 186 ++++++++
 drivers/net/nfp/flower/nfp_flower_cmsg.h        | 173 +++++++
 drivers/net/nfp/flower/nfp_flower_representor.c | 602 ++++++++++++++++++++++++
 drivers/net/nfp/flower/nfp_flower_representor.h |  39 ++
 drivers/net/nfp/meson.build                     |   2 +
 7 files changed, 1079 insertions(+)
 create mode 100644 drivers/net/nfp/flower/nfp_flower_cmsg.c
 create mode 100644 drivers/net/nfp/flower/nfp_flower_cmsg.h
 create mode 100644 drivers/net/nfp/flower/nfp_flower_representor.c
 create mode 100644 drivers/net/nfp/flower/nfp_flower_representor.h

diff --git a/drivers/net/nfp/flower/nfp_flower.c b/drivers/net/nfp/flower/nfp_flower.c
index de91fb4..8af18d0 100644
--- a/drivers/net/nfp/flower/nfp_flower.c
+++ b/drivers/net/nfp/flower/nfp_flower.c
@@ -22,6 +22,7 @@
 #include "nfp_flower.h"
 #include "nfp_flower_ovs_compat.h"
 #include "nfp_flower_ctrl.h"
+#include "nfp_flower_representor.h"
 
 #define MAX_PKT_BURST 32
 #define MEMPOOL_CACHE_SIZE 512
@@ -720,8 +721,13 @@
 	unsigned int numa_node;
 	struct nfp_net_hw *pf_hw;
 	struct nfp_net_hw *ctrl_hw;
+	struct rte_pci_device *pci_dev;
 	struct nfp_app_flower *app_flower;
+	struct rte_eth_devargs eth_da = {
+		.nb_representor_ports = 0
+	};
 
+	pci_dev = pf_dev->pci_dev;
 	numa_node = rte_socket_id();
 
 	/* Allocate memory for the Flower app */
@@ -815,6 +821,59 @@
 		goto ctrl_vnic_cleanup;
 	}
 
+	/* Allocate a switch domain for the flower app */
+	if (app_flower->switch_domain_id == RTE_ETH_DEV_SWITCH_DOMAIN_ID_INVALID &&
+			rte_eth_switch_domain_alloc(&app_flower->switch_domain_id)) {
+		PMD_INIT_LOG(WARNING,
+				"failed to allocate switch domain for device");
+	}
+
+	/* Now parse PCI device args passed for representor info */
+	if (pci_dev->device.devargs != NULL) {
+		ret = rte_eth_devargs_parse(pci_dev->device.devargs->args,
+				&eth_da);
+		if (ret != 0) {
+			PMD_INIT_LOG(ERR, "devarg parse failed");
+			goto ctrl_vnic_cleanup;
+		}
+	}
+
+	if (eth_da.nb_representor_ports == 0) {
+		PMD_INIT_LOG(DEBUG, "No representor port need to create.");
+		ret = 0;
+		goto done;
+	}
+
+	/* There always exist phy repr */
+	if (eth_da.nb_representor_ports < app_flower->nfp_eth_table->count) {
+		PMD_INIT_LOG(DEBUG, "Should also create phy representor port.");
+		ret = -ERANGE;
+		goto ctrl_vnic_cleanup;
+	}
+
+	/* Only support VF representor creation via the command line */
+	if (eth_da.type != RTE_ETH_REPRESENTOR_VF) {
+		PMD_INIT_LOG(ERR, "Unsupported representor type: %s",
+				pci_dev->device.devargs->args);
+		ret = -ENOTSUP;
+		goto ctrl_vnic_cleanup;
+	}
+
+	/* Fill in flower app with repr counts */
+	app_flower->num_phyport_reprs = (uint8_t)app_flower->nfp_eth_table->count;
+	app_flower->num_vf_reprs = eth_da.nb_representor_ports -
+			app_flower->nfp_eth_table->count;
+
+	PMD_INIT_LOG(INFO, "%d number of VF reprs", app_flower->num_vf_reprs);
+	PMD_INIT_LOG(INFO, "%d number of phyport reprs", app_flower->num_phyport_reprs);
+
+	ret = nfp_flower_repr_alloc(app_flower);
+	if (ret != 0) {
+		PMD_INIT_LOG(ERR,
+			"representors allocation for NFP_REPR_TYPE_VF error");
+		goto ctrl_vnic_cleanup;
+	}
+
 	return 0;
 
 ctrl_vnic_cleanup:
diff --git a/drivers/net/nfp/flower/nfp_flower.h b/drivers/net/nfp/flower/nfp_flower.h
index bdc64e3..24fced3 100644
--- a/drivers/net/nfp/flower/nfp_flower.h
+++ b/drivers/net/nfp/flower/nfp_flower.h
@@ -19,8 +19,20 @@ enum nfp_flower_service {
  */
 #define FLOWER_PKT_DATA_OFFSET 8
 
+#define MAX_FLOWER_PHYPORTS 8
+#define MAX_FLOWER_VFS 64
+
 /* The flower application's private structure */
 struct nfp_app_flower {
+	/* switch domain for this app */
+	uint16_t switch_domain_id;
+
+	/* Number of VF representors */
+	uint8_t num_vf_reprs;
+
+	/* Number of phyport representors */
+	uint8_t num_phyport_reprs;
+
 	/* List of rte_service ID's for the flower app */
 	uint32_t flower_services_ids[NFP_FLOWER_SERVICE_MAX];
 
@@ -44,6 +56,12 @@ struct nfp_app_flower {
 
 	/* Ctrl vNIC Tx counter */
 	uint64_t ctrl_vnic_tx_count;
+
+	/* Array of phyport representors */
+	struct nfp_flower_representor *phy_reprs[MAX_FLOWER_PHYPORTS];
+
+	/* Array of VF representors */
+	struct nfp_flower_representor *vf_reprs[MAX_FLOWER_VFS];
 };
 
 int nfp_init_app_flower(struct nfp_pf_dev *pf_dev);
diff --git a/drivers/net/nfp/flower/nfp_flower_cmsg.c b/drivers/net/nfp/flower/nfp_flower_cmsg.c
new file mode 100644
index 0000000..5ce547c
--- /dev/null
+++ b/drivers/net/nfp/flower/nfp_flower_cmsg.c
@@ -0,0 +1,186 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2022 Corigine, Inc.
+ * All rights reserved.
+ */
+
+#include "../nfpcore/nfp_nsp.h"
+#include "../nfp_logs.h"
+#include "../nfp_common.h"
+#include "nfp_flower.h"
+#include "nfp_flower_cmsg.h"
+#include "nfp_flower_ctrl.h"
+#include "nfp_flower_representor.h"
+
+static void *
+nfp_flower_cmsg_init(struct rte_mbuf *m,
+		enum nfp_flower_cmsg_type type,
+		uint32_t size)
+{
+	char *pkt;
+	uint32_t data;
+	uint32_t new_size = size;
+	struct nfp_flower_cmsg_hdr *hdr;
+
+	pkt = rte_pktmbuf_mtod(m, char *);
+	PMD_DRV_LOG(DEBUG, "flower_cmsg_init using pkt at %p", pkt);
+
+	data = rte_cpu_to_be_32(NFP_NET_META_PORTID);
+	rte_memcpy(pkt, &data, 4);
+	pkt += 4;
+	new_size += 4;
+
+	/* First the metadata as flower requires it */
+	data = rte_cpu_to_be_32(NFP_META_PORT_ID_CTRL);
+	rte_memcpy(pkt, &data, 4);
+	pkt += 4;
+	new_size += 4;
+
+	/* Now the ctrl header */
+	hdr = (struct nfp_flower_cmsg_hdr *)pkt;
+	hdr->pad     = 0;
+	hdr->type    = type;
+	hdr->version = NFP_FLOWER_CMSG_VER1;
+
+	pkt = (char *)hdr + NFP_FLOWER_CMSG_HLEN;
+	new_size += NFP_FLOWER_CMSG_HLEN;
+
+	m->pkt_len = new_size;
+	m->data_len = m->pkt_len;
+
+	return pkt;
+}
+
+static void
+nfp_flower_cmsg_mac_repr_init(struct rte_mbuf *m, int num_ports)
+{
+	uint32_t size;
+	struct nfp_flower_cmsg_mac_repr *msg;
+	enum nfp_flower_cmsg_type type = NFP_FLOWER_CMSG_TYPE_MAC_REPR;
+
+	size = sizeof(*msg) + (num_ports * sizeof(msg->ports[0]));
+	PMD_DRV_LOG(DEBUG, "mac repr cmsg init with size: %u", size);
+	msg = (struct nfp_flower_cmsg_mac_repr *)nfp_flower_cmsg_init(m,
+			type, size);
+
+	memset(msg->reserved, 0, sizeof(msg->reserved));
+	msg->num_ports = num_ports;
+}
+
+static void
+nfp_flower_cmsg_mac_repr_fill(struct rte_mbuf *m,
+		unsigned int idx,
+		unsigned int nbi,
+		unsigned int nbi_port,
+		unsigned int phys_port)
+{
+	struct nfp_flower_cmsg_mac_repr *msg;
+
+	msg = (struct nfp_flower_cmsg_mac_repr *)nfp_flower_cmsg_get_data(m);
+	msg->ports[idx].idx       = idx;
+	msg->ports[idx].info      = nbi & NFP_FLOWER_CMSG_MAC_REPR_NBI;
+	msg->ports[idx].nbi_port  = nbi_port;
+	msg->ports[idx].phys_port = phys_port;
+}
+
+int
+nfp_flower_cmsg_mac_repr(struct nfp_app_flower *app_flower)
+{
+	int i;
+	uint16_t cnt;
+	unsigned int nbi;
+	unsigned int nbi_port;
+	unsigned int phys_port;
+	struct rte_mbuf *mbuf;
+	struct nfp_eth_table *nfp_eth_table;
+
+	mbuf = rte_pktmbuf_alloc(app_flower->ctrl_pktmbuf_pool);
+	if (mbuf == NULL) {
+		PMD_DRV_LOG(ERR, "Could not allocate mac repr cmsg");
+		return -ENOMEM;
+	}
+
+	nfp_flower_cmsg_mac_repr_init(mbuf, app_flower->num_phyport_reprs);
+
+	/* Fill in the mac repr cmsg */
+	nfp_eth_table = app_flower->nfp_eth_table;
+	for (i = 0; i < app_flower->num_phyport_reprs; i++) {
+		nbi = nfp_eth_table->ports[i].nbi;
+		nbi_port = nfp_eth_table->ports[i].base;
+		phys_port = nfp_eth_table->ports[i].index;
+
+		nfp_flower_cmsg_mac_repr_fill(mbuf, i, nbi, nbi_port, phys_port);
+	}
+
+	/* Send the cmsg via the ctrl vNIC */
+	cnt = nfp_flower_ctrl_vnic_xmit(app_flower, mbuf);
+	if (cnt == 0) {
+		PMD_DRV_LOG(ERR, "Send cmsg through ctrl vnic failed.");
+		rte_pktmbuf_free(mbuf);
+		return -EIO;
+	}
+
+	return 0;
+}
+
+int
+nfp_flower_cmsg_repr_reify(struct nfp_app_flower *app_flower,
+		struct nfp_flower_representor *repr)
+{
+	uint16_t cnt;
+	struct rte_mbuf *mbuf;
+	struct nfp_flower_cmsg_port_reify *msg;
+
+	mbuf = rte_pktmbuf_alloc(app_flower->ctrl_pktmbuf_pool);
+	if (mbuf == NULL) {
+		PMD_DRV_LOG(DEBUG, "alloc mbuf for repr reify failed");
+		return -ENOMEM;
+	}
+
+	msg = (struct nfp_flower_cmsg_port_reify *)nfp_flower_cmsg_init(mbuf,
+			NFP_FLOWER_CMSG_TYPE_PORT_REIFY, sizeof(*msg));
+
+	msg->portnum  = rte_cpu_to_be_32(repr->port_id);
+	msg->reserved = 0;
+	msg->info     = rte_cpu_to_be_16(1);
+
+	cnt = nfp_flower_ctrl_vnic_xmit(app_flower, mbuf);
+	if (cnt == 0) {
+		PMD_DRV_LOG(ERR, "Send cmsg through ctrl vnic failed.");
+		rte_pktmbuf_free(mbuf);
+		return -EIO;
+	}
+
+	return 0;
+}
+
+int
+nfp_flower_cmsg_port_mod(struct nfp_app_flower *app_flower,
+		uint32_t port_id, bool carrier_ok)
+{
+	uint16_t cnt;
+	struct rte_mbuf *mbuf;
+	struct nfp_flower_cmsg_port_mod *msg;
+
+	mbuf = rte_pktmbuf_alloc(app_flower->ctrl_pktmbuf_pool);
+	if (mbuf == NULL) {
+		PMD_DRV_LOG(DEBUG, "alloc mbuf for repr portmod failed");
+		return -ENOMEM;
+	}
+
+	msg = (struct nfp_flower_cmsg_port_mod *)nfp_flower_cmsg_init(mbuf,
+			NFP_FLOWER_CMSG_TYPE_PORT_MOD, sizeof(*msg));
+
+	msg->portnum  = rte_cpu_to_be_32(port_id);
+	msg->reserved = 0;
+	msg->info     = carrier_ok;
+	msg->mtu      = 9000;
+
+	cnt = nfp_flower_ctrl_vnic_xmit(app_flower, mbuf);
+	if (cnt == 0) {
+		PMD_DRV_LOG(ERR, "Send cmsg through ctrl vnic failed.");
+		rte_pktmbuf_free(mbuf);
+		return -EIO;
+	}
+
+	return 0;
+}
diff --git a/drivers/net/nfp/flower/nfp_flower_cmsg.h b/drivers/net/nfp/flower/nfp_flower_cmsg.h
new file mode 100644
index 0000000..fce5163
--- /dev/null
+++ b/drivers/net/nfp/flower/nfp_flower_cmsg.h
@@ -0,0 +1,173 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2022 Corigine, Inc.
+ * All rights reserved.
+ */
+
+#ifndef _NFP_CMSG_H_
+#define _NFP_CMSG_H_
+
+#include <rte_byteorder.h>
+#include <rte_ether.h>
+
+struct nfp_flower_cmsg_hdr {
+	rte_be16_t pad;
+	uint8_t type;
+	uint8_t version;
+};
+
+/* Types defined for control messages */
+enum nfp_flower_cmsg_type {
+	NFP_FLOWER_CMSG_TYPE_FLOW_ADD       = 0,
+	NFP_FLOWER_CMSG_TYPE_FLOW_MOD       = 1,
+	NFP_FLOWER_CMSG_TYPE_FLOW_DEL       = 2,
+	NFP_FLOWER_CMSG_TYPE_LAG_CONFIG     = 4,
+	NFP_FLOWER_CMSG_TYPE_PORT_REIFY     = 6,
+	NFP_FLOWER_CMSG_TYPE_MAC_REPR       = 7,
+	NFP_FLOWER_CMSG_TYPE_PORT_MOD       = 8,
+	NFP_FLOWER_CMSG_TYPE_MERGE_HINT     = 9,
+	NFP_FLOWER_CMSG_TYPE_NO_NEIGH       = 10,
+	NFP_FLOWER_CMSG_TYPE_TUN_MAC        = 11,
+	NFP_FLOWER_CMSG_TYPE_ACTIVE_TUNS    = 12,
+	NFP_FLOWER_CMSG_TYPE_TUN_NEIGH      = 13,
+	NFP_FLOWER_CMSG_TYPE_TUN_IPS        = 14,
+	NFP_FLOWER_CMSG_TYPE_FLOW_STATS     = 15,
+	NFP_FLOWER_CMSG_TYPE_PORT_ECHO      = 16,
+	NFP_FLOWER_CMSG_TYPE_QOS_MOD        = 18,
+	NFP_FLOWER_CMSG_TYPE_QOS_DEL        = 19,
+	NFP_FLOWER_CMSG_TYPE_QOS_STATS      = 20,
+	NFP_FLOWER_CMSG_TYPE_PRE_TUN_RULE   = 21,
+	NFP_FLOWER_CMSG_TYPE_TUN_IPS_V6     = 22,
+	NFP_FLOWER_CMSG_TYPE_NO_NEIGH_V6    = 23,
+	NFP_FLOWER_CMSG_TYPE_TUN_NEIGH_V6   = 24,
+	NFP_FLOWER_CMSG_TYPE_ACTIVE_TUNS_V6 = 25,
+	NFP_FLOWER_CMSG_TYPE_MAX            = 32,
+};
+
+/*
+ * NFP_FLOWER_CMSG_TYPE_MAC_REPR
+ *    Bit    3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
+ *    -----\ 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
+ *     Word +---------------+-----------+---+---------------+---------------+
+ *       0  |                  spare                        |Number of ports|
+ *          +---------------+-----------+---+---------------+---------------+
+ *       1  |    Index      |   spare   |NBI|  Port on NBI  | Chip-wide port|
+ *          +---------------+-----------+---+---------------+---------------+
+ *                                        ....
+ *          +---------------+-----------+---+---------------+---------------+
+ *     N-1  |    Index      |   spare   |NBI|  Port on NBI  | Chip-wide port|
+ *          +---------------+-----------+---+---------------+---------------+
+ *     N    |    Index      |   spare   |NBI|  Port on NBI  | Chip-wide port|
+ *          +---------------+-----------+---+---------------+---------------+
+ *
+ *          Index: index into the eth table
+ *          NBI (bits 17-16): NBI number (0-3)
+ *          Port on NBI (bits 15-8): “base” in the driver
+ *            this forms NBIX.PortY notation as the NSP eth table.
+ *          "Chip-wide" port (bits 7-0):
+ */
+struct nfp_flower_cmsg_mac_repr {
+	uint8_t reserved[3];
+	uint8_t num_ports;
+	struct {
+		uint8_t idx;
+		uint8_t info;
+		uint8_t nbi_port;
+		uint8_t phys_port;
+	} ports[0];
+};
+
+/*
+ * NFP_FLOWER_CMSG_TYPE_PORT_REIFY
+ *    Bit    3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
+ *    -----\ 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
+ *    Word  +-------+-------+---+---+-------+---+---+-----------+-----------+
+ *       0  |Port Ty|Sys ID |NIC|Rsv| Spare |PCI|typ|    vNIC   |  queue    |
+ *          +-------+-----+-+---+---+-------+---+---+-----------+---------+-+
+ *       1  |                             Spare                           |E|
+ *          +-------------------------------------------------------------+-+
+ *          E: 1 = Representor exists, 0 = Representor does not exist
+ */
+struct nfp_flower_cmsg_port_reify {
+	rte_be32_t portnum;
+	rte_be16_t reserved;
+	rte_be16_t info;
+};
+
+/*
+ * NFP_FLOWER_CMSG_TYPE_PORT_MOD
+ *    Bit    3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
+ *    -----\ 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
+ *    Word  +-------+-------+---+---+-------+---+---+-------+---+-----------+
+ *       0  |Port Ty|Sys ID |NIC|Rsv|       Reserved        |    Port       |
+ *          +-------+-------+---+---+-----+-+---+---+-------+---+-----------+
+ *       1  |            Spare            |L|              MTU              |
+ *          +-----------------------------+-+-------------------------------+
+ *        L: Link or Admin state bit. When message is generated by host, this
+ *           bit indicates the admin state (0=down, 1=up). When generated by
+ *           NFP, it indicates the link state (0=down, 1=up)
+ *
+ *        Port Type (word 1, bits 31 to 28) = 1 (Physical Network)
+ *        Port: “Chip-wide number” as assigned by BSP
+ *
+ *    Bit    3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
+ *    -----\ 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0
+ *    Word  +-------+-------+---+---+-------+---+---+-------+---+-----------+
+ *       0  |Port Ty|Sys ID |NIC|Rsv| Spare |PCI|typ|    vNIC   |  queue    |
+ *          +-------+-----+-+---+---+---+-+-+---+---+-------+---+-----------+
+ *       1  |            Spare            |L|              MTU              |
+ *          +-----------------------------+-+-------------------------------+
+ *        L: Link or Admin state bit. When message is generated by host, this
+ *           bit indicates the admin state (0=down, 1=up). When generated by
+ *           NFP, it indicates the link state (0=down, 1=up)
+ *
+ *        Port Type (word 1, bits 31 to 28) = 2 (PCIE)
+ */
+struct nfp_flower_cmsg_port_mod {
+	rte_be32_t portnum;
+	uint8_t reserved;
+	uint8_t info;
+	rte_be16_t mtu;
+};
+
+enum nfp_flower_cmsg_port_type {
+	NFP_FLOWER_CMSG_PORT_TYPE_UNSPEC,
+	NFP_FLOWER_CMSG_PORT_TYPE_PHYS_PORT,
+	NFP_FLOWER_CMSG_PORT_TYPE_PCIE_PORT,
+	NFP_FLOWER_CMSG_PORT_TYPE_OTHER_PORT,
+};
+
+enum nfp_flower_cmsg_port_vnic_type {
+	NFP_FLOWER_CMSG_PORT_VNIC_TYPE_VF,
+	NFP_FLOWER_CMSG_PORT_VNIC_TYPE_PF,
+	NFP_FLOWER_CMSG_PORT_VNIC_TYPE_CTRL,
+};
+
+#define NFP_FLOWER_CMSG_MAC_REPR_NBI            (0x3)
+
+#define NFP_FLOWER_CMSG_HLEN            sizeof(struct nfp_flower_cmsg_hdr)
+#define NFP_FLOWER_CMSG_VER1            1
+#define NFP_NET_META_PORTID             5
+#define NFP_META_PORT_ID_CTRL           ~0U
+
+#define NFP_FLOWER_CMSG_PORT_TYPE(x)            (((x) >> 28) & 0xf)  /* [31,28] */
+#define NFP_FLOWER_CMSG_PORT_SYS_ID(x)          (((x) >> 24) & 0xf)  /* [24,27] */
+#define NFP_FLOWER_CMSG_PORT_NFP_ID(x)          (((x) >> 22) & 0x3)  /* [22,23] */
+#define NFP_FLOWER_CMSG_PORT_PCI(x)             (((x) >> 14) & 0x3)  /* [14,15] */
+#define NFP_FLOWER_CMSG_PORT_VNIC_TYPE(x)       (((x) >> 12) & 0x3)  /* [12,13] */
+#define NFP_FLOWER_CMSG_PORT_VNIC(x)            (((x) >> 6) & 0x3f)  /* [6,11] */
+#define NFP_FLOWER_CMSG_PORT_PCIE_Q(x)          ((x) & 0x3f)         /* [0,5] */
+#define NFP_FLOWER_CMSG_PORT_PHYS_PORT_NUM(x)   ((x) & 0xff)         /* [0,7] */
+
+static inline char*
+nfp_flower_cmsg_get_data(struct rte_mbuf *m)
+{
+	return rte_pktmbuf_mtod(m, char *) + 4 + 4 + NFP_FLOWER_CMSG_HLEN;
+}
+
+int nfp_flower_cmsg_mac_repr(struct nfp_app_flower *app_flower);
+int nfp_flower_cmsg_repr_reify(struct nfp_app_flower *app_flower,
+		struct nfp_flower_representor *repr);
+int nfp_flower_cmsg_port_mod(struct nfp_app_flower *app_flower,
+		uint32_t port_id, bool carrier_ok);
+
+#endif /* _NFP_CMSG_H_ */
diff --git a/drivers/net/nfp/flower/nfp_flower_representor.c b/drivers/net/nfp/flower/nfp_flower_representor.c
new file mode 100644
index 0000000..7408e63
--- /dev/null
+++ b/drivers/net/nfp/flower/nfp_flower_representor.c
@@ -0,0 +1,602 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright (c) 2022 Corigine, Inc.
+ * All rights reserved.
+ */
+
+#include <rte_common.h>
+#include <ethdev_pci.h>
+
+#include "../nfp_common.h"
+#include "../nfp_logs.h"
+#include "../nfp_ctrl.h"
+#include "../nfp_rxtx.h"
+#include "../nfpcore/nfp_mip.h"
+#include "../nfpcore/nfp_rtsym.h"
+#include "../nfpcore/nfp_nsp.h"
+#include "nfp_flower.h"
+#include "nfp_flower_representor.h"
+#include "nfp_flower_ctrl.h"
+#include "nfp_flower_cmsg.h"
+
+static int
+nfp_flower_repr_link_update(struct rte_eth_dev *dev,
+		__rte_unused int wait_to_complete)
+{
+	int ret;
+	uint32_t nn_link_status;
+	struct nfp_net_hw *pf_hw;
+	struct rte_eth_link *link;
+	struct nfp_flower_representor *repr;
+
+	static const uint32_t ls_to_ethtool[] = {
+		[NFP_NET_CFG_STS_LINK_RATE_UNSUPPORTED] = RTE_ETH_SPEED_NUM_NONE,
+		[NFP_NET_CFG_STS_LINK_RATE_UNKNOWN]     = RTE_ETH_SPEED_NUM_NONE,
+		[NFP_NET_CFG_STS_LINK_RATE_1G]          = RTE_ETH_SPEED_NUM_1G,
+		[NFP_NET_CFG_STS_LINK_RATE_10G]         = RTE_ETH_SPEED_NUM_10G,
+		[NFP_NET_CFG_STS_LINK_RATE_25G]         = RTE_ETH_SPEED_NUM_25G,
+		[NFP_NET_CFG_STS_LINK_RATE_40G]         = RTE_ETH_SPEED_NUM_40G,
+		[NFP_NET_CFG_STS_LINK_RATE_50G]         = RTE_ETH_SPEED_NUM_50G,
+		[NFP_NET_CFG_STS_LINK_RATE_100G]        = RTE_ETH_SPEED_NUM_100G,
+	};
+
+	repr = (struct nfp_flower_representor *)dev->data->dev_private;
+	link = &repr->link;
+	pf_hw = repr->app_flower->pf_hw;
+
+	memset(link, 0, sizeof(struct rte_eth_link));
+	nn_link_status = nn_cfg_readl(pf_hw, NFP_NET_CFG_STS);
+
+	if (nn_link_status & NFP_NET_CFG_STS_LINK)
+		link->link_status = RTE_ETH_LINK_UP;
+
+	link->link_duplex = RTE_ETH_LINK_FULL_DUPLEX;
+
+	nn_link_status = (nn_link_status >> NFP_NET_CFG_STS_LINK_RATE_SHIFT) &
+			 NFP_NET_CFG_STS_LINK_RATE_MASK;
+
+	if (nn_link_status >= RTE_DIM(ls_to_ethtool))
+		link->link_speed = RTE_ETH_SPEED_NUM_NONE;
+	else
+		link->link_speed = ls_to_ethtool[nn_link_status];
+
+	ret = rte_eth_linkstatus_set(dev, link);
+	if (ret == 0) {
+		if (link->link_status)
+			PMD_DRV_LOG(INFO, "NIC Link is Up");
+		else
+			PMD_DRV_LOG(INFO, "NIC Link is Down");
+	}
+
+	return ret;
+}
+
+static int
+nfp_flower_repr_dev_infos_get(__rte_unused struct rte_eth_dev *dev,
+		struct rte_eth_dev_info *dev_info)
+{
+	/* Hardcoded pktlen and queues for now */
+	dev_info->max_rx_queues = 1;
+	dev_info->max_tx_queues = 1;
+	dev_info->min_rx_bufsize = RTE_ETHER_MIN_MTU;
+	dev_info->max_rx_pktlen = 9000;
+
+	dev_info->rx_offload_capa = RTE_ETH_RX_OFFLOAD_VLAN_STRIP;
+	dev_info->rx_offload_capa |= RTE_ETH_RX_OFFLOAD_IPV4_CKSUM |
+			RTE_ETH_RX_OFFLOAD_UDP_CKSUM |
+			RTE_ETH_RX_OFFLOAD_TCP_CKSUM;
+
+	dev_info->tx_offload_capa = RTE_ETH_TX_OFFLOAD_VLAN_INSERT;
+	dev_info->tx_offload_capa |= RTE_ETH_TX_OFFLOAD_IPV4_CKSUM |
+			RTE_ETH_TX_OFFLOAD_UDP_CKSUM |
+			RTE_ETH_TX_OFFLOAD_TCP_CKSUM;
+	dev_info->tx_offload_capa |= RTE_ETH_TX_OFFLOAD_TCP_TSO;
+	dev_info->tx_offload_capa |= RTE_ETH_TX_OFFLOAD_MULTI_SEGS;
+
+	dev_info->max_mac_addrs = 1;
+
+	return 0;
+}
+
+static int
+nfp_flower_repr_dev_configure(struct rte_eth_dev *dev)
+{
+	struct nfp_net_hw *pf_hw;
+	struct rte_eth_conf *dev_conf;
+	struct rte_eth_rxmode *rxmode;
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)dev->data->dev_private;
+	pf_hw = repr->app_flower->pf_hw;
+
+	dev_conf = &dev->data->dev_conf;
+	rxmode = &dev_conf->rxmode;
+
+	/* Checking MTU set */
+	if (rxmode->mtu > pf_hw->flbufsz) {
+		PMD_DRV_LOG(INFO, "MTU (%u) larger then current mbufsize (%u) not supported",
+				rxmode->mtu, pf_hw->flbufsz);
+		return -ERANGE;
+	}
+
+	return 0;
+}
+
+static int
+nfp_flower_repr_dev_start(struct rte_eth_dev *dev)
+{
+	struct nfp_app_flower *app_flower;
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)dev->data->dev_private;
+	app_flower = repr->app_flower;
+
+	if (repr->repr_type == NFP_REPR_TYPE_PHYS_PORT)
+		nfp_eth_set_configured(app_flower->pf_hw->pf_dev->cpp,
+				repr->nfp_idx, 1);
+
+	nfp_flower_cmsg_port_mod(app_flower, repr->port_id, true);
+
+	return 0;
+}
+
+static int
+nfp_flower_repr_dev_stop(struct rte_eth_dev *dev)
+{
+	struct nfp_app_flower *app_flower;
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)dev->data->dev_private;
+	app_flower = repr->app_flower;
+
+	nfp_flower_cmsg_port_mod(app_flower, repr->port_id, false);
+
+	if (repr->repr_type == NFP_REPR_TYPE_PHYS_PORT)
+		nfp_eth_set_configured(app_flower->pf_hw->pf_dev->cpp,
+				repr->nfp_idx, 0);
+
+	return 0;
+}
+
+static int
+nfp_flower_repr_rx_queue_setup(struct rte_eth_dev *dev,
+		uint16_t rx_queue_id,
+		__rte_unused uint16_t nb_rx_desc,
+		unsigned int socket_id,
+		__rte_unused const struct rte_eth_rxconf *rx_conf,
+		__rte_unused struct rte_mempool *mb_pool)
+{
+	struct nfp_net_rxq *rxq;
+	struct nfp_net_hw *pf_hw;
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)dev->data->dev_private;
+	pf_hw = repr->app_flower->pf_hw;
+
+	/* Allocating rx queue data structure */
+	rxq = rte_zmalloc_socket("ethdev RX queue", sizeof(struct nfp_net_rxq),
+			RTE_CACHE_LINE_SIZE, socket_id);
+	if (rxq == NULL)
+		return -ENOMEM;
+
+	rxq->hw = pf_hw;
+	rxq->qidx = rx_queue_id;
+	rxq->port_id = dev->data->port_id;
+	dev->data->rx_queues[rx_queue_id] = rxq;
+
+	return 0;
+}
+
+static int
+nfp_flower_repr_tx_queue_setup(struct rte_eth_dev *dev,
+		uint16_t tx_queue_id,
+		__rte_unused uint16_t nb_tx_desc,
+		unsigned int socket_id,
+		__rte_unused const struct rte_eth_txconf *tx_conf)
+{
+	struct nfp_net_txq *txq;
+	struct nfp_net_hw *pf_hw;
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)dev->data->dev_private;
+	pf_hw = repr->app_flower->pf_hw;
+
+	/* Allocating tx queue data structure */
+	txq = rte_zmalloc_socket("ethdev TX queue", sizeof(struct nfp_net_txq),
+			RTE_CACHE_LINE_SIZE, socket_id);
+	if (txq == NULL)
+		return -ENOMEM;
+
+	txq->hw = pf_hw;
+	txq->qidx = tx_queue_id;
+	txq->port_id = dev->data->port_id;
+	dev->data->tx_queues[tx_queue_id] = txq;
+
+	return 0;
+}
+
+static int
+nfp_flower_repr_stats_get(struct rte_eth_dev *ethdev,
+		struct rte_eth_stats *stats)
+{
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)ethdev->data->dev_private;
+	rte_memcpy(stats, &repr->repr_stats, sizeof(struct rte_eth_stats));
+
+	return 0;
+}
+
+static int
+nfp_flower_repr_stats_reset(struct rte_eth_dev *ethdev)
+{
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)ethdev->data->dev_private;
+	memset(&repr->repr_stats, 0, sizeof(struct rte_eth_stats));
+
+	return 0;
+}
+
+static int
+nfp_flower_repr_promiscuous_enable(struct rte_eth_dev *dev)
+{
+	struct nfp_net_hw *pf_hw;
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)dev->data->dev_private;
+	pf_hw = repr->app_flower->pf_hw;
+
+	if (!(pf_hw->cap & NFP_NET_CFG_CTRL_PROMISC)) {
+		PMD_DRV_LOG(INFO, "Promiscuous mode not supported");
+		return -ENOTSUP;
+	}
+
+	if (pf_hw->ctrl & NFP_NET_CFG_CTRL_PROMISC) {
+		PMD_DRV_LOG(INFO, "Promiscuous mode already enabled");
+		return 0;
+	}
+
+	return nfp_net_promisc_enable(pf_hw->eth_dev);
+}
+
+static int
+nfp_flower_repr_promiscuous_disable(struct rte_eth_dev *dev)
+{
+	struct nfp_net_hw *pf_hw;
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)dev->data->dev_private;
+	pf_hw = repr->app_flower->pf_hw;
+
+	if ((pf_hw->ctrl & NFP_NET_CFG_CTRL_PROMISC) == 0) {
+		PMD_DRV_LOG(INFO, "Promiscuous mode already disabled");
+		return 0;
+	}
+
+	return nfp_net_promisc_disable(pf_hw->eth_dev);
+}
+
+static int
+nfp_flower_repr_mac_addr_set(struct rte_eth_dev *ethdev,
+		struct rte_ether_addr *mac_addr)
+{
+	struct nfp_flower_representor *repr;
+
+	repr = (struct nfp_flower_representor *)ethdev->data->dev_private;
+	rte_ether_addr_copy(mac_addr, &repr->mac_addr);
+	rte_ether_addr_copy(mac_addr, ethdev->data->mac_addrs);
+
+	return 0;
+}
+
+static uint16_t
+nfp_flower_repr_rx_burst(void *rx_queue,
+		struct rte_mbuf **rx_pkts,
+		uint16_t nb_pkts)
+{
+	unsigned int available = 0;
+	unsigned int total_dequeue;
+	struct nfp_net_rxq *rxq;
+	struct rte_eth_dev *dev;
+	struct nfp_flower_representor *repr;
+
+	rxq = rx_queue;
+	if (unlikely(rxq == NULL)) {
+		PMD_RX_LOG(ERR, "RX Bad queue");
+		return 0;
+	}
+
+	dev = &rte_eth_devices[rxq->port_id];
+	repr = dev->data->dev_private;
+	if (unlikely(repr->ring == NULL)) {
+		PMD_RX_LOG(ERR, "representor %s has no ring configured!",
+				repr->name);
+		return 0;
+	}
+
+	total_dequeue = rte_ring_dequeue_burst(repr->ring, (void *)rx_pkts,
+			nb_pkts, &available);
+	if (total_dequeue != 0) {
+		PMD_RX_LOG(DEBUG, "Representor Rx burst for %s, port_id: 0x%x, "
+				"received: %u, available: %u", repr->name,
+				repr->port_id, total_dequeue, available);
+
+		repr->repr_stats.ipackets += total_dequeue;
+	}
+
+	return total_dequeue;
+}
+
+static uint16_t
+nfp_flower_repr_tx_burst(void *tx_queue,
+		struct rte_mbuf **tx_pkts,
+		uint16_t nb_pkts)
+{
+	uint16_t i;
+	uint16_t sent;
+	char *meta_offset;
+	struct nfp_net_txq *txq;
+	struct nfp_net_hw *pf_hw;
+	struct rte_eth_dev *dev;
+	struct rte_eth_dev *repr_dev;
+	struct nfp_flower_representor *repr;
+
+	txq = tx_queue;
+	if (unlikely(txq == NULL)) {
+		PMD_TX_LOG(ERR, "TX Bad queue");
+		return 0;
+	}
+
+	/* This points to the PF vNIC that owns this representor */
+	pf_hw = txq->hw;
+	dev = pf_hw->eth_dev;
+
+	/* Grab a handle to the representor struct */
+	repr_dev = &rte_eth_devices[txq->port_id];
+	repr = repr_dev->data->dev_private;
+
+	for (i = 0; i < nb_pkts; i++) {
+		meta_offset = rte_pktmbuf_prepend(tx_pkts[i], FLOWER_PKT_DATA_OFFSET);
+		*(uint32_t *)meta_offset = rte_cpu_to_be_32(NFP_NET_META_PORTID);
+		meta_offset += 4;
+		*(uint32_t *)meta_offset = rte_cpu_to_be_32(repr->port_id);
+	}
+
+	/* Only using Tx queue 0 for now. */
+	sent = rte_eth_tx_burst(dev->data->port_id, 0, tx_pkts, nb_pkts);
+	if (sent != 0) {
+		PMD_TX_LOG(DEBUG, "Representor Tx burst for %s, port_id: 0x%x "
+			"transmitted: %u", repr->name, repr->port_id, sent);
+		repr->repr_stats.opackets += sent;
+	}
+
+	return sent;
+}
+
+static const struct eth_dev_ops nfp_flower_repr_dev_ops = {
+	.dev_infos_get        = nfp_flower_repr_dev_infos_get,
+
+	.dev_start            = nfp_flower_repr_dev_start,
+	.dev_configure        = nfp_flower_repr_dev_configure,
+	.dev_stop             = nfp_flower_repr_dev_stop,
+
+	.rx_queue_setup       = nfp_flower_repr_rx_queue_setup,
+	.tx_queue_setup       = nfp_flower_repr_tx_queue_setup,
+
+	.link_update          = nfp_flower_repr_link_update,
+
+	.stats_get            = nfp_flower_repr_stats_get,
+	.stats_reset          = nfp_flower_repr_stats_reset,
+
+	.promiscuous_enable   = nfp_flower_repr_promiscuous_enable,
+	.promiscuous_disable  = nfp_flower_repr_promiscuous_disable,
+
+	.mac_addr_set         = nfp_flower_repr_mac_addr_set,
+};
+
+static uint32_t
+nfp_flower_get_phys_port_id(uint8_t port)
+{
+	return (NFP_FLOWER_CMSG_PORT_TYPE_PHYS_PORT << 28) | port;
+}
+
+static uint32_t
+nfp_get_pcie_port_id(struct nfp_cpp *cpp,
+		int type,
+		uint8_t vnic,
+		uint8_t queue)
+{
+	uint8_t nfp_pcie;
+	uint32_t port_id;
+
+	nfp_pcie = NFP_CPP_INTERFACE_UNIT_of(nfp_cpp_interface(cpp));
+	port_id = ((nfp_pcie & 0x3) << 14) |
+		  ((type & 0x3) << 12) |
+		  ((vnic & 0x3f) << 6) |
+		  (queue & 0x3f) |
+		  ((NFP_FLOWER_CMSG_PORT_TYPE_PCIE_PORT & 0xf) << 28);
+
+	return port_id;
+}
+
+static int
+nfp_flower_repr_init(struct rte_eth_dev *eth_dev,
+		void *init_params)
+{
+	int ret;
+	unsigned int numa_node;
+	char ring_name[RTE_ETH_NAME_MAX_LEN];
+	struct nfp_app_flower *app_flower;
+	struct nfp_flower_representor *repr;
+	struct nfp_flower_representor *init_repr_data;
+
+	/* Cast the input representor data to the correct struct here */
+	init_repr_data = (struct nfp_flower_representor *)init_params;
+
+	app_flower = init_repr_data->app_flower;
+
+	/* Memory has been allocated in the eth_dev_create() function */
+	repr = eth_dev->data->dev_private;
+
+	/*
+	 * We need multiproduce rings as we can have multiple PF ports.
+	 * On the other hand, we need single consumer rings, as just one
+	 * representor PMD will try to read from the ring.
+	 */
+	snprintf(ring_name, sizeof(ring_name), "%s_%s",
+		init_repr_data->name, "ring");
+	numa_node = rte_socket_id();
+	repr->ring = rte_ring_create(ring_name, 256, numa_node, RING_F_SC_DEQ);
+	if (repr->ring == NULL) {
+		PMD_DRV_LOG(ERR, "rte_ring_create failed for %s", ring_name);
+		return -ENOMEM;
+	}
+
+	/* Copy data here from the input representor template*/
+	repr->vf_id            = init_repr_data->vf_id;
+	repr->switch_domain_id = init_repr_data->switch_domain_id;
+	repr->port_id          = init_repr_data->port_id;
+	repr->nfp_idx          = init_repr_data->nfp_idx;
+	repr->repr_type        = init_repr_data->repr_type;
+	repr->app_flower       = init_repr_data->app_flower;
+
+	snprintf(repr->name, sizeof(repr->name), "%s", init_repr_data->name);
+
+	eth_dev->dev_ops = &nfp_flower_repr_dev_ops;
+
+	eth_dev->rx_pkt_burst = nfp_flower_repr_rx_burst;
+	eth_dev->tx_pkt_burst = nfp_flower_repr_tx_burst;
+
+	eth_dev->data->dev_flags |= RTE_ETH_DEV_REPRESENTOR;
+
+	if (repr->repr_type == NFP_REPR_TYPE_PHYS_PORT)
+		eth_dev->data->representor_id = repr->vf_id;
+	else
+		eth_dev->data->representor_id = repr->vf_id +
+				app_flower->num_phyport_reprs;
+
+	/* This backer port is that of the eth_device created for the PF vNIC */
+	eth_dev->data->backer_port_id = app_flower->pf_hw->eth_dev->data->port_id;
+
+	/* Only single queues for representor devices */
+	eth_dev->data->nb_rx_queues = 1;
+	eth_dev->data->nb_tx_queues = 1;
+
+	/* Allocating memory for mac addr */
+	eth_dev->data->mac_addrs = rte_zmalloc("mac_addr",
+		RTE_ETHER_ADDR_LEN, 0);
+	if (eth_dev->data->mac_addrs == NULL) {
+		PMD_INIT_LOG(ERR, "Failed to allocate memory for repr MAC");
+		ret = -ENOMEM;
+		goto ring_cleanup;
+	}
+
+	rte_ether_addr_copy(&init_repr_data->mac_addr, &repr->mac_addr);
+	rte_ether_addr_copy(&init_repr_data->mac_addr, eth_dev->data->mac_addrs);
+
+	/* Send reify message to hardware to inform it about the new repr */
+	ret = nfp_flower_cmsg_repr_reify(app_flower, repr);
+	if (ret != 0) {
+		PMD_INIT_LOG(WARNING, "Failed to send repr reify message");
+		goto mac_cleanup;
+	}
+
+	/* Add repr to correct array */
+	if (repr->repr_type == NFP_REPR_TYPE_PHYS_PORT)
+		app_flower->phy_reprs[repr->nfp_idx] = repr;
+	else
+		app_flower->vf_reprs[repr->vf_id] = repr;
+
+	return 0;
+
+mac_cleanup:
+	rte_free(eth_dev->data->mac_addrs);
+ring_cleanup:
+	rte_ring_free(repr->ring);
+
+	return ret;
+}
+
+int
+nfp_flower_repr_alloc(struct nfp_app_flower *app_flower)
+{
+	int i;
+	int ret;
+	struct rte_eth_dev *eth_dev;
+	struct nfp_eth_table *nfp_eth_table;
+	struct nfp_eth_table_port *eth_port;
+	struct nfp_flower_representor flower_repr = {
+		.switch_domain_id = app_flower->switch_domain_id,
+		.app_flower       = app_flower,
+	};
+
+	nfp_eth_table = app_flower->nfp_eth_table;
+	eth_dev = app_flower->pf_hw->eth_dev;
+
+	/* Send a NFP_FLOWER_CMSG_TYPE_MAC_REPR cmsg to hardware*/
+	ret = nfp_flower_cmsg_mac_repr(app_flower);
+	if (ret != 0) {
+		PMD_INIT_LOG(ERR, "Cloud not send mac repr cmsgs");
+		return ret;
+	}
+
+	/* Create a rte_eth_dev for every phyport representor */
+	for (i = 0; i < app_flower->num_phyport_reprs; i++) {
+		eth_port = &nfp_eth_table->ports[i];
+		flower_repr.repr_type = NFP_REPR_TYPE_PHYS_PORT;
+		flower_repr.port_id = nfp_flower_get_phys_port_id(eth_port->index);
+		flower_repr.nfp_idx = eth_port->eth_index;
+		flower_repr.vf_id = i;
+
+		/* Copy the real mac of the interface to the representor struct */
+		rte_ether_addr_copy((struct rte_ether_addr *)eth_port->mac_addr,
+				&flower_repr.mac_addr);
+		sprintf(flower_repr.name, "flower_repr_p%d", i);
+
+		/*
+		 * Create a eth_dev for this representor
+		 * This will also allocate private memory for the device
+		 */
+		ret = rte_eth_dev_create(eth_dev->device, flower_repr.name,
+				sizeof(struct nfp_flower_representor),
+				NULL, NULL, nfp_flower_repr_init, &flower_repr);
+		if (ret != 0) {
+			PMD_INIT_LOG(ERR, "Cloud not create eth_dev for repr");
+			break;
+		}
+	}
+
+	if (i < app_flower->num_phyport_reprs)
+		return ret;
+
+	/*
+	 * Now allocate eth_dev's for VF representors.
+	 * Also send reify messages
+	 */
+	for (i = 0; i < app_flower->num_vf_reprs; i++) {
+		flower_repr.repr_type = NFP_REPR_TYPE_VF;
+		flower_repr.port_id = nfp_get_pcie_port_id(app_flower->pf_hw->cpp,
+				NFP_FLOWER_CMSG_PORT_VNIC_TYPE_VF, i, 0);
+		flower_repr.nfp_idx = 0;
+		flower_repr.vf_id = i;
+
+		/* VF reprs get a random MAC address */
+		rte_eth_random_addr(flower_repr.mac_addr.addr_bytes);
+
+		sprintf(flower_repr.name, "flower_repr_vf%d", i);
+
+		 /* This will also allocate private memory for the device*/
+		ret = rte_eth_dev_create(eth_dev->device, flower_repr.name,
+				sizeof(struct nfp_flower_representor),
+				NULL, NULL, nfp_flower_repr_init, &flower_repr);
+		if (ret != 0) {
+			PMD_INIT_LOG(ERR, "Cloud not create eth_dev for repr");
+			break;
+		}
+	}
+
+	if (i < app_flower->num_vf_reprs)
+		return ret;
+
+	return 0;
+}
diff --git a/drivers/net/nfp/flower/nfp_flower_representor.h b/drivers/net/nfp/flower/nfp_flower_representor.h
new file mode 100644
index 0000000..713dfc4
--- /dev/null
+++ b/drivers/net/nfp/flower/nfp_flower_representor.h
@@ -0,0 +1,39 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2022 Corigine, Inc.
+ * All rights reserved.
+ */
+
+#ifndef _NFP_FLOWER_REPRESENTOR_H_
+#define _NFP_FLOWER_REPRESENTOR_H_
+
+/*
+ * enum nfp_repr_type - type of representor
+ * @NFP_REPR_TYPE_PHYS_PORT:   external NIC port
+ * @NFP_REPR_TYPE_PF:          physical function
+ * @NFP_REPR_TYPE_VF:          virtual function
+ * @NFP_REPR_TYPE_MAX:         number of representor types
+ */
+enum nfp_repr_type {
+	NFP_REPR_TYPE_PHYS_PORT = 0,
+	NFP_REPR_TYPE_PF,
+	NFP_REPR_TYPE_VF,
+	NFP_REPR_TYPE_MAX,
+};
+
+struct nfp_flower_representor {
+	uint16_t vf_id;
+	uint16_t switch_domain_id;
+	uint32_t repr_type;
+	uint32_t port_id;
+	uint32_t nfp_idx;    /* only valid for the repr of physical port */
+	char name[RTE_ETH_NAME_MAX_LEN];
+	struct rte_ether_addr mac_addr;
+	struct nfp_app_flower *app_flower;
+	struct rte_ring *ring;
+	struct rte_eth_link link;
+	struct rte_eth_stats repr_stats;
+};
+
+int nfp_flower_repr_alloc(struct nfp_app_flower *app_flower);
+
+#endif /* _NFP_FLOWER_REPRESENTOR_H_ */
diff --git a/drivers/net/nfp/meson.build b/drivers/net/nfp/meson.build
index 8710213..8a63979 100644
--- a/drivers/net/nfp/meson.build
+++ b/drivers/net/nfp/meson.build
@@ -7,7 +7,9 @@ if not is_linux or not dpdk_conf.get('RTE_ARCH_64')
 endif
 sources = files(
         'flower/nfp_flower.c',
+        'flower/nfp_flower_cmsg.c',
         'flower/nfp_flower_ctrl.c',
+        'flower/nfp_flower_representor.c',
         'nfpcore/nfp_cpp_pcie_ops.c',
         'nfpcore/nfp_nsp.c',
         'nfpcore/nfp_cppcore.c',
-- 
1.8.3.1