diff --git a/doc/guides/eventdevs/octeontx.rst b/doc/guides/eventdevs/octeontx.rst
index ab36a36e0a..9a3646db0d 100644
--- a/doc/guides/eventdevs/octeontx.rst
+++ b/doc/guides/eventdevs/octeontx.rst
@@ -139,3 +139,10 @@ follows:
 
 When timvf is used as Event timer adapter event schedule type
 ``RTE_SCHED_TYPE_PARALLEL`` is not supported.
+
+Max mempool size
+~~~~~~~~~~~~~~~~
+
+Max mempool size when using OCTEON TX Eventdev (SSO) should be limited to 128K.
+When running dpdk-test-eventdev on OCTEON TX the application can limit the
+number of mbufs by using the option ``--pool_sz 131072``
diff --git a/doc/guides/nics/octeontx.rst b/doc/guides/nics/octeontx.rst
index 3c19c912db..8fc53810b2 100644
--- a/doc/guides/nics/octeontx.rst
+++ b/doc/guides/nics/octeontx.rst
@@ -174,3 +174,10 @@ The OCTEON TX SoC family NICs support a maximum of a 32K jumbo frame. The value
 is fixed and cannot be changed. So, even when the ``rxmode.max_rx_pkt_len``
 member of ``struct rte_eth_conf`` is set to a value lower than 32k, frames
 up to 32k bytes can still reach the host interface.
+
+Maximum mempool size
+~~~~~~~~~~~~~~~~~~~~
+
+The maximum mempool size supplied to Rx queue setup should be less than 128K.
+When running testpmd on OCTEON TX the application can limit the number of mbufs
+by using the option ``--total-num-mbufs=131072``.
diff --git a/drivers/common/octeontx/octeontx_mbox.c b/drivers/common/octeontx/octeontx_mbox.c
index 880f8a40f0..68cb0351f0 100644
--- a/drivers/common/octeontx/octeontx_mbox.c
+++ b/drivers/common/octeontx/octeontx_mbox.c
@@ -31,6 +31,7 @@ enum {
 
 struct mbox {
 	int init_once;
+	uint8_t ready;
 	uint8_t *ram_mbox_base; /* Base address of mbox message stored in ram */
 	uint8_t *reg; /* Store to this register triggers PF mbox interrupt */
 	uint16_t tag_own; /* Last tag which was written to own channel */
@@ -59,6 +60,13 @@ struct mbox_ram_hdr {
 	};
 };
 
+/* MBOX interface version message */
+struct mbox_intf_ver {
+	uint32_t platform:12;
+	uint32_t major:10;
+	uint32_t minor:10;
+};
+
 int octeontx_logtype_mbox;
 
 RTE_INIT(otx_init_log)
@@ -247,3 +255,92 @@ octeontx_mbox_send(struct octeontx_mbox_hdr *hdr, void *txdata,
 
 	return mbox_send(m, hdr, txdata, txlen, rxdata, rxlen);
 }
+
+static int
+octeontx_start_domain(void)
+{
+	struct octeontx_mbox_hdr hdr = {0};
+	int result = -EINVAL;
+
+	hdr.coproc = NO_COPROC;
+	hdr.msg = RM_START_APP;
+
+	result = octeontx_mbox_send(&hdr, NULL, 0, NULL, 0);
+	if (result != 0) {
+		mbox_log_err("Could not start domain. Err=%d. FuncErr=%d\n",
+			     result, hdr.res_code);
+		result = -EINVAL;
+	}
+
+	return result;
+}
+
+static int
+octeontx_check_mbox_version(struct mbox_intf_ver app_intf_ver,
+			    struct mbox_intf_ver *intf_ver)
+{
+	struct mbox_intf_ver kernel_intf_ver = {0};
+	struct octeontx_mbox_hdr hdr = {0};
+	int result = 0;
+
+
+	hdr.coproc = NO_COPROC;
+	hdr.msg = RM_INTERFACE_VERSION;
+
+	result = octeontx_mbox_send(&hdr, &app_intf_ver, sizeof(app_intf_ver),
+			&kernel_intf_ver, sizeof(kernel_intf_ver));
+	if (result != sizeof(kernel_intf_ver)) {
+		mbox_log_err("Could not send interface version. Err=%d. FuncErr=%d\n",
+			     result, hdr.res_code);
+		result = -EINVAL;
+	}
+
+	if (intf_ver)
+		*intf_ver = kernel_intf_ver;
+
+	if (app_intf_ver.platform != kernel_intf_ver.platform ||
+			app_intf_ver.major != kernel_intf_ver.major ||
+			app_intf_ver.minor != kernel_intf_ver.minor)
+		result = -EINVAL;
+
+	return result;
+}
+
+int
+octeontx_mbox_init(void)
+{
+	const struct mbox_intf_ver MBOX_INTERFACE_VERSION = {
+		.platform = 0x01,
+		.major = 0x01,
+		.minor = 0x03
+	};
+	struct mbox_intf_ver rm_intf_ver = {0};
+	struct mbox *m = &octeontx_mbox;
+	int ret;
+
+	if (m->ready)
+		return 0;
+
+	ret = octeontx_start_domain();
+	if (ret < 0) {
+		m->init_once = 0;
+		return ret;
+	}
+
+	ret = octeontx_check_mbox_version(MBOX_INTERFACE_VERSION,
+					  &rm_intf_ver);
+	if (ret < 0) {
+		mbox_log_err("MBOX version: Kernel(%d.%d.%d) != DPDK(%d.%d.%d)",
+			     rm_intf_ver.platform, rm_intf_ver.major,
+			     rm_intf_ver.minor, MBOX_INTERFACE_VERSION.platform,
+			     MBOX_INTERFACE_VERSION.major,
+			     MBOX_INTERFACE_VERSION.minor);
+		m->init_once = 0;
+		return -EINVAL;
+	}
+
+	m->ready = 1;
+	rte_mb();
+
+	return 0;
+}
diff --git a/drivers/common/octeontx/octeontx_mbox.h b/drivers/common/octeontx/octeontx_mbox.h
index 43fbda282d..1f794c7f71 100644
--- a/drivers/common/octeontx/octeontx_mbox.h
+++ b/drivers/common/octeontx/octeontx_mbox.h
@@ -11,6 +11,11 @@
 #define SSOW_BAR4_LEN			(64 * 1024)
 #define SSO_VHGRP_PF_MBOX(x)		(0x200ULL | ((x) << 3))
 
+#define NO_COPROC               0x0
+#define RM_START_APP            0x1
+#define RM_INTERFACE_VERSION    0x2
+
+
 #define MBOX_LOG(level, fmt, args...) \
 	rte_log(RTE_LOG_ ## level, octeontx_logtype_mbox,\
 			"%s() line %u: " fmt "\n", __func__, __LINE__, ## args)
@@ -26,9 +31,11 @@ struct octeontx_mbox_hdr {
 	uint16_t vfid;  /* VF index or pf resource index local to the domain */
 	uint8_t coproc; /* Coprocessor id */
 	uint8_t msg;    /* Message id */
+	uint8_t oob;	/* out of band data */
 	uint8_t res_code; /* Functional layer response code */
 };
 
+int octeontx_mbox_init(void);
 int octeontx_mbox_set_ram_mbox_base(uint8_t *ram_mbox_base);
 int octeontx_mbox_set_reg(uint8_t *reg);
 int octeontx_mbox_send(struct octeontx_mbox_hdr *hdr,
diff --git a/drivers/common/octeontx/rte_common_octeontx_version.map b/drivers/common/octeontx/rte_common_octeontx_version.map
index c15fb89112..5f416d33d6 100644
--- a/drivers/common/octeontx/rte_common_octeontx_version.map
+++ b/drivers/common/octeontx/rte_common_octeontx_version.map
@@ -2,6 +2,7 @@ DPDK_20.0 {
 	global:
 
 	octeontx_logtype_mbox;
+	octeontx_mbox_init;
 	octeontx_mbox_send;
 	octeontx_mbox_set_ram_mbox_base;
 	octeontx_mbox_set_reg;
diff --git a/drivers/event/octeontx/ssovf_evdev.c b/drivers/event/octeontx/ssovf_evdev.c
index e4e7c44ed8..f9e93244fd 100644
--- a/drivers/event/octeontx/ssovf_evdev.c
+++ b/drivers/event/octeontx/ssovf_evdev.c
@@ -77,6 +77,7 @@ ssovf_mbox_getwork_tmo_set(uint32_t timeout_ns)
 }
 
 struct ssovf_mbox_grp_pri {
+	uint8_t vhgrp_id;
 	uint8_t wgt_left; /* Read only */
 	uint8_t weight;
 	uint8_t affinity;
@@ -95,6 +96,7 @@ ssovf_mbox_priority_set(uint8_t queue, uint8_t prio)
 	hdr.msg = SSO_GRP_SET_PRIORITY;
 	hdr.vfid = queue;
 
+	grp.vhgrp_id = queue;
 	grp.weight = 0xff;
 	grp.affinity = 0xff;
 	grp.priority = prio / 32; /* Normalize to 0 to 7 */
@@ -433,7 +435,7 @@ ssovf_eth_rx_adapter_queue_add(const struct rte_eventdev *dev,
 	pki_qos.mmask.f_grptag_ok = 1;
 	pki_qos.mmask.f_grptag_bad = 1;
 
-	pki_qos.tag_type = queue_conf->ev.sched_type;
+	pki_qos.qos_entry.tag_type = queue_conf->ev.sched_type;
 	pki_qos.qos_entry.port_add = 0;
 	pki_qos.qos_entry.ggrp_ok = queue_conf->ev.queue_id;
 	pki_qos.qos_entry.ggrp_bad = queue_conf->ev.queue_id;
@@ -780,6 +782,7 @@ ssovf_vdev_probe(struct rte_vdev_device *vdev)
 		return 0;
 	}
 
+	octeontx_mbox_init();
 	ret = ssovf_info(&oinfo);
 	if (ret) {
 		ssovf_log_err("Failed to probe and validate ssovfs %d", ret);
diff --git a/drivers/event/octeontx/ssovf_probe.c b/drivers/event/octeontx/ssovf_probe.c
index b3db596d41..9252998c16 100644
--- a/drivers/event/octeontx/ssovf_probe.c
+++ b/drivers/event/octeontx/ssovf_probe.c
@@ -19,8 +19,6 @@
 #define SSO_MAX_VHGRP                     (64)
 #define SSO_MAX_VHWS                      (32)
 
-#define SSO_VHGRP_AQ_THR                  (0x1E0ULL)
-
 struct ssovf_res {
 	uint16_t domain;
 	uint16_t vfid;
diff --git a/drivers/mempool/octeontx/octeontx_fpavf.c b/drivers/mempool/octeontx/octeontx_fpavf.c
index baabc0152e..ec84a5cff2 100644
--- a/drivers/mempool/octeontx/octeontx_fpavf.c
+++ b/drivers/mempool/octeontx/octeontx_fpavf.c
@@ -507,6 +507,7 @@ octeontx_fpa_bufpool_create(unsigned int object_size, unsigned int object_count,
 	RTE_SET_USED(node_id);
 	RTE_BUILD_BUG_ON(sizeof(struct rte_mbuf) > OCTEONTX_FPAVF_BUF_OFFSET);
 
+	octeontx_mbox_init();
 	object_size = RTE_CACHE_LINE_ROUNDUP(object_size);
 	if (object_size > FPA_MAX_OBJ_SIZE) {
 		errno = EINVAL;
diff --git a/drivers/net/octeontx/base/octeontx_bgx.h b/drivers/net/octeontx/base/octeontx_bgx.h
index ff26514974..a9bbaf2c3c 100644
--- a/drivers/net/octeontx/base/octeontx_bgx.h
+++ b/drivers/net/octeontx/base/octeontx_bgx.h
@@ -27,6 +27,7 @@
 #define MBOX_BGX_PORT_SET_BP            11
 #define MBOX_BGX_PORT_SET_BCAST         12
 #define MBOX_BGX_PORT_SET_MCAST         13
+#define MBOX_BGX_PORT_SET_MTU		14
 
 /* BGX port configuration parameters: */
 typedef struct octeontx_mbox_bgx_port_conf {
@@ -51,6 +52,8 @@ typedef struct octeontx_mbox_bgx_port_conf {
 typedef struct octeontx_mbox_bgx_port_status {
 	uint8_t link_up;
 	uint8_t bp;
+	uint8_t duplex;
+	uint32_t speed;
 } octeontx_mbox_bgx_port_status_t;
 
 /* BGX port statistics: */
diff --git a/drivers/net/octeontx/base/octeontx_pkivf.h b/drivers/net/octeontx/base/octeontx_pkivf.h
index 764aff53f7..7f19a4bb8f 100644
--- a/drivers/net/octeontx/base/octeontx_pkivf.h
+++ b/drivers/net/octeontx/base/octeontx_pkivf.h
@@ -33,6 +33,9 @@
 #define MBOX_PKI_PORT_RESET_STATS		18
 #define MBOX_PKI_GET_PORT_CONFIG		19
 #define MBOX_PKI_GET_PORT_QOS_CONFIG		20
+#define MBOX_PKI_PORT_ALLOC_QPG			21
+#define MBOX_PKI_PORT_FREE_QPG			22
+#define MBOX_PKI_SET_PORT_CONFIG		23
 
 #define MBOX_PKI_MAX_QOS_ENTRY 64
 
@@ -64,6 +67,7 @@ typedef struct mbox_pki_port_cfg {
 	struct {
 		uint8_t fcs_pres:1;
 		uint8_t fcs_skip:1;
+		uint8_t inst_skip:1;
 		uint8_t parse_mode:1;
 		uint8_t mpls_parse:1;
 		uint8_t inst_hdr_parse:1;
@@ -74,6 +78,7 @@ typedef struct mbox_pki_port_cfg {
 	} mmask;
 	uint8_t fcs_pres;
 	uint8_t fcs_skip;
+	uint8_t inst_skip;
 	uint8_t parse_mode;
 	uint8_t mpls_parse;
 	uint8_t inst_hdr_parse;
@@ -189,6 +194,9 @@ struct mbox_pki_qos_entry {
 	uint16_t gaura;
 	uint8_t grptag_ok;
 	uint8_t grptag_bad;
+	uint8_t ena_red;
+	uint8_t ena_drop;
+	uint8_t tag_type;
 };
 
 /* pki flow/style enable qos */
@@ -201,7 +209,7 @@ typedef struct mbox_pki_port_create_qos {
 	struct mbox_pki_qos_entry qos_entry[MBOX_PKI_MAX_QOS_ENTRY];
 } mbox_pki_qos_cfg_t;
 
-/* pki flow/style enable qos */
+/* pki flow/style modify qos */
 typedef struct mbox_pki_port_modify_qos_entry {
 	uint8_t port_type;
 	uint16_t index;
@@ -214,11 +222,10 @@ typedef struct mbox_pki_port_modify_qos_entry {
 		uint8_t f_grptag_bad:1;
 		uint8_t f_tag_type:1;
 	} mmask;
-	uint8_t tag_type;
 	struct mbox_pki_qos_entry qos_entry;
 } mbox_pki_mod_qos_t;
 
-/* pki flow/style enable qos */
+/* pki flow/style delete qos */
 typedef struct mbox_pki_port_delete_qos_entry {
 	uint8_t port_type;
 	uint16_t index;
@@ -372,6 +379,7 @@ struct pki_qos_entry {
 	uint8_t grptag_bad;
 	uint8_t ena_red;
 	uint8_t ena_drop;
+	uint8_t tag_type;
 };
 
 #define PKO_MAX_QOS_ENTRY 64
@@ -405,7 +413,6 @@ typedef struct pki_port_modify_qos_entry {
 		uint8_t f_grptag_bad:1;
 		uint8_t f_tag_type:1;
 	} mmask;
-	uint8_t tag_type;
 	struct pki_qos_entry qos_entry;
 } pki_mod_qos_t;