DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN
@ 2020-12-21  3:50 yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 1/8] gro: support TCP/IPv6 yang_y_yi
                   ` (8 more replies)
  0 siblings, 9 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

In production system, both IPv4 and IPv6 are required, this patch
series added TCP/IPv6 and UDP/IPv6 gro support for both VLAN and
VXLAN use cases, with this patch seiries applied, DPDK can support gro
for all the below cases:

VLAN TCP/IPv4
VLAN UDP/IPv4
VLAN TCP/IPv6
VLAN UDP/IPv6
IPv4 VXLAN TCP/IPv4
IPv4 VXLAN UDP/IPv4
IPv4 VXLAN TCP/IPv6
IPv4 VXLAN UDP/IPv6
IPv6 VXLAN TCP/IPv4
IPv6 VXLAN UDP/IPv4
IPv6 VXLAN TCP/IPv6
IPv4 VXLAN UDP/IPv6

Yi Yang (8):
  gro: support TCP/IPv6
  gro: support IPv4 VXLAN TCP/IPv6
  gro: support IPv6 VXLAN TCP/IPv4
  gro: support IPv6 VXLAN TCP/IPv6
  gro: support UDP/IPv6
  gro: support IPv4 VXLAN UDP/IPv6
  gro: support IPv6 VXLAN UDP/IPv4
  gro: support IPv6 VXLAN UDP/IPv6

 .../prog_guide/generic_receive_offload_lib.rst     |  16 +-
 doc/guides/rel_notes/release_21_02.rst             |  39 ++
 lib/librte_gro/gro_tcp6.c                          | 368 +++++++++++++
 lib/librte_gro/gro_tcp6.h                          | 314 +++++++++++
 lib/librte_gro/gro_udp6.c                          | 487 +++++++++++++++++
 lib/librte_gro/gro_udp6.h                          | 284 ++++++++++
 lib/librte_gro/gro_vxlan6_tcp4.c                   | 507 +++++++++++++++++
 lib/librte_gro/gro_vxlan6_tcp4.h                   | 153 ++++++
 lib/librte_gro/gro_vxlan6_tcp6.c                   | 464 ++++++++++++++++
 lib/librte_gro/gro_vxlan6_tcp6.h                   | 153 ++++++
 lib/librte_gro/gro_vxlan6_udp4.c                   | 559 +++++++++++++++++++
 lib/librte_gro/gro_vxlan6_udp4.h                   | 148 +++++
 lib/librte_gro/gro_vxlan6_udp6.c                   | 589 ++++++++++++++++++++
 lib/librte_gro/gro_vxlan6_udp6.h                   | 148 +++++
 lib/librte_gro/gro_vxlan_tcp6.c                    | 496 +++++++++++++++++
 lib/librte_gro/gro_vxlan_tcp6.h                    | 156 ++++++
 lib/librte_gro/gro_vxlan_udp6.c                    | 607 +++++++++++++++++++++
 lib/librte_gro/gro_vxlan_udp6.h                    | 152 ++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           | 537 +++++++++++++++++-
 lib/librte_gro/rte_gro.h                           |  24 +
 21 files changed, 6186 insertions(+), 17 deletions(-)
 create mode 100644 lib/librte_gro/gro_tcp6.c
 create mode 100644 lib/librte_gro/gro_tcp6.h
 create mode 100644 lib/librte_gro/gro_udp6.c
 create mode 100644 lib/librte_gro/gro_udp6.h
 create mode 100644 lib/librte_gro/gro_vxlan6_tcp4.c
 create mode 100644 lib/librte_gro/gro_vxlan6_tcp4.h
 create mode 100644 lib/librte_gro/gro_vxlan6_tcp6.c
 create mode 100644 lib/librte_gro/gro_vxlan6_tcp6.h
 create mode 100644 lib/librte_gro/gro_vxlan6_udp4.c
 create mode 100644 lib/librte_gro/gro_vxlan6_udp4.h
 create mode 100644 lib/librte_gro/gro_vxlan6_udp6.c
 create mode 100644 lib/librte_gro/gro_vxlan6_udp6.h
 create mode 100644 lib/librte_gro/gro_vxlan_tcp6.c
 create mode 100644 lib/librte_gro/gro_vxlan_tcp6.h
 create mode 100644 lib/librte_gro/gro_vxlan_udp6.c
 create mode 100644 lib/librte_gro/gro_vxlan_udp6.h

-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-dev] [PATCH v1 1/8] gro: support TCP/IPv6
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
@ 2020-12-21  3:50 ` yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 2/8] gro: support IPv4 VXLAN TCP/IPv6 yang_y_yi
                   ` (7 subsequent siblings)
  8 siblings, 0 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

TCP/IPv6 GRO can help improve TCP/IPv6 performance.

With this enabled in DPDK, OVS DPDK can leverage it
to improve VM-to-VM TCP/IPv6 performance, it will merge
small adjacent TCP/IPv6 segments to a big TCP/IPv6 segment
immediate after they are received from a physical NIC. It
is very helpful in OVS DPDK VLAN use case.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
---
 .../prog_guide/generic_receive_offload_lib.rst     |  10 +-
 doc/guides/rel_notes/release_21_02.rst             |   5 +
 lib/librte_gro/gro_tcp6.c                          | 368 +++++++++++++++++++++
 lib/librte_gro/gro_tcp6.h                          | 314 ++++++++++++++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           |  74 ++++-
 lib/librte_gro/rte_gro.h                           |   3 +
 7 files changed, 762 insertions(+), 14 deletions(-)
 create mode 100644 lib/librte_gro/gro_tcp6.c
 create mode 100644 lib/librte_gro/gro_tcp6.h

diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst b/doc/guides/prog_guide/generic_receive_offload_lib.rst
index 98a5d29..fee856e 100644
--- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
+++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
@@ -30,9 +30,9 @@ assumes the packets are complete (i.e., MF==0 && frag_off==0), when IP
 fragmentation is possible (i.e., DF==0). Additionally, it complies RFC
 6864 to process the IPv4 ID field.
 
-Currently, the GRO library provides GRO supports for TCP/IPv4 and UDP/IPv4
-packets as well as VxLAN packets which contain an outer IPv4 header and an
-inner TCP/IPv4 or UDP/IPv4 packet.
+Currently, the GRO library provides GRO supports for TCP/IPv4, UDP/IPv4,
+and TCP/IPv6 packets as well as VxLAN packets which contain an outer IPv4
+header and an inner TCP/IPv4 or UDP/IPv4 packet.
 
 Two Sets of API
 ---------------
@@ -83,8 +83,8 @@ Reassembly Algorithm
 
 The reassembly algorithm is used for reassembling packets. In the GRO
 library, different GRO types can use different algorithms. In this
-section, we will introduce an algorithm, which is used by TCP/IPv4 GRO
-and VxLAN GRO.
+section, we will introduce an algorithm, which is used by TCP/IPv4 GRO,
+TCP/IPv6 GRO and VxLAN GRO.
 
 Challenges
 ~~~~~~~~~~
diff --git a/doc/guides/rel_notes/release_21_02.rst b/doc/guides/rel_notes/release_21_02.rst
index 638f981..1a1aeff 100644
--- a/doc/guides/rel_notes/release_21_02.rst
+++ b/doc/guides/rel_notes/release_21_02.rst
@@ -55,6 +55,11 @@ New Features
      Also, make sure to start the actual text at the margin.
      =======================================================
 
+* **Added TCP/IPv6 GRO support for non-VXLAN packets.**
+
+   Added TCP/IPv6 GRO support for non-VXLAN packets, this enabled TCP/IPv6
+   GRO support for VLAN use case.
+
 
 Removed Items
 -------------
diff --git a/lib/librte_gro/gro_tcp6.c b/lib/librte_gro/gro_tcp6.c
new file mode 100644
index 0000000..4a4df34
--- /dev/null
+++ b/lib/librte_gro/gro_tcp6.c
@@ -0,0 +1,368 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+
+#include "gro_tcp6.h"
+
+void *
+gro_tcp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow)
+{
+	struct gro_tcp6_tbl *tbl;
+	size_t size;
+	uint32_t entries_num, i;
+
+	entries_num = max_flow_num * max_item_per_flow;
+	entries_num = RTE_MIN(entries_num, GRO_TCP6_TBL_MAX_ITEM_NUM);
+
+	if (entries_num == 0)
+		return NULL;
+
+	tbl = rte_zmalloc_socket(__func__,
+			sizeof(struct gro_tcp6_tbl),
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl == NULL)
+		return NULL;
+
+	size = sizeof(struct gro_tcp6_item) * entries_num;
+	tbl->items = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->items == NULL) {
+		rte_free(tbl);
+		return NULL;
+	}
+	tbl->max_item_num = entries_num;
+
+	size = sizeof(struct gro_tcp6_flow) * entries_num;
+	tbl->flows = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->flows == NULL) {
+		rte_free(tbl->items);
+		rte_free(tbl);
+		return NULL;
+	}
+	/* INVALID_ARRAY_INDEX indicates an empty flow */
+	for (i = 0; i < entries_num; i++)
+		tbl->flows[i].start_index = INVALID_ARRAY_INDEX;
+	tbl->max_flow_num = entries_num;
+
+	return tbl;
+}
+
+void
+gro_tcp6_tbl_destroy(void *tbl)
+{
+	struct gro_tcp6_tbl *tcp_tbl = tbl;
+
+	if (tcp_tbl) {
+		rte_free(tcp_tbl->items);
+		rte_free(tcp_tbl->flows);
+	}
+	rte_free(tcp_tbl);
+}
+
+static inline uint32_t
+find_an_empty_item(struct gro_tcp6_tbl *tbl)
+{
+	uint32_t i;
+	uint32_t max_item_num = tbl->max_item_num;
+
+	for (i = 0; i < max_item_num; i++)
+		if (tbl->items[i].firstseg == NULL)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+find_an_empty_flow(struct gro_tcp6_tbl *tbl)
+{
+	uint32_t i;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++)
+		if (tbl->flows[i].start_index == INVALID_ARRAY_INDEX)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_item(struct gro_tcp6_tbl *tbl,
+		struct rte_mbuf *pkt,
+		uint64_t start_time,
+		uint32_t prev_idx,
+		uint32_t sent_seq)
+{
+	uint32_t item_idx;
+
+	item_idx = find_an_empty_item(tbl);
+	if (item_idx == INVALID_ARRAY_INDEX)
+		return INVALID_ARRAY_INDEX;
+
+	tbl->items[item_idx].firstseg = pkt;
+	tbl->items[item_idx].lastseg = rte_pktmbuf_lastseg(pkt);
+	tbl->items[item_idx].start_time = start_time;
+	tbl->items[item_idx].next_pkt_idx = INVALID_ARRAY_INDEX;
+	tbl->items[item_idx].sent_seq = sent_seq;
+	tbl->items[item_idx].nb_merged = 1;
+	tbl->item_num++;
+
+	/* if the previous packet exists, chain them together. */
+	if (prev_idx != INVALID_ARRAY_INDEX) {
+		tbl->items[item_idx].next_pkt_idx =
+			tbl->items[prev_idx].next_pkt_idx;
+		tbl->items[prev_idx].next_pkt_idx = item_idx;
+	}
+
+	return item_idx;
+}
+
+static inline uint32_t
+delete_item(struct gro_tcp6_tbl *tbl, uint32_t item_idx,
+		uint32_t prev_item_idx)
+{
+	uint32_t next_idx = tbl->items[item_idx].next_pkt_idx;
+
+	/* NULL indicates an empty item */
+	tbl->items[item_idx].firstseg = NULL;
+	tbl->item_num--;
+	if (prev_item_idx != INVALID_ARRAY_INDEX)
+		tbl->items[prev_item_idx].next_pkt_idx = next_idx;
+
+	return next_idx;
+}
+
+static inline uint32_t
+insert_new_flow(struct gro_tcp6_tbl *tbl,
+		struct tcp6_flow_key *src,
+		uint32_t item_idx)
+{
+	struct tcp6_flow_key *dst;
+	uint32_t flow_idx;
+
+	flow_idx = find_an_empty_flow(tbl);
+	if (unlikely(flow_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	dst = &(tbl->flows[flow_idx].key);
+
+	rte_ether_addr_copy(&(src->eth_saddr), &(dst->eth_saddr));
+	rte_ether_addr_copy(&(src->eth_daddr), &(dst->eth_daddr));
+	gro_ipv6_addr_copy(src->ip_saddr, dst->ip_saddr);
+	gro_ipv6_addr_copy(src->ip_daddr, dst->ip_daddr);
+	dst->recv_ack = src->recv_ack;
+	dst->src_port = src->src_port;
+	dst->dst_port = src->dst_port;
+
+	tbl->flows[flow_idx].start_index = item_idx;
+	tbl->flow_num++;
+
+	return flow_idx;
+}
+
+/*
+ * update the packet length for the flushed packet.
+ */
+static inline void
+update_header(struct gro_tcp6_item *item)
+{
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_mbuf *pkt = item->firstseg;
+
+	ipv6_hdr = (struct rte_ipv6_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+			pkt->l2_len);
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(pkt->pkt_len -
+			pkt->l2_len - pkt->l3_len);
+}
+
+int32_t
+gro_tcp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_tcp6_tbl *tbl,
+		uint64_t start_time)
+{
+	struct rte_ether_hdr *eth_hdr;
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_tcp_hdr *tcp_hdr;
+	uint32_t sent_seq;
+	int32_t tcp_dl;
+	uint16_t hdr_len;
+
+	struct tcp6_flow_key key;
+	uint32_t cur_idx, prev_idx, item_idx;
+	uint32_t i, max_flow_num, remaining_flow_num;
+	int cmp;
+	uint8_t find;
+
+	/*
+	 * Don't process the packet whose TCP header length is greater
+	 * than 60 bytes or less than 20 bytes.
+	 */
+	if (unlikely(INVALID_TCP_HDRLEN(pkt->l4_len)))
+		return -1;
+
+	eth_hdr = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *);
+	ipv6_hdr = (struct rte_ipv6_hdr *)((char *)eth_hdr + pkt->l2_len);
+	tcp_hdr = (struct rte_tcp_hdr *)((char *)ipv6_hdr + pkt->l3_len);
+	hdr_len = pkt->l2_len + pkt->l3_len + pkt->l4_len;
+
+	/*
+	 * Don't process the packet which has FIN, SYN, RST, PSH, URG, ECE
+	 * or CWR set.
+	 */
+	if (tcp_hdr->tcp_flags != RTE_TCP_ACK_FLAG)
+		return -1;
+	/*
+	 * Don't process the packet whose payload length is less than or
+	 * equal to 0.
+	 */
+	if (pkt->pkt_len <= hdr_len)
+		return -1;
+	tcp_dl = pkt->pkt_len - hdr_len;
+
+	/*
+	 * Save IPv6 ID for the packet whose DF bit is 0. For the packet
+	 * whose DF bit is 1, IPv6 ID is ignored.
+	 */
+	sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
+
+	rte_ether_addr_copy(&(eth_hdr->s_addr), &(key.eth_saddr));
+	rte_ether_addr_copy(&(eth_hdr->d_addr), &(key.eth_daddr));
+	gro_ipv6_addr_copy(ipv6_hdr->src_addr, key.ip_saddr);
+	gro_ipv6_addr_copy(ipv6_hdr->dst_addr, key.ip_daddr);
+	key.src_port = tcp_hdr->src_port;
+	key.dst_port = tcp_hdr->dst_port;
+	key.recv_ack = tcp_hdr->recv_ack;
+
+	/* Search for a matched flow. */
+	max_flow_num = tbl->max_flow_num;
+	remaining_flow_num = tbl->flow_num;
+	find = 0;
+	for (i = 0; i < max_flow_num && remaining_flow_num; i++) {
+		if (tbl->flows[i].start_index != INVALID_ARRAY_INDEX) {
+			if (is_same_tcp6_flow(tbl->flows[i].key, key)) {
+				find = 1;
+				break;
+			}
+			remaining_flow_num--;
+		}
+	}
+
+	/*
+	 * Fail to find a matched flow. Insert a new flow and store the
+	 * packet into the flow.
+	 */
+	if (find == 0) {
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, sent_seq);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		if (insert_new_flow(tbl, &key, item_idx) ==
+				INVALID_ARRAY_INDEX) {
+			/*
+			 * Fail to insert a new flow, so delete the
+			 * stored packet.
+			 */
+			delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+			return -1;
+		}
+		return 0;
+	}
+
+	/*
+	 * Check all packets in the flow and try to find a neighbor for
+	 * the input packet.
+	 */
+	cur_idx = tbl->flows[i].start_index;
+	prev_idx = cur_idx;
+	do {
+		cmp = check_seq_option_v6(&(tbl->items[cur_idx]), tcp_hdr,
+				sent_seq, pkt->l4_len, tcp_dl, 0);
+		if (cmp) {
+			if (merge_two_tcp6_packets(&(tbl->items[cur_idx]),
+						pkt, cmp, sent_seq, 0))
+				return 1;
+			/*
+			 * Fail to merge the two packets, as the packet
+			 * length is greater than the max value. Store
+			 * the packet into the flow.
+			 */
+			if (insert_new_item(tbl, pkt, start_time, prev_idx,
+						sent_seq) ==
+					INVALID_ARRAY_INDEX)
+				return -1;
+			return 0;
+		}
+		prev_idx = cur_idx;
+		cur_idx = tbl->items[cur_idx].next_pkt_idx;
+	} while (cur_idx != INVALID_ARRAY_INDEX);
+
+	/* Fail to find a neighbor, so store the packet into the flow. */
+	if (insert_new_item(tbl, pkt, start_time, prev_idx, sent_seq
+				) == INVALID_ARRAY_INDEX)
+		return -1;
+
+	return 0;
+}
+
+uint16_t
+gro_tcp6_tbl_timeout_flush(struct gro_tcp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out)
+{
+	uint16_t k = 0;
+	uint32_t i, j;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++) {
+		if (unlikely(tbl->flow_num == 0))
+			return k;
+
+		j = tbl->flows[i].start_index;
+		while (j != INVALID_ARRAY_INDEX) {
+			if (tbl->items[j].start_time <= flush_timestamp) {
+				out[k++] = tbl->items[j].firstseg;
+				if (tbl->items[j].nb_merged > 1)
+					update_header(&(tbl->items[j]));
+				/*
+				 * Delete the packet and get the next
+				 * packet in the flow.
+				 */
+				j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+				tbl->flows[i].start_index = j;
+				if (j == INVALID_ARRAY_INDEX)
+					tbl->flow_num--;
+
+				if (unlikely(k == nb_out))
+					return k;
+			} else
+				/*
+				 * The left packets in this flow won't be
+				 * timeout. Go to check other flows.
+				 */
+				break;
+		}
+	}
+	return k;
+}
+
+uint32_t
+gro_tcp6_tbl_pkt_count(void *tbl)
+{
+	struct gro_tcp6_tbl *gro_tbl = tbl;
+
+	if (gro_tbl)
+		return gro_tbl->item_num;
+
+	return 0;
+}
diff --git a/lib/librte_gro/gro_tcp6.h b/lib/librte_gro/gro_tcp6.h
new file mode 100644
index 0000000..a9dec7b
--- /dev/null
+++ b/lib/librte_gro/gro_tcp6.h
@@ -0,0 +1,314 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#ifndef _GRO_TCP6_H_
+#define _GRO_TCP6_H_
+
+#include <rte_ip.h>
+#include <rte_tcp.h>
+#include <rte_vxlan.h>
+
+#define INVALID_ARRAY_INDEX 0xffffffffUL
+#define GRO_TCP6_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/*
+ * The max length of a IPv6 packet, which includes the length of the L3
+ * header, the L4 header and the data payload.
+ */
+#define MAX_IPV6_PKT_LENGTH UINT16_MAX
+
+/* The maximum TCP header length */
+#define MAX_TCP_HLEN 60
+#define INVALID_TCP_HDRLEN(len) \
+	(((len) < sizeof(struct rte_tcp_hdr)) || ((len) > MAX_TCP_HLEN))
+
+/* Header fields representing a TCP/IPv6 flow */
+struct tcp6_flow_key {
+	struct rte_ether_addr eth_saddr;
+	struct rte_ether_addr eth_daddr;
+	uint8_t  ip_saddr[16];	/**< IPv6 address of source host. */
+	uint8_t  ip_daddr[16];	/**< IPv6 address of destination host(s). */
+
+	uint32_t recv_ack;
+	uint16_t src_port;
+	uint16_t dst_port;
+};
+
+struct gro_tcp6_flow {
+	struct tcp6_flow_key key;
+	/*
+	 * The index of the first packet in the flow.
+	 * INVALID_ARRAY_INDEX indicates an empty flow.
+	 */
+	uint32_t start_index;
+};
+
+struct gro_tcp6_item {
+	/*
+	 * The first MBUF segment of the packet. If the value
+	 * is NULL, it means the item is empty.
+	 */
+	struct rte_mbuf *firstseg;
+	/* The last MBUF segment of the packet */
+	struct rte_mbuf *lastseg;
+	/*
+	 * The time when the first packet is inserted into the table.
+	 * This value won't be updated, even if the packet is merged
+	 * with other packets.
+	 */
+	uint64_t start_time;
+	/*
+	 * next_pkt_idx is used to chain the packets that
+	 * are in the same flow but can't be merged together
+	 * (e.g. caused by packet reordering).
+	 */
+	uint32_t next_pkt_idx;
+	/* TCP sequence number of the packet */
+	uint32_t sent_seq;
+	/* the number of merged packets */
+	uint16_t nb_merged;
+};
+
+/*
+ * TCP/IPv6 reassembly table structure.
+ */
+struct gro_tcp6_tbl {
+	/* item array */
+	struct gro_tcp6_item *items;
+	/* flow array */
+	struct gro_tcp6_flow *flows;
+	/* current item number */
+	uint32_t item_num;
+	/* current flow num */
+	uint32_t flow_num;
+	/* item array size */
+	uint32_t max_item_num;
+	/* flow array size */
+	uint32_t max_flow_num;
+};
+
+/**
+ * This function creates a TCP/IPv6 reassembly table.
+ *
+ * @param socket_id
+ *  Socket index for allocating the TCP/IPv6 reassemble table
+ * @param max_flow_num
+ *  The maximum number of flows in the TCP/IPv6 GRO table
+ * @param max_item_per_flow
+ *  The maximum number of packets per flow
+ *
+ * @return
+ *  - Return the table pointer on success.
+ *  - Return NULL on failure.
+ */
+void *gro_tcp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow);
+
+/**
+ * This function destroys a TCP/IPv6 reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the TCP/IPv6 reassembly table.
+ */
+void gro_tcp6_tbl_destroy(void *tbl);
+
+/**
+ * This function merges a TCP/IPv6 packet. It does not process the packet,
+ * which has SYN, FIN, RST, PSH, CWR, ECE or URG set, or does not have
+ * payload.
+ *
+ * This function does not check if the packet has correct checksums and
+ * does not re-calculate checksums for the merged packet. Additionally,
+ * it assumes the packets are complete (i.e., MF==0 && frag_off==0),
+ * when IP fragmentation is possible (i.e., DF==0). It returns the
+ * packet, if the packet has invalid parameters (e.g. SYN bit is set)
+ * or there is no available space in the table.
+ *
+ * @param pkt
+ *  Packet to reassemble
+ * @param tbl
+ *  Pointer pointing to the TCP/IPv6 reassembly table
+ * @start_time
+ *  The time when the packet is inserted into the table
+ *
+ * @return
+ *  - Return a positive value if the packet is merged.
+ *  - Return zero if the packet isn't merged but stored in the table.
+ *  - Return a negative value for invalid parameters or no available
+ *    space in the table.
+ */
+int32_t gro_tcp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_tcp6_tbl *tbl,
+		uint64_t start_time);
+
+/**
+ * This function flushes timeout packets in a TCP/IPv6 reassembly table,
+ * and without updating checksums.
+ *
+ * @param tbl
+ *  TCP/IPv6 reassembly table pointer
+ * @param flush_timestamp
+ *  Flush packets which are inserted into the table before or at the
+ *  flush_timestamp.
+ * @param out
+ *  Pointer array used to keep flushed packets
+ * @param nb_out
+ *  The element number in 'out'. It also determines the maximum number of
+ *  packets that can be flushed finally.
+ *
+ * @return
+ *  The number of flushed packets
+ */
+uint16_t gro_tcp6_tbl_timeout_flush(struct gro_tcp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out);
+
+/**
+ * This function returns the number of the packets in a TCP/IPv6
+ * reassembly table.
+ *
+ * @param tbl
+ *  TCP/IPv6 reassembly table pointer
+ *
+ * @return
+ *  The number of packets in the table
+ */
+uint32_t gro_tcp6_tbl_pkt_count(void *tbl);
+
+/*
+ * Check if two IPv6 address is same.
+ */
+static inline int rte_is_same_ipv6_addr(const uint8_t *ipv6_a1,
+					const uint8_t *ipv6_a2)
+{
+	const uint64_t *w1 = (const uint64_t *)ipv6_a1;
+	const uint64_t *w2 = (const uint64_t *)ipv6_a2;
+
+	return ((w1[0] ^ w2[0]) | (w1[1] ^ w2[1])) == 0;
+}
+
+/* Copy IPv6 addr */
+static inline void gro_ipv6_addr_copy(const uint8_t *ipv6_from,
+				      uint8_t *ipv6_to)
+{
+	const uint64_t *from_words = (const uint64_t *)ipv6_from;
+	uint64_t *to_words   = (uint64_t *)ipv6_to;
+
+	to_words[0] = from_words[0];
+	to_words[1] = from_words[1];
+}
+
+/*
+ * Check if two TCP/IPv6 packets belong to the same flow.
+ */
+static inline int
+is_same_tcp6_flow(struct tcp6_flow_key k1, struct tcp6_flow_key k2)
+{
+	return (rte_is_same_ether_addr(&k1.eth_saddr, &k2.eth_saddr) &&
+			rte_is_same_ether_addr(&k1.eth_daddr, &k2.eth_daddr) &&
+			rte_is_same_ipv6_addr(k1.ip_saddr, k2.ip_saddr) &&
+			rte_is_same_ipv6_addr(k1.ip_daddr, k2.ip_daddr) &&
+			(k1.recv_ack == k2.recv_ack) &&
+			(k1.src_port == k2.src_port) &&
+			(k1.dst_port == k2.dst_port));
+}
+
+/*
+ * Merge two TCP/IPv6, packets without updating checksums.
+ * If cmp is larger than 0, append the new packet to the
+ * original packet. Otherwise, pre-pend the new packet to
+ * the original packet.
+ */
+static inline int
+merge_two_tcp6_packets(struct gro_tcp6_item *item,
+		struct rte_mbuf *pkt,
+		int cmp,
+		uint32_t sent_seq,
+		uint16_t l2_offset)
+{
+	struct rte_mbuf *pkt_head, *pkt_tail, *lastseg;
+	uint16_t hdr_len, l2_len;
+
+	if (cmp > 0) {
+		pkt_head = item->firstseg;
+		pkt_tail = pkt;
+	} else {
+		pkt_head = pkt;
+		pkt_tail = item->firstseg;
+	}
+
+	/* check if the IPv6 packet length is greater than the max value */
+	hdr_len = l2_offset + pkt_head->l2_len + pkt_head->l3_len +
+		pkt_head->l4_len;
+	l2_len = l2_offset > 0 ? pkt_head->outer_l2_len : pkt_head->l2_len;
+	if (unlikely(pkt_head->pkt_len - l2_len + pkt_tail->pkt_len -
+				hdr_len > MAX_IPV6_PKT_LENGTH))
+		return 0;
+
+	/* remove the packet header for the tail packet */
+	rte_pktmbuf_adj(pkt_tail, hdr_len);
+
+	/* chain two packets together */
+	if (cmp > 0) {
+		item->lastseg->next = pkt;
+		item->lastseg = rte_pktmbuf_lastseg(pkt);
+	} else {
+		lastseg = rte_pktmbuf_lastseg(pkt);
+		lastseg->next = item->firstseg;
+		item->firstseg = pkt;
+		/* update sent_seq to the smaller value */
+		item->sent_seq = sent_seq;
+	}
+	item->nb_merged++;
+
+	/* update MBUF metadata for the merged packet */
+	pkt_head->nb_segs += pkt_tail->nb_segs;
+	pkt_head->pkt_len += pkt_tail->pkt_len;
+
+	return 1;
+}
+
+/*
+ * Check if two TCP/IPv6 packets are neighbors.
+ */
+static inline int
+check_seq_option_v6(struct gro_tcp6_item *item,
+		struct rte_tcp_hdr *tcph,
+		uint32_t sent_seq,
+		uint16_t tcp_hl,
+		uint16_t tcp_dl,
+		uint16_t l2_offset)
+{
+	struct rte_mbuf *pkt_orig = item->firstseg;
+	struct rte_ipv6_hdr *iph_orig;
+	struct rte_tcp_hdr *tcph_orig;
+	uint16_t len, tcp_hl_orig;
+
+	iph_orig = (struct rte_ipv6_hdr *)(rte_pktmbuf_mtod(pkt_orig, char *) +
+			l2_offset + pkt_orig->l2_len);
+	tcph_orig = (struct rte_tcp_hdr *)((char *)iph_orig + pkt_orig->l3_len);
+	tcp_hl_orig = pkt_orig->l4_len;
+
+	/* Check if TCP option fields equal */
+	len = RTE_MAX(tcp_hl, tcp_hl_orig) - sizeof(struct rte_tcp_hdr);
+	if ((tcp_hl != tcp_hl_orig) || ((len > 0) &&
+				(memcmp(tcph + 1, tcph_orig + 1,
+					len) != 0)))
+		return 0;
+
+	/* check if the two packets are neighbors */
+	len = pkt_orig->pkt_len - l2_offset - pkt_orig->l2_len -
+		pkt_orig->l3_len - tcp_hl_orig;
+	if (sent_seq == item->sent_seq + len)
+		/* append the new packet */
+		return 1;
+	else if (sent_seq + tcp_dl == item->sent_seq)
+		/* pre-pend the new packet */
+		return -1;
+
+	return 0;
+}
+#endif
diff --git a/lib/librte_gro/meson.build b/lib/librte_gro/meson.build
index ea8b45c..6143dd0 100644
--- a/lib/librte_gro/meson.build
+++ b/lib/librte_gro/meson.build
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c')
+sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c')
 headers = files('rte_gro.h')
 deps += ['ethdev']
diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
index 8ca4da6..86958d5 100644
--- a/lib/librte_gro/rte_gro.c
+++ b/lib/librte_gro/rte_gro.c
@@ -9,6 +9,7 @@
 
 #include "rte_gro.h"
 #include "gro_tcp4.h"
+#include "gro_tcp6.h"
 #include "gro_udp4.h"
 #include "gro_vxlan_tcp4.h"
 #include "gro_vxlan_udp4.h"
@@ -21,15 +22,16 @@
 
 static gro_tbl_create_fn tbl_create_fn[RTE_GRO_TYPE_MAX_NUM] = {
 		gro_tcp4_tbl_create, gro_vxlan_tcp4_tbl_create,
-		gro_udp4_tbl_create, gro_vxlan_udp4_tbl_create, NULL};
+		gro_udp4_tbl_create, gro_vxlan_udp4_tbl_create,
+		gro_tcp6_tbl_create, NULL};
 static gro_tbl_destroy_fn tbl_destroy_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_destroy, gro_vxlan_tcp4_tbl_destroy,
 			gro_udp4_tbl_destroy, gro_vxlan_udp4_tbl_destroy,
-			NULL};
+			gro_tcp6_tbl_destroy, NULL};
 static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_pkt_count, gro_vxlan_tcp4_tbl_pkt_count,
 			gro_udp4_tbl_pkt_count, gro_vxlan_udp4_tbl_pkt_count,
-			NULL};
+			gro_tcp6_tbl_pkt_count, NULL};
 
 #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
@@ -65,6 +67,10 @@
 		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
 		  RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN)))
 
+#define IS_IPV6_TCP_PKT(ptype) (RTE_ETH_IS_IPV6_HDR(ptype) && \
+		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
+		(RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
+
 /*
  * GRO context structure. It keeps the table structures, which are
  * used to merge packets, for different GRO types. Before using
@@ -165,17 +171,24 @@ struct gro_ctx {
 	struct gro_vxlan_udp4_item vxlan_udp_items[RTE_GRO_MAX_BURST_ITEM_NUM]
 			= {{{0}} };
 
+	/* allocate a reassembly table for TCP/IPv6 GRO */
+	struct gro_tcp6_tbl tcp6_tbl;
+	struct gro_tcp6_flow tcp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+	struct gro_tcp6_item tcp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
+
+
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	uint32_t item_num;
 	int32_t ret;
 	uint16_t i, unprocess_num = 0, nb_after_gro = nb_pkts;
 	uint8_t do_tcp4_gro = 0, do_vxlan_tcp_gro = 0, do_udp4_gro = 0,
-		do_vxlan_udp_gro = 0;
+		do_vxlan_udp_gro = 0, do_tcp6_gro = 0;
 
 	if (unlikely((param->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
 					RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
-					RTE_GRO_UDP_IPV4)) == 0))
+					RTE_GRO_UDP_IPV4 |
+					RTE_GRO_TCP_IPV6)) == 0))
 		return nb_pkts;
 
 	/* Get the maximum number of packets */
@@ -235,6 +248,20 @@ struct gro_ctx {
 		do_udp4_gro = 1;
 	}
 
+	if (param->gro_types & RTE_GRO_TCP_IPV6) {
+		for (i = 0; i < item_num; i++)
+			tcp6_flows[i].start_index = INVALID_ARRAY_INDEX;
+
+		tcp6_tbl.flows = tcp6_flows;
+		tcp6_tbl.items = tcp6_items;
+		tcp6_tbl.flow_num = 0;
+		tcp6_tbl.item_num = 0;
+		tcp6_tbl.max_flow_num = item_num;
+		tcp6_tbl.max_item_num = item_num;
+		do_tcp6_gro = 1;
+	}
+
+
 
 	for (i = 0; i < nb_pkts; i++) {
 		/*
@@ -275,6 +302,14 @@ struct gro_ctx {
 				nb_after_gro--;
 			else if (ret < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_TCP_PKT(pkts[i]->packet_type) &&
+				do_tcp6_gro) {
+			ret = gro_tcp6_reassemble(pkts[i], &tcp6_tbl, 0);
+			if (ret > 0)
+				/* merge successfully */
+				nb_after_gro--;
+			else if (ret < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -303,6 +338,12 @@ struct gro_ctx {
 			i += gro_udp4_tbl_timeout_flush(&udp_tbl, 0,
 					&pkts[i], nb_pkts - i);
 		}
+
+		if (do_tcp6_gro) {
+			i += gro_tcp6_tbl_timeout_flush(&tcp6_tbl, 0,
+					&pkts[i], nb_pkts - i);
+		}
+
 		/* Copy unprocessed packets */
 		if (unprocess_num > 0) {
 			memcpy(&pkts[i], unprocess_pkts,
@@ -322,21 +363,24 @@ struct gro_ctx {
 {
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	struct gro_ctx *gro_ctx = ctx;
-	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl;
+	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl;
 	uint64_t current_time;
 	uint16_t i, unprocess_num = 0;
-	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro;
+	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro,
+		do_tcp6_gro;
 
 	if (unlikely((gro_ctx->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
 					RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
-					RTE_GRO_UDP_IPV4)) == 0))
+					RTE_GRO_UDP_IPV4 |
+					RTE_GRO_TCP_IPV6)) == 0))
 		return nb_pkts;
 
 	tcp_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV4_INDEX];
 	vxlan_tcp_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_TCP_IPV4_INDEX];
 	udp_tbl = gro_ctx->tbls[RTE_GRO_UDP_IPV4_INDEX];
 	vxlan_udp_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_UDP_IPV4_INDEX];
+	tcp6_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV6_INDEX];
 
 	do_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV4) ==
 		RTE_GRO_TCP_IPV4;
@@ -346,6 +390,8 @@ struct gro_ctx {
 		RTE_GRO_UDP_IPV4;
 	do_vxlan_udp_gro = (gro_ctx->gro_types & RTE_GRO_IPV4_VXLAN_UDP_IPV4) ==
 		RTE_GRO_IPV4_VXLAN_UDP_IPV4;
+	do_tcp6_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV6) ==
+		RTE_GRO_TCP_IPV6;
 
 	current_time = rte_rdtsc();
 
@@ -370,6 +416,11 @@ struct gro_ctx {
 			if (gro_udp4_reassemble(pkts[i], udp_tbl,
 						current_time) < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_TCP_PKT(pkts[i]->packet_type) &&
+				do_tcp6_gro) {
+			if (gro_tcp6_reassemble(pkts[i], tcp6_tbl,
+						current_time) < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -425,6 +476,13 @@ struct gro_ctx {
 				gro_ctx->tbls[RTE_GRO_UDP_IPV4_INDEX],
 				flush_timestamp,
 				&out[num], left_nb_out);
+		left_nb_out = max_nb_out - num;
+	}
+
+	if ((gro_types & RTE_GRO_TCP_IPV6) && left_nb_out > 0) {
+		num += gro_tcp6_tbl_timeout_flush(gro_ctx->tbls[
+				RTE_GRO_TCP_IPV6_INDEX],
+				flush_timestamp, &out[num], left_nb_out);
 	}
 
 	return num;
diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
index 9f9ed49..39b57e8 100644
--- a/lib/librte_gro/rte_gro.h
+++ b/lib/librte_gro/rte_gro.h
@@ -38,6 +38,9 @@
 #define RTE_GRO_IPV4_VXLAN_UDP_IPV4_INDEX 3
 #define RTE_GRO_IPV4_VXLAN_UDP_IPV4 (1ULL << RTE_GRO_IPV4_VXLAN_UDP_IPV4_INDEX)
 /**< VxLAN UDP/IPv4 GRO flag. */
+#define RTE_GRO_TCP_IPV6_INDEX 4
+#define RTE_GRO_TCP_IPV6 (1ULL << RTE_GRO_TCP_IPV6_INDEX)
+/**< TCP/IPv6 GRO flag */
 
 /**
  * Structure used to create GRO context objects or used to pass
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-dev] [PATCH v1 2/8] gro: support IPv4 VXLAN TCP/IPv6
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 1/8] gro: support TCP/IPv6 yang_y_yi
@ 2020-12-21  3:50 ` yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 3/8] gro: support IPv6 VXLAN TCP/IPv4 yang_y_yi
                   ` (6 subsequent siblings)
  8 siblings, 0 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

IPv4 VXLAN TCP/IPv6 GRO can help improve TCP/IPv6 performance
in IPv4 VXLAN use case.

With this enabled in DPDK, OVS DPDK can leverage it to improve
VM-to-VM TCP/IPv6 performance, it will merge small adjacent
IPv4 VXLAN TCP/IPv6 packets to a big IPv4 VXLAN TCP/IPv6 packet
immediate after they are received from a physical NIC. It is
very helpful in OVS DPDK IPv4 VXLAN use case.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
---
 .../prog_guide/generic_receive_offload_lib.rst     |   6 +-
 doc/guides/rel_notes/release_21_02.rst             |   5 +
 lib/librte_gro/gro_vxlan_tcp6.c                    | 496 +++++++++++++++++++++
 lib/librte_gro/gro_vxlan_tcp6.h                    | 156 +++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           |  81 +++-
 lib/librte_gro/rte_gro.h                           |   3 +
 7 files changed, 736 insertions(+), 13 deletions(-)
 create mode 100644 lib/librte_gro/gro_vxlan_tcp6.c
 create mode 100644 lib/librte_gro/gro_vxlan_tcp6.h

diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst b/doc/guides/prog_guide/generic_receive_offload_lib.rst
index fee856e..ecb5909 100644
--- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
+++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
@@ -32,7 +32,7 @@ fragmentation is possible (i.e., DF==0). Additionally, it complies RFC
 
 Currently, the GRO library provides GRO supports for TCP/IPv4, UDP/IPv4,
 and TCP/IPv6 packets as well as VxLAN packets which contain an outer IPv4
-header and an inner TCP/IPv4 or UDP/IPv4 packet.
+header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6 packet.
 
 Two Sets of API
 ---------------
@@ -161,8 +161,8 @@ Header fields deciding if two packets are neighbors include:
 - IPv4 ID. The IPv4 ID fields of the packets, whose DF bit is 0, should
   be increased by 1.
 
-VxLAN GRO
----------
+IPv4 VxLAN TCP/IPv4 GRO
+-----------------------
 
 The table structure used by VxLAN GRO, which is in charge of processing
 VxLAN packets with an outer IPv4 header and inner TCP/IPv4 packet, is
diff --git a/doc/guides/rel_notes/release_21_02.rst b/doc/guides/rel_notes/release_21_02.rst
index 1a1aeff..035d774 100644
--- a/doc/guides/rel_notes/release_21_02.rst
+++ b/doc/guides/rel_notes/release_21_02.rst
@@ -60,6 +60,11 @@ New Features
    Added TCP/IPv6 GRO support for non-VXLAN packets, this enabled TCP/IPv6
    GRO support for VLAN use case.
 
+* **Added TCP/IPv6 GRO support for IPv4 VXLAN packets.**
+
+   Added TCP/IPv6 GRO support for IPv4 VXLAN packets, this enabled TCP/IPv6
+   GRO support for IPv4 VXLAN use case.
+
 
 Removed Items
 -------------
diff --git a/lib/librte_gro/gro_vxlan_tcp6.c b/lib/librte_gro/gro_vxlan_tcp6.c
new file mode 100644
index 0000000..d49a2dd
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan_tcp6.c
@@ -0,0 +1,496 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+#include <rte_udp.h>
+
+#include "gro_vxlan_tcp6.h"
+
+void *
+gro_vxlan_tcp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow)
+{
+	struct gro_vxlan_tcp6_tbl *tbl;
+	size_t size;
+	uint32_t entries_num, i;
+
+	entries_num = max_flow_num * max_item_per_flow;
+	entries_num = RTE_MIN(entries_num, GRO_VXLAN_TCP6_TBL_MAX_ITEM_NUM);
+
+	if (entries_num == 0)
+		return NULL;
+
+	tbl = rte_zmalloc_socket(__func__,
+			sizeof(struct gro_vxlan_tcp6_tbl),
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl == NULL)
+		return NULL;
+
+	size = sizeof(struct gro_vxlan_tcp6_item) * entries_num;
+	tbl->items = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->items == NULL) {
+		rte_free(tbl);
+		return NULL;
+	}
+	tbl->max_item_num = entries_num;
+
+	size = sizeof(struct gro_vxlan_tcp6_flow) * entries_num;
+	tbl->flows = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->flows == NULL) {
+		rte_free(tbl->items);
+		rte_free(tbl);
+		return NULL;
+	}
+
+	for (i = 0; i < entries_num; i++)
+		tbl->flows[i].start_index = INVALID_ARRAY_INDEX;
+	tbl->max_flow_num = entries_num;
+
+	return tbl;
+}
+
+void
+gro_vxlan_tcp6_tbl_destroy(void *tbl)
+{
+	struct gro_vxlan_tcp6_tbl *vxlan_tbl = tbl;
+
+	if (vxlan_tbl) {
+		rte_free(vxlan_tbl->items);
+		rte_free(vxlan_tbl->flows);
+	}
+	rte_free(vxlan_tbl);
+}
+
+static inline uint32_t
+find_an_empty_item(struct gro_vxlan_tcp6_tbl *tbl)
+{
+	uint32_t max_item_num = tbl->max_item_num, i;
+
+	for (i = 0; i < max_item_num; i++)
+		if (tbl->items[i].inner_item.firstseg == NULL)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+find_an_empty_flow(struct gro_vxlan_tcp6_tbl *tbl)
+{
+	uint32_t max_flow_num = tbl->max_flow_num, i;
+
+	for (i = 0; i < max_flow_num; i++)
+		if (tbl->flows[i].start_index == INVALID_ARRAY_INDEX)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_item(struct gro_vxlan_tcp6_tbl *tbl,
+		struct rte_mbuf *pkt,
+		uint64_t start_time,
+		uint32_t prev_idx,
+		uint32_t sent_seq,
+		uint16_t outer_ip_id,
+		uint8_t outer_is_atomic)
+{
+	uint32_t item_idx;
+
+	item_idx = find_an_empty_item(tbl);
+	if (unlikely(item_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	tbl->items[item_idx].inner_item.firstseg = pkt;
+	tbl->items[item_idx].inner_item.lastseg = rte_pktmbuf_lastseg(pkt);
+	tbl->items[item_idx].inner_item.start_time = start_time;
+	tbl->items[item_idx].inner_item.next_pkt_idx = INVALID_ARRAY_INDEX;
+	tbl->items[item_idx].inner_item.sent_seq = sent_seq;
+	tbl->items[item_idx].inner_item.nb_merged = 1;
+	tbl->items[item_idx].outer_ip_id = outer_ip_id;
+	tbl->items[item_idx].outer_is_atomic = outer_is_atomic;
+	tbl->item_num++;
+
+	/* If the previous packet exists, chain the new one with it. */
+	if (prev_idx != INVALID_ARRAY_INDEX) {
+		tbl->items[item_idx].inner_item.next_pkt_idx =
+			tbl->items[prev_idx].inner_item.next_pkt_idx;
+		tbl->items[prev_idx].inner_item.next_pkt_idx = item_idx;
+	}
+
+	return item_idx;
+}
+
+static inline uint32_t
+delete_item(struct gro_vxlan_tcp6_tbl *tbl,
+		uint32_t item_idx,
+		uint32_t prev_item_idx)
+{
+	uint32_t next_idx = tbl->items[item_idx].inner_item.next_pkt_idx;
+
+	/* NULL indicates an empty item. */
+	tbl->items[item_idx].inner_item.firstseg = NULL;
+	tbl->item_num--;
+	if (prev_item_idx != INVALID_ARRAY_INDEX)
+		tbl->items[prev_item_idx].inner_item.next_pkt_idx = next_idx;
+
+	return next_idx;
+}
+
+static inline uint32_t
+insert_new_flow(struct gro_vxlan_tcp6_tbl *tbl,
+		struct vxlan_tcp6_flow_key *src,
+		uint32_t item_idx)
+{
+	struct vxlan_tcp6_flow_key *dst;
+	uint32_t flow_idx;
+
+	flow_idx = find_an_empty_flow(tbl);
+	if (unlikely(flow_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	dst = &(tbl->flows[flow_idx].key);
+
+	rte_ether_addr_copy(&(src->inner_key.eth_saddr),
+			&(dst->inner_key.eth_saddr));
+	rte_ether_addr_copy(&(src->inner_key.eth_daddr),
+			&(dst->inner_key.eth_daddr));
+	gro_ipv6_addr_copy(src->inner_key.ip_saddr, dst->inner_key.ip_saddr);
+	gro_ipv6_addr_copy(src->inner_key.ip_daddr, dst->inner_key.ip_daddr);
+	dst->inner_key.recv_ack = src->inner_key.recv_ack;
+	dst->inner_key.src_port = src->inner_key.src_port;
+	dst->inner_key.dst_port = src->inner_key.dst_port;
+
+	dst->vxlan_hdr.vx_flags = src->vxlan_hdr.vx_flags;
+	dst->vxlan_hdr.vx_vni = src->vxlan_hdr.vx_vni;
+	rte_ether_addr_copy(&(src->outer_eth_saddr), &(dst->outer_eth_saddr));
+	rte_ether_addr_copy(&(src->outer_eth_daddr), &(dst->outer_eth_daddr));
+	dst->outer_ip_src_addr = src->outer_ip_src_addr;
+	dst->outer_ip_dst_addr = src->outer_ip_dst_addr;
+	dst->outer_src_port = src->outer_src_port;
+	dst->outer_dst_port = src->outer_dst_port;
+
+	tbl->flows[flow_idx].start_index = item_idx;
+	tbl->flow_num++;
+
+	return flow_idx;
+}
+
+static inline int
+is_same_vxlan_tcp6_flow(struct vxlan_tcp6_flow_key k1,
+		struct vxlan_tcp6_flow_key k2)
+{
+	return (rte_is_same_ether_addr(&k1.outer_eth_saddr,
+					&k2.outer_eth_saddr) &&
+			rte_is_same_ether_addr(&k1.outer_eth_daddr,
+				&k2.outer_eth_daddr) &&
+			(k1.outer_ip_src_addr == k2.outer_ip_src_addr) &&
+			(k1.outer_ip_dst_addr == k2.outer_ip_dst_addr) &&
+			(k1.outer_src_port == k2.outer_src_port) &&
+			(k1.outer_dst_port == k2.outer_dst_port) &&
+			(k1.vxlan_hdr.vx_flags == k2.vxlan_hdr.vx_flags) &&
+			(k1.vxlan_hdr.vx_vni == k2.vxlan_hdr.vx_vni) &&
+			is_same_tcp6_flow(k1.inner_key, k2.inner_key));
+}
+
+static inline int
+check_vxlan_seq_option_v6(struct gro_vxlan_tcp6_item *item,
+		struct rte_tcp_hdr *tcp_hdr,
+		uint32_t sent_seq,
+		uint16_t outer_ip_id,
+		uint16_t tcp_hl,
+		uint16_t tcp_dl,
+		uint8_t outer_is_atomic)
+{
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	int cmp;
+	uint16_t l2_offset;
+
+	/* Don't merge packets whose outer DF bits are different. */
+	if (unlikely(item->outer_is_atomic ^ outer_is_atomic))
+		return 0;
+
+	l2_offset = pkt->outer_l2_len + pkt->outer_l3_len;
+	cmp = check_seq_option_v6(&item->inner_item, tcp_hdr, sent_seq,
+			tcp_hl, tcp_dl, l2_offset);
+	if ((cmp > 0) && (outer_is_atomic ||
+				(outer_ip_id == item->outer_ip_id + 1)))
+		/* Append the new packet. */
+		return 1;
+	else if ((cmp < 0) && (outer_is_atomic ||
+				(outer_ip_id + item->inner_item.nb_merged ==
+				 item->outer_ip_id)))
+		/* Prepend the new packet. */
+		return -1;
+
+	return 0;
+}
+
+static inline int
+merge_two_vxlan_tcp6_packets(struct gro_vxlan_tcp6_item *item,
+		struct rte_mbuf *pkt,
+		int cmp,
+		uint32_t sent_seq,
+		uint16_t outer_ip_id)
+{
+	if (merge_two_tcp6_packets(&item->inner_item, pkt, cmp, sent_seq,
+				pkt->outer_l2_len +
+				pkt->outer_l3_len)) {
+		/* Update the outer IPv4 ID to the large value. */
+		item->outer_ip_id = cmp > 0 ? outer_ip_id : item->outer_ip_id;
+		return 1;
+	}
+
+	return 0;
+}
+
+static inline void
+update_vxlan_header(struct gro_vxlan_tcp6_item *item)
+{
+	struct rte_ipv4_hdr *ipv4_hdr;
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	uint16_t len;
+
+	/* Update the outer IPv4 header. */
+	len = pkt->pkt_len - pkt->outer_l2_len;
+	ipv4_hdr = (struct rte_ipv4_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+			pkt->outer_l2_len);
+	ipv4_hdr->total_length = rte_cpu_to_be_16(len);
+
+	/* Update the outer UDP header. */
+	len -= pkt->outer_l3_len;
+	udp_hdr = (struct rte_udp_hdr *)((char *)ipv4_hdr + pkt->outer_l3_len);
+	udp_hdr->dgram_len = rte_cpu_to_be_16(len);
+
+	/* Update the inner IPv6 header. */
+	len -= pkt->l2_len;
+	ipv6_hdr = (struct rte_ipv6_hdr *)((char *)udp_hdr + pkt->l2_len);
+	/* IPv6 payload_len does not include IPv6 header. */
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(len - pkt->l3_len);
+}
+
+int32_t
+gro_vxlan_tcp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan_tcp6_tbl *tbl,
+		uint64_t start_time)
+{
+	struct rte_ether_hdr *outer_eth_hdr, *eth_hdr;
+	struct rte_ipv4_hdr *outer_ipv4_hdr;
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_tcp_hdr *tcp_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_vxlan_hdr *vxlan_hdr;
+	uint32_t sent_seq;
+	int32_t tcp_dl;
+	uint16_t frag_off, outer_ip_id;
+	uint8_t outer_is_atomic;
+
+	struct vxlan_tcp6_flow_key key;
+	uint32_t cur_idx, prev_idx, item_idx;
+	uint32_t i, max_flow_num, remaining_flow_num;
+	int cmp;
+	uint16_t hdr_len;
+	uint8_t find;
+
+	/*
+	 * Don't process the packet whose TCP header length is greater
+	 * than 60 bytes or less than 20 bytes.
+	 */
+	if (unlikely(INVALID_TCP_HDRLEN(pkt->l4_len)))
+		return -1;
+
+	outer_eth_hdr = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *);
+	outer_ipv4_hdr = (struct rte_ipv4_hdr *)((char *)outer_eth_hdr +
+			pkt->outer_l2_len);
+	udp_hdr = (struct rte_udp_hdr *)((char *)outer_ipv4_hdr +
+			pkt->outer_l3_len);
+	vxlan_hdr = (struct rte_vxlan_hdr *)((char *)udp_hdr +
+			sizeof(struct rte_udp_hdr));
+	eth_hdr = (struct rte_ether_hdr *)((char *)vxlan_hdr +
+			sizeof(struct rte_vxlan_hdr));
+	ipv6_hdr = (struct rte_ipv6_hdr *)((char *)udp_hdr + pkt->l2_len);
+	tcp_hdr = (struct rte_tcp_hdr *)((char *)ipv6_hdr + pkt->l3_len);
+
+	/*
+	 * Don't process the packet which has FIN, SYN, RST, PSH, URG,
+	 * ECE or CWR set.
+	 */
+	if (tcp_hdr->tcp_flags != RTE_TCP_ACK_FLAG)
+		return -1;
+
+	hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+		pkt->l3_len + pkt->l4_len;
+	/*
+	 * Don't process the packet whose payload length is less than or
+	 * equal to 0.
+	 */
+	if (pkt->pkt_len <= hdr_len)
+		return -1;
+	tcp_dl = pkt->pkt_len - hdr_len;
+
+	/*
+	 * Save IPv4 ID for the packet whose DF bit is 0. For the packet
+	 * whose DF bit is 1, IPv4 ID is ignored.
+	 */
+	frag_off = rte_be_to_cpu_16(outer_ipv4_hdr->fragment_offset);
+	outer_is_atomic =
+		(frag_off & RTE_IPV4_HDR_DF_FLAG) == RTE_IPV4_HDR_DF_FLAG;
+	outer_ip_id = outer_is_atomic ? 0 :
+		rte_be_to_cpu_16(outer_ipv4_hdr->packet_id);
+
+	sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
+
+	rte_ether_addr_copy(&(eth_hdr->s_addr), &(key.inner_key.eth_saddr));
+	rte_ether_addr_copy(&(eth_hdr->d_addr), &(key.inner_key.eth_daddr));
+	gro_ipv6_addr_copy(ipv6_hdr->src_addr, key.inner_key.ip_saddr);
+	gro_ipv6_addr_copy(ipv6_hdr->dst_addr, key.inner_key.ip_daddr);
+	key.inner_key.recv_ack = tcp_hdr->recv_ack;
+	key.inner_key.src_port = tcp_hdr->src_port;
+	key.inner_key.dst_port = tcp_hdr->dst_port;
+
+	key.vxlan_hdr.vx_flags = vxlan_hdr->vx_flags;
+	key.vxlan_hdr.vx_vni = vxlan_hdr->vx_vni;
+	rte_ether_addr_copy(&(outer_eth_hdr->s_addr), &(key.outer_eth_saddr));
+	rte_ether_addr_copy(&(outer_eth_hdr->d_addr), &(key.outer_eth_daddr));
+	key.outer_ip_src_addr = outer_ipv4_hdr->src_addr;
+	key.outer_ip_dst_addr = outer_ipv4_hdr->dst_addr;
+	key.outer_src_port = udp_hdr->src_port;
+	key.outer_dst_port = udp_hdr->dst_port;
+
+	/* Search for a matched flow. */
+	max_flow_num = tbl->max_flow_num;
+	remaining_flow_num = tbl->flow_num;
+	find = 0;
+	for (i = 0; i < max_flow_num && remaining_flow_num; i++) {
+		if (tbl->flows[i].start_index != INVALID_ARRAY_INDEX) {
+			if (is_same_vxlan_tcp6_flow(tbl->flows[i].key, key)) {
+				find = 1;
+				break;
+			}
+			remaining_flow_num--;
+		}
+	}
+
+	/*
+	 * Can't find a matched flow. Insert a new flow and store the
+	 * packet into the flow.
+	 */
+	if (find == 0) {
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, sent_seq, outer_ip_id,
+				outer_is_atomic);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		if (insert_new_flow(tbl, &key, item_idx) ==
+				INVALID_ARRAY_INDEX) {
+			/*
+			 * Fail to insert a new flow, so
+			 * delete the inserted packet.
+			 */
+			delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+			return -1;
+		}
+		return 0;
+	}
+
+	/* Check all packets in the flow and try to find a neighbor. */
+	cur_idx = tbl->flows[i].start_index;
+	prev_idx = cur_idx;
+	do {
+		cmp = check_vxlan_seq_option_v6(&(tbl->items[cur_idx]), tcp_hdr,
+				sent_seq, outer_ip_id, pkt->l4_len,
+				tcp_dl, outer_is_atomic);
+		if (cmp) {
+			if (merge_two_vxlan_tcp6_packets(&(tbl->items[cur_idx]),
+						pkt, cmp, sent_seq,
+						outer_ip_id))
+				return 1;
+			/*
+			 * Can't merge two packets, as the packet
+			 * length will be greater than the max value.
+			 * Insert the packet into the flow.
+			 */
+			if (insert_new_item(tbl, pkt, start_time, prev_idx,
+						sent_seq, outer_ip_id,
+						outer_is_atomic) ==
+					INVALID_ARRAY_INDEX)
+				return -1;
+			return 0;
+		}
+		prev_idx = cur_idx;
+		cur_idx = tbl->items[cur_idx].inner_item.next_pkt_idx;
+	} while (cur_idx != INVALID_ARRAY_INDEX);
+
+	/* Can't find neighbor. Insert the packet into the flow. */
+	if (insert_new_item(tbl, pkt, start_time, prev_idx, sent_seq,
+				outer_ip_id, outer_is_atomic
+				) == INVALID_ARRAY_INDEX)
+		return -1;
+
+	return 0;
+}
+
+uint16_t
+gro_vxlan_tcp6_tbl_timeout_flush(struct gro_vxlan_tcp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out)
+{
+	uint16_t k = 0;
+	uint32_t i, j;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++) {
+		if (unlikely(tbl->flow_num == 0))
+			return k;
+
+		j = tbl->flows[i].start_index;
+		while (j != INVALID_ARRAY_INDEX) {
+			if (tbl->items[j].inner_item.start_time <=
+					flush_timestamp) {
+				out[k++] = tbl->items[j].inner_item.firstseg;
+				if (tbl->items[j].inner_item.nb_merged > 1)
+					update_vxlan_header(&(tbl->items[j]));
+				/*
+				 * Delete the item and get the next packet
+				 * index.
+				 */
+				j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+				tbl->flows[i].start_index = j;
+				if (j == INVALID_ARRAY_INDEX)
+					tbl->flow_num--;
+
+				if (unlikely(k == nb_out))
+					return k;
+			} else
+				/*
+				 * The left packets in the flow won't be
+				 * timeout. Go to check other flows.
+				 */
+				break;
+		}
+	}
+	return k;
+}
+
+uint32_t
+gro_vxlan_tcp6_tbl_pkt_count(void *tbl)
+{
+	struct gro_vxlan_tcp6_tbl *gro_tbl = tbl;
+
+	if (gro_tbl)
+		return gro_tbl->item_num;
+
+	return 0;
+}
diff --git a/lib/librte_gro/gro_vxlan_tcp6.h b/lib/librte_gro/gro_vxlan_tcp6.h
new file mode 100644
index 0000000..87d03d8
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan_tcp6.h
@@ -0,0 +1,156 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#ifndef _GRO_VXLAN_TCP6_H_
+#define _GRO_VXLAN_TCP6_H_
+
+#include "gro_tcp6.h"
+
+#define GRO_VXLAN_TCP6_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/* Header fields representing a VxLAN flow */
+struct vxlan_tcp6_flow_key {
+	struct tcp6_flow_key inner_key;
+	struct rte_vxlan_hdr vxlan_hdr;
+
+	struct rte_ether_addr outer_eth_saddr;
+	struct rte_ether_addr outer_eth_daddr;
+
+	uint32_t outer_ip_src_addr;
+	uint32_t outer_ip_dst_addr;
+
+	/* Outer UDP ports */
+	uint16_t outer_src_port;
+	uint16_t outer_dst_port;
+
+};
+
+struct gro_vxlan_tcp6_flow {
+	struct vxlan_tcp6_flow_key key;
+	/*
+	 * The index of the first packet in the flow. INVALID_ARRAY_INDEX
+	 * indicates an empty flow.
+	 */
+	uint32_t start_index;
+};
+
+struct gro_vxlan_tcp6_item {
+	struct gro_tcp6_item inner_item;
+	/* IPv4 ID in the outer IPv4 header */
+	uint16_t outer_ip_id;
+	/* Indicate if outer IPv4 ID can be ignored */
+	uint8_t outer_is_atomic;
+};
+
+/*
+ * VxLAN (with an outer IPv4 header and an inner TCP/IPv6 packet)
+ * reassembly table structure
+ */
+struct gro_vxlan_tcp6_tbl {
+	/* item array */
+	struct gro_vxlan_tcp6_item *items;
+	/* flow array */
+	struct gro_vxlan_tcp6_flow *flows;
+	/* current item number */
+	uint32_t item_num;
+	/* current flow number */
+	uint32_t flow_num;
+	/* the maximum item number */
+	uint32_t max_item_num;
+	/* the maximum flow number */
+	uint32_t max_flow_num;
+};
+
+/**
+ * This function creates a VxLAN reassembly table for VxLAN packets
+ * which have an outer IPv4 header and an inner TCP/IPv6 packet.
+ *
+ * @param socket_id
+ *  Socket index for allocating the table
+ * @param max_flow_num
+ *  The maximum number of flows in the table
+ * @param max_item_per_flow
+ *  The maximum number of packets per flow
+ *
+ * @return
+ *  - Return the table pointer on success.
+ *  - Return NULL on failure.
+ */
+void *gro_vxlan_tcp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow);
+
+/**
+ * This function destroys a VxLAN reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the VxLAN reassembly table
+ */
+void gro_vxlan_tcp6_tbl_destroy(void *tbl);
+
+/**
+ * This function merges a VxLAN packet which has an outer IPv4 header and
+ * an inner TCP/IPv6 packet. It does not process the packet, whose TCP
+ * header has SYN, FIN, RST, PSH, CWR, ECE or URG bit set, or which
+ * does not have payload.
+ *
+ * This function does not check if the packet has correct checksums and
+ * does not re-calculate checksums for the merged packet. Additionally,
+ * it assumes the packets are complete (i.e., MF==0 && frag_off==0), when
+ * IP fragmentation is possible (i.e., DF==0). It returns the packet, if
+ * the packet has invalid parameters (e.g. SYN bit is set) or there is no
+ * available space in the table.
+ *
+ * @param pkt
+ *  Packet to reassemble
+ * @param tbl
+ *  Pointer pointing to the VxLAN reassembly table
+ * @start_time
+ *  The time when the packet is inserted into the table
+ *
+ * @return
+ *  - Return a positive value if the packet is merged.
+ *  - Return zero if the packet isn't merged but stored in the table.
+ *  - Return a negative value for invalid parameters or no available
+ *    space in the table.
+ */
+int32_t gro_vxlan_tcp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan_tcp6_tbl *tbl,
+		uint64_t start_time);
+
+/**
+ * This function flushes timeout packets in the VxLAN reassembly table,
+ * and without updating checksums.
+ *
+ * @param tbl
+ *  Pointer pointing to a VxLAN GRO table
+ * @param flush_timestamp
+ *  This function flushes packets which are inserted into the table
+ *  before or at the flush_timestamp.
+ * @param out
+ *  Pointer array used to keep flushed packets
+ * @param nb_out
+ *  The element number in 'out'. It also determines the maximum number of
+ *  packets that can be flushed finally.
+ *
+ * @return
+ *  The number of flushed packets
+ */
+uint16_t gro_vxlan_tcp6_tbl_timeout_flush(struct gro_vxlan_tcp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out);
+
+/**
+ * This function returns the number of the packets in a VxLAN
+ * reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the VxLAN reassembly table
+ *
+ * @return
+ *  The number of packets in the table
+ */
+uint32_t gro_vxlan_tcp6_tbl_pkt_count(void *tbl);
+#endif
diff --git a/lib/librte_gro/meson.build b/lib/librte_gro/meson.build
index 6143dd0..020e7e8 100644
--- a/lib/librte_gro/meson.build
+++ b/lib/librte_gro/meson.build
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c')
+sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c')
 headers = files('rte_gro.h')
 deps += ['ethdev']
diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
index 86958d5..89fed3c 100644
--- a/lib/librte_gro/rte_gro.c
+++ b/lib/librte_gro/rte_gro.c
@@ -12,6 +12,7 @@
 #include "gro_tcp6.h"
 #include "gro_udp4.h"
 #include "gro_vxlan_tcp4.h"
+#include "gro_vxlan_tcp6.h"
 #include "gro_vxlan_udp4.h"
 
 typedef void *(*gro_tbl_create_fn)(uint16_t socket_id,
@@ -23,15 +24,16 @@
 static gro_tbl_create_fn tbl_create_fn[RTE_GRO_TYPE_MAX_NUM] = {
 		gro_tcp4_tbl_create, gro_vxlan_tcp4_tbl_create,
 		gro_udp4_tbl_create, gro_vxlan_udp4_tbl_create,
-		gro_tcp6_tbl_create, NULL};
+		gro_tcp6_tbl_create, gro_vxlan_tcp6_tbl_create, NULL};
 static gro_tbl_destroy_fn tbl_destroy_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_destroy, gro_vxlan_tcp4_tbl_destroy,
 			gro_udp4_tbl_destroy, gro_vxlan_udp4_tbl_destroy,
-			gro_tcp6_tbl_destroy, NULL};
+			gro_tcp6_tbl_destroy, gro_vxlan_tcp6_tbl_destroy, NULL};
 static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_pkt_count, gro_vxlan_tcp4_tbl_pkt_count,
 			gro_udp4_tbl_pkt_count, gro_vxlan_udp4_tbl_pkt_count,
-			gro_tcp6_tbl_pkt_count, NULL};
+			gro_tcp6_tbl_pkt_count, gro_vxlan_tcp6_tbl_pkt_count,
+			NULL};
 
 #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
@@ -71,6 +73,19 @@
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
 		(RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
 
+#define IS_IPV4_VXLAN_TCP6_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
+		((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
+		((ptype & RTE_PTYPE_TUNNEL_VXLAN) == \
+		 RTE_PTYPE_TUNNEL_VXLAN) && \
+		((ptype & RTE_PTYPE_INNER_L4_TCP) == \
+		 RTE_PTYPE_INNER_L4_TCP) && \
+		(((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6_EXT) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN)))
+
 /*
  * GRO context structure. It keeps the table structures, which are
  * used to merge packets, for different GRO types. Before using
@@ -176,19 +191,25 @@ struct gro_ctx {
 	struct gro_tcp6_flow tcp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
 	struct gro_tcp6_item tcp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
 
+	/* Allocate a reassembly table for VXLAN TCP/IPv6 GRO */
+	struct gro_vxlan_tcp6_tbl vxlan_tcp6_tbl;
+	struct gro_vxlan_tcp6_flow vxlan_tcp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+	struct gro_vxlan_tcp6_item vxlan_tcp6_items[RTE_GRO_MAX_BURST_ITEM_NUM]
+			= {{{0}, 0, 0} };
 
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	uint32_t item_num;
 	int32_t ret;
 	uint16_t i, unprocess_num = 0, nb_after_gro = nb_pkts;
 	uint8_t do_tcp4_gro = 0, do_vxlan_tcp_gro = 0, do_udp4_gro = 0,
-		do_vxlan_udp_gro = 0, do_tcp6_gro = 0;
+		do_vxlan_udp_gro = 0, do_tcp6_gro = 0, do_vxlan_tcp6_gro = 0;
 
 	if (unlikely((param->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
 					RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
 					RTE_GRO_UDP_IPV4 |
-					RTE_GRO_TCP_IPV6)) == 0))
+					RTE_GRO_TCP_IPV6 |
+					RTE_GRO_IPV4_VXLAN_TCP_IPV6)) == 0))
 		return nb_pkts;
 
 	/* Get the maximum number of packets */
@@ -261,7 +282,18 @@ struct gro_ctx {
 		do_tcp6_gro = 1;
 	}
 
-
+	if (param->gro_types & RTE_GRO_IPV4_VXLAN_TCP_IPV6) {
+		for (i = 0; i < item_num; i++)
+			vxlan_tcp6_flows[i].start_index = INVALID_ARRAY_INDEX;
+
+		vxlan_tcp6_tbl.flows = vxlan_tcp6_flows;
+		vxlan_tcp6_tbl.items = vxlan_tcp6_items;
+		vxlan_tcp6_tbl.flow_num = 0;
+		vxlan_tcp6_tbl.item_num = 0;
+		vxlan_tcp6_tbl.max_flow_num = item_num;
+		vxlan_tcp6_tbl.max_item_num = item_num;
+		do_vxlan_tcp6_gro = 1;
+	}
 
 	for (i = 0; i < nb_pkts; i++) {
 		/*
@@ -310,6 +342,15 @@ struct gro_ctx {
 				nb_after_gro--;
 			else if (ret < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV4_VXLAN_TCP6_PKT(pkts[i]->packet_type) &&
+				do_vxlan_tcp6_gro) {
+			ret = gro_vxlan_tcp6_reassemble(pkts[i],
+							&vxlan_tcp6_tbl, 0);
+			if (ret > 0)
+				/* Merge successfully */
+				nb_after_gro--;
+			else if (ret < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -344,6 +385,11 @@ struct gro_ctx {
 					&pkts[i], nb_pkts - i);
 		}
 
+		if (do_vxlan_tcp6_gro) {
+			i += gro_vxlan_tcp6_tbl_timeout_flush(&vxlan_tcp6_tbl,
+					0, &pkts[i], nb_pkts - i);
+		}
+
 		/* Copy unprocessed packets */
 		if (unprocess_num > 0) {
 			memcpy(&pkts[i], unprocess_pkts,
@@ -363,17 +409,19 @@ struct gro_ctx {
 {
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	struct gro_ctx *gro_ctx = ctx;
-	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl;
+	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl,
+		*vxlan_tcp6_tbl;
 	uint64_t current_time;
 	uint16_t i, unprocess_num = 0;
 	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro,
-		do_tcp6_gro;
+		do_tcp6_gro, do_vxlan_tcp6_gro;
 
 	if (unlikely((gro_ctx->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
 					RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
 					RTE_GRO_UDP_IPV4 |
-					RTE_GRO_TCP_IPV6)) == 0))
+					RTE_GRO_TCP_IPV6 |
+					RTE_GRO_IPV4_VXLAN_TCP_IPV6)) == 0))
 		return nb_pkts;
 
 	tcp_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV4_INDEX];
@@ -381,6 +429,7 @@ struct gro_ctx {
 	udp_tbl = gro_ctx->tbls[RTE_GRO_UDP_IPV4_INDEX];
 	vxlan_udp_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_UDP_IPV4_INDEX];
 	tcp6_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV6_INDEX];
+	vxlan_tcp6_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX];
 
 	do_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV4) ==
 		RTE_GRO_TCP_IPV4;
@@ -392,6 +441,8 @@ struct gro_ctx {
 		RTE_GRO_IPV4_VXLAN_UDP_IPV4;
 	do_tcp6_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV6) ==
 		RTE_GRO_TCP_IPV6;
+	do_vxlan_tcp6_gro = (gro_ctx->gro_types & RTE_GRO_IPV4_VXLAN_TCP_IPV6)
+				== RTE_GRO_IPV4_VXLAN_TCP_IPV6;
 
 	current_time = rte_rdtsc();
 
@@ -421,6 +472,11 @@ struct gro_ctx {
 			if (gro_tcp6_reassemble(pkts[i], tcp6_tbl,
 						current_time) < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV4_VXLAN_TCP6_PKT(pkts[i]->packet_type) &&
+				do_vxlan_tcp6_gro) {
+			if (gro_vxlan_tcp6_reassemble(pkts[i], vxlan_tcp6_tbl,
+						current_time) < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -483,6 +539,13 @@ struct gro_ctx {
 		num += gro_tcp6_tbl_timeout_flush(gro_ctx->tbls[
 				RTE_GRO_TCP_IPV6_INDEX],
 				flush_timestamp, &out[num], left_nb_out);
+		left_nb_out = max_nb_out - num;
+	}
+
+	if ((gro_types & RTE_GRO_IPV4_VXLAN_TCP_IPV6) && left_nb_out > 0) {
+		num += gro_vxlan_tcp6_tbl_timeout_flush(gro_ctx->tbls[
+				RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX],
+				flush_timestamp, &out[num], left_nb_out);
 	}
 
 	return num;
diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
index 39b57e8..41a2bec 100644
--- a/lib/librte_gro/rte_gro.h
+++ b/lib/librte_gro/rte_gro.h
@@ -41,6 +41,9 @@
 #define RTE_GRO_TCP_IPV6_INDEX 4
 #define RTE_GRO_TCP_IPV6 (1ULL << RTE_GRO_TCP_IPV6_INDEX)
 /**< TCP/IPv6 GRO flag */
+#define RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX 5
+#define RTE_GRO_IPV4_VXLAN_TCP_IPV6 (1ULL << RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX)
+/**< IPv4 VxLAN TCP/IPv6 GRO flag. */
 
 /**
  * Structure used to create GRO context objects or used to pass
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-dev] [PATCH v1 3/8] gro: support IPv6 VXLAN TCP/IPv4
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 1/8] gro: support TCP/IPv6 yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 2/8] gro: support IPv4 VXLAN TCP/IPv6 yang_y_yi
@ 2020-12-21  3:50 ` yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 4/8] gro: support IPv6 VXLAN TCP/IPv6 yang_y_yi
                   ` (5 subsequent siblings)
  8 siblings, 0 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

IPv6 VXLAN TCP/IPv4 GRO can help improve TCP/IPv4 performance
in IPv6 VXLAN use case.

With this enabled in DPDK, OVS DPDK can leverage it to improve
VM-to-VM TCP/IPv4 performance, it will merge small adjacent
IPv6 VXLAN TCP/IPv4 packets to a big IPv6 VXLAN TCP/IPv4 packet
immediate after they are received from a physical NIC. It is
very helpful in OVS DPDK IPv6 VXLAN use case.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
---
 .../prog_guide/generic_receive_offload_lib.rst     |   3 +-
 doc/guides/rel_notes/release_21_02.rst             |   4 +
 lib/librte_gro/gro_vxlan6_tcp4.c                   | 507 +++++++++++++++++++++
 lib/librte_gro/gro_vxlan6_tcp4.h                   | 153 +++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           |  86 +++-
 lib/librte_gro/rte_gro.h                           |   3 +
 7 files changed, 747 insertions(+), 11 deletions(-)
 create mode 100644 lib/librte_gro/gro_vxlan6_tcp4.c
 create mode 100644 lib/librte_gro/gro_vxlan6_tcp4.h

diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst b/doc/guides/prog_guide/generic_receive_offload_lib.rst
index ecb5909..a822eaf 100644
--- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
+++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
@@ -32,7 +32,8 @@ fragmentation is possible (i.e., DF==0). Additionally, it complies RFC
 
 Currently, the GRO library provides GRO supports for TCP/IPv4, UDP/IPv4,
 and TCP/IPv6 packets as well as VxLAN packets which contain an outer IPv4
-header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6 packet.
+header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6 packet or an outer IPv6
+header and an inner TCP/IPv4 packet.
 
 Two Sets of API
 ---------------
diff --git a/doc/guides/rel_notes/release_21_02.rst b/doc/guides/rel_notes/release_21_02.rst
index 035d774..3b98915 100644
--- a/doc/guides/rel_notes/release_21_02.rst
+++ b/doc/guides/rel_notes/release_21_02.rst
@@ -65,6 +65,10 @@ New Features
    Added TCP/IPv6 GRO support for IPv4 VXLAN packets, this enabled TCP/IPv6
    GRO support for IPv4 VXLAN use case.
 
+* **Added TCP/IPv4 GRO support for IPv6 VXLAN packets.**
+
+   Added TCP/IPv4 GRO support for IPv6 VXLAN packets, this enabled TCP/IPv4
+   GRO support for IPv6 VXLAN use case.
 
 Removed Items
 -------------
diff --git a/lib/librte_gro/gro_vxlan6_tcp4.c b/lib/librte_gro/gro_vxlan6_tcp4.c
new file mode 100644
index 0000000..1e17c1a
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan6_tcp4.c
@@ -0,0 +1,507 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+#include <rte_udp.h>
+
+#include "gro_vxlan6_tcp4.h"
+
+/*
+ * Check if two IPv6 address is same.
+ */
+static inline int rte_is_same_ipv6_addr(const uint8_t *ipv6_a1,
+					const uint8_t *ipv6_a2)
+{
+	const uint64_t *w1 = (const uint64_t *)ipv6_a1;
+	const uint64_t *w2 = (const uint64_t *)ipv6_a2;
+
+	return ((w1[0] ^ w2[0]) | (w1[1] ^ w2[1])) == 0;
+}
+
+/* Copy IPv6 addr */
+static inline void gro_ipv6_addr_copy(const uint8_t *ipv6_from,
+				      uint8_t *ipv6_to)
+{
+	const uint64_t *from_words = (const uint64_t *)ipv6_from;
+	uint64_t *to_words   = (uint64_t *)ipv6_to;
+
+	to_words[0] = from_words[0];
+	to_words[1] = from_words[1];
+}
+
+void *
+gro_vxlan6_tcp4_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow)
+{
+	struct gro_vxlan6_tcp4_tbl *tbl;
+	size_t size;
+	uint32_t entries_num, i;
+
+	entries_num = max_flow_num * max_item_per_flow;
+	entries_num = RTE_MIN(entries_num, GRO_VXLAN6_TCP4_TBL_MAX_ITEM_NUM);
+
+	if (entries_num == 0)
+		return NULL;
+
+	tbl = rte_zmalloc_socket(__func__,
+			sizeof(struct gro_vxlan6_tcp4_tbl),
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl == NULL)
+		return NULL;
+
+	size = sizeof(struct gro_vxlan6_tcp4_item) * entries_num;
+	tbl->items = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->items == NULL) {
+		rte_free(tbl);
+		return NULL;
+	}
+	tbl->max_item_num = entries_num;
+
+	size = sizeof(struct gro_vxlan6_tcp4_flow) * entries_num;
+	tbl->flows = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->flows == NULL) {
+		rte_free(tbl->items);
+		rte_free(tbl);
+		return NULL;
+	}
+
+	for (i = 0; i < entries_num; i++)
+		tbl->flows[i].start_index = INVALID_ARRAY_INDEX;
+	tbl->max_flow_num = entries_num;
+
+	return tbl;
+}
+
+void
+gro_vxlan6_tcp4_tbl_destroy(void *tbl)
+{
+	struct gro_vxlan6_tcp4_tbl *vxlan6_tbl = tbl;
+
+	if (vxlan6_tbl) {
+		rte_free(vxlan6_tbl->items);
+		rte_free(vxlan6_tbl->flows);
+	}
+	rte_free(vxlan6_tbl);
+}
+
+static inline uint32_t
+find_an_empty_item(struct gro_vxlan6_tcp4_tbl *tbl)
+{
+	uint32_t max_item_num = tbl->max_item_num, i;
+
+	for (i = 0; i < max_item_num; i++)
+		if (tbl->items[i].inner_item.firstseg == NULL)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+find_an_empty_flow(struct gro_vxlan6_tcp4_tbl *tbl)
+{
+	uint32_t max_flow_num = tbl->max_flow_num, i;
+
+	for (i = 0; i < max_flow_num; i++)
+		if (tbl->flows[i].start_index == INVALID_ARRAY_INDEX)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_item(struct gro_vxlan6_tcp4_tbl *tbl,
+		struct rte_mbuf *pkt,
+		uint64_t start_time,
+		uint32_t prev_idx,
+		uint32_t sent_seq,
+		uint16_t ip_id,
+		uint8_t is_atomic)
+{
+	uint32_t item_idx;
+
+	item_idx = find_an_empty_item(tbl);
+	if (unlikely(item_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	tbl->items[item_idx].inner_item.firstseg = pkt;
+	tbl->items[item_idx].inner_item.lastseg = rte_pktmbuf_lastseg(pkt);
+	tbl->items[item_idx].inner_item.start_time = start_time;
+	tbl->items[item_idx].inner_item.next_pkt_idx = INVALID_ARRAY_INDEX;
+	tbl->items[item_idx].inner_item.sent_seq = sent_seq;
+	tbl->items[item_idx].inner_item.ip_id = ip_id;
+	tbl->items[item_idx].inner_item.nb_merged = 1;
+	tbl->items[item_idx].inner_item.is_atomic = is_atomic;
+	tbl->item_num++;
+
+	/* If the previous packet exists, chain the new one with it. */
+	if (prev_idx != INVALID_ARRAY_INDEX) {
+		tbl->items[item_idx].inner_item.next_pkt_idx =
+			tbl->items[prev_idx].inner_item.next_pkt_idx;
+		tbl->items[prev_idx].inner_item.next_pkt_idx = item_idx;
+	}
+
+	return item_idx;
+}
+
+static inline uint32_t
+delete_item(struct gro_vxlan6_tcp4_tbl *tbl,
+		uint32_t item_idx,
+		uint32_t prev_item_idx)
+{
+	uint32_t next_idx = tbl->items[item_idx].inner_item.next_pkt_idx;
+
+	/* NULL indicates an empty item. */
+	tbl->items[item_idx].inner_item.firstseg = NULL;
+	tbl->item_num--;
+	if (prev_item_idx != INVALID_ARRAY_INDEX)
+		tbl->items[prev_item_idx].inner_item.next_pkt_idx = next_idx;
+
+	return next_idx;
+}
+
+static inline uint32_t
+insert_new_flow(struct gro_vxlan6_tcp4_tbl *tbl,
+		struct vxlan6_tcp4_flow_key *src,
+		uint32_t item_idx)
+{
+	struct vxlan6_tcp4_flow_key *dst;
+	uint32_t flow_idx;
+
+	flow_idx = find_an_empty_flow(tbl);
+	if (unlikely(flow_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	dst = &(tbl->flows[flow_idx].key);
+
+	rte_ether_addr_copy(&(src->inner_key.eth_saddr),
+			&(dst->inner_key.eth_saddr));
+	rte_ether_addr_copy(&(src->inner_key.eth_daddr),
+			&(dst->inner_key.eth_daddr));
+	dst->inner_key.ip_src_addr = src->inner_key.ip_src_addr;
+	dst->inner_key.ip_dst_addr = src->inner_key.ip_dst_addr;
+	dst->inner_key.recv_ack = src->inner_key.recv_ack;
+	dst->inner_key.src_port = src->inner_key.src_port;
+	dst->inner_key.dst_port = src->inner_key.dst_port;
+
+	dst->vxlan_hdr.vx_flags = src->vxlan_hdr.vx_flags;
+	dst->vxlan_hdr.vx_vni = src->vxlan_hdr.vx_vni;
+	rte_ether_addr_copy(&(src->outer_eth_saddr), &(dst->outer_eth_saddr));
+	rte_ether_addr_copy(&(src->outer_eth_daddr), &(dst->outer_eth_daddr));
+	gro_ipv6_addr_copy(src->outer_ip_saddr, dst->outer_ip_saddr);
+	gro_ipv6_addr_copy(src->outer_ip_daddr, dst->outer_ip_daddr);
+	dst->outer_src_port = src->outer_src_port;
+	dst->outer_dst_port = src->outer_dst_port;
+
+	tbl->flows[flow_idx].start_index = item_idx;
+	tbl->flow_num++;
+
+	return flow_idx;
+}
+
+static inline int
+is_same_vxlan6_tcp4_flow(struct vxlan6_tcp4_flow_key k1,
+		struct vxlan6_tcp4_flow_key k2)
+{
+	return (rte_is_same_ether_addr(&k1.outer_eth_saddr,
+					&k2.outer_eth_saddr) &&
+			rte_is_same_ether_addr(&k1.outer_eth_daddr,
+				&k2.outer_eth_daddr) &&
+			rte_is_same_ipv6_addr(k1.outer_ip_saddr,
+				k2.outer_ip_saddr) &&
+			rte_is_same_ipv6_addr(k1.outer_ip_daddr,
+				k2.outer_ip_daddr) &&
+			(k1.outer_src_port == k2.outer_src_port) &&
+			(k1.outer_dst_port == k2.outer_dst_port) &&
+			(k1.vxlan_hdr.vx_flags == k2.vxlan_hdr.vx_flags) &&
+			(k1.vxlan_hdr.vx_vni == k2.vxlan_hdr.vx_vni) &&
+			is_same_tcp4_flow(k1.inner_key, k2.inner_key));
+}
+
+static inline int
+check_vxlan6_seq_option(struct gro_vxlan6_tcp4_item *item,
+		struct rte_tcp_hdr *tcp_hdr,
+		uint32_t sent_seq,
+		uint16_t ip_id,
+		uint16_t tcp_hl,
+		uint16_t tcp_dl,
+		uint8_t is_atomic)
+{
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	int cmp;
+	uint16_t l2_offset;
+
+	l2_offset = pkt->outer_l2_len + pkt->outer_l3_len;
+	cmp = check_seq_option(&item->inner_item, tcp_hdr, sent_seq, ip_id,
+			tcp_hl, tcp_dl, l2_offset, is_atomic);
+	if (cmp > 0)
+		/* Append the new packet. */
+		return 1;
+	else if (cmp < 0)
+		/* Prepend the new packet. */
+		return -1;
+
+	return 0;
+}
+
+static inline int
+merge_two_vxlan6_tcp4_packets(struct gro_vxlan6_tcp4_item *item,
+		struct rte_mbuf *pkt,
+		int cmp,
+		uint32_t sent_seq,
+		uint16_t ip_id)
+{
+	if (merge_two_tcp4_packets(&item->inner_item, pkt, cmp, sent_seq,
+				ip_id, pkt->outer_l2_len +
+				pkt->outer_l3_len))
+		return 1;
+
+	return 0;
+}
+
+static inline void
+update_vxlan6_header(struct gro_vxlan6_tcp4_item *item)
+{
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_ipv4_hdr *ipv4_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	uint16_t len;
+
+	/* Update the outer IPv6 header. */
+	len = pkt->pkt_len - pkt->outer_l2_len;
+	ipv6_hdr = (struct rte_ipv6_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+			pkt->outer_l2_len);
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(len - pkt->outer_l3_len);
+
+	/* Update the outer UDP header. */
+	len -= pkt->outer_l3_len;
+	udp_hdr = (struct rte_udp_hdr *)((char *)ipv6_hdr + pkt->outer_l3_len);
+	udp_hdr->dgram_len = rte_cpu_to_be_16(len);
+
+	/* Update the inner IPv4 header. */
+	len -= pkt->l2_len;
+	ipv4_hdr = (struct rte_ipv4_hdr *)((char *)udp_hdr + pkt->l2_len);
+	ipv4_hdr->total_length = rte_cpu_to_be_16(len);
+}
+
+int32_t
+gro_vxlan6_tcp4_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan6_tcp4_tbl *tbl,
+		uint64_t start_time)
+{
+	struct rte_ether_hdr *outer_eth_hdr, *eth_hdr;
+	struct rte_ipv6_hdr *outer_ipv6_hdr;
+	struct rte_ipv4_hdr *ipv4_hdr;
+	struct rte_tcp_hdr *tcp_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_vxlan_hdr *vxlan_hdr;
+	uint32_t sent_seq;
+	int32_t tcp_dl;
+	uint16_t frag_off, ip_id;
+	uint8_t is_atomic;
+
+	struct vxlan6_tcp4_flow_key key;
+	uint32_t cur_idx, prev_idx, item_idx;
+	uint32_t i, max_flow_num, remaining_flow_num;
+	int cmp;
+	uint16_t hdr_len;
+	uint8_t find;
+
+	/*
+	 * Don't process the packet whose TCP header length is greater
+	 * than 60 bytes or less than 20 bytes.
+	 */
+	if (unlikely(INVALID_TCP_HDRLEN(pkt->l4_len)))
+		return -1;
+
+	outer_eth_hdr = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *);
+	outer_ipv6_hdr = (struct rte_ipv6_hdr *)((char *)outer_eth_hdr +
+			pkt->outer_l2_len);
+	udp_hdr = (struct rte_udp_hdr *)((char *)outer_ipv6_hdr +
+			pkt->outer_l3_len);
+	vxlan_hdr = (struct rte_vxlan_hdr *)((char *)udp_hdr +
+			sizeof(struct rte_udp_hdr));
+	eth_hdr = (struct rte_ether_hdr *)((char *)vxlan_hdr +
+			sizeof(struct rte_vxlan_hdr));
+	ipv4_hdr = (struct rte_ipv4_hdr *)((char *)udp_hdr + pkt->l2_len);
+	tcp_hdr = (struct rte_tcp_hdr *)((char *)ipv4_hdr + pkt->l3_len);
+
+	/*
+	 * Don't process the packet which has FIN, SYN, RST, PSH, URG,
+	 * ECE or CWR set.
+	 */
+	if (tcp_hdr->tcp_flags != RTE_TCP_ACK_FLAG)
+		return -1;
+
+	hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+		pkt->l3_len + pkt->l4_len;
+	/*
+	 * Don't process the packet whose payload length is less than or
+	 * equal to 0.
+	 */
+	if (pkt->pkt_len <= hdr_len)
+		return -1;
+	tcp_dl = pkt->pkt_len - hdr_len;
+
+	/*
+	 * Save IPv4 ID for the packet whose DF bit is 0. For the packet
+	 * whose DF bit is 1, IPv4 ID is ignored.
+	 */
+	frag_off = rte_be_to_cpu_16(ipv4_hdr->fragment_offset);
+	is_atomic = (frag_off & RTE_IPV4_HDR_DF_FLAG) == RTE_IPV4_HDR_DF_FLAG;
+	ip_id = is_atomic ? 0 : rte_be_to_cpu_16(ipv4_hdr->packet_id);
+
+	sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
+
+	rte_ether_addr_copy(&(eth_hdr->s_addr), &(key.inner_key.eth_saddr));
+	rte_ether_addr_copy(&(eth_hdr->d_addr), &(key.inner_key.eth_daddr));
+	key.inner_key.ip_src_addr = ipv4_hdr->src_addr;
+	key.inner_key.ip_dst_addr = ipv4_hdr->dst_addr;
+	key.inner_key.recv_ack = tcp_hdr->recv_ack;
+	key.inner_key.src_port = tcp_hdr->src_port;
+	key.inner_key.dst_port = tcp_hdr->dst_port;
+
+	key.vxlan_hdr.vx_flags = vxlan_hdr->vx_flags;
+	key.vxlan_hdr.vx_vni = vxlan_hdr->vx_vni;
+	rte_ether_addr_copy(&(outer_eth_hdr->s_addr), &(key.outer_eth_saddr));
+	rte_ether_addr_copy(&(outer_eth_hdr->d_addr), &(key.outer_eth_daddr));
+	gro_ipv6_addr_copy(outer_ipv6_hdr->src_addr, key.outer_ip_saddr);
+	gro_ipv6_addr_copy(outer_ipv6_hdr->dst_addr, key.outer_ip_daddr);
+	key.outer_src_port = udp_hdr->src_port;
+	key.outer_dst_port = udp_hdr->dst_port;
+
+	/* Search for a matched flow. */
+	max_flow_num = tbl->max_flow_num;
+	remaining_flow_num = tbl->flow_num;
+	find = 0;
+	for (i = 0; i < max_flow_num && remaining_flow_num; i++) {
+		if (tbl->flows[i].start_index != INVALID_ARRAY_INDEX) {
+			if (is_same_vxlan6_tcp4_flow(tbl->flows[i].key, key)) {
+				find = 1;
+				break;
+			}
+			remaining_flow_num--;
+		}
+	}
+
+	/*
+	 * Can't find a matched flow. Insert a new flow and store the
+	 * packet into the flow.
+	 */
+	if (find == 0) {
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, sent_seq,
+				ip_id, is_atomic);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		if (insert_new_flow(tbl, &key, item_idx) ==
+				INVALID_ARRAY_INDEX) {
+			/*
+			 * Fail to insert a new flow, so
+			 * delete the inserted packet.
+			 */
+			delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+			return -1;
+		}
+		return 0;
+	}
+
+	/* Check all packets in the flow and try to find a neighbor. */
+	cur_idx = tbl->flows[i].start_index;
+	prev_idx = cur_idx;
+	do {
+		cmp = check_vxlan6_seq_option(&(tbl->items[cur_idx]), tcp_hdr,
+				sent_seq, ip_id, pkt->l4_len,
+				tcp_dl, is_atomic);
+		if (cmp) {
+			if (merge_two_vxlan6_tcp4_packets(
+						&(tbl->items[cur_idx]),
+						pkt, cmp, sent_seq, ip_id))
+				return 1;
+			/*
+			 * Can't merge two packets, as the packet
+			 * length will be greater than the max value.
+			 * Insert the packet into the flow.
+			 */
+			if (insert_new_item(tbl, pkt, start_time, prev_idx,
+					    sent_seq, ip_id, is_atomic) ==
+					INVALID_ARRAY_INDEX)
+				return -1;
+			return 0;
+		}
+		prev_idx = cur_idx;
+		cur_idx = tbl->items[cur_idx].inner_item.next_pkt_idx;
+	} while (cur_idx != INVALID_ARRAY_INDEX);
+
+	/* Can't find neighbor. Insert the packet into the flow. */
+	if (insert_new_item(tbl, pkt, start_time, prev_idx, sent_seq,
+				ip_id, is_atomic)
+				== INVALID_ARRAY_INDEX)
+		return -1;
+
+	return 0;
+}
+
+uint16_t
+gro_vxlan6_tcp4_tbl_timeout_flush(struct gro_vxlan6_tcp4_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out)
+{
+	uint16_t k = 0;
+	uint32_t i, j;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++) {
+		if (unlikely(tbl->flow_num == 0))
+			return k;
+
+		j = tbl->flows[i].start_index;
+		while (j != INVALID_ARRAY_INDEX) {
+			if (tbl->items[j].inner_item.start_time <=
+					flush_timestamp) {
+				out[k++] = tbl->items[j].inner_item.firstseg;
+				if (tbl->items[j].inner_item.nb_merged > 1)
+					update_vxlan6_header(&(tbl->items[j]));
+				/*
+				 * Delete the item and get the next packet
+				 * index.
+				 */
+				j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+				tbl->flows[i].start_index = j;
+				if (j == INVALID_ARRAY_INDEX)
+					tbl->flow_num--;
+
+				if (unlikely(k == nb_out))
+					return k;
+			} else
+				/*
+				 * The left packets in the flow won't be
+				 * timeout. Go to check other flows.
+				 */
+				break;
+		}
+	}
+	return k;
+}
+
+uint32_t
+gro_vxlan6_tcp4_tbl_pkt_count(void *tbl)
+{
+	struct gro_vxlan6_tcp4_tbl *gro_tbl = tbl;
+
+	if (gro_tbl)
+		return gro_tbl->item_num;
+
+	return 0;
+}
diff --git a/lib/librte_gro/gro_vxlan6_tcp4.h b/lib/librte_gro/gro_vxlan6_tcp4.h
new file mode 100644
index 0000000..2516429
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan6_tcp4.h
@@ -0,0 +1,153 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#ifndef _GRO_VXLAN6_TCP4_H_
+#define _GRO_VXLAN6_TCP4_H_
+
+#include "gro_tcp4.h"
+
+#define GRO_VXLAN6_TCP4_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/* Header fields representing a IPv6 VxLAN flow */
+struct vxlan6_tcp4_flow_key {
+	struct tcp4_flow_key inner_key;
+	struct rte_vxlan_hdr vxlan_hdr;
+
+	struct rte_ether_addr outer_eth_saddr;
+	struct rte_ether_addr outer_eth_daddr;
+
+	uint8_t outer_ip_saddr[16];
+	uint8_t outer_ip_daddr[16];
+
+	/* Outer UDP ports */
+	uint16_t outer_src_port;
+	uint16_t outer_dst_port;
+
+};
+
+struct gro_vxlan6_tcp4_flow {
+	struct vxlan6_tcp4_flow_key key;
+	/*
+	 * The index of the first packet in the flow. INVALID_ARRAY_INDEX
+	 * indicates an empty flow.
+	 */
+	uint32_t start_index;
+};
+
+struct gro_vxlan6_tcp4_item {
+	struct gro_tcp4_item inner_item;
+};
+
+/*
+ * IPv6 VxLAN (with an outer IPv6 header and an inner TCP/IPv4 packet)
+ * reassembly table structure
+ */
+struct gro_vxlan6_tcp4_tbl {
+	/* item array */
+	struct gro_vxlan6_tcp4_item *items;
+	/* flow array */
+	struct gro_vxlan6_tcp4_flow *flows;
+	/* current item number */
+	uint32_t item_num;
+	/* current flow number */
+	uint32_t flow_num;
+	/* the maximum item number */
+	uint32_t max_item_num;
+	/* the maximum flow number */
+	uint32_t max_flow_num;
+};
+
+
+/**
+ * This function creates a IPv6 VxLAN reassembly table for IPv6 VxLAN packets
+ * which have an outer IPv6 header and an inner TCP/IPv4 packet.
+ *
+ * @param socket_id
+ *  Socket index for allocating the table
+ * @param max_flow_num
+ *  The maximum number of flows in the table
+ * @param max_item_per_flow
+ *  The maximum number of packets per flow
+ *
+ * @return
+ *  - Return the table pointer on success.
+ *  - Return NULL on failure.
+ */
+void *gro_vxlan6_tcp4_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow);
+
+/**
+ * This function destroys a IPv6 VxLAN reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ */
+void gro_vxlan6_tcp4_tbl_destroy(void *tbl);
+
+/**
+ * This function merges a IPv6 VxLAN packet which has an outer IPv6 header and
+ * an inner TCP/IPv4 packet. It does not process the packet, whose TCP
+ * header has SYN, FIN, RST, PSH, CWR, ECE or URG bit set, or which
+ * does not have payload.
+ *
+ * This function doesi not check if the packet has correct checksums and
+ * does not re-calculate checksums for the merged packet. Additionally,
+ * it assumes the packets are complete (i.e., MF==0 && frag_off==0), when
+ * IP fragmentation is possible (i.e., DF==0). It returns the packet, if
+ * the packet has invalid parameters (e.g. SYN bit is set) or there is no
+ * available space in the table.
+ *
+ * @param pkt
+ *  Packet to reassemble
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ * @start_time
+ *  The time when the packet is inserted into the table
+ *
+ * @return
+ *  - Return a positive value if the packet is merged.
+ *  - Return zero if the packet isn't merged but stored in the table.
+ *  - Return a negative value for invalid parameters or no available
+ *    space in the table.
+ */
+int32_t gro_vxlan6_tcp4_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan6_tcp4_tbl *tbl,
+		uint64_t start_time);
+
+/**
+ * This function flushes timeout packets in the IPv6 VxLAN reassembly table,
+ * and without updating checksums.
+ *
+ * @param tbl
+ *  Pointer pointing to a IPv6 VxLAN GRO table
+ * @param flush_timestamp
+ *  This function flushes packets which are inserted into the table
+ *  before or at the flush_timestamp.
+ * @param out
+ *  Pointer array used to keep flushed packets
+ * @param nb_out
+ *  The element number in 'out'. It also determines the maximum number of
+ *  packets that can be flushed finally.
+ *
+ * @return
+ *  The number of flushed packets
+ */
+uint16_t gro_vxlan6_tcp4_tbl_timeout_flush(struct gro_vxlan6_tcp4_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out);
+
+/**
+ * This function returns the number of the packets in a IPv6 VxLAN
+ * reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ *
+ * @return
+ *  The number of packets in the table
+ */
+uint32_t gro_vxlan6_tcp4_tbl_pkt_count(void *tbl);
+#endif
diff --git a/lib/librte_gro/meson.build b/lib/librte_gro/meson.build
index 020e7e8..994ee06 100644
--- a/lib/librte_gro/meson.build
+++ b/lib/librte_gro/meson.build
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c')
+sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c')
 headers = files('rte_gro.h')
 deps += ['ethdev']
diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
index 89fed3c..12ed3b2 100644
--- a/lib/librte_gro/rte_gro.c
+++ b/lib/librte_gro/rte_gro.c
@@ -11,6 +11,7 @@
 #include "gro_tcp4.h"
 #include "gro_tcp6.h"
 #include "gro_udp4.h"
+#include "gro_vxlan6_tcp4.h"
 #include "gro_vxlan_tcp4.h"
 #include "gro_vxlan_tcp6.h"
 #include "gro_vxlan_udp4.h"
@@ -24,16 +25,18 @@
 static gro_tbl_create_fn tbl_create_fn[RTE_GRO_TYPE_MAX_NUM] = {
 		gro_tcp4_tbl_create, gro_vxlan_tcp4_tbl_create,
 		gro_udp4_tbl_create, gro_vxlan_udp4_tbl_create,
-		gro_tcp6_tbl_create, gro_vxlan_tcp6_tbl_create, NULL};
+		gro_tcp6_tbl_create, gro_vxlan_tcp6_tbl_create,
+		gro_vxlan6_tcp4_tbl_create, NULL};
 static gro_tbl_destroy_fn tbl_destroy_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_destroy, gro_vxlan_tcp4_tbl_destroy,
 			gro_udp4_tbl_destroy, gro_vxlan_udp4_tbl_destroy,
-			gro_tcp6_tbl_destroy, gro_vxlan_tcp6_tbl_destroy, NULL};
+			gro_tcp6_tbl_destroy, gro_vxlan_tcp6_tbl_destroy,
+			gro_vxlan6_tcp4_tbl_destroy, NULL};
 static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_pkt_count, gro_vxlan_tcp4_tbl_pkt_count,
 			gro_udp4_tbl_pkt_count, gro_vxlan_udp4_tbl_pkt_count,
 			gro_tcp6_tbl_pkt_count, gro_vxlan_tcp6_tbl_pkt_count,
-			NULL};
+			gro_vxlan6_tcp4_tbl_pkt_count, NULL};
 
 #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
@@ -86,6 +89,19 @@
 		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
 		  RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN)))
 
+#define IS_IPV6_VXLAN_TCP4_PKT(ptype) (RTE_ETH_IS_IPV6_HDR(ptype) && \
+		((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
+		((ptype & RTE_PTYPE_TUNNEL_VXLAN) == \
+		 RTE_PTYPE_TUNNEL_VXLAN) && \
+		((ptype & RTE_PTYPE_INNER_L4_TCP) == \
+		 RTE_PTYPE_INNER_L4_TCP) && \
+		(((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV4) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV4_EXT) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN)))
+
 /*
  * GRO context structure. It keeps the table structures, which are
  * used to merge packets, for different GRO types. Before using
@@ -174,7 +190,7 @@ struct gro_ctx {
 	struct gro_udp4_flow udp_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
 	struct gro_udp4_item udp_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
 
-	/* Allocate a reassembly table for VXLAN TCP GRO */
+	/* Allocate a reassembly table for IPv4 VXLAN TCP/IPv4 GRO */
 	struct gro_vxlan_tcp4_tbl vxlan_tcp_tbl;
 	struct gro_vxlan_tcp4_flow vxlan_tcp_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
 	struct gro_vxlan_tcp4_item vxlan_tcp_items[RTE_GRO_MAX_BURST_ITEM_NUM]
@@ -197,19 +213,28 @@ struct gro_ctx {
 	struct gro_vxlan_tcp6_item vxlan_tcp6_items[RTE_GRO_MAX_BURST_ITEM_NUM]
 			= {{{0}, 0, 0} };
 
+	/* Allocate a reassembly table for IPv6 VXLAN TCP/IPv4 GRO */
+	struct gro_vxlan6_tcp4_tbl vxlan6_tcp4_tbl;
+	struct gro_vxlan6_tcp4_flow
+		vxlan6_tcp4_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+	struct gro_vxlan6_tcp4_item
+		vxlan6_tcp4_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{{0}} };
+
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	uint32_t item_num;
 	int32_t ret;
 	uint16_t i, unprocess_num = 0, nb_after_gro = nb_pkts;
 	uint8_t do_tcp4_gro = 0, do_vxlan_tcp_gro = 0, do_udp4_gro = 0,
-		do_vxlan_udp_gro = 0, do_tcp6_gro = 0, do_vxlan_tcp6_gro = 0;
+		do_vxlan_udp_gro = 0, do_tcp6_gro = 0, do_vxlan_tcp6_gro = 0,
+		do_vxlan6_tcp4_gro = 0;
 
 	if (unlikely((param->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
 					RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
 					RTE_GRO_UDP_IPV4 |
 					RTE_GRO_TCP_IPV6 |
-					RTE_GRO_IPV4_VXLAN_TCP_IPV6)) == 0))
+					RTE_GRO_IPV4_VXLAN_TCP_IPV6 |
+					RTE_GRO_IPV6_VXLAN_TCP_IPV4)) == 0))
 		return nb_pkts;
 
 	/* Get the maximum number of packets */
@@ -295,6 +320,19 @@ struct gro_ctx {
 		do_vxlan_tcp6_gro = 1;
 	}
 
+	if (param->gro_types & RTE_GRO_IPV6_VXLAN_TCP_IPV4) {
+		for (i = 0; i < item_num; i++)
+			vxlan6_tcp4_flows[i].start_index = INVALID_ARRAY_INDEX;
+
+		vxlan6_tcp4_tbl.flows = vxlan6_tcp4_flows;
+		vxlan6_tcp4_tbl.items = vxlan6_tcp4_items;
+		vxlan6_tcp4_tbl.flow_num = 0;
+		vxlan6_tcp4_tbl.item_num = 0;
+		vxlan6_tcp4_tbl.max_flow_num = item_num;
+		vxlan6_tcp4_tbl.max_item_num = item_num;
+		do_vxlan6_tcp4_gro = 1;
+	}
+
 	for (i = 0; i < nb_pkts; i++) {
 		/*
 		 * The timestamp is ignored, since all packets
@@ -351,6 +389,15 @@ struct gro_ctx {
 				nb_after_gro--;
 			else if (ret < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_VXLAN_TCP4_PKT(pkts[i]->packet_type) &&
+				do_vxlan6_tcp4_gro) {
+			ret = gro_vxlan6_tcp4_reassemble(pkts[i],
+							&vxlan6_tcp4_tbl, 0);
+			if (ret > 0)
+				/* Merge successfully */
+				nb_after_gro--;
+			else if (ret < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -390,6 +437,11 @@ struct gro_ctx {
 					0, &pkts[i], nb_pkts - i);
 		}
 
+		if (do_vxlan6_tcp4_gro) {
+			i += gro_vxlan6_tcp4_tbl_timeout_flush(&vxlan6_tcp4_tbl,
+					0, &pkts[i], nb_pkts - i);
+		}
+
 		/* Copy unprocessed packets */
 		if (unprocess_num > 0) {
 			memcpy(&pkts[i], unprocess_pkts,
@@ -410,18 +462,19 @@ struct gro_ctx {
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	struct gro_ctx *gro_ctx = ctx;
 	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl,
-		*vxlan_tcp6_tbl;
+		*vxlan_tcp6_tbl, *vxlan6_tcp4_tbl;
 	uint64_t current_time;
 	uint16_t i, unprocess_num = 0;
 	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro,
-		do_tcp6_gro, do_vxlan_tcp6_gro;
+		do_tcp6_gro, do_vxlan_tcp6_gro, do_vxlan6_tcp4_gro;
 
 	if (unlikely((gro_ctx->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
 					RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
 					RTE_GRO_UDP_IPV4 |
 					RTE_GRO_TCP_IPV6 |
-					RTE_GRO_IPV4_VXLAN_TCP_IPV6)) == 0))
+					RTE_GRO_IPV4_VXLAN_TCP_IPV6 |
+					RTE_GRO_IPV6_VXLAN_TCP_IPV4)) == 0))
 		return nb_pkts;
 
 	tcp_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV4_INDEX];
@@ -430,6 +483,7 @@ struct gro_ctx {
 	vxlan_udp_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_UDP_IPV4_INDEX];
 	tcp6_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV6_INDEX];
 	vxlan_tcp6_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX];
+	vxlan6_tcp4_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX];
 
 	do_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV4) ==
 		RTE_GRO_TCP_IPV4;
@@ -443,6 +497,8 @@ struct gro_ctx {
 		RTE_GRO_TCP_IPV6;
 	do_vxlan_tcp6_gro = (gro_ctx->gro_types & RTE_GRO_IPV4_VXLAN_TCP_IPV6)
 				== RTE_GRO_IPV4_VXLAN_TCP_IPV6;
+	do_vxlan6_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_IPV6_VXLAN_TCP_IPV4)
+				== RTE_GRO_IPV6_VXLAN_TCP_IPV4;
 
 	current_time = rte_rdtsc();
 
@@ -477,6 +533,11 @@ struct gro_ctx {
 			if (gro_vxlan_tcp6_reassemble(pkts[i], vxlan_tcp6_tbl,
 						current_time) < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_VXLAN_TCP4_PKT(pkts[i]->packet_type) &&
+				do_vxlan6_tcp4_gro) {
+			if (gro_vxlan6_tcp4_reassemble(pkts[i], vxlan6_tcp4_tbl,
+						current_time) < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -546,6 +607,13 @@ struct gro_ctx {
 		num += gro_vxlan_tcp6_tbl_timeout_flush(gro_ctx->tbls[
 				RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX],
 				flush_timestamp, &out[num], left_nb_out);
+		left_nb_out = max_nb_out - num;
+	}
+
+	if ((gro_types & RTE_GRO_IPV6_VXLAN_TCP_IPV4) && left_nb_out > 0) {
+		num += gro_vxlan6_tcp4_tbl_timeout_flush(gro_ctx->tbls[
+				RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX],
+				flush_timestamp, &out[num], left_nb_out);
 	}
 
 	return num;
diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
index 41a2bec..2949980 100644
--- a/lib/librte_gro/rte_gro.h
+++ b/lib/librte_gro/rte_gro.h
@@ -44,6 +44,9 @@
 #define RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX 5
 #define RTE_GRO_IPV4_VXLAN_TCP_IPV6 (1ULL << RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX)
 /**< IPv4 VxLAN TCP/IPv6 GRO flag. */
+#define RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX 6
+#define RTE_GRO_IPV6_VXLAN_TCP_IPV4 (1ULL << RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX)
+/**< IPv6 VxLAN TCP/IPv4 GRO flag. */
 
 /**
  * Structure used to create GRO context objects or used to pass
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-dev] [PATCH v1 4/8] gro: support IPv6 VXLAN TCP/IPv6
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
                   ` (2 preceding siblings ...)
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 3/8] gro: support IPv6 VXLAN TCP/IPv4 yang_y_yi
@ 2020-12-21  3:50 ` yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 5/8] gro: support UDP/IPv6 yang_y_yi
                   ` (4 subsequent siblings)
  8 siblings, 0 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

IPv6 VXLAN TCP/IPv6 GRO can help improve TCP/IPv6 performance
in IPv6 VXLAN use case.

With this enabled in DPDK, OVS DPDK can leverage it to improve
VM-to-VM TCP/IPv6 performance, it will merge small adjacent
IPv6 VXLAN TCP/IPv6 packets to a big IPv6 VXLAN TCP/IPv6 packet
immediate after they are received from a physical NIC. It is
very helpful in OVS DPDK IPv6 VXLAN use case.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
---
 .../prog_guide/generic_receive_offload_lib.rst     |   2 +-
 doc/guides/rel_notes/release_21_02.rst             |   5 +
 lib/librte_gro/gro_vxlan6_tcp6.c                   | 464 +++++++++++++++++++++
 lib/librte_gro/gro_vxlan6_tcp6.h                   | 153 +++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           |  87 +++-
 lib/librte_gro/rte_gro.h                           |   3 +
 7 files changed, 706 insertions(+), 10 deletions(-)
 create mode 100644 lib/librte_gro/gro_vxlan6_tcp6.c
 create mode 100644 lib/librte_gro/gro_vxlan6_tcp6.h

diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst b/doc/guides/prog_guide/generic_receive_offload_lib.rst
index a822eaf..0efade3 100644
--- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
+++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
@@ -33,7 +33,7 @@ fragmentation is possible (i.e., DF==0). Additionally, it complies RFC
 Currently, the GRO library provides GRO supports for TCP/IPv4, UDP/IPv4,
 and TCP/IPv6 packets as well as VxLAN packets which contain an outer IPv4
 header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6 packet or an outer IPv6
-header and an inner TCP/IPv4 packet.
+header and an inner TCP/IPv4 or TCP/IPv6 packet.
 
 Two Sets of API
 ---------------
diff --git a/doc/guides/rel_notes/release_21_02.rst b/doc/guides/rel_notes/release_21_02.rst
index 3b98915..a5bc2f1 100644
--- a/doc/guides/rel_notes/release_21_02.rst
+++ b/doc/guides/rel_notes/release_21_02.rst
@@ -70,6 +70,11 @@ New Features
    Added TCP/IPv4 GRO support for IPv6 VXLAN packets, this enabled TCP/IPv4
    GRO support for IPv6 VXLAN use case.
 
+* **Added TCP/IPv6 GRO support for IPv6 VXLAN packets.**
+
+   Added TCP/IPv6 GRO support for IPv6 VXLAN packets, this enabled TCP/IPv6
+   GRO support for IPv6 VXLAN use case.
+
 Removed Items
 -------------
 
diff --git a/lib/librte_gro/gro_vxlan6_tcp6.c b/lib/librte_gro/gro_vxlan6_tcp6.c
new file mode 100644
index 0000000..5b9dee4
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan6_tcp6.c
@@ -0,0 +1,464 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+#include <rte_udp.h>
+
+#include "gro_vxlan6_tcp6.h"
+
+void *
+gro_vxlan6_tcp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow)
+{
+	struct gro_vxlan6_tcp6_tbl *tbl;
+	size_t size;
+	uint32_t entries_num, i;
+
+	entries_num = max_flow_num * max_item_per_flow;
+	entries_num = RTE_MIN(entries_num, GRO_VXLAN6_TCP6_TBL_MAX_ITEM_NUM);
+
+	if (entries_num == 0)
+		return NULL;
+
+	tbl = rte_zmalloc_socket(__func__,
+			sizeof(struct gro_vxlan6_tcp6_tbl),
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl == NULL)
+		return NULL;
+
+	size = sizeof(struct gro_vxlan6_tcp6_item) * entries_num;
+	tbl->items = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->items == NULL) {
+		rte_free(tbl);
+		return NULL;
+	}
+	tbl->max_item_num = entries_num;
+
+	size = sizeof(struct gro_vxlan6_tcp6_flow) * entries_num;
+	tbl->flows = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->flows == NULL) {
+		rte_free(tbl->items);
+		rte_free(tbl);
+		return NULL;
+	}
+
+	for (i = 0; i < entries_num; i++)
+		tbl->flows[i].start_index = INVALID_ARRAY_INDEX;
+	tbl->max_flow_num = entries_num;
+
+	return tbl;
+}
+
+void
+gro_vxlan6_tcp6_tbl_destroy(void *tbl)
+{
+	struct gro_vxlan6_tcp6_tbl *vxlan6_tbl = tbl;
+
+	if (vxlan6_tbl) {
+		rte_free(vxlan6_tbl->items);
+		rte_free(vxlan6_tbl->flows);
+	}
+	rte_free(vxlan6_tbl);
+}
+
+static inline uint32_t
+find_an_empty_item(struct gro_vxlan6_tcp6_tbl *tbl)
+{
+	uint32_t max_item_num = tbl->max_item_num, i;
+
+	for (i = 0; i < max_item_num; i++)
+		if (tbl->items[i].inner_item.firstseg == NULL)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+find_an_empty_flow(struct gro_vxlan6_tcp6_tbl *tbl)
+{
+	uint32_t max_flow_num = tbl->max_flow_num, i;
+
+	for (i = 0; i < max_flow_num; i++)
+		if (tbl->flows[i].start_index == INVALID_ARRAY_INDEX)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_item(struct gro_vxlan6_tcp6_tbl *tbl,
+		struct rte_mbuf *pkt,
+		uint64_t start_time,
+		uint32_t prev_idx,
+		uint32_t sent_seq)
+{
+	uint32_t item_idx;
+
+	item_idx = find_an_empty_item(tbl);
+	if (unlikely(item_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	tbl->items[item_idx].inner_item.firstseg = pkt;
+	tbl->items[item_idx].inner_item.lastseg = rte_pktmbuf_lastseg(pkt);
+	tbl->items[item_idx].inner_item.start_time = start_time;
+	tbl->items[item_idx].inner_item.next_pkt_idx = INVALID_ARRAY_INDEX;
+	tbl->items[item_idx].inner_item.sent_seq = sent_seq;
+	tbl->items[item_idx].inner_item.nb_merged = 1;
+	tbl->item_num++;
+
+	/* If the previous packet exists, chain the new one with it. */
+	if (prev_idx != INVALID_ARRAY_INDEX) {
+		tbl->items[item_idx].inner_item.next_pkt_idx =
+			tbl->items[prev_idx].inner_item.next_pkt_idx;
+		tbl->items[prev_idx].inner_item.next_pkt_idx = item_idx;
+	}
+
+	return item_idx;
+}
+
+static inline uint32_t
+delete_item(struct gro_vxlan6_tcp6_tbl *tbl,
+		uint32_t item_idx,
+		uint32_t prev_item_idx)
+{
+	uint32_t next_idx = tbl->items[item_idx].inner_item.next_pkt_idx;
+
+	/* NULL indicates an empty item. */
+	tbl->items[item_idx].inner_item.firstseg = NULL;
+	tbl->item_num--;
+	if (prev_item_idx != INVALID_ARRAY_INDEX)
+		tbl->items[prev_item_idx].inner_item.next_pkt_idx = next_idx;
+
+	return next_idx;
+}
+
+static inline uint32_t
+insert_new_flow(struct gro_vxlan6_tcp6_tbl *tbl,
+		struct vxlan6_tcp6_flow_key *src,
+		uint32_t item_idx)
+{
+	struct vxlan6_tcp6_flow_key *dst;
+	uint32_t flow_idx;
+
+	flow_idx = find_an_empty_flow(tbl);
+	if (unlikely(flow_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	dst = &(tbl->flows[flow_idx].key);
+
+	rte_ether_addr_copy(&(src->inner_key.eth_saddr),
+			&(dst->inner_key.eth_saddr));
+	rte_ether_addr_copy(&(src->inner_key.eth_daddr),
+			&(dst->inner_key.eth_daddr));
+	gro_ipv6_addr_copy(src->inner_key.ip_saddr, dst->inner_key.ip_saddr);
+	gro_ipv6_addr_copy(src->inner_key.ip_daddr, dst->inner_key.ip_daddr);
+	dst->inner_key.recv_ack = src->inner_key.recv_ack;
+	dst->inner_key.src_port = src->inner_key.src_port;
+	dst->inner_key.dst_port = src->inner_key.dst_port;
+
+	dst->vxlan_hdr.vx_flags = src->vxlan_hdr.vx_flags;
+	dst->vxlan_hdr.vx_vni = src->vxlan_hdr.vx_vni;
+	rte_ether_addr_copy(&(src->outer_eth_saddr), &(dst->outer_eth_saddr));
+	rte_ether_addr_copy(&(src->outer_eth_daddr), &(dst->outer_eth_daddr));
+	gro_ipv6_addr_copy(src->outer_ip_saddr, dst->outer_ip_saddr);
+	gro_ipv6_addr_copy(src->outer_ip_daddr, dst->outer_ip_daddr);
+	dst->outer_src_port = src->outer_src_port;
+	dst->outer_dst_port = src->outer_dst_port;
+
+	tbl->flows[flow_idx].start_index = item_idx;
+	tbl->flow_num++;
+
+	return flow_idx;
+}
+
+static inline int
+is_same_vxlan6_tcp6_flow(struct vxlan6_tcp6_flow_key k1,
+		struct vxlan6_tcp6_flow_key k2)
+{
+	return (rte_is_same_ether_addr(&k1.outer_eth_saddr,
+					&k2.outer_eth_saddr) &&
+			rte_is_same_ether_addr(&k1.outer_eth_daddr,
+				&k2.outer_eth_daddr) &&
+			rte_is_same_ipv6_addr(k1.outer_ip_saddr,
+					      k2.outer_ip_saddr) &&
+			rte_is_same_ipv6_addr(k1.outer_ip_daddr,
+					      k2.outer_ip_daddr) &&
+			(k1.outer_src_port == k2.outer_src_port) &&
+			(k1.outer_dst_port == k2.outer_dst_port) &&
+			(k1.vxlan_hdr.vx_flags == k2.vxlan_hdr.vx_flags) &&
+			(k1.vxlan_hdr.vx_vni == k2.vxlan_hdr.vx_vni) &&
+			is_same_tcp6_flow(k1.inner_key, k2.inner_key));
+}
+
+static inline int
+check_vxlan_seq_option_v6(struct gro_vxlan6_tcp6_item *item,
+		struct rte_tcp_hdr *tcp_hdr,
+		uint32_t sent_seq,
+		uint16_t tcp_hl,
+		uint16_t tcp_dl)
+{
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	int cmp;
+	uint16_t l2_offset;
+
+	l2_offset = pkt->outer_l2_len + pkt->outer_l3_len;
+	cmp = check_seq_option_v6(&item->inner_item, tcp_hdr, sent_seq,
+			tcp_hl, tcp_dl, l2_offset);
+	if (cmp > 0)
+		/* Append the new packet. */
+		return 1;
+	else if (cmp < 0)
+		/* Prepend the new packet. */
+		return -1;
+
+	return 0;
+}
+
+static inline int
+merge_two_vxlan6_tcp6_packets(struct gro_vxlan6_tcp6_item *item,
+		struct rte_mbuf *pkt,
+		int cmp,
+		uint32_t sent_seq)
+{
+	if (merge_two_tcp6_packets(&item->inner_item, pkt, cmp, sent_seq,
+				pkt->outer_l2_len +
+				pkt->outer_l3_len))
+		return 1;
+
+	return 0;
+}
+
+static inline void
+update_vxlan6_header(struct gro_vxlan6_tcp6_item *item)
+{
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	uint16_t len;
+
+	/* Update the outer IPv6 header. */
+	len = pkt->pkt_len - pkt->outer_l2_len;
+	ipv6_hdr = (struct rte_ipv6_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+			pkt->outer_l2_len);
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(len - pkt->outer_l3_len);
+
+	/* Update the outer UDP header. */
+	len -= pkt->outer_l3_len;
+	udp_hdr = (struct rte_udp_hdr *)((char *)ipv6_hdr + pkt->outer_l3_len);
+	udp_hdr->dgram_len = rte_cpu_to_be_16(len);
+
+	/* Update the inner IPv6 header. */
+	len -= pkt->l2_len;
+	ipv6_hdr = (struct rte_ipv6_hdr *)((char *)udp_hdr + pkt->l2_len);
+	/* IPv6 payload_len does not include IPv6 header. */
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(len - pkt->l3_len);
+}
+
+int32_t
+gro_vxlan6_tcp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan6_tcp6_tbl *tbl,
+		uint64_t start_time)
+{
+	struct rte_ether_hdr *outer_eth_hdr, *eth_hdr;
+	struct rte_ipv6_hdr *outer_ipv6_hdr;
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_tcp_hdr *tcp_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_vxlan_hdr *vxlan_hdr;
+	uint32_t sent_seq;
+	int32_t tcp_dl;
+
+	struct vxlan6_tcp6_flow_key key;
+	uint32_t cur_idx, prev_idx, item_idx;
+	uint32_t i, max_flow_num, remaining_flow_num;
+	int cmp;
+	uint16_t hdr_len;
+	uint8_t find;
+
+	/*
+	 * Don't process the packet whose TCP header length is greater
+	 * than 60 bytes or less than 20 bytes.
+	 */
+	if (unlikely(INVALID_TCP_HDRLEN(pkt->l4_len)))
+		return -1;
+
+	outer_eth_hdr = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *);
+	outer_ipv6_hdr = (struct rte_ipv6_hdr *)((char *)outer_eth_hdr +
+			pkt->outer_l2_len);
+	udp_hdr = (struct rte_udp_hdr *)((char *)outer_ipv6_hdr +
+			pkt->outer_l3_len);
+	vxlan_hdr = (struct rte_vxlan_hdr *)((char *)udp_hdr +
+			sizeof(struct rte_udp_hdr));
+	eth_hdr = (struct rte_ether_hdr *)((char *)vxlan_hdr +
+			sizeof(struct rte_vxlan_hdr));
+	ipv6_hdr = (struct rte_ipv6_hdr *)((char *)udp_hdr + pkt->l2_len);
+	tcp_hdr = (struct rte_tcp_hdr *)((char *)ipv6_hdr + pkt->l3_len);
+
+	/*
+	 * Don't process the packet which has FIN, SYN, RST, PSH, URG,
+	 * ECE or CWR set.
+	 */
+	if (tcp_hdr->tcp_flags != RTE_TCP_ACK_FLAG)
+		return -1;
+
+	hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+		pkt->l3_len + pkt->l4_len;
+	/*
+	 * Don't process the packet whose payload length is less than or
+	 * equal to 0.
+	 */
+	if (pkt->pkt_len <= hdr_len)
+		return -1;
+	tcp_dl = pkt->pkt_len - hdr_len;
+
+	sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
+
+	rte_ether_addr_copy(&(eth_hdr->s_addr), &(key.inner_key.eth_saddr));
+	rte_ether_addr_copy(&(eth_hdr->d_addr), &(key.inner_key.eth_daddr));
+	gro_ipv6_addr_copy(ipv6_hdr->src_addr, key.inner_key.ip_saddr);
+	gro_ipv6_addr_copy(ipv6_hdr->dst_addr, key.inner_key.ip_daddr);
+	key.inner_key.recv_ack = tcp_hdr->recv_ack;
+	key.inner_key.src_port = tcp_hdr->src_port;
+	key.inner_key.dst_port = tcp_hdr->dst_port;
+
+	key.vxlan_hdr.vx_flags = vxlan_hdr->vx_flags;
+	key.vxlan_hdr.vx_vni = vxlan_hdr->vx_vni;
+	rte_ether_addr_copy(&(outer_eth_hdr->s_addr), &(key.outer_eth_saddr));
+	rte_ether_addr_copy(&(outer_eth_hdr->d_addr), &(key.outer_eth_daddr));
+	gro_ipv6_addr_copy(outer_ipv6_hdr->src_addr, key.outer_ip_saddr);
+	gro_ipv6_addr_copy(outer_ipv6_hdr->dst_addr, key.outer_ip_daddr);
+	key.outer_src_port = udp_hdr->src_port;
+	key.outer_dst_port = udp_hdr->dst_port;
+
+	/* Search for a matched flow. */
+	max_flow_num = tbl->max_flow_num;
+	remaining_flow_num = tbl->flow_num;
+	find = 0;
+	for (i = 0; i < max_flow_num && remaining_flow_num; i++) {
+		if (tbl->flows[i].start_index != INVALID_ARRAY_INDEX) {
+			if (is_same_vxlan6_tcp6_flow(tbl->flows[i].key, key)) {
+				find = 1;
+				break;
+			}
+			remaining_flow_num--;
+		}
+	}
+
+	/*
+	 * Can't find a matched flow. Insert a new flow and store the
+	 * packet into the flow.
+	 */
+	if (find == 0) {
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, sent_seq);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		if (insert_new_flow(tbl, &key, item_idx) ==
+				INVALID_ARRAY_INDEX) {
+			/*
+			 * Fail to insert a new flow, so
+			 * delete the inserted packet.
+			 */
+			delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+			return -1;
+		}
+		return 0;
+	}
+
+	/* Check all packets in the flow and try to find a neighbor. */
+	cur_idx = tbl->flows[i].start_index;
+	prev_idx = cur_idx;
+	do {
+		cmp = check_vxlan_seq_option_v6(&(tbl->items[cur_idx]), tcp_hdr,
+				sent_seq, pkt->l4_len, tcp_dl);
+		if (cmp) {
+			if (merge_two_vxlan6_tcp6_packets(
+						&(tbl->items[cur_idx]),
+						pkt, cmp, sent_seq))
+				return 1;
+			/*
+			 * Can't merge two packets, as the packet
+			 * length will be greater than the max value.
+			 * Insert the packet into the flow.
+			 */
+			if (insert_new_item(tbl, pkt, start_time, prev_idx,
+						sent_seq) ==
+					INVALID_ARRAY_INDEX)
+				return -1;
+			return 0;
+		}
+		prev_idx = cur_idx;
+		cur_idx = tbl->items[cur_idx].inner_item.next_pkt_idx;
+	} while (cur_idx != INVALID_ARRAY_INDEX);
+
+	/* Can't find neighbor. Insert the packet into the flow. */
+	if (insert_new_item(tbl, pkt, start_time, prev_idx, sent_seq
+				) == INVALID_ARRAY_INDEX)
+		return -1;
+
+	return 0;
+}
+
+uint16_t
+gro_vxlan6_tcp6_tbl_timeout_flush(struct gro_vxlan6_tcp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out)
+{
+	uint16_t k = 0;
+	uint32_t i, j;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++) {
+		if (unlikely(tbl->flow_num == 0))
+			return k;
+
+		j = tbl->flows[i].start_index;
+		while (j != INVALID_ARRAY_INDEX) {
+			if (tbl->items[j].inner_item.start_time <=
+					flush_timestamp) {
+				out[k++] = tbl->items[j].inner_item.firstseg;
+				if (tbl->items[j].inner_item.nb_merged > 1)
+					update_vxlan6_header(&(tbl->items[j]));
+				/*
+				 * Delete the item and get the next packet
+				 * index.
+				 */
+				j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+				tbl->flows[i].start_index = j;
+				if (j == INVALID_ARRAY_INDEX)
+					tbl->flow_num--;
+
+				if (unlikely(k == nb_out))
+					return k;
+			} else
+				/*
+				 * The left packets in the flow won't be
+				 * timeout. Go to check other flows.
+				 */
+				break;
+		}
+	}
+	return k;
+}
+
+uint32_t
+gro_vxlan6_tcp6_tbl_pkt_count(void *tbl)
+{
+	struct gro_vxlan6_tcp6_tbl *gro_tbl = tbl;
+
+	if (gro_tbl)
+		return gro_tbl->item_num;
+
+	return 0;
+}
diff --git a/lib/librte_gro/gro_vxlan6_tcp6.h b/lib/librte_gro/gro_vxlan6_tcp6.h
new file mode 100644
index 0000000..0659ae4
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan6_tcp6.h
@@ -0,0 +1,153 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#ifndef _GRO_VXLAN6_TCP6_H_
+#define _GRO_VXLAN6_TCP6_H_
+
+#include "gro_tcp6.h"
+
+#define GRO_VXLAN6_TCP6_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/* Header fields representing a IPv6 VxLAN flow */
+struct vxlan6_tcp6_flow_key {
+	struct tcp6_flow_key inner_key;
+	struct rte_vxlan_hdr vxlan_hdr;
+
+	struct rte_ether_addr outer_eth_saddr;
+	struct rte_ether_addr outer_eth_daddr;
+
+	uint8_t outer_ip_saddr[16];
+	uint8_t outer_ip_daddr[16];
+
+	/* Outer UDP ports */
+	uint16_t outer_src_port;
+	uint16_t outer_dst_port;
+
+};
+
+struct gro_vxlan6_tcp6_flow {
+	struct vxlan6_tcp6_flow_key key;
+	/*
+	 * The index of the first packet in the flow. INVALID_ARRAY_INDEX
+	 * indicates an empty flow.
+	 */
+	uint32_t start_index;
+};
+
+struct gro_vxlan6_tcp6_item {
+	struct gro_tcp6_item inner_item;
+};
+
+/*
+ * IPv6 VxLAN (with an outer IPv6 header and an inner TCP/IPv6 packet)
+ * reassembly table structure
+ */
+struct gro_vxlan6_tcp6_tbl {
+	/* item array */
+	struct gro_vxlan6_tcp6_item *items;
+	/* flow array */
+	struct gro_vxlan6_tcp6_flow *flows;
+	/* current item number */
+	uint32_t item_num;
+	/* current flow number */
+	uint32_t flow_num;
+	/* the maximum item number */
+	uint32_t max_item_num;
+	/* the maximum flow number */
+	uint32_t max_flow_num;
+};
+
+
+/**
+ * This function creates a IPv6 VxLAN reassembly table for IPv6 VxLAN packets
+ * which have an outer IPv6 header and an inner TCP/IPv6 packet.
+ *
+ * @param socket_id
+ *  Socket index for allocating the table
+ * @param max_flow_num
+ *  The maximum number of flows in the table
+ * @param max_item_per_flow
+ *  The maximum number of packets per flow
+ *
+ * @return
+ *  - Return the table pointer on success.
+ *  - Return NULL on failure.
+ */
+void *gro_vxlan6_tcp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow);
+
+/**
+ * This function destroys a IPv6 VxLAN reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ */
+void gro_vxlan6_tcp6_tbl_destroy(void *tbl);
+
+/**
+ * This function merges a IPv6 VxLAN packet which has an outer IPv6 header and
+ * an inner TCP/IPv6 packet. It does not process the packet, whose TCP
+ * header has SYN, FIN, RST, PSH, CWR, ECE or URG bit set, or which
+ * does not have payload.
+ *
+ * This function does not check if the packet has correct checksums and
+ * does not re-calculate checksums for the merged packet. Additionally,
+ * it assumes the packets are complete (i.e., MF==0 && frag_off==0), when
+ * IP fragmentation is possible (i.e., DF==0). It returns the packet, if
+ * the packet has invalid parameters (e.g. SYN bit is set) or there is no
+ * available space in the table.
+ *
+ * @param pkt
+ *  Packet to reassemble
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ * @start_time
+ *  The time when the packet is inserted into the table
+ *
+ * @return
+ *  - Return a positive value if the packet is merged.
+ *  - Return zero if the packet isn't merged but stored in the table.
+ *  - Return a negative value for invalid parameters or no available
+ *    space in the table.
+ */
+int32_t gro_vxlan6_tcp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan6_tcp6_tbl *tbl,
+		uint64_t start_time);
+
+/**
+ * This function flushes timeout packets in the IPv6 VxLAN reassembly table,
+ * and without updating checksums.
+ *
+ * @param tbl
+ *  Pointer pointing to a IPv6 VxLAN GRO table
+ * @param flush_timestamp
+ *  This function flushes packets which are inserted into the table
+ *  before or at the flush_timestamp.
+ * @param out
+ *  Pointer array used to keep flushed packets
+ * @param nb_out
+ *  The element number in 'out'. It also determines the maximum number of
+ *  packets that can be flushed finally.
+ *
+ * @return
+ *  The number of flushed packets
+ */
+uint16_t gro_vxlan6_tcp6_tbl_timeout_flush(struct gro_vxlan6_tcp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out);
+
+/**
+ * This function returns the number of the packets in a IPv6 VxLAN
+ * reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ *
+ * @return
+ *  The number of packets in the table
+ */
+uint32_t gro_vxlan6_tcp6_tbl_pkt_count(void *tbl);
+#endif
diff --git a/lib/librte_gro/meson.build b/lib/librte_gro/meson.build
index 994ee06..3eff89c 100644
--- a/lib/librte_gro/meson.build
+++ b/lib/librte_gro/meson.build
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c')
+sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c')
 headers = files('rte_gro.h')
 deps += ['ethdev']
diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
index 12ed3b2..0c5584e 100644
--- a/lib/librte_gro/rte_gro.c
+++ b/lib/librte_gro/rte_gro.c
@@ -12,6 +12,7 @@
 #include "gro_tcp6.h"
 #include "gro_udp4.h"
 #include "gro_vxlan6_tcp4.h"
+#include "gro_vxlan6_tcp6.h"
 #include "gro_vxlan_tcp4.h"
 #include "gro_vxlan_tcp6.h"
 #include "gro_vxlan_udp4.h"
@@ -26,17 +27,22 @@
 		gro_tcp4_tbl_create, gro_vxlan_tcp4_tbl_create,
 		gro_udp4_tbl_create, gro_vxlan_udp4_tbl_create,
 		gro_tcp6_tbl_create, gro_vxlan_tcp6_tbl_create,
-		gro_vxlan6_tcp4_tbl_create, NULL};
+		gro_vxlan6_tcp4_tbl_create, gro_vxlan6_tcp6_tbl_create,
+		NULL};
 static gro_tbl_destroy_fn tbl_destroy_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_destroy, gro_vxlan_tcp4_tbl_destroy,
 			gro_udp4_tbl_destroy, gro_vxlan_udp4_tbl_destroy,
 			gro_tcp6_tbl_destroy, gro_vxlan_tcp6_tbl_destroy,
-			gro_vxlan6_tcp4_tbl_destroy, NULL};
+			gro_vxlan6_tcp4_tbl_destroy,
+			gro_vxlan6_tcp6_tbl_destroy,
+			NULL};
 static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_pkt_count, gro_vxlan_tcp4_tbl_pkt_count,
 			gro_udp4_tbl_pkt_count, gro_vxlan_udp4_tbl_pkt_count,
 			gro_tcp6_tbl_pkt_count, gro_vxlan_tcp6_tbl_pkt_count,
-			gro_vxlan6_tcp4_tbl_pkt_count, NULL};
+			gro_vxlan6_tcp4_tbl_pkt_count,
+			gro_vxlan6_tcp6_tbl_pkt_count,
+			NULL};
 
 #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
@@ -102,6 +108,19 @@
 		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
 		  RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN)))
 
+#define IS_IPV6_VXLAN_TCP6_PKT(ptype) (RTE_ETH_IS_IPV6_HDR(ptype) && \
+		((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
+		((ptype & RTE_PTYPE_TUNNEL_VXLAN) == \
+		 RTE_PTYPE_TUNNEL_VXLAN) && \
+		((ptype & RTE_PTYPE_INNER_L4_TCP) == \
+		 RTE_PTYPE_INNER_L4_TCP) && \
+		(((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6_EXT) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN)))
+
 /*
  * GRO context structure. It keeps the table structures, which are
  * used to merge packets, for different GRO types. Before using
@@ -220,13 +239,20 @@ struct gro_ctx {
 	struct gro_vxlan6_tcp4_item
 		vxlan6_tcp4_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{{0}} };
 
+	/* Allocate a reassembly table for IPv6 VXLAN TCP/IPv6 GRO */
+	struct gro_vxlan6_tcp6_tbl vxlan6_tcp6_tbl;
+	struct gro_vxlan6_tcp6_flow
+		vxlan6_tcp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+	struct gro_vxlan6_tcp6_item
+		vxlan6_tcp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{{0}} };
+
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	uint32_t item_num;
 	int32_t ret;
 	uint16_t i, unprocess_num = 0, nb_after_gro = nb_pkts;
 	uint8_t do_tcp4_gro = 0, do_vxlan_tcp_gro = 0, do_udp4_gro = 0,
 		do_vxlan_udp_gro = 0, do_tcp6_gro = 0, do_vxlan_tcp6_gro = 0,
-		do_vxlan6_tcp4_gro = 0;
+		do_vxlan6_tcp4_gro = 0, do_vxlan6_tcp6_gro = 0;
 
 	if (unlikely((param->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -234,7 +260,8 @@ struct gro_ctx {
 					RTE_GRO_UDP_IPV4 |
 					RTE_GRO_TCP_IPV6 |
 					RTE_GRO_IPV4_VXLAN_TCP_IPV6 |
-					RTE_GRO_IPV6_VXLAN_TCP_IPV4)) == 0))
+					RTE_GRO_IPV6_VXLAN_TCP_IPV4 |
+					RTE_GRO_IPV6_VXLAN_TCP_IPV6)) == 0))
 		return nb_pkts;
 
 	/* Get the maximum number of packets */
@@ -333,6 +360,19 @@ struct gro_ctx {
 		do_vxlan6_tcp4_gro = 1;
 	}
 
+	if (param->gro_types & RTE_GRO_IPV6_VXLAN_TCP_IPV6) {
+		for (i = 0; i < item_num; i++)
+			vxlan6_tcp6_flows[i].start_index = INVALID_ARRAY_INDEX;
+
+		vxlan6_tcp6_tbl.flows = vxlan6_tcp6_flows;
+		vxlan6_tcp6_tbl.items = vxlan6_tcp6_items;
+		vxlan6_tcp6_tbl.flow_num = 0;
+		vxlan6_tcp6_tbl.item_num = 0;
+		vxlan6_tcp6_tbl.max_flow_num = item_num;
+		vxlan6_tcp6_tbl.max_item_num = item_num;
+		do_vxlan6_tcp6_gro = 1;
+	}
+
 	for (i = 0; i < nb_pkts; i++) {
 		/*
 		 * The timestamp is ignored, since all packets
@@ -398,6 +438,15 @@ struct gro_ctx {
 				nb_after_gro--;
 			else if (ret < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_VXLAN_TCP6_PKT(pkts[i]->packet_type) &&
+				do_vxlan6_tcp6_gro) {
+			ret = gro_vxlan6_tcp6_reassemble(pkts[i],
+							&vxlan6_tcp6_tbl, 0);
+			if (ret > 0)
+				/* Merge successfully */
+				nb_after_gro--;
+			else if (ret < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -442,6 +491,11 @@ struct gro_ctx {
 					0, &pkts[i], nb_pkts - i);
 		}
 
+		if (do_vxlan6_tcp6_gro) {
+			i += gro_vxlan6_tcp6_tbl_timeout_flush(&vxlan6_tcp6_tbl,
+					0, &pkts[i], nb_pkts - i);
+		}
+
 		/* Copy unprocessed packets */
 		if (unprocess_num > 0) {
 			memcpy(&pkts[i], unprocess_pkts,
@@ -462,11 +516,12 @@ struct gro_ctx {
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	struct gro_ctx *gro_ctx = ctx;
 	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl,
-		*vxlan_tcp6_tbl, *vxlan6_tcp4_tbl;
+		*vxlan_tcp6_tbl, *vxlan6_tcp4_tbl, *vxlan6_tcp6_tbl;
 	uint64_t current_time;
 	uint16_t i, unprocess_num = 0;
 	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro,
-		do_tcp6_gro, do_vxlan_tcp6_gro, do_vxlan6_tcp4_gro;
+		do_tcp6_gro, do_vxlan_tcp6_gro, do_vxlan6_tcp4_gro,
+		do_vxlan6_tcp6_gro;
 
 	if (unlikely((gro_ctx->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -474,7 +529,8 @@ struct gro_ctx {
 					RTE_GRO_UDP_IPV4 |
 					RTE_GRO_TCP_IPV6 |
 					RTE_GRO_IPV4_VXLAN_TCP_IPV6 |
-					RTE_GRO_IPV6_VXLAN_TCP_IPV4)) == 0))
+					RTE_GRO_IPV6_VXLAN_TCP_IPV4 |
+					RTE_GRO_IPV6_VXLAN_TCP_IPV6)) == 0))
 		return nb_pkts;
 
 	tcp_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV4_INDEX];
@@ -484,6 +540,7 @@ struct gro_ctx {
 	tcp6_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV6_INDEX];
 	vxlan_tcp6_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX];
 	vxlan6_tcp4_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX];
+	vxlan6_tcp6_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX];
 
 	do_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV4) ==
 		RTE_GRO_TCP_IPV4;
@@ -499,6 +556,8 @@ struct gro_ctx {
 				== RTE_GRO_IPV4_VXLAN_TCP_IPV6;
 	do_vxlan6_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_IPV6_VXLAN_TCP_IPV4)
 				== RTE_GRO_IPV6_VXLAN_TCP_IPV4;
+	do_vxlan6_tcp6_gro = (gro_ctx->gro_types & RTE_GRO_IPV6_VXLAN_TCP_IPV6)
+				== RTE_GRO_IPV6_VXLAN_TCP_IPV6;
 
 	current_time = rte_rdtsc();
 
@@ -538,6 +597,11 @@ struct gro_ctx {
 			if (gro_vxlan6_tcp4_reassemble(pkts[i], vxlan6_tcp4_tbl,
 						current_time) < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_VXLAN_TCP6_PKT(pkts[i]->packet_type) &&
+				do_vxlan6_tcp6_gro) {
+			if (gro_vxlan6_tcp6_reassemble(pkts[i], vxlan6_tcp6_tbl,
+						current_time) < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -614,6 +678,13 @@ struct gro_ctx {
 		num += gro_vxlan6_tcp4_tbl_timeout_flush(gro_ctx->tbls[
 				RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX],
 				flush_timestamp, &out[num], left_nb_out);
+		left_nb_out = max_nb_out - num;
+	}
+
+	if ((gro_types & RTE_GRO_IPV6_VXLAN_TCP_IPV6) && left_nb_out > 0) {
+		num += gro_vxlan6_tcp6_tbl_timeout_flush(gro_ctx->tbls[
+				RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX],
+				flush_timestamp, &out[num], left_nb_out);
 	}
 
 	return num;
diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
index 2949980..cb0895f 100644
--- a/lib/librte_gro/rte_gro.h
+++ b/lib/librte_gro/rte_gro.h
@@ -47,6 +47,9 @@
 #define RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX 6
 #define RTE_GRO_IPV6_VXLAN_TCP_IPV4 (1ULL << RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX)
 /**< IPv6 VxLAN TCP/IPv4 GRO flag. */
+#define RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX 7
+#define RTE_GRO_IPV6_VXLAN_TCP_IPV6 (1ULL << RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX)
+/**< IPv6 VxLAN TCP/IPv6 GRO flag. */
 
 /**
  * Structure used to create GRO context objects or used to pass
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-dev] [PATCH v1 5/8] gro: support UDP/IPv6
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
                   ` (3 preceding siblings ...)
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 4/8] gro: support IPv6 VXLAN TCP/IPv6 yang_y_yi
@ 2020-12-21  3:50 ` yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 6/8] gro: support IPv4 VXLAN UDP/IPv6 yang_y_yi
                   ` (3 subsequent siblings)
  8 siblings, 0 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

UDP/IPv6 GRO can help improve UDP/IPv6 performance.

With this enabled in DPDK, OVS DPDK can leverage it
to improve VM-to-VM UDP/IPv6 performance, it will merge
small adjacent UDP/IPv6 fragments to a big UDP/IPv6 packet
immediate after they are received from a physical NIC. It
is very helpful in OVS DPDK VLAN use case.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
---
 .../prog_guide/generic_receive_offload_lib.rst     |   6 +-
 doc/guides/rel_notes/release_21_02.rst             |   5 +
 lib/librte_gro/gro_udp6.c                          | 487 +++++++++++++++++++++
 lib/librte_gro/gro_udp6.h                          | 284 ++++++++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           |  72 ++-
 lib/librte_gro/rte_gro.h                           |   3 +
 7 files changed, 847 insertions(+), 12 deletions(-)
 create mode 100644 lib/librte_gro/gro_udp6.c
 create mode 100644 lib/librte_gro/gro_udp6.h

diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst b/doc/guides/prog_guide/generic_receive_offload_lib.rst
index 0efade3..0ea3076 100644
--- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
+++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
@@ -31,9 +31,9 @@ fragmentation is possible (i.e., DF==0). Additionally, it complies RFC
 6864 to process the IPv4 ID field.
 
 Currently, the GRO library provides GRO supports for TCP/IPv4, UDP/IPv4,
-and TCP/IPv6 packets as well as VxLAN packets which contain an outer IPv4
-header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6 packet or an outer IPv6
-header and an inner TCP/IPv4 or TCP/IPv6 packet.
+UDP/IPv6, and TCP/IPv6 packets as well as VxLAN packets which contain an
+outer IPv4 header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6 packet or
+an outer IPv6 header and an inner TCP/IPv4 or TCP/IPv6 packet.
 
 Two Sets of API
 ---------------
diff --git a/doc/guides/rel_notes/release_21_02.rst b/doc/guides/rel_notes/release_21_02.rst
index a5bc2f1..d8974b7 100644
--- a/doc/guides/rel_notes/release_21_02.rst
+++ b/doc/guides/rel_notes/release_21_02.rst
@@ -75,6 +75,11 @@ New Features
    Added TCP/IPv6 GRO support for IPv6 VXLAN packets, this enabled TCP/IPv6
    GRO support for IPv6 VXLAN use case.
 
+* **Added UDP/IPv6 GRO support for non-VXLAN packets.**
+
+   Added UDP/IPv6 GRO support for non-VXLAN packets, this enabled UDP/IPv6
+   GRO support for IPv6 VLAN use case.
+
 Removed Items
 -------------
 
diff --git a/lib/librte_gro/gro_udp6.c b/lib/librte_gro/gro_udp6.c
new file mode 100644
index 0000000..193cf2e
--- /dev/null
+++ b/lib/librte_gro/gro_udp6.c
@@ -0,0 +1,487 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+#include <rte_ip.h>
+
+#include "gro_udp6.h"
+
+void *
+gro_udp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow)
+{
+	struct gro_udp6_tbl *tbl;
+	size_t size;
+	uint32_t entries_num, i;
+
+	entries_num = max_flow_num * max_item_per_flow;
+	entries_num = RTE_MIN(entries_num, GRO_UDP6_TBL_MAX_ITEM_NUM);
+
+	if (entries_num == 0)
+		return NULL;
+
+	tbl = rte_zmalloc_socket(__func__,
+			sizeof(struct gro_udp6_tbl),
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl == NULL)
+		return NULL;
+
+	size = sizeof(struct gro_udp6_item) * entries_num;
+	tbl->items = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->items == NULL) {
+		rte_free(tbl);
+		return NULL;
+	}
+	tbl->max_item_num = entries_num;
+
+	size = sizeof(struct gro_udp6_flow) * entries_num;
+	tbl->flows = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->flows == NULL) {
+		rte_free(tbl->items);
+		rte_free(tbl);
+		return NULL;
+	}
+	/* INVALID_ARRAY_INDEX indicates an empty flow */
+	for (i = 0; i < entries_num; i++)
+		tbl->flows[i].start_index = INVALID_ARRAY_INDEX;
+	tbl->max_flow_num = entries_num;
+
+	return tbl;
+}
+
+void
+gro_udp6_tbl_destroy(void *tbl)
+{
+	struct gro_udp6_tbl *udp6_tbl = tbl;
+
+	if (udp6_tbl) {
+		rte_free(udp6_tbl->items);
+		rte_free(udp6_tbl->flows);
+	}
+	rte_free(udp6_tbl);
+}
+
+static inline uint32_t
+find_an_empty_item(struct gro_udp6_tbl *tbl)
+{
+	uint32_t i;
+	uint32_t max_item_num = tbl->max_item_num;
+
+	for (i = 0; i < max_item_num; i++)
+		if (tbl->items[i].firstseg == NULL)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+find_an_empty_flow(struct gro_udp6_tbl *tbl)
+{
+	uint32_t i;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++)
+		if (tbl->flows[i].start_index == INVALID_ARRAY_INDEX)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_item(struct gro_udp6_tbl *tbl,
+		struct rte_mbuf *pkt,
+		uint64_t start_time,
+		uint32_t prev_idx,
+		uint16_t frag_offset,
+		uint8_t is_last_frag)
+{
+	uint32_t item_idx;
+
+	item_idx = find_an_empty_item(tbl);
+	if (item_idx == INVALID_ARRAY_INDEX)
+		return INVALID_ARRAY_INDEX;
+
+	tbl->items[item_idx].firstseg = pkt;
+	tbl->items[item_idx].lastseg = rte_pktmbuf_lastseg(pkt);
+	tbl->items[item_idx].start_time = start_time;
+	tbl->items[item_idx].next_pkt_idx = INVALID_ARRAY_INDEX;
+	tbl->items[item_idx].frag_offset = frag_offset;
+	tbl->items[item_idx].is_last_frag = is_last_frag;
+	tbl->items[item_idx].nb_merged = 1;
+	tbl->item_num++;
+
+	/* if the previous packet exists, chain them together. */
+	if (prev_idx != INVALID_ARRAY_INDEX) {
+		tbl->items[item_idx].next_pkt_idx =
+			tbl->items[prev_idx].next_pkt_idx;
+		tbl->items[prev_idx].next_pkt_idx = item_idx;
+	}
+
+	return item_idx;
+}
+
+static inline uint32_t
+delete_item(struct gro_udp6_tbl *tbl, uint32_t item_idx,
+		uint32_t prev_item_idx)
+{
+	uint32_t next_idx = tbl->items[item_idx].next_pkt_idx;
+
+	/* NULL indicates an empty item */
+	tbl->items[item_idx].firstseg = NULL;
+	tbl->item_num--;
+	if (prev_item_idx != INVALID_ARRAY_INDEX)
+		tbl->items[prev_item_idx].next_pkt_idx = next_idx;
+
+	return next_idx;
+}
+
+/* Copy IPv6 addr */
+static inline void gro_ipv6_addr_copy(const uint8_t *ipv6_from,
+				      uint8_t *ipv6_to)
+{
+	const uint64_t *from_words = (const uint64_t *)ipv6_from;
+	uint64_t *to_words   = (uint64_t *)ipv6_to;
+
+	to_words[0] = from_words[0];
+	to_words[1] = from_words[1];
+}
+
+static inline uint32_t
+insert_new_flow(struct gro_udp6_tbl *tbl,
+		struct udp6_flow_key *src,
+		uint32_t item_idx)
+{
+	struct udp6_flow_key *dst;
+	uint32_t flow_idx;
+
+	flow_idx = find_an_empty_flow(tbl);
+	if (unlikely(flow_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	dst = &(tbl->flows[flow_idx].key);
+
+	rte_ether_addr_copy(&(src->eth_saddr), &(dst->eth_saddr));
+	rte_ether_addr_copy(&(src->eth_daddr), &(dst->eth_daddr));
+	gro_ipv6_addr_copy(src->ip_saddr, dst->ip_saddr);
+	gro_ipv6_addr_copy(src->ip_daddr, dst->ip_daddr);
+	dst->ip_id = src->ip_id;
+
+	tbl->flows[flow_idx].start_index = item_idx;
+	tbl->flow_num++;
+
+	return flow_idx;
+}
+
+static inline uint16_t
+get_ipv6_frag_offset(struct rte_ipv6_fragment_ext *ipv6_frag_hdr)
+{
+	return ((rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) >> 3) * 8);
+}
+
+static inline uint8_t
+is_last_ipv6_frag(struct rte_ipv6_fragment_ext *ipv6_frag_hdr)
+{
+	return (rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) & 0x0001);
+}
+
+/*
+ * update the packet length for the flushed packet.
+ */
+static inline void
+update_header(struct gro_udp6_item *item)
+{
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_mbuf *pkt = item->firstseg;
+	struct rte_ipv6_fragment_ext *ipv6_frag_hdr;
+	size_t fh_len = sizeof(*ipv6_frag_hdr);
+
+	ipv6_hdr = (struct rte_ipv6_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+			pkt->l2_len);
+	/* Note: payload_len includes extension headers and upper layers
+	 * data, but l3_len also includes extension headers, so need to
+	 * add length of extension headers if they exist.
+	 */
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(pkt->pkt_len -
+			pkt->l2_len - pkt->l3_len);
+
+	/* Remove fragment extension header or clear MF flag */
+	if (item->is_last_frag && (ipv6_hdr->proto == IPPROTO_FRAGMENT)) {
+		uint16_t ip_ofs;
+
+		ipv6_frag_hdr = (struct rte_ipv6_fragment_ext *)(ipv6_hdr + 1);
+		ip_ofs = get_ipv6_frag_offset(ipv6_frag_hdr);
+		if (ip_ofs == 0) {
+			ipv6_hdr->proto = ipv6_frag_hdr->next_header;
+			pkt->l3_len -= fh_len;
+
+			/* Remove IPv6 fragment extension header */
+			memmove(rte_pktmbuf_mtod_offset(pkt, char *, fh_len),
+				rte_pktmbuf_mtod(pkt, char*),
+				pkt->outer_l2_len + pkt->outer_l3_len +
+					pkt->l2_len + pkt->l3_len);
+			rte_pktmbuf_adj(pkt, fh_len);
+		} else {
+			/* clear MF flag */
+			ipv6_frag_hdr->frag_data = rte_cpu_to_be_16(
+				(rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) &
+					0xFFFE));
+			ipv6_hdr->payload_len += fh_len;
+		}
+	}
+}
+
+int32_t
+gro_udp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_udp6_tbl *tbl,
+		uint64_t start_time)
+{
+	struct rte_ether_hdr *eth_hdr;
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_ipv6_fragment_ext *ipv6_frag_hdr;
+	uint16_t fh_len = sizeof(*ipv6_frag_hdr);
+	uint16_t ip_dl;
+	uint32_t ip_id;
+	uint16_t hdr_len;
+	uint16_t frag_offset = 0;
+	uint8_t is_last_frag;
+
+	struct udp6_flow_key key;
+	uint32_t cur_idx, prev_idx, item_idx;
+	uint32_t i, max_flow_num, remaining_flow_num;
+	int cmp;
+	uint8_t find;
+
+	/*
+	 * Don't process the packet whose UDP header length is not equal
+	 * to 20.
+	 */
+	if (unlikely(pkt->l4_len != UDP_HDRLEN))
+		return -1;
+
+	eth_hdr = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *);
+	ipv6_hdr = (struct rte_ipv6_hdr *)((char *)eth_hdr + pkt->l2_len);
+
+	/* Note: l3_len includes length of extension headers */
+	hdr_len = pkt->l2_len + pkt->l3_len;
+
+	/*
+	 * Don't process non-fragment packet.
+	 */
+	if (ipv6_hdr->proto != IPPROTO_FRAGMENT)
+		return -1;
+
+	/*
+	 * Don't process the packet whose payload length is less than or
+	 * equal to 0.
+	 */
+	if (pkt->pkt_len <= hdr_len)
+		return -1;
+
+	ipv6_frag_hdr = (struct rte_ipv6_fragment_ext *)(ipv6_hdr + 1);
+	ip_dl = rte_be_to_cpu_16(ipv6_hdr->payload_len) - fh_len;
+	ip_id = rte_be_to_cpu_32(ipv6_frag_hdr->id);
+	frag_offset = get_ipv6_frag_offset(ipv6_frag_hdr);
+	is_last_frag = is_last_ipv6_frag(ipv6_frag_hdr);
+
+	rte_ether_addr_copy(&(eth_hdr->s_addr), &(key.eth_saddr));
+	rte_ether_addr_copy(&(eth_hdr->d_addr), &(key.eth_daddr));
+	gro_ipv6_addr_copy(ipv6_hdr->src_addr, key.ip_saddr);
+	gro_ipv6_addr_copy(ipv6_hdr->dst_addr, key.ip_daddr);
+	key.ip_id = ip_id;
+
+	/* Search for a matched flow. */
+	max_flow_num = tbl->max_flow_num;
+	remaining_flow_num = tbl->flow_num;
+	find = 0;
+	for (i = 0; i < max_flow_num && remaining_flow_num; i++) {
+		if (tbl->flows[i].start_index != INVALID_ARRAY_INDEX) {
+			if (is_same_udp6_flow(tbl->flows[i].key, key)) {
+				find = 1;
+				break;
+			}
+			remaining_flow_num--;
+		}
+	}
+
+	/*
+	 * Fail to find a matched flow. Insert a new flow and store the
+	 * packet into the flow.
+	 */
+	if (find == 0) {
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, frag_offset,
+				is_last_frag);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		if (insert_new_flow(tbl, &key, item_idx) ==
+				INVALID_ARRAY_INDEX) {
+			/*
+			 * Fail to insert a new flow, so delete the
+			 * stored packet.
+			 */
+			delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+			return -1;
+		}
+		return 0;
+	}
+
+	/*
+	 * Check all packets in the flow and try to find a neighbor for
+	 * the input packet.
+	 */
+	cur_idx = tbl->flows[i].start_index;
+	prev_idx = cur_idx;
+	do {
+		cmp = udp6_check_neighbor(&(tbl->items[cur_idx]),
+				frag_offset, ip_dl, 0);
+		if (cmp) {
+			if (merge_two_udp6_packets(&(tbl->items[cur_idx]),
+						pkt, cmp, frag_offset,
+						is_last_frag, 0))
+				return 1;
+			/*
+			 * Fail to merge the two packets, as the packet
+			 * length is greater than the max value. Store
+			 * the packet into the flow.
+			 */
+			if (insert_new_item(tbl, pkt, start_time, prev_idx,
+						frag_offset, is_last_frag) ==
+					INVALID_ARRAY_INDEX)
+				return -1;
+			return 0;
+		}
+
+		/* Ensure inserted items are ordered by frag_offset */
+		if (frag_offset
+			< tbl->items[cur_idx].frag_offset) {
+			break;
+		}
+
+		prev_idx = cur_idx;
+		cur_idx = tbl->items[cur_idx].next_pkt_idx;
+	} while (cur_idx != INVALID_ARRAY_INDEX);
+
+	/* Fail to find a neighbor, so store the packet into the flow. */
+	if (cur_idx == tbl->flows[i].start_index) {
+		/* Insert it before the first packet of the flow */
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, frag_offset,
+				is_last_frag);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		tbl->items[item_idx].next_pkt_idx = cur_idx;
+		tbl->flows[i].start_index = item_idx;
+	} else {
+		if (insert_new_item(tbl, pkt, start_time, prev_idx,
+			frag_offset, is_last_frag) == INVALID_ARRAY_INDEX)
+			return -1;
+	}
+
+	return 0;
+}
+
+static int
+gro_udp6_merge_items(struct gro_udp6_tbl *tbl,
+			   uint32_t start_idx)
+{
+	uint16_t frag_offset;
+	uint8_t is_last_frag;
+	int16_t ip_dl;
+	struct rte_mbuf *pkt;
+	int cmp;
+	uint32_t item_idx;
+	uint16_t hdr_len;
+
+	item_idx = tbl->items[start_idx].next_pkt_idx;
+	while (item_idx != INVALID_ARRAY_INDEX) {
+		pkt = tbl->items[item_idx].firstseg;
+		hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+			pkt->l3_len;
+		ip_dl = pkt->pkt_len - hdr_len;
+		frag_offset = tbl->items[item_idx].frag_offset;
+		is_last_frag = tbl->items[item_idx].is_last_frag;
+		cmp = udp6_check_neighbor(&(tbl->items[start_idx]),
+					frag_offset, ip_dl, 0);
+		if (cmp) {
+			if (merge_two_udp6_packets(
+					&(tbl->items[start_idx]),
+					pkt, cmp, frag_offset,
+					is_last_frag, 0)) {
+				item_idx = delete_item(tbl, item_idx,
+							INVALID_ARRAY_INDEX);
+				tbl->items[start_idx].next_pkt_idx
+					= item_idx;
+			} else {
+				return 0;
+			}
+		} else {
+			return 0;
+		}
+	}
+
+	return 0;
+}
+
+uint16_t
+gro_udp6_tbl_timeout_flush(struct gro_udp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out)
+{
+	uint16_t k = 0;
+	uint32_t i, j;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++) {
+		if (unlikely(tbl->flow_num == 0))
+			return k;
+
+		j = tbl->flows[i].start_index;
+		while (j != INVALID_ARRAY_INDEX) {
+			if (tbl->items[j].start_time <= flush_timestamp) {
+				gro_udp6_merge_items(tbl, j);
+				out[k++] = tbl->items[j].firstseg;
+				if (tbl->items[j].nb_merged > 1)
+					update_header(&(tbl->items[j]));
+				/*
+				 * Delete the packet and get the next
+				 * packet in the flow.
+				 */
+				j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+				tbl->flows[i].start_index = j;
+				if (j == INVALID_ARRAY_INDEX)
+					tbl->flow_num--;
+
+				if (unlikely(k == nb_out))
+					return k;
+			} else
+				/*
+				 * The left packets in this flow won't be
+				 * timeout. Go to check other flows.
+				 */
+				break;
+		}
+	}
+	return k;
+}
+
+uint32_t
+gro_udp6_tbl_pkt_count(void *tbl)
+{
+	struct gro_udp6_tbl *gro_tbl = tbl;
+
+	if (gro_tbl)
+		return gro_tbl->item_num;
+
+	return 0;
+}
diff --git a/lib/librte_gro/gro_udp6.h b/lib/librte_gro/gro_udp6.h
new file mode 100644
index 0000000..92696e4
--- /dev/null
+++ b/lib/librte_gro/gro_udp6.h
@@ -0,0 +1,284 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#ifndef _GRO_UDP6_H_
+#define _GRO_UDP6_H_
+
+#include <rte_ip.h>
+#include <rte_udp.h>
+#include <rte_vxlan.h>
+
+#define INVALID_ARRAY_INDEX 0xffffffffUL
+#define GRO_UDP6_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/*
+ * The max length of a IPv6 packet, which includes the length of the L3
+ * header, the L4 header and the data payload.
+ */
+#define MAX_IPV6_PKT_LENGTH UINT16_MAX
+
+#define UDP_HDRLEN 8
+
+/* Header fields representing a UDP/IPv6 flow */
+struct udp6_flow_key {
+	struct rte_ether_addr eth_saddr;
+	struct rte_ether_addr eth_daddr;
+	uint8_t ip_saddr[16];
+	uint8_t ip_daddr[16];
+
+	/* IP fragment for UDP does not contain UDP header
+	 * except the first one. But IP ID must be same.
+	 * ip_id of fragment is 32 bit for IPv6
+	 */
+	uint32_t ip_id;
+};
+
+struct gro_udp6_flow {
+	struct udp6_flow_key key;
+	/*
+	 * The index of the first packet in the flow.
+	 * INVALID_ARRAY_INDEX indicates an empty flow.
+	 */
+	uint32_t start_index;
+};
+
+struct gro_udp6_item {
+	/*
+	 * The first MBUF segment of the packet. If the value
+	 * is NULL, it means the item is empty.
+	 */
+	struct rte_mbuf *firstseg;
+	/* The last MBUF segment of the packet */
+	struct rte_mbuf *lastseg;
+	/*
+	 * The time when the first packet is inserted into the table.
+	 * This value won't be updated, even if the packet is merged
+	 * with other packets.
+	 */
+	uint64_t start_time;
+	/*
+	 * next_pkt_idx is used to chain the packets that
+	 * are in the same flow but can't be merged together
+	 * (e.g. caused by packet reordering).
+	 */
+	uint32_t next_pkt_idx;
+	/* offset of IP fragment packet */
+	uint16_t frag_offset;
+	/* is last IP fragment? */
+	uint8_t is_last_frag;
+	/* the number of merged packets */
+	uint16_t nb_merged;
+};
+
+/*
+ * UDP/IPv6 reassembly table structure.
+ */
+struct gro_udp6_tbl {
+	/* item array */
+	struct gro_udp6_item *items;
+	/* flow array */
+	struct gro_udp6_flow *flows;
+	/* current item number */
+	uint32_t item_num;
+	/* current flow num */
+	uint32_t flow_num;
+	/* item array size */
+	uint32_t max_item_num;
+	/* flow array size */
+	uint32_t max_flow_num;
+};
+
+/**
+ * This function creates a UDP/IPv6 reassembly table.
+ *
+ * @param socket_id
+ *  Socket index for allocating the UDP/IPv6 reassemble table
+ * @param max_flow_num
+ *  The maximum number of flows in the UDP/IPv6 GRO table
+ * @param max_item_per_flow
+ *  The maximum number of packets per flow
+ *
+ * @return
+ *  - Return the table pointer on success.
+ *  - Return NULL on failure.
+ */
+void *gro_udp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow);
+
+/**
+ * This function destroys a UDP/IPv6 reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the UDP/IPv6 reassembly table.
+ */
+void gro_udp6_tbl_destroy(void *tbl);
+
+/**
+ * This function merges a UDP/IPv6 packet.
+ *
+ * This function does not check if the packet has correct checksums and
+ * does not re-calculate checksums for the merged packet. It returns the
+ * packet if it isn't UDP fragment or there is no available space in
+ * the table.
+ *
+ * @param pkt
+ *  Packet to reassemble
+ * @param tbl
+ *  Pointer pointing to the UDP/IPv6 reassembly table
+ * @start_time
+ *  The time when the packet is inserted into the table
+ *
+ * @return
+ *  - Return a positive value if the packet is merged.
+ *  - Return zero if the packet isn't merged but stored in the table.
+ *  - Return a negative value for invalid parameters or no available
+ *    space in the table.
+ */
+int32_t gro_udp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_udp6_tbl *tbl,
+		uint64_t start_time);
+
+/**
+ * This function flushes timeout packets in a UDP/IPv6 reassembly table,
+ * and without updating checksums.
+ *
+ * @param tbl
+ *  UDP/IPv6 reassembly table pointer
+ * @param flush_timestamp
+ *  Flush packets which are inserted into the table before or at the
+ *  flush_timestamp.
+ * @param out
+ *  Pointer array used to keep flushed packets
+ * @param nb_out
+ *  The element number in 'out'. It also determines the maximum number of
+ *  packets that can be flushed finally.
+ *
+ * @return
+ *  The number of flushed packets
+ */
+uint16_t gro_udp6_tbl_timeout_flush(struct gro_udp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out);
+
+/**
+ * This function returns the number of the packets in a UDP/IPv6
+ * reassembly table.
+ *
+ * @param tbl
+ *  UDP/IPv6 reassembly table pointer
+ *
+ * @return
+ *  The number of packets in the table
+ */
+uint32_t gro_udp6_tbl_pkt_count(void *tbl);
+
+#ifndef _GRO_TCP6_H_
+static inline int rte_is_same_ipv6_addr(const uint8_t *ipv6_a1,
+					const uint8_t *ipv6_a2)
+{
+	const uint64_t *w1 = (const uint64_t *)ipv6_a1;
+	const uint64_t *w2 = (const uint64_t *)ipv6_a2;
+
+	return ((w1[0] ^ w2[0]) | (w1[1] ^ w2[1])) == 0;
+}
+#endif
+
+/*
+ * Check if two UDP/IPv6 packets belong to the same flow.
+ */
+static inline int
+is_same_udp6_flow(struct udp6_flow_key k1, struct udp6_flow_key k2)
+{
+	return (rte_is_same_ether_addr(&k1.eth_saddr, &k2.eth_saddr) &&
+			rte_is_same_ether_addr(&k1.eth_daddr, &k2.eth_daddr) &&
+			rte_is_same_ipv6_addr(k1.ip_saddr, k2.ip_saddr) &&
+			rte_is_same_ipv6_addr(k1.ip_daddr, k2.ip_daddr) &&
+			(k1.ip_id == k2.ip_id));
+}
+
+/*
+ * Merge two UDP/IPv6 packets without updating checksums.
+ * If cmp is larger than 0, append the new packet to the
+ * original packet. Otherwise, pre-pend the new packet to
+ * the original packet.
+ */
+static inline int
+merge_two_udp6_packets(struct gro_udp6_item *item,
+		struct rte_mbuf *pkt,
+		int cmp,
+		uint16_t frag_offset,
+		uint8_t is_last_frag,
+		uint16_t l2_offset)
+{
+	struct rte_mbuf *pkt_head, *pkt_tail, *lastseg;
+	uint16_t hdr_len, l2_len;
+	uint32_t ip_len;
+
+	if (cmp > 0) {
+		pkt_head = item->firstseg;
+		pkt_tail = pkt;
+	} else {
+		pkt_head = pkt;
+		pkt_tail = item->firstseg;
+	}
+
+	/* check if the IPv6 packet length is greater than the max value */
+	hdr_len = l2_offset + pkt_head->l2_len + pkt_head->l3_len;
+	l2_len = l2_offset > 0 ? pkt_head->outer_l2_len : pkt_head->l2_len;
+	ip_len = pkt_head->pkt_len - l2_len
+		 + pkt_tail->pkt_len - hdr_len;
+	if (unlikely(ip_len > MAX_IPV6_PKT_LENGTH))
+		return 0;
+
+	/* remove the packet header for the tail packet */
+	rte_pktmbuf_adj(pkt_tail, hdr_len);
+
+	/* chain two packets together */
+	if (cmp > 0) {
+		item->lastseg->next = pkt;
+		item->lastseg = rte_pktmbuf_lastseg(pkt);
+	} else {
+		lastseg = rte_pktmbuf_lastseg(pkt);
+		lastseg->next = item->firstseg;
+		item->firstseg = pkt;
+		item->frag_offset = frag_offset;
+	}
+	item->nb_merged++;
+	if (is_last_frag)
+		item->is_last_frag = is_last_frag;
+
+	/* update MBUF metadata for the merged packet */
+	pkt_head->nb_segs += pkt_tail->nb_segs;
+	pkt_head->pkt_len += pkt_tail->pkt_len;
+
+	return 1;
+}
+
+/*
+ * Check if two UDP/IPv6 packets are neighbors.
+ */
+static inline int
+udp6_check_neighbor(struct gro_udp6_item *item,
+		uint16_t frag_offset,
+		uint16_t ip_dl,
+		uint16_t l2_offset)
+{
+	struct rte_mbuf *pkt_orig = item->firstseg;
+	uint16_t len;
+
+	/* check if the two packets are neighbors */
+	len = pkt_orig->pkt_len - l2_offset - pkt_orig->l2_len -
+		pkt_orig->l3_len;
+	if (frag_offset == item->frag_offset + len)
+		/* append the new packet */
+		return 1;
+	else if (frag_offset + ip_dl == item->frag_offset)
+		/* pre-pend the new packet */
+		return -1;
+
+	return 0;
+}
+#endif
diff --git a/lib/librte_gro/meson.build b/lib/librte_gro/meson.build
index 3eff89c..7939081 100644
--- a/lib/librte_gro/meson.build
+++ b/lib/librte_gro/meson.build
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c')
+sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c', 'gro_udp6.c')
 headers = files('rte_gro.h')
 deps += ['ethdev']
diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
index 0c5584e..13b38cc 100644
--- a/lib/librte_gro/rte_gro.c
+++ b/lib/librte_gro/rte_gro.c
@@ -11,6 +11,7 @@
 #include "gro_tcp4.h"
 #include "gro_tcp6.h"
 #include "gro_udp4.h"
+#include "gro_udp6.h"
 #include "gro_vxlan6_tcp4.h"
 #include "gro_vxlan6_tcp6.h"
 #include "gro_vxlan_tcp4.h"
@@ -28,21 +29,21 @@
 		gro_udp4_tbl_create, gro_vxlan_udp4_tbl_create,
 		gro_tcp6_tbl_create, gro_vxlan_tcp6_tbl_create,
 		gro_vxlan6_tcp4_tbl_create, gro_vxlan6_tcp6_tbl_create,
-		NULL};
+		gro_udp6_tbl_create, NULL};
 static gro_tbl_destroy_fn tbl_destroy_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_destroy, gro_vxlan_tcp4_tbl_destroy,
 			gro_udp4_tbl_destroy, gro_vxlan_udp4_tbl_destroy,
 			gro_tcp6_tbl_destroy, gro_vxlan_tcp6_tbl_destroy,
 			gro_vxlan6_tcp4_tbl_destroy,
 			gro_vxlan6_tcp6_tbl_destroy,
-			NULL};
+			gro_udp6_tbl_destroy, NULL};
 static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_pkt_count, gro_vxlan_tcp4_tbl_pkt_count,
 			gro_udp4_tbl_pkt_count, gro_vxlan_udp4_tbl_pkt_count,
 			gro_tcp6_tbl_pkt_count, gro_vxlan_tcp6_tbl_pkt_count,
 			gro_vxlan6_tcp4_tbl_pkt_count,
 			gro_vxlan6_tcp6_tbl_pkt_count,
-			NULL};
+			gro_udp6_tbl_pkt_count, NULL};
 
 #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
@@ -121,6 +122,10 @@
 		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
 		  RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN)))
 
+#define IS_IPV6_UDP_PKT(ptype) (RTE_ETH_IS_IPV6_HDR(ptype) && \
+		((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
+		(RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
+
 /*
  * GRO context structure. It keeps the table structures, which are
  * used to merge packets, for different GRO types. Before using
@@ -246,13 +251,19 @@ struct gro_ctx {
 	struct gro_vxlan6_tcp6_item
 		vxlan6_tcp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{{0}} };
 
+	/* allocate a reassembly table for UDP/IPv6 GRO */
+	struct gro_udp6_tbl udp6_tbl;
+	struct gro_udp6_flow udp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+	struct gro_udp6_item udp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
+
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	uint32_t item_num;
 	int32_t ret;
 	uint16_t i, unprocess_num = 0, nb_after_gro = nb_pkts;
 	uint8_t do_tcp4_gro = 0, do_vxlan_tcp_gro = 0, do_udp4_gro = 0,
 		do_vxlan_udp_gro = 0, do_tcp6_gro = 0, do_vxlan_tcp6_gro = 0,
-		do_vxlan6_tcp4_gro = 0, do_vxlan6_tcp6_gro = 0;
+		do_vxlan6_tcp4_gro = 0, do_vxlan6_tcp6_gro = 0,
+		do_udp6_gro = 0;
 
 	if (unlikely((param->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -261,7 +272,8 @@ struct gro_ctx {
 					RTE_GRO_TCP_IPV6 |
 					RTE_GRO_IPV4_VXLAN_TCP_IPV6 |
 					RTE_GRO_IPV6_VXLAN_TCP_IPV4 |
-					RTE_GRO_IPV6_VXLAN_TCP_IPV6)) == 0))
+					RTE_GRO_IPV6_VXLAN_TCP_IPV6 |
+					RTE_GRO_UDP_IPV6)) == 0))
 		return nb_pkts;
 
 	/* Get the maximum number of packets */
@@ -373,6 +385,19 @@ struct gro_ctx {
 		do_vxlan6_tcp6_gro = 1;
 	}
 
+	if (param->gro_types & RTE_GRO_UDP_IPV6) {
+		for (i = 0; i < item_num; i++)
+			udp6_flows[i].start_index = INVALID_ARRAY_INDEX;
+
+		udp6_tbl.flows = udp6_flows;
+		udp6_tbl.items = udp6_items;
+		udp6_tbl.flow_num = 0;
+		udp6_tbl.item_num = 0;
+		udp6_tbl.max_flow_num = item_num;
+		udp6_tbl.max_item_num = item_num;
+		do_udp6_gro = 1;
+	}
+
 	for (i = 0; i < nb_pkts; i++) {
 		/*
 		 * The timestamp is ignored, since all packets
@@ -447,6 +472,14 @@ struct gro_ctx {
 				nb_after_gro--;
 			else if (ret < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_UDP_PKT(pkts[i]->packet_type) &&
+				do_udp6_gro) {
+			ret = gro_udp6_reassemble(pkts[i], &udp6_tbl, 0);
+			if (ret > 0)
+				/* merge successfully */
+				nb_after_gro--;
+			else if (ret < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -496,6 +529,11 @@ struct gro_ctx {
 					0, &pkts[i], nb_pkts - i);
 		}
 
+		if (do_udp6_gro) {
+			i += gro_udp6_tbl_timeout_flush(&udp6_tbl, 0,
+					&pkts[i], nb_pkts - i);
+		}
+
 		/* Copy unprocessed packets */
 		if (unprocess_num > 0) {
 			memcpy(&pkts[i], unprocess_pkts,
@@ -516,12 +554,13 @@ struct gro_ctx {
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	struct gro_ctx *gro_ctx = ctx;
 	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl,
-		*vxlan_tcp6_tbl, *vxlan6_tcp4_tbl, *vxlan6_tcp6_tbl;
+		*vxlan_tcp6_tbl, *vxlan6_tcp4_tbl, *vxlan6_tcp6_tbl,
+		*udp6_tbl;
 	uint64_t current_time;
 	uint16_t i, unprocess_num = 0;
 	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro,
 		do_tcp6_gro, do_vxlan_tcp6_gro, do_vxlan6_tcp4_gro,
-		do_vxlan6_tcp6_gro;
+		do_vxlan6_tcp6_gro, do_udp6_gro;
 
 	if (unlikely((gro_ctx->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -530,7 +569,8 @@ struct gro_ctx {
 					RTE_GRO_TCP_IPV6 |
 					RTE_GRO_IPV4_VXLAN_TCP_IPV6 |
 					RTE_GRO_IPV6_VXLAN_TCP_IPV4 |
-					RTE_GRO_IPV6_VXLAN_TCP_IPV6)) == 0))
+					RTE_GRO_IPV6_VXLAN_TCP_IPV6 |
+					RTE_GRO_UDP_IPV6)) == 0))
 		return nb_pkts;
 
 	tcp_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV4_INDEX];
@@ -541,6 +581,7 @@ struct gro_ctx {
 	vxlan_tcp6_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_TCP_IPV6_INDEX];
 	vxlan6_tcp4_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX];
 	vxlan6_tcp6_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX];
+	udp6_tbl = gro_ctx->tbls[RTE_GRO_UDP_IPV6_INDEX];
 
 	do_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV4) ==
 		RTE_GRO_TCP_IPV4;
@@ -558,6 +599,8 @@ struct gro_ctx {
 				== RTE_GRO_IPV6_VXLAN_TCP_IPV4;
 	do_vxlan6_tcp6_gro = (gro_ctx->gro_types & RTE_GRO_IPV6_VXLAN_TCP_IPV6)
 				== RTE_GRO_IPV6_VXLAN_TCP_IPV6;
+	do_udp6_gro = (gro_ctx->gro_types & RTE_GRO_UDP_IPV6) ==
+		RTE_GRO_UDP_IPV6;
 
 	current_time = rte_rdtsc();
 
@@ -602,6 +645,11 @@ struct gro_ctx {
 			if (gro_vxlan6_tcp6_reassemble(pkts[i], vxlan6_tcp6_tbl,
 						current_time) < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_UDP_PKT(pkts[i]->packet_type) &&
+				do_udp6_gro) {
+			if (gro_udp6_reassemble(pkts[i], udp6_tbl,
+						current_time) < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -685,6 +733,14 @@ struct gro_ctx {
 		num += gro_vxlan6_tcp6_tbl_timeout_flush(gro_ctx->tbls[
 				RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX],
 				flush_timestamp, &out[num], left_nb_out);
+		left_nb_out = max_nb_out - num;
+	}
+
+	if ((gro_types & RTE_GRO_UDP_IPV6) && left_nb_out > 0) {
+		num += gro_udp6_tbl_timeout_flush(
+				gro_ctx->tbls[RTE_GRO_UDP_IPV6_INDEX],
+				flush_timestamp,
+				&out[num], left_nb_out);
 	}
 
 	return num;
diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
index cb0895f..94ed3d3 100644
--- a/lib/librte_gro/rte_gro.h
+++ b/lib/librte_gro/rte_gro.h
@@ -50,6 +50,9 @@
 #define RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX 7
 #define RTE_GRO_IPV6_VXLAN_TCP_IPV6 (1ULL << RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX)
 /**< IPv6 VxLAN TCP/IPv6 GRO flag. */
+#define RTE_GRO_UDP_IPV6_INDEX 8
+#define RTE_GRO_UDP_IPV6 (1ULL << RTE_GRO_UDP_IPV6_INDEX)
+/**< UDP/IPv6 GRO flag */
 
 /**
  * Structure used to create GRO context objects or used to pass
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-dev] [PATCH v1 6/8] gro: support IPv4 VXLAN UDP/IPv6
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
                   ` (4 preceding siblings ...)
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 5/8] gro: support UDP/IPv6 yang_y_yi
@ 2020-12-21  3:50 ` yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 7/8] gro: support IPv6 VXLAN UDP/IPv4 yang_y_yi
                   ` (2 subsequent siblings)
  8 siblings, 0 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

IPv4 VXLAN UDP/IPv6 GRO can help improve UDP/IPv6
performance in IPv4 VXLAN use case.

With this enabled in DPDK, OVS DPDK can leverage it
to improve VM-to-VM UDP/IPv6 performance, it will merge
small adjacent IPv4 VXLAN UDP/IPv6 fragments to a big
IPv4 VXLAN UDP/IPv6 packet immediate after they are
received from a physical NIC. It is very helpful in OVS
DPDK IPv4 VXLAN use case.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
---
 .../prog_guide/generic_receive_offload_lib.rst     |   4 +-
 doc/guides/rel_notes/release_21_02.rst             |   5 +
 lib/librte_gro/gro_vxlan_udp6.c                    | 607 +++++++++++++++++++++
 lib/librte_gro/gro_vxlan_udp6.h                    | 152 ++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           |  83 ++-
 lib/librte_gro/rte_gro.h                           |   3 +
 7 files changed, 845 insertions(+), 11 deletions(-)
 create mode 100644 lib/librte_gro/gro_vxlan_udp6.c
 create mode 100644 lib/librte_gro/gro_vxlan_udp6.h

diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst b/doc/guides/prog_guide/generic_receive_offload_lib.rst
index 0ea3076..906e85f 100644
--- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
+++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
@@ -32,8 +32,8 @@ fragmentation is possible (i.e., DF==0). Additionally, it complies RFC
 
 Currently, the GRO library provides GRO supports for TCP/IPv4, UDP/IPv4,
 UDP/IPv6, and TCP/IPv6 packets as well as VxLAN packets which contain an
-outer IPv4 header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6 packet or
-an outer IPv6 header and an inner TCP/IPv4 or TCP/IPv6 packet.
+outer IPv4 header and an inner TCP/IPv4, UDP/IPv4, TCP/IPv6 or UDP/IPv6
+packet or an outer IPv6 header and an inner TCP/IPv4 or TCP/IPv6 packet.
 
 Two Sets of API
 ---------------
diff --git a/doc/guides/rel_notes/release_21_02.rst b/doc/guides/rel_notes/release_21_02.rst
index d8974b7..4758cea 100644
--- a/doc/guides/rel_notes/release_21_02.rst
+++ b/doc/guides/rel_notes/release_21_02.rst
@@ -80,6 +80,11 @@ New Features
    Added UDP/IPv6 GRO support for non-VXLAN packets, this enabled UDP/IPv6
    GRO support for IPv6 VLAN use case.
 
+* **Added UDP/IPv6 GRO support for IPv4 VXLAN packets.**
+
+   Added UDP/IPv6 GRO support for IPv4 VXLAN packets, this enabled UDP/IPv6
+   GRO support for IPv4 VXLAN use case.
+
 Removed Items
 -------------
 
diff --git a/lib/librte_gro/gro_vxlan_udp6.c b/lib/librte_gro/gro_vxlan_udp6.c
new file mode 100644
index 0000000..8c76b41
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan_udp6.c
@@ -0,0 +1,607 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+#include <rte_udp.h>
+#include <rte_ip.h>
+
+#include "gro_vxlan_udp6.h"
+
+void *
+gro_vxlan_udp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow)
+{
+	struct gro_vxlan_udp6_tbl *tbl;
+	size_t size;
+	uint32_t entries_num, i;
+
+	entries_num = max_flow_num * max_item_per_flow;
+	entries_num = RTE_MIN(entries_num, GRO_VXLAN_UDP6_TBL_MAX_ITEM_NUM);
+
+	if (entries_num == 0)
+		return NULL;
+
+	tbl = rte_zmalloc_socket(__func__,
+			sizeof(struct gro_vxlan_udp6_tbl),
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl == NULL)
+		return NULL;
+
+	size = sizeof(struct gro_vxlan_udp6_item) * entries_num;
+	tbl->items = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->items == NULL) {
+		rte_free(tbl);
+		return NULL;
+	}
+	tbl->max_item_num = entries_num;
+
+	size = sizeof(struct gro_vxlan_udp6_flow) * entries_num;
+	tbl->flows = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->flows == NULL) {
+		rte_free(tbl->items);
+		rte_free(tbl);
+		return NULL;
+	}
+
+	for (i = 0; i < entries_num; i++)
+		tbl->flows[i].start_index = INVALID_ARRAY_INDEX;
+	tbl->max_flow_num = entries_num;
+
+	return tbl;
+}
+
+void
+gro_vxlan_udp6_tbl_destroy(void *tbl)
+{
+	struct gro_vxlan_udp6_tbl *vxlan_tbl = tbl;
+
+	if (vxlan_tbl) {
+		rte_free(vxlan_tbl->items);
+		rte_free(vxlan_tbl->flows);
+	}
+	rte_free(vxlan_tbl);
+}
+
+static inline uint32_t
+find_an_empty_item(struct gro_vxlan_udp6_tbl *tbl)
+{
+	uint32_t max_item_num = tbl->max_item_num, i;
+
+	for (i = 0; i < max_item_num; i++)
+		if (tbl->items[i].inner_item.firstseg == NULL)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+find_an_empty_flow(struct gro_vxlan_udp6_tbl *tbl)
+{
+	uint32_t max_flow_num = tbl->max_flow_num, i;
+
+	for (i = 0; i < max_flow_num; i++)
+		if (tbl->flows[i].start_index == INVALID_ARRAY_INDEX)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_item(struct gro_vxlan_udp6_tbl *tbl,
+		struct rte_mbuf *pkt,
+		uint64_t start_time,
+		uint32_t prev_idx,
+		uint16_t frag_offset,
+		uint8_t is_last_frag,
+		uint16_t outer_ip_id,
+		uint8_t outer_is_atomic)
+{
+	uint32_t item_idx;
+
+	item_idx = find_an_empty_item(tbl);
+	if (unlikely(item_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	tbl->items[item_idx].inner_item.firstseg = pkt;
+	tbl->items[item_idx].inner_item.lastseg = rte_pktmbuf_lastseg(pkt);
+	tbl->items[item_idx].inner_item.start_time = start_time;
+	tbl->items[item_idx].inner_item.next_pkt_idx = INVALID_ARRAY_INDEX;
+	tbl->items[item_idx].inner_item.frag_offset = frag_offset;
+	tbl->items[item_idx].inner_item.is_last_frag = is_last_frag;
+	tbl->items[item_idx].inner_item.nb_merged = 1;
+	tbl->items[item_idx].outer_ip_id = outer_ip_id;
+	tbl->items[item_idx].outer_is_atomic = outer_is_atomic;
+	tbl->item_num++;
+
+	/* If the previous packet exists, chain the new one with it. */
+	if (prev_idx != INVALID_ARRAY_INDEX) {
+		tbl->items[item_idx].inner_item.next_pkt_idx =
+			tbl->items[prev_idx].inner_item.next_pkt_idx;
+		tbl->items[prev_idx].inner_item.next_pkt_idx = item_idx;
+	}
+
+	return item_idx;
+}
+
+static inline uint32_t
+delete_item(struct gro_vxlan_udp6_tbl *tbl,
+		uint32_t item_idx,
+		uint32_t prev_item_idx)
+{
+	uint32_t next_idx = tbl->items[item_idx].inner_item.next_pkt_idx;
+
+	/* NULL indicates an empty item. */
+	tbl->items[item_idx].inner_item.firstseg = NULL;
+	tbl->item_num--;
+	if (prev_item_idx != INVALID_ARRAY_INDEX)
+		tbl->items[prev_item_idx].inner_item.next_pkt_idx = next_idx;
+
+	return next_idx;
+}
+
+/* Copy IPv6 addr */
+static inline void gro_ipv6_addr_copy(const uint8_t *ipv6_from,
+				      uint8_t *ipv6_to)
+{
+	const uint64_t *from_words = (const uint64_t *)ipv6_from;
+	uint64_t *to_words   = (uint64_t *)ipv6_to;
+
+	to_words[0] = from_words[0];
+	to_words[1] = from_words[1];
+}
+
+
+static inline uint32_t
+insert_new_flow(struct gro_vxlan_udp6_tbl *tbl,
+		struct vxlan_udp6_flow_key *src,
+		uint32_t item_idx)
+{
+	struct vxlan_udp6_flow_key *dst;
+	uint32_t flow_idx;
+
+	flow_idx = find_an_empty_flow(tbl);
+	if (unlikely(flow_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	dst = &(tbl->flows[flow_idx].key);
+
+	rte_ether_addr_copy(&(src->inner_key.eth_saddr),
+			&(dst->inner_key.eth_saddr));
+	rte_ether_addr_copy(&(src->inner_key.eth_daddr),
+			&(dst->inner_key.eth_daddr));
+	gro_ipv6_addr_copy(src->inner_key.ip_saddr, dst->inner_key.ip_saddr);
+	gro_ipv6_addr_copy(src->inner_key.ip_daddr, dst->inner_key.ip_daddr);
+	dst->inner_key.ip_id = src->inner_key.ip_id;
+
+	dst->vxlan_hdr.vx_flags = src->vxlan_hdr.vx_flags;
+	dst->vxlan_hdr.vx_vni = src->vxlan_hdr.vx_vni;
+	rte_ether_addr_copy(&(src->outer_eth_saddr), &(dst->outer_eth_saddr));
+	rte_ether_addr_copy(&(src->outer_eth_daddr), &(dst->outer_eth_daddr));
+	dst->outer_ip_src_addr = src->outer_ip_src_addr;
+	dst->outer_ip_dst_addr = src->outer_ip_dst_addr;
+	dst->outer_src_port = src->outer_src_port;
+	dst->outer_dst_port = src->outer_dst_port;
+
+	tbl->flows[flow_idx].start_index = item_idx;
+	tbl->flow_num++;
+
+	return flow_idx;
+}
+
+static inline int
+is_same_vxlan_udp6_flow(struct vxlan_udp6_flow_key k1,
+		struct vxlan_udp6_flow_key k2)
+{
+	/* For VxLAN packet, outer udp src port is calculated from
+	 * inner packet RSS hash, udp src port of the first UDP
+	 * fragment is different from one of other UDP fragments
+	 * even if they are same flow, so we have to skip outer udp
+	 * src port comparison here.
+	 */
+	return (rte_is_same_ether_addr(&k1.outer_eth_saddr,
+					&k2.outer_eth_saddr) &&
+			rte_is_same_ether_addr(&k1.outer_eth_daddr,
+				&k2.outer_eth_daddr) &&
+			(k1.outer_ip_src_addr == k2.outer_ip_src_addr) &&
+			(k1.outer_ip_dst_addr == k2.outer_ip_dst_addr) &&
+			(k1.outer_dst_port == k2.outer_dst_port) &&
+			(k1.vxlan_hdr.vx_flags == k2.vxlan_hdr.vx_flags) &&
+			(k1.vxlan_hdr.vx_vni == k2.vxlan_hdr.vx_vni) &&
+			is_same_udp6_flow(k1.inner_key, k2.inner_key));
+}
+
+static inline int
+udp6_check_vxlan_neighbor(struct gro_vxlan_udp6_item *item,
+		uint16_t frag_offset,
+		uint16_t ip_dl)
+{
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	int cmp;
+	uint16_t l2_offset;
+	int ret = 0;
+
+	l2_offset = pkt->outer_l2_len + pkt->outer_l3_len;
+	cmp = udp6_check_neighbor(&item->inner_item, frag_offset,
+			ip_dl, l2_offset);
+	if (cmp > 0)
+		/* Append the new packet. */
+		ret = 1;
+	else if (cmp < 0)
+		/* Prepend the new packet. */
+		ret = -1;
+
+	return ret;
+}
+
+static inline int
+merge_two_vxlan_udp6_packets(struct gro_vxlan_udp6_item *item,
+		struct rte_mbuf *pkt,
+		int cmp,
+		uint16_t frag_offset,
+		uint8_t is_last_frag)
+{
+	if (merge_two_udp6_packets(&item->inner_item, pkt, cmp, frag_offset,
+				is_last_frag,
+				pkt->outer_l2_len + pkt->outer_l3_len)) {
+		return 1;
+	}
+
+	return 0;
+}
+
+static inline uint16_t
+get_ipv6_frag_offset(struct rte_ipv6_fragment_ext *ipv6_frag_hdr)
+{
+	return ((rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) >> 3) * 8);
+}
+
+static inline uint8_t
+is_last_ipv6_frag(struct rte_ipv6_fragment_ext *ipv6_frag_hdr)
+{
+	return (rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) & 0x0001);
+}
+
+static inline void
+update_vxlan_header(struct gro_vxlan_udp6_item *item)
+{
+	struct rte_ipv4_hdr *outer_ipv4_hdr;
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_ipv6_fragment_ext *ipv6_frag_hdr;
+	size_t fh_len = sizeof(*ipv6_frag_hdr);
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	uint16_t len;
+
+	/* Position to inner IPv6 header first */
+	ipv6_hdr = (struct rte_ipv6_hdr *)rte_pktmbuf_mtod_offset(pkt, char *,
+			pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len);
+	len = pkt->pkt_len - pkt->outer_l2_len - pkt->outer_l3_len
+		- pkt->l2_len;
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(len - pkt->l3_len);
+
+	/* Remove fragment extension header or clear MF flag */
+	if (item->inner_item.is_last_frag
+		&& (ipv6_hdr->proto == IPPROTO_FRAGMENT)) {
+		uint16_t ip_ofs;
+
+		ipv6_frag_hdr = (struct rte_ipv6_fragment_ext *)(ipv6_hdr + 1);
+		ip_ofs = get_ipv6_frag_offset(ipv6_frag_hdr);
+		if (ip_ofs == 0) {
+			ipv6_hdr->proto = ipv6_frag_hdr->next_header;
+			pkt->l3_len -= fh_len;
+
+			/* Remove IPv6 fragment extension header */
+			memmove(rte_pktmbuf_mtod_offset(pkt, char *, fh_len),
+				rte_pktmbuf_mtod(pkt, char*),
+				pkt->outer_l2_len + pkt->outer_l3_len +
+					pkt->l2_len + pkt->l3_len);
+			rte_pktmbuf_adj(pkt, fh_len);
+		} else {
+			/* clear MF flag */
+			ipv6_frag_hdr->frag_data = rte_cpu_to_be_16(
+				(rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) &
+					0xFFFE));
+			ipv6_hdr->payload_len += fh_len;
+		}
+	}
+
+	/* Must adjust outer header after inner IPv6 is handled. */
+	/* Update the outer IPv4 header. */
+	len = pkt->pkt_len - pkt->outer_l2_len;
+	outer_ipv4_hdr = (struct rte_ipv4_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+			pkt->outer_l2_len);
+	outer_ipv4_hdr->total_length = rte_cpu_to_be_16(len);
+
+	/* Point to the outer UDP header. */
+	len -= pkt->outer_l3_len;
+	udp_hdr = (struct rte_udp_hdr *)((char *)outer_ipv4_hdr
+						+ pkt->outer_l3_len);
+	udp_hdr->dgram_len = rte_cpu_to_be_16(len);
+}
+
+int32_t
+gro_vxlan_udp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan_udp6_tbl *tbl,
+		uint64_t start_time)
+{
+	struct rte_ether_hdr *outer_eth_hdr, *eth_hdr;
+	struct rte_ipv4_hdr *outer_ipv4_hdr;
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_ipv6_fragment_ext *ipv6_frag_hdr;
+	uint16_t fh_len = sizeof(*ipv6_frag_hdr);
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_vxlan_hdr *vxlan_hdr;
+	uint16_t frag_offset;
+	uint8_t is_last_frag;
+	int16_t ip_dl;
+	uint32_t ip_id;
+	uint16_t outer_ip_id;
+	uint8_t outer_is_atomic;
+
+	struct vxlan_udp6_flow_key key;
+	uint32_t cur_idx, prev_idx, item_idx;
+	uint32_t i, max_flow_num, remaining_flow_num;
+	int cmp;
+	uint16_t hdr_len;
+	uint8_t find;
+
+	outer_eth_hdr = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *);
+	outer_ipv4_hdr = (struct rte_ipv4_hdr *)((char *)outer_eth_hdr +
+			pkt->outer_l2_len);
+
+	udp_hdr = (struct rte_udp_hdr *)((char *)outer_ipv4_hdr +
+			pkt->outer_l3_len);
+	vxlan_hdr = (struct rte_vxlan_hdr *)((char *)udp_hdr +
+			sizeof(struct rte_udp_hdr));
+	eth_hdr = (struct rte_ether_hdr *)((char *)vxlan_hdr +
+			sizeof(struct rte_vxlan_hdr));
+	/* l2_len = outer udp hdr len + vxlan hdr len + inner l2 len */
+	ipv6_hdr = (struct rte_ipv6_hdr *)((char *)udp_hdr + pkt->l2_len);
+
+	/*
+	 * Don't process the packet which has non-fragment inner IP.
+	 */
+	if (ipv6_hdr->proto != IPPROTO_FRAGMENT)
+		return -1;
+
+	/* Note: l3_len includes length of extension headers */
+	hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+		pkt->l3_len;
+	/*
+	 * Don't process the packet whose payload length is less than or
+	 * equal to 0.
+	 */
+	if (pkt->pkt_len <= hdr_len)
+		return -1;
+
+	/*
+	 * Save IPv4 ID for the packet whose DF bit is 0. For the packet
+	 * whose DF bit is 1, IPv4 ID is ignored.
+	 */
+	frag_offset = rte_be_to_cpu_16(outer_ipv4_hdr->fragment_offset);
+	outer_is_atomic =
+		((frag_offset & RTE_IPV4_HDR_DF_FLAG) == RTE_IPV4_HDR_DF_FLAG);
+	outer_ip_id = outer_is_atomic ? 0 :
+		rte_be_to_cpu_16(outer_ipv4_hdr->packet_id);
+	ipv6_frag_hdr = (struct rte_ipv6_fragment_ext *)(ipv6_hdr + 1);
+	ip_dl = rte_be_to_cpu_16(ipv6_hdr->payload_len) - fh_len;
+	ip_id = rte_be_to_cpu_32(ipv6_frag_hdr->id);
+	frag_offset = get_ipv6_frag_offset(ipv6_frag_hdr);
+	is_last_frag = is_last_ipv6_frag(ipv6_frag_hdr);
+
+	rte_ether_addr_copy(&(eth_hdr->s_addr), &(key.inner_key.eth_saddr));
+	rte_ether_addr_copy(&(eth_hdr->d_addr), &(key.inner_key.eth_daddr));
+	gro_ipv6_addr_copy(ipv6_hdr->src_addr, key.inner_key.ip_saddr);
+	gro_ipv6_addr_copy(ipv6_hdr->dst_addr, key.inner_key.ip_daddr);
+	key.inner_key.ip_id = ip_id;
+
+	key.vxlan_hdr.vx_flags = vxlan_hdr->vx_flags;
+	key.vxlan_hdr.vx_vni = vxlan_hdr->vx_vni;
+	rte_ether_addr_copy(&(outer_eth_hdr->s_addr), &(key.outer_eth_saddr));
+	rte_ether_addr_copy(&(outer_eth_hdr->d_addr), &(key.outer_eth_daddr));
+	key.outer_ip_src_addr = outer_ipv4_hdr->src_addr;
+	key.outer_ip_dst_addr = outer_ipv4_hdr->dst_addr;
+	key.outer_src_port = udp_hdr->src_port;
+	key.outer_dst_port = udp_hdr->dst_port;
+
+	/* Search for a matched flow. */
+	max_flow_num = tbl->max_flow_num;
+	remaining_flow_num = tbl->flow_num;
+	find = 0;
+	for (i = 0; i < max_flow_num && remaining_flow_num; i++) {
+		if (tbl->flows[i].start_index != INVALID_ARRAY_INDEX) {
+			if (is_same_vxlan_udp6_flow(tbl->flows[i].key, key)) {
+				find = 1;
+				break;
+			}
+			remaining_flow_num--;
+		}
+	}
+
+	/*
+	 * Can't find a matched flow. Insert a new flow and store the
+	 * packet into the flow.
+	 */
+	if (find == 0) {
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, frag_offset,
+				is_last_frag, outer_ip_id,
+				outer_is_atomic);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		if (insert_new_flow(tbl, &key, item_idx) ==
+				INVALID_ARRAY_INDEX) {
+			/*
+			 * Fail to insert a new flow, so
+			 * delete the inserted packet.
+			 */
+			delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+			return -1;
+		}
+		return 0;
+	}
+
+	/* Check all packets in the flow and try to find a neighbor. */
+	cur_idx = tbl->flows[i].start_index;
+	prev_idx = cur_idx;
+	do {
+		cmp = udp6_check_vxlan_neighbor(&(tbl->items[cur_idx]),
+				frag_offset, ip_dl);
+		if (cmp) {
+			if (merge_two_vxlan_udp6_packets(
+						&(tbl->items[cur_idx]),
+						pkt, cmp, frag_offset,
+						is_last_frag)) {
+				return 1;
+			}
+			/*
+			 * Can't merge two packets, as the packet
+			 * length will be greater than the max value.
+			 * Insert the packet into the flow.
+			 */
+			if (insert_new_item(tbl, pkt, start_time, prev_idx,
+						frag_offset, is_last_frag,
+						outer_ip_id,
+						outer_is_atomic)
+				== INVALID_ARRAY_INDEX)
+				return -1;
+			return 0;
+		}
+
+		/* Ensure inserted items are ordered by frag_offset */
+		if (frag_offset
+			< tbl->items[cur_idx].inner_item.frag_offset) {
+			break;
+		}
+
+		prev_idx = cur_idx;
+		cur_idx = tbl->items[cur_idx].inner_item.next_pkt_idx;
+	} while (cur_idx != INVALID_ARRAY_INDEX);
+
+	/* Can't find neighbor. Insert the packet into the flow. */
+	if (cur_idx == tbl->flows[i].start_index) {
+		/* Insert it before the first packet of the flow */
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, frag_offset,
+				is_last_frag, outer_ip_id,
+				outer_is_atomic);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		tbl->items[item_idx].inner_item.next_pkt_idx = cur_idx;
+		tbl->flows[i].start_index = item_idx;
+	} else {
+		if (insert_new_item(tbl, pkt, start_time, prev_idx,
+				frag_offset, is_last_frag, outer_ip_id,
+				outer_is_atomic) == INVALID_ARRAY_INDEX)
+			return -1;
+	}
+
+	return 0;
+}
+
+static int
+gro_vxlan_udp6_merge_items(struct gro_vxlan_udp6_tbl *tbl,
+			   uint32_t start_idx)
+{
+	uint16_t frag_offset;
+	uint8_t is_last_frag;
+	int16_t ip_dl;
+	struct rte_mbuf *pkt;
+	int cmp;
+	uint32_t item_idx;
+	uint16_t hdr_len;
+
+	item_idx = tbl->items[start_idx].inner_item.next_pkt_idx;
+	while (item_idx != INVALID_ARRAY_INDEX) {
+		pkt = tbl->items[item_idx].inner_item.firstseg;
+		hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+			pkt->l3_len;
+		ip_dl = pkt->pkt_len - hdr_len;
+		frag_offset = tbl->items[item_idx].inner_item.frag_offset;
+		is_last_frag = tbl->items[item_idx].inner_item.is_last_frag;
+		cmp = udp6_check_vxlan_neighbor(&(tbl->items[start_idx]),
+					frag_offset, ip_dl);
+		if (cmp) {
+			if (merge_two_vxlan_udp6_packets(
+					&(tbl->items[start_idx]),
+					pkt, cmp, frag_offset,
+					is_last_frag)) {
+				item_idx = delete_item(tbl, item_idx,
+							INVALID_ARRAY_INDEX);
+				tbl->items[start_idx].inner_item.next_pkt_idx
+					= item_idx;
+			} else {
+				return 0;
+			}
+		} else {
+			return 0;
+		}
+	}
+
+	return 0;
+}
+
+uint16_t
+gro_vxlan_udp6_tbl_timeout_flush(struct gro_vxlan_udp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out)
+{
+	uint16_t k = 0;
+	uint32_t i, j;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++) {
+		if (unlikely(tbl->flow_num == 0))
+			return k;
+
+		j = tbl->flows[i].start_index;
+		while (j != INVALID_ARRAY_INDEX) {
+			if (tbl->items[j].inner_item.start_time <=
+					flush_timestamp) {
+				gro_vxlan_udp6_merge_items(tbl, j);
+				out[k++] = tbl->items[j].inner_item.firstseg;
+				if (tbl->items[j].inner_item.nb_merged > 1)
+					update_vxlan_header(&(tbl->items[j]));
+				/*
+				 * Delete the item and get the next packet
+				 * index.
+				 */
+				j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+				tbl->flows[i].start_index = j;
+				if (j == INVALID_ARRAY_INDEX)
+					tbl->flow_num--;
+
+				if (unlikely(k == nb_out))
+					return k;
+			} else
+				/*
+				 * The left packets in the flow won't be
+				 * timeout. Go to check other flows.
+				 */
+				break;
+		}
+	}
+	return k;
+}
+
+uint32_t
+gro_vxlan_udp6_tbl_pkt_count(void *tbl)
+{
+	struct gro_vxlan_udp6_tbl *gro_tbl = tbl;
+
+	if (gro_tbl)
+		return gro_tbl->item_num;
+
+	return 0;
+}
diff --git a/lib/librte_gro/gro_vxlan_udp6.h b/lib/librte_gro/gro_vxlan_udp6.h
new file mode 100644
index 0000000..648733d
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan_udp6.h
@@ -0,0 +1,152 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#ifndef _GRO_VXLAN_UDP6_H_
+#define _GRO_VXLAN_UDP6_H_
+
+#include "gro_udp6.h"
+
+#define GRO_VXLAN_UDP6_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/* Header fields representing a VxLAN flow */
+struct vxlan_udp6_flow_key {
+	struct udp6_flow_key inner_key;
+	struct rte_vxlan_hdr vxlan_hdr;
+
+	struct rte_ether_addr outer_eth_saddr;
+	struct rte_ether_addr outer_eth_daddr;
+
+	uint32_t outer_ip_src_addr;
+	uint32_t outer_ip_dst_addr;
+
+	/* Outer UDP ports */
+	uint16_t outer_src_port;
+	uint16_t outer_dst_port;
+
+};
+
+struct gro_vxlan_udp6_flow {
+	struct vxlan_udp6_flow_key key;
+	/*
+	 * The index of the first packet in the flow. INVALID_ARRAY_INDEX
+	 * indicates an empty flow.
+	 */
+	uint32_t start_index;
+};
+
+struct gro_vxlan_udp6_item {
+	struct gro_udp6_item inner_item;
+	/* IPv4 ID in the outer IPv4 header */
+	uint16_t outer_ip_id;
+	/* Indicate if outer IPv4 ID can be ignored */
+	uint8_t outer_is_atomic;
+};
+
+/*
+ * VxLAN (with an outer IPv4 header and an inner UDP/IPv6 packet)
+ * reassembly table structure
+ */
+struct gro_vxlan_udp6_tbl {
+	/* item array */
+	struct gro_vxlan_udp6_item *items;
+	/* flow array */
+	struct gro_vxlan_udp6_flow *flows;
+	/* current item number */
+	uint32_t item_num;
+	/* current flow number */
+	uint32_t flow_num;
+	/* the maximum item number */
+	uint32_t max_item_num;
+	/* the maximum flow number */
+	uint32_t max_flow_num;
+};
+
+/**
+ * This function creates a IPv4 VxLAN reassembly table for IPv4 VxLAN packets
+ * which have an outer IPv4 header and an inner UDP/IPv6 packet.
+ *
+ * @param socket_id
+ *  Socket index for allocating the table
+ * @param max_flow_num
+ *  The maximum number of flows in the table
+ * @param max_item_per_flow
+ *  The maximum number of packets per flow
+ *
+ * @return
+ *  - Return the table pointer on success.
+ *  - Return NULL on failure.
+ */
+void *gro_vxlan_udp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow);
+
+/**
+ * This function destroys a IPv4 VxLAN reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv4 VxLAN reassembly table
+ */
+void gro_vxlan_udp6_tbl_destroy(void *tbl);
+
+/**
+ * This function merges a IPv4 VxLAN packet which has an outer IPv4 header and
+ * an inner UDP/IPv6 packet. It does not process the packet which does not
+ * have payload.
+ *
+ * This function does not check if the packet has correct checksums and
+ * does not re-calculate checksums for the merged packet. It returns the
+ * packet if there is no available space in the table.
+ *
+ * @param pkt
+ *  Packet to reassemble
+ * @param tbl
+ *  Pointer pointing to the IPv4 VxLAN reassembly table
+ * @start_time
+ *  The time when the packet is inserted into the table
+ *
+ * @return
+ *  - Return a positive value if the packet is merged.
+ *  - Return zero if the packet isn't merged but stored in the table.
+ *  - Return a negative value for invalid parameters or no available
+ *    space in the table.
+ */
+int32_t gro_vxlan_udp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan_udp6_tbl *tbl,
+		uint64_t start_time);
+
+/**
+ * This function flushes timeout packets in the IPv4 VxLAN reassembly table,
+ * and without updating checksums.
+ *
+ * @param tbl
+ *  Pointer pointing to a IPv4 VxLAN GRO table
+ * @param flush_timestamp
+ *  This function flushes packets which are inserted into the table
+ *  before or at the flush_timestamp.
+ * @param out
+ *  Pointer array used to keep flushed packets
+ * @param nb_out
+ *  The element number in 'out'. It also determines the maximum number of
+ *  packets that can be flushed finally.
+ *
+ * @return
+ *  The number of flushed packets
+ */
+uint16_t gro_vxlan_udp6_tbl_timeout_flush(struct gro_vxlan_udp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out);
+
+/**
+ * This function returns the number of the packets in a IPv4 VxLAN
+ * reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv4 VxLAN reassembly table
+ *
+ * @return
+ *  The number of packets in the table
+ */
+uint32_t gro_vxlan_udp6_tbl_pkt_count(void *tbl);
+#endif
diff --git a/lib/librte_gro/meson.build b/lib/librte_gro/meson.build
index 7939081..14b6b64 100644
--- a/lib/librte_gro/meson.build
+++ b/lib/librte_gro/meson.build
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c', 'gro_udp6.c')
+sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c', 'gro_udp6.c', 'gro_vxlan_udp6.c')
 headers = files('rte_gro.h')
 deps += ['ethdev']
diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
index 13b38cc..3bb7ea1 100644
--- a/lib/librte_gro/rte_gro.c
+++ b/lib/librte_gro/rte_gro.c
@@ -17,6 +17,7 @@
 #include "gro_vxlan_tcp4.h"
 #include "gro_vxlan_tcp6.h"
 #include "gro_vxlan_udp4.h"
+#include "gro_vxlan_udp6.h"
 
 typedef void *(*gro_tbl_create_fn)(uint16_t socket_id,
 		uint16_t max_flow_num,
@@ -29,21 +30,24 @@
 		gro_udp4_tbl_create, gro_vxlan_udp4_tbl_create,
 		gro_tcp6_tbl_create, gro_vxlan_tcp6_tbl_create,
 		gro_vxlan6_tcp4_tbl_create, gro_vxlan6_tcp6_tbl_create,
-		gro_udp6_tbl_create, NULL};
+		gro_udp6_tbl_create, gro_vxlan_udp6_tbl_create,
+		NULL};
 static gro_tbl_destroy_fn tbl_destroy_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_destroy, gro_vxlan_tcp4_tbl_destroy,
 			gro_udp4_tbl_destroy, gro_vxlan_udp4_tbl_destroy,
 			gro_tcp6_tbl_destroy, gro_vxlan_tcp6_tbl_destroy,
 			gro_vxlan6_tcp4_tbl_destroy,
 			gro_vxlan6_tcp6_tbl_destroy,
-			gro_udp6_tbl_destroy, NULL};
+			gro_udp6_tbl_destroy, gro_vxlan_udp6_tbl_destroy,
+			NULL};
 static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_pkt_count, gro_vxlan_tcp4_tbl_pkt_count,
 			gro_udp4_tbl_pkt_count, gro_vxlan_udp4_tbl_pkt_count,
 			gro_tcp6_tbl_pkt_count, gro_vxlan_tcp6_tbl_pkt_count,
 			gro_vxlan6_tcp4_tbl_pkt_count,
 			gro_vxlan6_tcp6_tbl_pkt_count,
-			gro_udp6_tbl_pkt_count, NULL};
+			gro_udp6_tbl_pkt_count, gro_vxlan_udp6_tbl_pkt_count,
+			NULL};
 
 #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
@@ -126,6 +130,19 @@
 		((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
 		(RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
 
+#define IS_IPV4_VXLAN_UDP6_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
+		((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
+		((ptype & RTE_PTYPE_TUNNEL_VXLAN) == \
+		 RTE_PTYPE_TUNNEL_VXLAN) && \
+		((ptype & RTE_PTYPE_INNER_L4_UDP) == \
+		 RTE_PTYPE_INNER_L4_UDP) && \
+		(((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6_EXT) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN)))
+
 /*
  * GRO context structure. It keeps the table structures, which are
  * used to merge packets, for different GRO types. Before using
@@ -256,6 +273,12 @@ struct gro_ctx {
 	struct gro_udp6_flow udp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
 	struct gro_udp6_item udp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
 
+	/* Allocate a reassembly table for IPv4 VXLAN UDP/IPv6 GRO */
+	struct gro_vxlan_udp6_tbl vxlan_udp6_tbl;
+	struct gro_vxlan_udp6_flow vxlan_udp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+	struct gro_vxlan_udp6_item vxlan_udp6_items[RTE_GRO_MAX_BURST_ITEM_NUM]
+			= {{{0}, 0, 0} };
+
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	uint32_t item_num;
 	int32_t ret;
@@ -263,7 +286,7 @@ struct gro_ctx {
 	uint8_t do_tcp4_gro = 0, do_vxlan_tcp_gro = 0, do_udp4_gro = 0,
 		do_vxlan_udp_gro = 0, do_tcp6_gro = 0, do_vxlan_tcp6_gro = 0,
 		do_vxlan6_tcp4_gro = 0, do_vxlan6_tcp6_gro = 0,
-		do_udp6_gro = 0;
+		do_udp6_gro = 0, do_vxlan_udp6_gro = 0;
 
 	if (unlikely((param->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -273,7 +296,8 @@ struct gro_ctx {
 					RTE_GRO_IPV4_VXLAN_TCP_IPV6 |
 					RTE_GRO_IPV6_VXLAN_TCP_IPV4 |
 					RTE_GRO_IPV6_VXLAN_TCP_IPV6 |
-					RTE_GRO_UDP_IPV6)) == 0))
+					RTE_GRO_UDP_IPV6 |
+					RTE_GRO_IPV4_VXLAN_UDP_IPV6)) == 0))
 		return nb_pkts;
 
 	/* Get the maximum number of packets */
@@ -398,6 +422,19 @@ struct gro_ctx {
 		do_udp6_gro = 1;
 	}
 
+	if (param->gro_types & RTE_GRO_IPV4_VXLAN_UDP_IPV6) {
+		for (i = 0; i < item_num; i++)
+			vxlan_udp6_flows[i].start_index = INVALID_ARRAY_INDEX;
+
+		vxlan_udp6_tbl.flows = vxlan_udp6_flows;
+		vxlan_udp6_tbl.items = vxlan_udp6_items;
+		vxlan_udp6_tbl.flow_num = 0;
+		vxlan_udp6_tbl.item_num = 0;
+		vxlan_udp6_tbl.max_flow_num = item_num;
+		vxlan_udp6_tbl.max_item_num = item_num;
+		do_vxlan_udp6_gro = 1;
+	}
+
 	for (i = 0; i < nb_pkts; i++) {
 		/*
 		 * The timestamp is ignored, since all packets
@@ -480,6 +517,15 @@ struct gro_ctx {
 				nb_after_gro--;
 			else if (ret < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV4_VXLAN_UDP6_PKT(pkts[i]->packet_type) &&
+				do_vxlan_udp6_gro) {
+			ret = gro_vxlan_udp6_reassemble(pkts[i],
+							&vxlan_udp6_tbl, 0);
+			if (ret > 0)
+				/* Merge successfully */
+				nb_after_gro--;
+			else if (ret < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -534,6 +580,11 @@ struct gro_ctx {
 					&pkts[i], nb_pkts - i);
 		}
 
+		if (do_vxlan_udp6_gro) {
+			i += gro_vxlan_udp6_tbl_timeout_flush(&vxlan_udp6_tbl,
+					0, &pkts[i], nb_pkts - i);
+		}
+
 		/* Copy unprocessed packets */
 		if (unprocess_num > 0) {
 			memcpy(&pkts[i], unprocess_pkts,
@@ -555,12 +606,12 @@ struct gro_ctx {
 	struct gro_ctx *gro_ctx = ctx;
 	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl,
 		*vxlan_tcp6_tbl, *vxlan6_tcp4_tbl, *vxlan6_tcp6_tbl,
-		*udp6_tbl;
+		*udp6_tbl, *vxlan_udp6_tbl;
 	uint64_t current_time;
 	uint16_t i, unprocess_num = 0;
 	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro,
 		do_tcp6_gro, do_vxlan_tcp6_gro, do_vxlan6_tcp4_gro,
-		do_vxlan6_tcp6_gro, do_udp6_gro;
+		do_vxlan6_tcp6_gro, do_udp6_gro, do_vxlan_udp6_gro;
 
 	if (unlikely((gro_ctx->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -570,7 +621,8 @@ struct gro_ctx {
 					RTE_GRO_IPV4_VXLAN_TCP_IPV6 |
 					RTE_GRO_IPV6_VXLAN_TCP_IPV4 |
 					RTE_GRO_IPV6_VXLAN_TCP_IPV6 |
-					RTE_GRO_UDP_IPV6)) == 0))
+					RTE_GRO_UDP_IPV6 |
+					RTE_GRO_IPV4_VXLAN_UDP_IPV6)) == 0))
 		return nb_pkts;
 
 	tcp_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV4_INDEX];
@@ -582,6 +634,7 @@ struct gro_ctx {
 	vxlan6_tcp4_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_TCP_IPV4_INDEX];
 	vxlan6_tcp6_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX];
 	udp6_tbl = gro_ctx->tbls[RTE_GRO_UDP_IPV6_INDEX];
+	vxlan_udp6_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX];
 
 	do_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV4) ==
 		RTE_GRO_TCP_IPV4;
@@ -601,6 +654,8 @@ struct gro_ctx {
 				== RTE_GRO_IPV6_VXLAN_TCP_IPV6;
 	do_udp6_gro = (gro_ctx->gro_types & RTE_GRO_UDP_IPV6) ==
 		RTE_GRO_UDP_IPV6;
+	do_vxlan_udp6_gro = (gro_ctx->gro_types & RTE_GRO_IPV4_VXLAN_UDP_IPV6)
+				== RTE_GRO_IPV4_VXLAN_UDP_IPV6;
 
 	current_time = rte_rdtsc();
 
@@ -650,6 +705,11 @@ struct gro_ctx {
 			if (gro_udp6_reassemble(pkts[i], udp6_tbl,
 						current_time) < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV4_VXLAN_UDP6_PKT(pkts[i]->packet_type) &&
+				do_vxlan_udp6_gro) {
+			if (gro_vxlan_udp6_reassemble(pkts[i], vxlan_udp6_tbl,
+						current_time) < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -741,6 +801,13 @@ struct gro_ctx {
 				gro_ctx->tbls[RTE_GRO_UDP_IPV6_INDEX],
 				flush_timestamp,
 				&out[num], left_nb_out);
+		left_nb_out = max_nb_out - num;
+	}
+
+	if ((gro_types & RTE_GRO_IPV4_VXLAN_UDP_IPV6) && left_nb_out > 0) {
+		num += gro_vxlan_udp6_tbl_timeout_flush(gro_ctx->tbls[
+				RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX],
+				flush_timestamp, &out[num], left_nb_out);
 	}
 
 	return num;
diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
index 94ed3d3..1824ce3 100644
--- a/lib/librte_gro/rte_gro.h
+++ b/lib/librte_gro/rte_gro.h
@@ -53,6 +53,9 @@
 #define RTE_GRO_UDP_IPV6_INDEX 8
 #define RTE_GRO_UDP_IPV6 (1ULL << RTE_GRO_UDP_IPV6_INDEX)
 /**< UDP/IPv6 GRO flag */
+#define RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX 9
+#define RTE_GRO_IPV4_VXLAN_UDP_IPV6 (1ULL << RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX)
+/**< IPv4 VxLAN UDP/IPv6 GRO flag. */
 
 /**
  * Structure used to create GRO context objects or used to pass
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-dev] [PATCH v1 7/8] gro: support IPv6 VXLAN UDP/IPv4
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
                   ` (5 preceding siblings ...)
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 6/8] gro: support IPv4 VXLAN UDP/IPv6 yang_y_yi
@ 2020-12-21  3:50 ` yang_y_yi
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 8/8] gro: support IPv6 VXLAN UDP/IPv6 yang_y_yi
  2021-03-24 21:22 ` [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN Thomas Monjalon
  8 siblings, 0 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

IPv6 VXLAN UDP/IPv4 GRO can help improve UDP/IPv4
performance in IPv6 VXLAN use case.

With this enabled in DPDK, OVS DPDK can leverage it
to improve VM-to-VM UDP/IPv4 performance, it will merge
small adjacent IPv6 VXLAN UDP/IPv4 fragments to a big
IPv6 VXLAN UDP/IPv4 packet immediate after they are
received from a physical NIC. It is very helpful in OVS
DPDK VXLAN use case.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
---
 .../prog_guide/generic_receive_offload_lib.rst     |   3 +-
 doc/guides/rel_notes/release_21_02.rst             |   5 +
 lib/librte_gro/gro_vxlan6_udp4.c                   | 559 +++++++++++++++++++++
 lib/librte_gro/gro_vxlan6_udp4.h                   | 148 ++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           |  84 +++-
 lib/librte_gro/rte_gro.h                           |   3 +
 7 files changed, 794 insertions(+), 10 deletions(-)
 create mode 100644 lib/librte_gro/gro_vxlan6_udp4.c
 create mode 100644 lib/librte_gro/gro_vxlan6_udp4.h

diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst b/doc/guides/prog_guide/generic_receive_offload_lib.rst
index 906e85f..8f48809 100644
--- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
+++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
@@ -33,7 +33,8 @@ fragmentation is possible (i.e., DF==0). Additionally, it complies RFC
 Currently, the GRO library provides GRO supports for TCP/IPv4, UDP/IPv4,
 UDP/IPv6, and TCP/IPv6 packets as well as VxLAN packets which contain an
 outer IPv4 header and an inner TCP/IPv4, UDP/IPv4, TCP/IPv6 or UDP/IPv6
-packet or an outer IPv6 header and an inner TCP/IPv4 or TCP/IPv6 packet.
+packet or an outer IPv6 header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6
+packet.
 
 Two Sets of API
 ---------------
diff --git a/doc/guides/rel_notes/release_21_02.rst b/doc/guides/rel_notes/release_21_02.rst
index 4758cea..5e3d10b 100644
--- a/doc/guides/rel_notes/release_21_02.rst
+++ b/doc/guides/rel_notes/release_21_02.rst
@@ -85,6 +85,11 @@ New Features
    Added UDP/IPv6 GRO support for IPv4 VXLAN packets, this enabled UDP/IPv6
    GRO support for IPv4 VXLAN use case.
 
+* **Added UDP/IPv4 GRO support for IPv6 VXLAN packets.**
+
+   Added UDP/IPv4 GRO support for IPv6 VXLAN packets, this enabled UDP/IPv4
+   GRO support for IPv6 VXLAN use case.
+
 Removed Items
 -------------
 
diff --git a/lib/librte_gro/gro_vxlan6_udp4.c b/lib/librte_gro/gro_vxlan6_udp4.c
new file mode 100644
index 0000000..c9e7af4
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan6_udp4.c
@@ -0,0 +1,559 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+#include <rte_udp.h>
+#include <rte_ip.h>
+
+#include "gro_vxlan6_udp4.h"
+
+void *
+gro_vxlan6_udp4_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow)
+{
+	struct gro_vxlan6_udp4_tbl *tbl;
+	size_t size;
+	uint32_t entries_num, i;
+
+	entries_num = max_flow_num * max_item_per_flow;
+	entries_num = RTE_MIN(entries_num, GRO_VXLAN6_UDP4_TBL_MAX_ITEM_NUM);
+
+	if (entries_num == 0)
+		return NULL;
+
+	tbl = rte_zmalloc_socket(__func__,
+			sizeof(struct gro_vxlan6_udp4_tbl),
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl == NULL)
+		return NULL;
+
+	size = sizeof(struct gro_vxlan6_udp4_item) * entries_num;
+	tbl->items = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->items == NULL) {
+		rte_free(tbl);
+		return NULL;
+	}
+	tbl->max_item_num = entries_num;
+
+	size = sizeof(struct gro_vxlan6_udp4_flow) * entries_num;
+	tbl->flows = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->flows == NULL) {
+		rte_free(tbl->items);
+		rte_free(tbl);
+		return NULL;
+	}
+
+	for (i = 0; i < entries_num; i++)
+		tbl->flows[i].start_index = INVALID_ARRAY_INDEX;
+	tbl->max_flow_num = entries_num;
+
+	return tbl;
+}
+
+void
+gro_vxlan6_udp4_tbl_destroy(void *tbl)
+{
+	struct gro_vxlan6_udp4_tbl *vxlan6_tbl = tbl;
+
+	if (vxlan6_tbl) {
+		rte_free(vxlan6_tbl->items);
+		rte_free(vxlan6_tbl->flows);
+	}
+	rte_free(vxlan6_tbl);
+}
+
+static inline uint32_t
+find_an_empty_item(struct gro_vxlan6_udp4_tbl *tbl)
+{
+	uint32_t max_item_num = tbl->max_item_num, i;
+
+	for (i = 0; i < max_item_num; i++)
+		if (tbl->items[i].inner_item.firstseg == NULL)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+find_an_empty_flow(struct gro_vxlan6_udp4_tbl *tbl)
+{
+	uint32_t max_flow_num = tbl->max_flow_num, i;
+
+	for (i = 0; i < max_flow_num; i++)
+		if (tbl->flows[i].start_index == INVALID_ARRAY_INDEX)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_item(struct gro_vxlan6_udp4_tbl *tbl,
+		struct rte_mbuf *pkt,
+		uint64_t start_time,
+		uint32_t prev_idx,
+		uint16_t frag_offset,
+		uint8_t is_last_frag)
+{
+	uint32_t item_idx;
+
+	item_idx = find_an_empty_item(tbl);
+	if (unlikely(item_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	tbl->items[item_idx].inner_item.firstseg = pkt;
+	tbl->items[item_idx].inner_item.lastseg = rte_pktmbuf_lastseg(pkt);
+	tbl->items[item_idx].inner_item.start_time = start_time;
+	tbl->items[item_idx].inner_item.next_pkt_idx = INVALID_ARRAY_INDEX;
+	tbl->items[item_idx].inner_item.frag_offset = frag_offset;
+	tbl->items[item_idx].inner_item.is_last_frag = is_last_frag;
+	tbl->items[item_idx].inner_item.nb_merged = 1;
+	tbl->item_num++;
+
+	/* If the previous packet exists, chain the new one with it. */
+	if (prev_idx != INVALID_ARRAY_INDEX) {
+		tbl->items[item_idx].inner_item.next_pkt_idx =
+			tbl->items[prev_idx].inner_item.next_pkt_idx;
+		tbl->items[prev_idx].inner_item.next_pkt_idx = item_idx;
+	}
+
+	return item_idx;
+}
+
+static inline uint32_t
+delete_item(struct gro_vxlan6_udp4_tbl *tbl,
+		uint32_t item_idx,
+		uint32_t prev_item_idx)
+{
+	uint32_t next_idx = tbl->items[item_idx].inner_item.next_pkt_idx;
+
+	/* NULL indicates an empty item. */
+	tbl->items[item_idx].inner_item.firstseg = NULL;
+	tbl->item_num--;
+	if (prev_item_idx != INVALID_ARRAY_INDEX)
+		tbl->items[prev_item_idx].inner_item.next_pkt_idx = next_idx;
+
+	return next_idx;
+}
+
+/* Copy IPv6 addr */
+static inline void gro_ipv6_addr_copy(const uint8_t *ipv6_from,
+				      uint8_t *ipv6_to)
+{
+	const uint64_t *from_words = (const uint64_t *)ipv6_from;
+	uint64_t *to_words   = (uint64_t *)ipv6_to;
+
+	to_words[0] = from_words[0];
+	to_words[1] = from_words[1];
+}
+
+static inline int rte_is_same_ipv6_addr(const uint8_t *ipv6_a1,
+					const uint8_t *ipv6_a2)
+{
+	const uint64_t *w1 = (const uint64_t *)ipv6_a1;
+	const uint64_t *w2 = (const uint64_t *)ipv6_a2;
+
+	return ((w1[0] ^ w2[0]) | (w1[1] ^ w2[1])) == 0;
+}
+
+static inline uint32_t
+insert_new_flow(struct gro_vxlan6_udp4_tbl *tbl,
+		struct vxlan6_udp4_flow_key *src,
+		uint32_t item_idx)
+{
+	struct vxlan6_udp4_flow_key *dst;
+	uint32_t flow_idx;
+
+	flow_idx = find_an_empty_flow(tbl);
+	if (unlikely(flow_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	dst = &(tbl->flows[flow_idx].key);
+
+	rte_ether_addr_copy(&(src->inner_key.eth_saddr),
+			&(dst->inner_key.eth_saddr));
+	rte_ether_addr_copy(&(src->inner_key.eth_daddr),
+			&(dst->inner_key.eth_daddr));
+	dst->inner_key.ip_src_addr = src->inner_key.ip_src_addr;
+	dst->inner_key.ip_dst_addr = src->inner_key.ip_dst_addr;
+	dst->inner_key.ip_id = src->inner_key.ip_id;
+
+	dst->vxlan_hdr.vx_flags = src->vxlan_hdr.vx_flags;
+	dst->vxlan_hdr.vx_vni = src->vxlan_hdr.vx_vni;
+	rte_ether_addr_copy(&(src->outer_eth_saddr), &(dst->outer_eth_saddr));
+	rte_ether_addr_copy(&(src->outer_eth_daddr), &(dst->outer_eth_daddr));
+	gro_ipv6_addr_copy(src->outer_ip_saddr, dst->outer_ip_saddr);
+	gro_ipv6_addr_copy(src->outer_ip_daddr, dst->outer_ip_daddr);
+	dst->outer_src_port = src->outer_src_port;
+	dst->outer_dst_port = src->outer_dst_port;
+
+	tbl->flows[flow_idx].start_index = item_idx;
+	tbl->flow_num++;
+
+	return flow_idx;
+}
+
+static inline int
+is_same_vxlan6_udp4_flow(struct vxlan6_udp4_flow_key k1,
+		struct vxlan6_udp4_flow_key k2)
+{
+	/* For VxLAN packet, outer udp src port is calculated from
+	 * inner packet RSS hash, udp src port of the first UDP
+	 * fragment is different from one of other UDP fragments
+	 * even if they are same flow, so we have to skip outer udp
+	 * src port comparison here.
+	 */
+	return (rte_is_same_ether_addr(&k1.outer_eth_saddr,
+					&k2.outer_eth_saddr) &&
+			rte_is_same_ether_addr(&k1.outer_eth_daddr,
+				&k2.outer_eth_daddr) &&
+			rte_is_same_ipv6_addr(k1.outer_ip_saddr,
+					      k2.outer_ip_saddr) &&
+			rte_is_same_ipv6_addr(k1.outer_ip_daddr,
+					      k2.outer_ip_daddr) &&
+			(k1.outer_dst_port == k2.outer_dst_port) &&
+			(k1.vxlan_hdr.vx_flags == k2.vxlan_hdr.vx_flags) &&
+			(k1.vxlan_hdr.vx_vni == k2.vxlan_hdr.vx_vni) &&
+			is_same_udp4_flow(k1.inner_key, k2.inner_key));
+}
+
+static inline int
+udp_check_vxlan6_neighbor(struct gro_vxlan6_udp4_item *item,
+		uint16_t frag_offset,
+		uint16_t ip_dl)
+{
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	int cmp;
+	uint16_t l2_offset;
+	int ret = 0;
+
+	l2_offset = pkt->outer_l2_len + pkt->outer_l3_len;
+	cmp = udp4_check_neighbor(&item->inner_item, frag_offset,
+			ip_dl, l2_offset);
+	if (cmp > 0)
+		/* Append the new packet. */
+		ret = 1;
+	else if (cmp < 0)
+		/* Prepend the new packet. */
+		ret = -1;
+
+	return ret;
+}
+
+static inline int
+merge_two_vxlan6_udp4_packets(struct gro_vxlan6_udp4_item *item,
+		struct rte_mbuf *pkt,
+		int cmp,
+		uint16_t frag_offset,
+		uint8_t is_last_frag)
+{
+	if (merge_two_udp4_packets(&item->inner_item, pkt, cmp, frag_offset,
+				is_last_frag,
+				pkt->outer_l2_len + pkt->outer_l3_len)) {
+		return 1;
+	}
+
+	return 0;
+}
+
+static inline void
+update_vxlan6_header(struct gro_vxlan6_udp4_item *item)
+{
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_ipv4_hdr *ipv4_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	uint16_t len;
+	uint16_t frag_offset;
+
+	/* Update the outer IPv6 header. */
+	len = pkt->pkt_len - pkt->outer_l2_len;
+	ipv6_hdr = (struct rte_ipv6_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+			pkt->outer_l2_len);
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(len - pkt->outer_l3_len);
+
+	/* Update the outer UDP header. */
+	len -= pkt->outer_l3_len;
+	udp_hdr = (struct rte_udp_hdr *)((char *)ipv6_hdr + pkt->outer_l3_len);
+	udp_hdr->dgram_len = rte_cpu_to_be_16(len);
+
+	/* Update the inner IPv4 header. */
+	len -= pkt->l2_len;
+	ipv4_hdr = (struct rte_ipv4_hdr *)((char *)udp_hdr + pkt->l2_len);
+	ipv4_hdr->total_length = rte_cpu_to_be_16(len);
+
+	/* Clear MF bit if it is last fragment */
+	if (item->inner_item.is_last_frag) {
+		frag_offset = rte_be_to_cpu_16(ipv4_hdr->fragment_offset);
+		ipv4_hdr->fragment_offset =
+			rte_cpu_to_be_16(frag_offset & ~RTE_IPV4_HDR_MF_FLAG);
+	}
+}
+
+int32_t
+gro_vxlan6_udp4_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan6_udp4_tbl *tbl,
+		uint64_t start_time)
+{
+	struct rte_ether_hdr *outer_eth_hdr, *eth_hdr;
+	struct rte_ipv6_hdr *outer_ipv6_hdr;
+	struct rte_ipv4_hdr *ipv4_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_vxlan_hdr *vxlan_hdr;
+	uint16_t frag_offset;
+	uint8_t is_last_frag;
+	int16_t ip_dl;
+	uint16_t ip_id;
+
+	struct vxlan6_udp4_flow_key key;
+	uint32_t cur_idx, prev_idx, item_idx;
+	uint32_t i, max_flow_num, remaining_flow_num;
+	int cmp;
+	uint16_t hdr_len;
+	uint8_t find;
+
+	outer_eth_hdr = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *);
+	outer_ipv6_hdr = (struct rte_ipv6_hdr *)((char *)outer_eth_hdr +
+			pkt->outer_l2_len);
+
+	udp_hdr = (struct rte_udp_hdr *)((char *)outer_ipv6_hdr +
+			pkt->outer_l3_len);
+	vxlan_hdr = (struct rte_vxlan_hdr *)((char *)udp_hdr +
+			sizeof(struct rte_udp_hdr));
+	eth_hdr = (struct rte_ether_hdr *)((char *)vxlan_hdr +
+			sizeof(struct rte_vxlan_hdr));
+	/* l2_len = outer udp hdr len + vxlan hdr len + inner l2 len */
+	ipv4_hdr = (struct rte_ipv4_hdr *)((char *)udp_hdr + pkt->l2_len);
+
+	/*
+	 * Don't process the packet which has non-fragment inner IP.
+	 */
+	if (!is_ipv4_fragment(ipv4_hdr))
+		return -1;
+
+	hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+		pkt->l3_len;
+	/*
+	 * Don't process the packet whose payload length is less than or
+	 * equal to 0.
+	 */
+	if (pkt->pkt_len <= hdr_len)
+		return -1;
+	ip_dl = pkt->pkt_len - hdr_len;
+
+	ip_id = rte_be_to_cpu_16(ipv4_hdr->packet_id);
+	frag_offset = rte_be_to_cpu_16(ipv4_hdr->fragment_offset);
+	is_last_frag = ((frag_offset & RTE_IPV4_HDR_MF_FLAG) == 0) ? 1 : 0;
+	frag_offset = (uint16_t)(frag_offset & RTE_IPV4_HDR_OFFSET_MASK) << 3;
+
+	rte_ether_addr_copy(&(eth_hdr->s_addr), &(key.inner_key.eth_saddr));
+	rte_ether_addr_copy(&(eth_hdr->d_addr), &(key.inner_key.eth_daddr));
+	key.inner_key.ip_src_addr = ipv4_hdr->src_addr;
+	key.inner_key.ip_dst_addr = ipv4_hdr->dst_addr;
+	key.inner_key.ip_id = ip_id;
+
+	key.vxlan_hdr.vx_flags = vxlan_hdr->vx_flags;
+	key.vxlan_hdr.vx_vni = vxlan_hdr->vx_vni;
+	rte_ether_addr_copy(&(outer_eth_hdr->s_addr), &(key.outer_eth_saddr));
+	rte_ether_addr_copy(&(outer_eth_hdr->d_addr), &(key.outer_eth_daddr));
+	gro_ipv6_addr_copy(outer_ipv6_hdr->src_addr, key.outer_ip_saddr);
+	gro_ipv6_addr_copy(outer_ipv6_hdr->dst_addr, key.outer_ip_daddr);
+	key.outer_src_port = udp_hdr->src_port;
+	key.outer_dst_port = udp_hdr->dst_port;
+
+	/* Search for a matched flow. */
+	max_flow_num = tbl->max_flow_num;
+	remaining_flow_num = tbl->flow_num;
+	find = 0;
+	for (i = 0; i < max_flow_num && remaining_flow_num; i++) {
+		if (tbl->flows[i].start_index != INVALID_ARRAY_INDEX) {
+			if (is_same_vxlan6_udp4_flow(tbl->flows[i].key, key)) {
+				find = 1;
+				break;
+			}
+			remaining_flow_num--;
+		}
+	}
+
+	/*
+	 * Can't find a matched flow. Insert a new flow and store the
+	 * packet into the flow.
+	 */
+	if (find == 0) {
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, frag_offset,
+				is_last_frag);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		if (insert_new_flow(tbl, &key, item_idx) ==
+				INVALID_ARRAY_INDEX) {
+			/*
+			 * Fail to insert a new flow, so
+			 * delete the inserted packet.
+			 */
+			delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+			return -1;
+		}
+		return 0;
+	}
+
+	/* Check all packets in the flow and try to find a neighbor. */
+	cur_idx = tbl->flows[i].start_index;
+	prev_idx = cur_idx;
+	do {
+		cmp = udp_check_vxlan6_neighbor(&(tbl->items[cur_idx]),
+				frag_offset, ip_dl);
+		if (cmp) {
+			if (merge_two_vxlan6_udp4_packets(
+						&(tbl->items[cur_idx]),
+						pkt, cmp, frag_offset,
+						is_last_frag)) {
+				return 1;
+			}
+			/*
+			 * Can't merge two packets, as the packet
+			 * length will be greater than the max value.
+			 * Insert the packet into the flow.
+			 */
+			if (insert_new_item(tbl, pkt, start_time, prev_idx,
+						frag_offset, is_last_frag)
+				== INVALID_ARRAY_INDEX)
+				return -1;
+			return 0;
+		}
+
+		/* Ensure inserted items are ordered by frag_offset */
+		if (frag_offset
+			< tbl->items[cur_idx].inner_item.frag_offset) {
+			break;
+		}
+
+		prev_idx = cur_idx;
+		cur_idx = tbl->items[cur_idx].inner_item.next_pkt_idx;
+	} while (cur_idx != INVALID_ARRAY_INDEX);
+
+	/* Can't find neighbor. Insert the packet into the flow. */
+	if (cur_idx == tbl->flows[i].start_index) {
+		/* Insert it before the first packet of the flow */
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, frag_offset,
+				is_last_frag);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		tbl->items[item_idx].inner_item.next_pkt_idx = cur_idx;
+		tbl->flows[i].start_index = item_idx;
+	} else {
+		if (insert_new_item(tbl, pkt, start_time, prev_idx,
+				frag_offset, is_last_frag)
+			== INVALID_ARRAY_INDEX)
+			return -1;
+	}
+
+	return 0;
+}
+
+static int
+gro_vxlan6_udp4_merge_items(struct gro_vxlan6_udp4_tbl *tbl,
+			   uint32_t start_idx)
+{
+	uint16_t frag_offset;
+	uint8_t is_last_frag;
+	int16_t ip_dl;
+	struct rte_mbuf *pkt;
+	int cmp;
+	uint32_t item_idx;
+	uint16_t hdr_len;
+
+	item_idx = tbl->items[start_idx].inner_item.next_pkt_idx;
+	while (item_idx != INVALID_ARRAY_INDEX) {
+		pkt = tbl->items[item_idx].inner_item.firstseg;
+		hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+			pkt->l3_len;
+		ip_dl = pkt->pkt_len - hdr_len;
+		frag_offset = tbl->items[item_idx].inner_item.frag_offset;
+		is_last_frag = tbl->items[item_idx].inner_item.is_last_frag;
+		cmp = udp_check_vxlan6_neighbor(&(tbl->items[start_idx]),
+					frag_offset, ip_dl);
+		if (cmp) {
+			if (merge_two_vxlan6_udp4_packets(
+					&(tbl->items[start_idx]),
+					pkt, cmp, frag_offset,
+					is_last_frag)) {
+				item_idx = delete_item(tbl, item_idx,
+							INVALID_ARRAY_INDEX);
+				tbl->items[start_idx].inner_item.next_pkt_idx
+					= item_idx;
+			} else {
+				return 0;
+			}
+		} else {
+			return 0;
+		}
+	}
+
+	return 0;
+}
+
+uint16_t
+gro_vxlan6_udp4_tbl_timeout_flush(struct gro_vxlan6_udp4_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out)
+{
+	uint16_t k = 0;
+	uint32_t i, j;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++) {
+		if (unlikely(tbl->flow_num == 0))
+			return k;
+
+		j = tbl->flows[i].start_index;
+		while (j != INVALID_ARRAY_INDEX) {
+			if (tbl->items[j].inner_item.start_time <=
+					flush_timestamp) {
+				gro_vxlan6_udp4_merge_items(tbl, j);
+				out[k++] = tbl->items[j].inner_item.firstseg;
+				if (tbl->items[j].inner_item.nb_merged > 1)
+					update_vxlan6_header(&(tbl->items[j]));
+				/*
+				 * Delete the item and get the next packet
+				 * index.
+				 */
+				j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+				tbl->flows[i].start_index = j;
+				if (j == INVALID_ARRAY_INDEX)
+					tbl->flow_num--;
+
+				if (unlikely(k == nb_out))
+					return k;
+			} else
+				/*
+				 * The left packets in the flow won't be
+				 * timeout. Go to check other flows.
+				 */
+				break;
+		}
+	}
+	return k;
+}
+
+uint32_t
+gro_vxlan6_udp4_tbl_pkt_count(void *tbl)
+{
+	struct gro_vxlan6_udp4_tbl *gro_tbl = tbl;
+
+	if (gro_tbl)
+		return gro_tbl->item_num;
+
+	return 0;
+}
diff --git a/lib/librte_gro/gro_vxlan6_udp4.h b/lib/librte_gro/gro_vxlan6_udp4.h
new file mode 100644
index 0000000..14dc2b6
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan6_udp4.h
@@ -0,0 +1,148 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#ifndef _GRO_VXLAN6_UDP4_H_
+#define _GRO_VXLAN6_UDP4_H_
+
+#include "gro_udp4.h"
+
+#define GRO_VXLAN6_UDP4_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/* Header fields representing a IPv6 VxLAN flow */
+struct vxlan6_udp4_flow_key {
+	struct udp4_flow_key inner_key;
+	struct rte_vxlan_hdr vxlan_hdr;
+
+	struct rte_ether_addr outer_eth_saddr;
+	struct rte_ether_addr outer_eth_daddr;
+
+	uint8_t outer_ip_saddr[16];
+	uint8_t outer_ip_daddr[16];
+
+	/* Outer UDP ports */
+	uint16_t outer_src_port;
+	uint16_t outer_dst_port;
+
+};
+
+struct gro_vxlan6_udp4_flow {
+	struct vxlan6_udp4_flow_key key;
+	/*
+	 * The index of the first packet in the flow. INVALID_ARRAY_INDEX
+	 * indicates an empty flow.
+	 */
+	uint32_t start_index;
+};
+
+struct gro_vxlan6_udp4_item {
+	struct gro_udp4_item inner_item;
+};
+
+/*
+ * IPv6 VxLAN (with an outer IPv6 header and an inner UDP/IPv4 packet)
+ * reassembly table structure
+ */
+struct gro_vxlan6_udp4_tbl {
+	/* item array */
+	struct gro_vxlan6_udp4_item *items;
+	/* flow array */
+	struct gro_vxlan6_udp4_flow *flows;
+	/* current item number */
+	uint32_t item_num;
+	/* current flow number */
+	uint32_t flow_num;
+	/* the maximum item number */
+	uint32_t max_item_num;
+	/* the maximum flow number */
+	uint32_t max_flow_num;
+};
+
+/**
+ * This function creates a IPv6 VxLAN reassembly table for IPv6 VxLAN packets
+ * which have an outer IPv6 header and an inner UDP/IPv4 packet.
+ *
+ * @param socket_id
+ *  Socket index for allocating the table
+ * @param max_flow_num
+ *  The maximum number of flows in the table
+ * @param max_item_per_flow
+ *  The maximum number of packets per flow
+ *
+ * @return
+ *  - Return the table pointer on success.
+ *  - Return NULL on failure.
+ */
+void *gro_vxlan6_udp4_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow);
+
+/**
+ * This function destroys a IPv6 VxLAN reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ */
+void gro_vxlan6_udp4_tbl_destroy(void *tbl);
+
+/**
+ * This function merges a IPv6 VxLAN packet which has an outer IPv6 header and
+ * an inner UDP/IPv4 packet. It does not process the packet which does not
+ * have payload.
+ *
+ * This function does not check if the packet has correct checksums and
+ * does not re-calculate checksums for the merged packet. It returns the
+ * packet if there is no available space in the table.
+ *
+ * @param pkt
+ *  Packet to reassemble
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ * @start_time
+ *  The time when the packet is inserted into the table
+ *
+ * @return
+ *  - Return a positive value if the packet is merged.
+ *  - Return zero if the packet isn't merged but stored in the table.
+ *  - Return a negative value for invalid parameters or no available
+ *    space in the table.
+ */
+int32_t gro_vxlan6_udp4_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan6_udp4_tbl *tbl,
+		uint64_t start_time);
+
+/**
+ * This function flushes timeout packets in the IPv6 VxLAN reassembly table,
+ * and without updating checksums.
+ *
+ * @param tbl
+ *  Pointer pointing to a IPv6 VxLAN GRO table
+ * @param flush_timestamp
+ *  This function flushes packets which are inserted into the table
+ *  before or at the flush_timestamp.
+ * @param out
+ *  Pointer array used to keep flushed packets
+ * @param nb_out
+ *  The element number in 'out'. It also determines the maximum number of
+ *  packets that can be flushed finally.
+ *
+ * @return
+ *  The number of flushed packets
+ */
+uint16_t gro_vxlan6_udp4_tbl_timeout_flush(struct gro_vxlan6_udp4_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out);
+
+/**
+ * This function returns the number of the packets in a IPv6 VxLAN
+ * reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ *
+ * @return
+ *  The number of packets in the table
+ */
+uint32_t gro_vxlan6_udp4_tbl_pkt_count(void *tbl);
+#endif
diff --git a/lib/librte_gro/meson.build b/lib/librte_gro/meson.build
index 14b6b64..e23197d 100644
--- a/lib/librte_gro/meson.build
+++ b/lib/librte_gro/meson.build
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c', 'gro_udp6.c', 'gro_vxlan_udp6.c')
+sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c', 'gro_udp6.c', 'gro_vxlan_udp6.c', 'gro_vxlan6_udp4.c')
 headers = files('rte_gro.h')
 deps += ['ethdev']
diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
index 3bb7ea1..021a49e 100644
--- a/lib/librte_gro/rte_gro.c
+++ b/lib/librte_gro/rte_gro.c
@@ -14,6 +14,7 @@
 #include "gro_udp6.h"
 #include "gro_vxlan6_tcp4.h"
 #include "gro_vxlan6_tcp6.h"
+#include "gro_vxlan6_udp4.h"
 #include "gro_vxlan_tcp4.h"
 #include "gro_vxlan_tcp6.h"
 #include "gro_vxlan_udp4.h"
@@ -31,7 +32,7 @@
 		gro_tcp6_tbl_create, gro_vxlan_tcp6_tbl_create,
 		gro_vxlan6_tcp4_tbl_create, gro_vxlan6_tcp6_tbl_create,
 		gro_udp6_tbl_create, gro_vxlan_udp6_tbl_create,
-		NULL};
+		gro_vxlan6_udp4_tbl_create, NULL};
 static gro_tbl_destroy_fn tbl_destroy_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_destroy, gro_vxlan_tcp4_tbl_destroy,
 			gro_udp4_tbl_destroy, gro_vxlan_udp4_tbl_destroy,
@@ -39,7 +40,7 @@
 			gro_vxlan6_tcp4_tbl_destroy,
 			gro_vxlan6_tcp6_tbl_destroy,
 			gro_udp6_tbl_destroy, gro_vxlan_udp6_tbl_destroy,
-			NULL};
+			gro_vxlan6_udp4_tbl_destroy, NULL};
 static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_pkt_count, gro_vxlan_tcp4_tbl_pkt_count,
 			gro_udp4_tbl_pkt_count, gro_vxlan_udp4_tbl_pkt_count,
@@ -47,7 +48,7 @@
 			gro_vxlan6_tcp4_tbl_pkt_count,
 			gro_vxlan6_tcp6_tbl_pkt_count,
 			gro_udp6_tbl_pkt_count, gro_vxlan_udp6_tbl_pkt_count,
-			NULL};
+			gro_vxlan6_udp4_tbl_pkt_count, NULL};
 
 #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
@@ -143,6 +144,19 @@
 		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
 		  RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN)))
 
+#define IS_IPV6_VXLAN_UDP4_PKT(ptype) (RTE_ETH_IS_IPV6_HDR(ptype) && \
+		((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
+		((ptype & RTE_PTYPE_TUNNEL_VXLAN) == \
+		 RTE_PTYPE_TUNNEL_VXLAN) && \
+		((ptype & RTE_PTYPE_INNER_L4_UDP) == \
+		 RTE_PTYPE_INNER_L4_UDP) && \
+		(((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV4) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV4_EXT) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN)))
+
 /*
  * GRO context structure. It keeps the table structures, which are
  * used to merge packets, for different GRO types. Before using
@@ -279,6 +293,13 @@ struct gro_ctx {
 	struct gro_vxlan_udp6_item vxlan_udp6_items[RTE_GRO_MAX_BURST_ITEM_NUM]
 			= {{{0}, 0, 0} };
 
+	/* Allocate a reassembly table for IPv6 VXLAN UDP/IPv4 GRO */
+	struct gro_vxlan6_udp4_tbl vxlan6_udp4_tbl;
+	struct gro_vxlan6_udp4_flow
+		vxlan6_udp4_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+	struct gro_vxlan6_udp4_item
+		vxlan6_udp4_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{{0}} };
+
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	uint32_t item_num;
 	int32_t ret;
@@ -286,7 +307,8 @@ struct gro_ctx {
 	uint8_t do_tcp4_gro = 0, do_vxlan_tcp_gro = 0, do_udp4_gro = 0,
 		do_vxlan_udp_gro = 0, do_tcp6_gro = 0, do_vxlan_tcp6_gro = 0,
 		do_vxlan6_tcp4_gro = 0, do_vxlan6_tcp6_gro = 0,
-		do_udp6_gro = 0, do_vxlan_udp6_gro = 0;
+		do_udp6_gro = 0, do_vxlan_udp6_gro = 0,
+		do_vxlan6_udp4_gro = 0;
 
 	if (unlikely((param->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -297,7 +319,8 @@ struct gro_ctx {
 					RTE_GRO_IPV6_VXLAN_TCP_IPV4 |
 					RTE_GRO_IPV6_VXLAN_TCP_IPV6 |
 					RTE_GRO_UDP_IPV6 |
-					RTE_GRO_IPV4_VXLAN_UDP_IPV6)) == 0))
+					RTE_GRO_IPV4_VXLAN_UDP_IPV6 |
+					RTE_GRO_IPV6_VXLAN_UDP_IPV4)) == 0))
 		return nb_pkts;
 
 	/* Get the maximum number of packets */
@@ -435,6 +458,19 @@ struct gro_ctx {
 		do_vxlan_udp6_gro = 1;
 	}
 
+	if (param->gro_types & RTE_GRO_IPV6_VXLAN_UDP_IPV4) {
+		for (i = 0; i < item_num; i++)
+			vxlan6_udp4_flows[i].start_index = INVALID_ARRAY_INDEX;
+
+		vxlan6_udp4_tbl.flows = vxlan6_udp4_flows;
+		vxlan6_udp4_tbl.items = vxlan6_udp4_items;
+		vxlan6_udp4_tbl.flow_num = 0;
+		vxlan6_udp4_tbl.item_num = 0;
+		vxlan6_udp4_tbl.max_flow_num = item_num;
+		vxlan6_udp4_tbl.max_item_num = item_num;
+		do_vxlan6_udp4_gro = 1;
+	}
+
 	for (i = 0; i < nb_pkts; i++) {
 		/*
 		 * The timestamp is ignored, since all packets
@@ -526,6 +562,15 @@ struct gro_ctx {
 				nb_after_gro--;
 			else if (ret < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_VXLAN_UDP4_PKT(pkts[i]->packet_type) &&
+				do_vxlan6_udp4_gro) {
+			ret = gro_vxlan6_udp4_reassemble(pkts[i],
+							&vxlan6_udp4_tbl, 0);
+			if (ret > 0)
+				/* Merge successfully */
+				nb_after_gro--;
+			else if (ret < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -585,6 +630,12 @@ struct gro_ctx {
 					0, &pkts[i], nb_pkts - i);
 		}
 
+		if (do_vxlan6_udp4_gro) {
+			i += gro_vxlan6_udp4_tbl_timeout_flush(&vxlan6_udp4_tbl,
+					0, &pkts[i], nb_pkts - i);
+
+		}
+
 		/* Copy unprocessed packets */
 		if (unprocess_num > 0) {
 			memcpy(&pkts[i], unprocess_pkts,
@@ -606,12 +657,13 @@ struct gro_ctx {
 	struct gro_ctx *gro_ctx = ctx;
 	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl,
 		*vxlan_tcp6_tbl, *vxlan6_tcp4_tbl, *vxlan6_tcp6_tbl,
-		*udp6_tbl, *vxlan_udp6_tbl;
+		*udp6_tbl, *vxlan_udp6_tbl, *vxlan6_udp4_tbl;
 	uint64_t current_time;
 	uint16_t i, unprocess_num = 0;
 	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro,
 		do_tcp6_gro, do_vxlan_tcp6_gro, do_vxlan6_tcp4_gro,
-		do_vxlan6_tcp6_gro, do_udp6_gro, do_vxlan_udp6_gro;
+		do_vxlan6_tcp6_gro, do_udp6_gro, do_vxlan_udp6_gro,
+		do_vxlan6_udp4_gro;
 
 	if (unlikely((gro_ctx->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -622,7 +674,8 @@ struct gro_ctx {
 					RTE_GRO_IPV6_VXLAN_TCP_IPV4 |
 					RTE_GRO_IPV6_VXLAN_TCP_IPV6 |
 					RTE_GRO_UDP_IPV6 |
-					RTE_GRO_IPV4_VXLAN_UDP_IPV6)) == 0))
+					RTE_GRO_IPV4_VXLAN_UDP_IPV6 |
+					RTE_GRO_IPV6_VXLAN_UDP_IPV4)) == 0))
 		return nb_pkts;
 
 	tcp_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV4_INDEX];
@@ -635,6 +688,7 @@ struct gro_ctx {
 	vxlan6_tcp6_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_TCP_IPV6_INDEX];
 	udp6_tbl = gro_ctx->tbls[RTE_GRO_UDP_IPV6_INDEX];
 	vxlan_udp6_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX];
+	vxlan6_udp4_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_UDP_IPV4_INDEX];
 
 	do_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV4) ==
 		RTE_GRO_TCP_IPV4;
@@ -656,6 +710,8 @@ struct gro_ctx {
 		RTE_GRO_UDP_IPV6;
 	do_vxlan_udp6_gro = (gro_ctx->gro_types & RTE_GRO_IPV4_VXLAN_UDP_IPV6)
 				== RTE_GRO_IPV4_VXLAN_UDP_IPV6;
+	do_vxlan6_udp4_gro = (gro_ctx->gro_types & RTE_GRO_IPV6_VXLAN_UDP_IPV4)
+				== RTE_GRO_IPV6_VXLAN_UDP_IPV4;
 
 	current_time = rte_rdtsc();
 
@@ -710,6 +766,11 @@ struct gro_ctx {
 			if (gro_vxlan_udp6_reassemble(pkts[i], vxlan_udp6_tbl,
 						current_time) < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_VXLAN_UDP4_PKT(pkts[i]->packet_type) &&
+				do_vxlan6_udp4_gro) {
+			if (gro_vxlan6_udp4_reassemble(pkts[i], vxlan6_udp4_tbl,
+						current_time) < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -808,6 +869,13 @@ struct gro_ctx {
 		num += gro_vxlan_udp6_tbl_timeout_flush(gro_ctx->tbls[
 				RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX],
 				flush_timestamp, &out[num], left_nb_out);
+		left_nb_out = max_nb_out - num;
+	}
+
+	if ((gro_types & RTE_GRO_IPV6_VXLAN_UDP_IPV4) && left_nb_out > 0) {
+		num += gro_vxlan6_udp4_tbl_timeout_flush(gro_ctx->tbls[
+				RTE_GRO_IPV6_VXLAN_UDP_IPV4_INDEX],
+				flush_timestamp, &out[num], left_nb_out);
 	}
 
 	return num;
diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
index 1824ce3..ddcded6 100644
--- a/lib/librte_gro/rte_gro.h
+++ b/lib/librte_gro/rte_gro.h
@@ -56,6 +56,9 @@
 #define RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX 9
 #define RTE_GRO_IPV4_VXLAN_UDP_IPV6 (1ULL << RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX)
 /**< IPv4 VxLAN UDP/IPv6 GRO flag. */
+#define RTE_GRO_IPV6_VXLAN_UDP_IPV4_INDEX 10
+#define RTE_GRO_IPV6_VXLAN_UDP_IPV4 (1ULL << RTE_GRO_IPV6_VXLAN_UDP_IPV4_INDEX)
+/**< IPv6 VxLAN UDP/IPv4 GRO flag. */
 
 /**
  * Structure used to create GRO context objects or used to pass
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [dpdk-dev] [PATCH v1 8/8] gro: support IPv6 VXLAN UDP/IPv6
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
                   ` (6 preceding siblings ...)
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 7/8] gro: support IPv6 VXLAN UDP/IPv4 yang_y_yi
@ 2020-12-21  3:50 ` yang_y_yi
  2021-03-24 21:22 ` [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN Thomas Monjalon
  8 siblings, 0 replies; 13+ messages in thread
From: yang_y_yi @ 2020-12-21  3:50 UTC (permalink / raw)
  To: dev; +Cc: jiayu.hu, konstantin.ananyev, thomas, yangyi01, yang_y_yi

From: Yi Yang <yangyi01@inspur.com>

IPv6 VXLAN UDP/IPv6 GRO can help improve UDP/IPv6
performance in IPv6 VXLAN use case.

With this enabled in DPDK, OVS DPDK can leverage it
to improve VM-to-VM UDP/IPv6 performance, it will merge
small adjacent IPv6 VXLAN UDP/IPv6 fragments to a big
IPv6 VXLAN UDP/IPv6 packet immediate after they are
received from a physical NIC. It is very helpful in OVS
DPDK IPv6 VXLAN use case.

Signed-off-by: Yi Yang <yangyi01@inspur.com>
---
 .../prog_guide/generic_receive_offload_lib.rst     |   4 +-
 doc/guides/rel_notes/release_21_02.rst             |   5 +
 lib/librte_gro/gro_vxlan6_udp6.c                   | 589 +++++++++++++++++++++
 lib/librte_gro/gro_vxlan6_udp6.h                   | 148 ++++++
 lib/librte_gro/meson.build                         |   2 +-
 lib/librte_gro/rte_gro.c                           |  84 ++-
 lib/librte_gro/rte_gro.h                           |   3 +
 7 files changed, 824 insertions(+), 11 deletions(-)
 create mode 100644 lib/librte_gro/gro_vxlan6_udp6.c
 create mode 100644 lib/librte_gro/gro_vxlan6_udp6.h

diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst b/doc/guides/prog_guide/generic_receive_offload_lib.rst
index 8f48809..971323d 100644
--- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
+++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
@@ -33,8 +33,8 @@ fragmentation is possible (i.e., DF==0). Additionally, it complies RFC
 Currently, the GRO library provides GRO supports for TCP/IPv4, UDP/IPv4,
 UDP/IPv6, and TCP/IPv6 packets as well as VxLAN packets which contain an
 outer IPv4 header and an inner TCP/IPv4, UDP/IPv4, TCP/IPv6 or UDP/IPv6
-packet or an outer IPv6 header and an inner TCP/IPv4, UDP/IPv4 or TCP/IPv6
-packet.
+packet or an outer IPv6 header and an inner TCP/IPv4, UDP/IPv4, TCP/IPv6
+or UDP/IPv6 packet.
 
 Two Sets of API
 ---------------
diff --git a/doc/guides/rel_notes/release_21_02.rst b/doc/guides/rel_notes/release_21_02.rst
index 5e3d10b..991aa51 100644
--- a/doc/guides/rel_notes/release_21_02.rst
+++ b/doc/guides/rel_notes/release_21_02.rst
@@ -90,6 +90,11 @@ New Features
    Added UDP/IPv4 GRO support for IPv6 VXLAN packets, this enabled UDP/IPv4
    GRO support for IPv6 VXLAN use case.
 
+* **Added UDP/IPv6 GRO support for IPv6 VXLAN packets.**
+
+   Added UDP/IPv6 GRO support for IPv6 VXLAN packets, this enabled UDP/IPv6
+   GRO support for IPv6 VXLAN use case.
+
 Removed Items
 -------------
 
diff --git a/lib/librte_gro/gro_vxlan6_udp6.c b/lib/librte_gro/gro_vxlan6_udp6.c
new file mode 100644
index 0000000..c038e13
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan6_udp6.c
@@ -0,0 +1,589 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_cycles.h>
+#include <rte_ethdev.h>
+#include <rte_udp.h>
+#include <rte_ip.h>
+
+#include "gro_vxlan6_udp6.h"
+
+void *
+gro_vxlan6_udp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow)
+{
+	struct gro_vxlan6_udp6_tbl *tbl;
+	size_t size;
+	uint32_t entries_num, i;
+
+	entries_num = max_flow_num * max_item_per_flow;
+	entries_num = RTE_MIN(entries_num, GRO_VXLAN6_UDP6_TBL_MAX_ITEM_NUM);
+
+	if (entries_num == 0)
+		return NULL;
+
+	tbl = rte_zmalloc_socket(__func__,
+			sizeof(struct gro_vxlan6_udp6_tbl),
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl == NULL)
+		return NULL;
+
+	size = sizeof(struct gro_vxlan6_udp6_item) * entries_num;
+	tbl->items = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->items == NULL) {
+		rte_free(tbl);
+		return NULL;
+	}
+	tbl->max_item_num = entries_num;
+
+	size = sizeof(struct gro_vxlan6_udp6_flow) * entries_num;
+	tbl->flows = rte_zmalloc_socket(__func__,
+			size,
+			RTE_CACHE_LINE_SIZE,
+			socket_id);
+	if (tbl->flows == NULL) {
+		rte_free(tbl->items);
+		rte_free(tbl);
+		return NULL;
+	}
+
+	for (i = 0; i < entries_num; i++)
+		tbl->flows[i].start_index = INVALID_ARRAY_INDEX;
+	tbl->max_flow_num = entries_num;
+
+	return tbl;
+}
+
+void
+gro_vxlan6_udp6_tbl_destroy(void *tbl)
+{
+	struct gro_vxlan6_udp6_tbl *vxlan6_tbl = tbl;
+
+	if (vxlan6_tbl) {
+		rte_free(vxlan6_tbl->items);
+		rte_free(vxlan6_tbl->flows);
+	}
+	rte_free(vxlan6_tbl);
+}
+
+static inline uint32_t
+find_an_empty_item(struct gro_vxlan6_udp6_tbl *tbl)
+{
+	uint32_t max_item_num = tbl->max_item_num, i;
+
+	for (i = 0; i < max_item_num; i++)
+		if (tbl->items[i].inner_item.firstseg == NULL)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+find_an_empty_flow(struct gro_vxlan6_udp6_tbl *tbl)
+{
+	uint32_t max_flow_num = tbl->max_flow_num, i;
+
+	for (i = 0; i < max_flow_num; i++)
+		if (tbl->flows[i].start_index == INVALID_ARRAY_INDEX)
+			return i;
+	return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_item(struct gro_vxlan6_udp6_tbl *tbl,
+		struct rte_mbuf *pkt,
+		uint64_t start_time,
+		uint32_t prev_idx,
+		uint16_t frag_offset,
+		uint8_t is_last_frag)
+{
+	uint32_t item_idx;
+
+	item_idx = find_an_empty_item(tbl);
+	if (unlikely(item_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	tbl->items[item_idx].inner_item.firstseg = pkt;
+	tbl->items[item_idx].inner_item.lastseg = rte_pktmbuf_lastseg(pkt);
+	tbl->items[item_idx].inner_item.start_time = start_time;
+	tbl->items[item_idx].inner_item.next_pkt_idx = INVALID_ARRAY_INDEX;
+	tbl->items[item_idx].inner_item.frag_offset = frag_offset;
+	tbl->items[item_idx].inner_item.is_last_frag = is_last_frag;
+	tbl->items[item_idx].inner_item.nb_merged = 1;
+	tbl->item_num++;
+
+	/* If the previous packet exists, chain the new one with it. */
+	if (prev_idx != INVALID_ARRAY_INDEX) {
+		tbl->items[item_idx].inner_item.next_pkt_idx =
+			tbl->items[prev_idx].inner_item.next_pkt_idx;
+		tbl->items[prev_idx].inner_item.next_pkt_idx = item_idx;
+	}
+
+	return item_idx;
+}
+
+static inline uint32_t
+delete_item(struct gro_vxlan6_udp6_tbl *tbl,
+		uint32_t item_idx,
+		uint32_t prev_item_idx)
+{
+	uint32_t next_idx = tbl->items[item_idx].inner_item.next_pkt_idx;
+
+	/* NULL indicates an empty item. */
+	tbl->items[item_idx].inner_item.firstseg = NULL;
+	tbl->item_num--;
+	if (prev_item_idx != INVALID_ARRAY_INDEX)
+		tbl->items[prev_item_idx].inner_item.next_pkt_idx = next_idx;
+
+	return next_idx;
+}
+
+/* Copy IPv6 addr */
+static inline void gro_ipv6_addr_copy(const uint8_t *ipv6_from,
+				      uint8_t *ipv6_to)
+{
+	const uint64_t *from_words = (const uint64_t *)ipv6_from;
+	uint64_t *to_words   = (uint64_t *)ipv6_to;
+
+	to_words[0] = from_words[0];
+	to_words[1] = from_words[1];
+}
+
+static inline uint32_t
+insert_new_flow(struct gro_vxlan6_udp6_tbl *tbl,
+		struct vxlan6_udp6_flow_key *src,
+		uint32_t item_idx)
+{
+	struct vxlan6_udp6_flow_key *dst;
+	uint32_t flow_idx;
+
+	flow_idx = find_an_empty_flow(tbl);
+	if (unlikely(flow_idx == INVALID_ARRAY_INDEX))
+		return INVALID_ARRAY_INDEX;
+
+	dst = &(tbl->flows[flow_idx].key);
+
+	rte_ether_addr_copy(&(src->inner_key.eth_saddr),
+			&(dst->inner_key.eth_saddr));
+	rte_ether_addr_copy(&(src->inner_key.eth_daddr),
+			&(dst->inner_key.eth_daddr));
+	gro_ipv6_addr_copy(src->inner_key.ip_saddr, dst->inner_key.ip_saddr);
+	gro_ipv6_addr_copy(src->inner_key.ip_daddr, dst->inner_key.ip_daddr);
+	dst->inner_key.ip_id = src->inner_key.ip_id;
+
+	dst->vxlan_hdr.vx_flags = src->vxlan_hdr.vx_flags;
+	dst->vxlan_hdr.vx_vni = src->vxlan_hdr.vx_vni;
+	rte_ether_addr_copy(&(src->outer_eth_saddr), &(dst->outer_eth_saddr));
+	rte_ether_addr_copy(&(src->outer_eth_daddr), &(dst->outer_eth_daddr));
+	gro_ipv6_addr_copy(src->outer_ip_saddr, dst->outer_ip_saddr);
+	gro_ipv6_addr_copy(src->outer_ip_daddr, dst->outer_ip_daddr);
+	dst->outer_src_port = src->outer_src_port;
+	dst->outer_dst_port = src->outer_dst_port;
+
+	tbl->flows[flow_idx].start_index = item_idx;
+	tbl->flow_num++;
+
+	return flow_idx;
+}
+
+static inline int
+is_same_vxlan6_udp6_flow(struct vxlan6_udp6_flow_key k1,
+		struct vxlan6_udp6_flow_key k2)
+{
+	/* For VxLAN packet, outer udp src port is calculated from
+	 * inner packet RSS hash, udp src port of the first UDP
+	 * fragment is different from one of other UDP fragments
+	 * even if they are same flow, so we have to skip outer udp
+	 * src port comparison here.
+	 */
+	return (rte_is_same_ether_addr(&k1.outer_eth_saddr,
+					&k2.outer_eth_saddr) &&
+			rte_is_same_ether_addr(&k1.outer_eth_daddr,
+				&k2.outer_eth_daddr) &&
+			rte_is_same_ipv6_addr(k1.outer_ip_saddr,
+					      k2.outer_ip_saddr) &&
+			rte_is_same_ipv6_addr(k1.outer_ip_daddr,
+					      k2.outer_ip_daddr) &&
+			(k1.outer_dst_port == k2.outer_dst_port) &&
+			(k1.vxlan_hdr.vx_flags == k2.vxlan_hdr.vx_flags) &&
+			(k1.vxlan_hdr.vx_vni == k2.vxlan_hdr.vx_vni) &&
+			is_same_udp6_flow(k1.inner_key, k2.inner_key));
+}
+
+static inline int
+udp6_check_vxlan6_neighbor(struct gro_vxlan6_udp6_item *item,
+		uint16_t frag_offset,
+		uint16_t ip_dl)
+{
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	int cmp;
+	uint16_t l2_offset;
+	int ret = 0;
+
+	l2_offset = pkt->outer_l2_len + pkt->outer_l3_len;
+	cmp = udp6_check_neighbor(&item->inner_item, frag_offset,
+			ip_dl, l2_offset);
+	if (cmp > 0)
+		/* Append the new packet. */
+		ret = 1;
+	else if (cmp < 0)
+		/* Prepend the new packet. */
+		ret = -1;
+
+	return ret;
+}
+
+static inline int
+merge_two_vxlan6_udp6_packets(struct gro_vxlan6_udp6_item *item,
+		struct rte_mbuf *pkt,
+		int cmp,
+		uint16_t frag_offset,
+		uint8_t is_last_frag)
+{
+	if (merge_two_udp6_packets(&item->inner_item, pkt, cmp, frag_offset,
+				is_last_frag,
+				pkt->outer_l2_len + pkt->outer_l3_len)) {
+		return 1;
+	}
+
+	return 0;
+}
+
+static inline uint16_t
+get_ipv6_frag_offset(struct rte_ipv6_fragment_ext *ipv6_frag_hdr)
+{
+	return ((rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) >> 3) * 8);
+}
+
+static inline uint8_t
+is_last_ipv6_frag(struct rte_ipv6_fragment_ext *ipv6_frag_hdr)
+{
+	return (rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) & 0x0001);
+}
+
+static inline void
+update_vxlan6_header(struct gro_vxlan6_udp6_item *item)
+{
+	struct rte_ipv6_hdr *outer_ipv6_hdr, *ipv6_hdr;
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_ipv6_fragment_ext *ipv6_frag_hdr;
+	size_t fh_len = sizeof(*ipv6_frag_hdr);
+	struct rte_mbuf *pkt = item->inner_item.firstseg;
+	uint16_t len;
+
+	/* Position to inner IPv6 header first */
+	ipv6_hdr = (struct rte_ipv6_hdr *)rte_pktmbuf_mtod_offset(pkt, char *,
+			pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len);
+	len = pkt->pkt_len - pkt->outer_l2_len - pkt->outer_l3_len
+		- pkt->l2_len;
+	ipv6_hdr->payload_len = rte_cpu_to_be_16(len - pkt->l3_len);
+
+	/* Remove fragment extension header or clear MF flag */
+	if (item->inner_item.is_last_frag
+		&& (ipv6_hdr->proto == IPPROTO_FRAGMENT)) {
+		uint16_t ip_ofs;
+
+		ipv6_frag_hdr = (struct rte_ipv6_fragment_ext *)(ipv6_hdr + 1);
+		ip_ofs = get_ipv6_frag_offset(ipv6_frag_hdr);
+		if (ip_ofs == 0) {
+			ipv6_hdr->proto = ipv6_frag_hdr->next_header;
+			pkt->l3_len -= fh_len;
+
+			/* Remove IPv6 fragment extension header */
+			memmove(rte_pktmbuf_mtod_offset(pkt, char *, fh_len),
+				rte_pktmbuf_mtod(pkt, char*),
+				pkt->outer_l2_len + pkt->outer_l3_len +
+					pkt->l2_len + pkt->l3_len);
+			rte_pktmbuf_adj(pkt, fh_len);
+		} else {
+			/* clear MF flag */
+			ipv6_frag_hdr->frag_data = rte_cpu_to_be_16(
+				(rte_be_to_cpu_16(ipv6_frag_hdr->frag_data) &
+					0xFFFE));
+			ipv6_hdr->payload_len += fh_len;
+		}
+	}
+
+	/* Must adjust outer payload_len after inner IPv6 is handled. */
+	/* Update the outer IPv6 header. */
+	len = pkt->pkt_len - pkt->outer_l2_len;
+	outer_ipv6_hdr = (struct rte_ipv6_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+			pkt->outer_l2_len);
+	outer_ipv6_hdr->payload_len = rte_cpu_to_be_16(len
+							- pkt->outer_l3_len);
+
+	/* Point to the outer UDP header. */
+	len -= pkt->outer_l3_len;
+	udp_hdr = (struct rte_udp_hdr *)((char *)outer_ipv6_hdr
+						+ pkt->outer_l3_len);
+	udp_hdr->dgram_len = rte_cpu_to_be_16(len);
+}
+
+int32_t
+gro_vxlan6_udp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan6_udp6_tbl *tbl,
+		uint64_t start_time)
+{
+	struct rte_ether_hdr *outer_eth_hdr, *eth_hdr;
+	struct rte_ipv6_hdr *outer_ipv6_hdr;
+	struct rte_ipv6_hdr *ipv6_hdr;
+	struct rte_ipv6_fragment_ext *ipv6_frag_hdr;
+	uint16_t fh_len = sizeof(*ipv6_frag_hdr);
+	struct rte_udp_hdr *udp_hdr;
+	struct rte_vxlan_hdr *vxlan_hdr;
+	uint16_t frag_offset;
+	uint8_t is_last_frag;
+	int16_t ip_dl;
+	uint32_t ip_id;
+
+	struct vxlan6_udp6_flow_key key;
+	uint32_t cur_idx, prev_idx, item_idx;
+	uint32_t i, max_flow_num, remaining_flow_num;
+	int cmp;
+	uint16_t hdr_len;
+	uint8_t find;
+
+	outer_eth_hdr = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *);
+	outer_ipv6_hdr = (struct rte_ipv6_hdr *)((char *)outer_eth_hdr +
+			pkt->outer_l2_len);
+
+	udp_hdr = (struct rte_udp_hdr *)((char *)outer_ipv6_hdr +
+			pkt->outer_l3_len);
+	vxlan_hdr = (struct rte_vxlan_hdr *)((char *)udp_hdr +
+			sizeof(struct rte_udp_hdr));
+	eth_hdr = (struct rte_ether_hdr *)((char *)vxlan_hdr +
+			sizeof(struct rte_vxlan_hdr));
+	/* l2_len = outer udp hdr len + vxlan hdr len + inner l2 len */
+	ipv6_hdr = (struct rte_ipv6_hdr *)((char *)udp_hdr + pkt->l2_len);
+
+	/*
+	 * Don't process the packet which has non-fragment inner IP.
+	 */
+	if (ipv6_hdr->proto != IPPROTO_FRAGMENT)
+		return -1;
+
+	/* Note: l3_len includes length of extension headers */
+	hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+		pkt->l3_len;
+	/*
+	 * Don't process the packet whose payload length is less than or
+	 * equal to 0.
+	 */
+	if (pkt->pkt_len <= hdr_len)
+		return -1;
+
+	ipv6_frag_hdr = (struct rte_ipv6_fragment_ext *)(ipv6_hdr + 1);
+	ip_dl = rte_be_to_cpu_16(ipv6_hdr->payload_len) - fh_len;
+	ip_id = rte_be_to_cpu_32(ipv6_frag_hdr->id);
+	frag_offset = get_ipv6_frag_offset(ipv6_frag_hdr);
+	is_last_frag = is_last_ipv6_frag(ipv6_frag_hdr);
+
+	rte_ether_addr_copy(&(eth_hdr->s_addr), &(key.inner_key.eth_saddr));
+	rte_ether_addr_copy(&(eth_hdr->d_addr), &(key.inner_key.eth_daddr));
+	gro_ipv6_addr_copy(ipv6_hdr->src_addr, key.inner_key.ip_saddr);
+	gro_ipv6_addr_copy(ipv6_hdr->dst_addr, key.inner_key.ip_daddr);
+	key.inner_key.ip_id = ip_id;
+
+	key.vxlan_hdr.vx_flags = vxlan_hdr->vx_flags;
+	key.vxlan_hdr.vx_vni = vxlan_hdr->vx_vni;
+	rte_ether_addr_copy(&(outer_eth_hdr->s_addr), &(key.outer_eth_saddr));
+	rte_ether_addr_copy(&(outer_eth_hdr->d_addr), &(key.outer_eth_daddr));
+	gro_ipv6_addr_copy(outer_ipv6_hdr->src_addr, key.outer_ip_saddr);
+	gro_ipv6_addr_copy(outer_ipv6_hdr->dst_addr, key.outer_ip_daddr);
+	key.outer_src_port = udp_hdr->src_port;
+	key.outer_dst_port = udp_hdr->dst_port;
+
+	/* Search for a matched flow. */
+	max_flow_num = tbl->max_flow_num;
+	remaining_flow_num = tbl->flow_num;
+	find = 0;
+	for (i = 0; i < max_flow_num && remaining_flow_num; i++) {
+		if (tbl->flows[i].start_index != INVALID_ARRAY_INDEX) {
+			if (is_same_vxlan6_udp6_flow(tbl->flows[i].key, key)) {
+				find = 1;
+				break;
+			}
+			remaining_flow_num--;
+		}
+	}
+
+	/*
+	 * Can't find a matched flow. Insert a new flow and store the
+	 * packet into the flow.
+	 */
+	if (find == 0) {
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, frag_offset,
+				is_last_frag);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		if (insert_new_flow(tbl, &key, item_idx) ==
+				INVALID_ARRAY_INDEX) {
+			/*
+			 * Fail to insert a new flow, so
+			 * delete the inserted packet.
+			 */
+			delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+			return -1;
+		}
+		return 0;
+	}
+
+	/* Check all packets in the flow and try to find a neighbor. */
+	cur_idx = tbl->flows[i].start_index;
+	prev_idx = cur_idx;
+	do {
+		cmp = udp6_check_vxlan6_neighbor(&(tbl->items[cur_idx]),
+				frag_offset, ip_dl);
+		if (cmp) {
+			if (merge_two_vxlan6_udp6_packets(
+						&(tbl->items[cur_idx]),
+						pkt, cmp, frag_offset,
+						is_last_frag)) {
+				return 1;
+			}
+			/*
+			 * Can't merge two packets, as the packet
+			 * length will be greater than the max value.
+			 * Insert the packet into the flow.
+			 */
+			if (insert_new_item(tbl, pkt, start_time, prev_idx,
+						frag_offset, is_last_frag)
+				== INVALID_ARRAY_INDEX)
+				return -1;
+			return 0;
+		}
+
+		/* Ensure inserted items are ordered by frag_offset */
+		if (frag_offset
+			< tbl->items[cur_idx].inner_item.frag_offset) {
+			break;
+		}
+
+		prev_idx = cur_idx;
+		cur_idx = tbl->items[cur_idx].inner_item.next_pkt_idx;
+	} while (cur_idx != INVALID_ARRAY_INDEX);
+
+	/* Can't find neighbor. Insert the packet into the flow. */
+	if (cur_idx == tbl->flows[i].start_index) {
+		/* Insert it before the first packet of the flow */
+		item_idx = insert_new_item(tbl, pkt, start_time,
+				INVALID_ARRAY_INDEX, frag_offset,
+				is_last_frag);
+		if (item_idx == INVALID_ARRAY_INDEX)
+			return -1;
+		tbl->items[item_idx].inner_item.next_pkt_idx = cur_idx;
+		tbl->flows[i].start_index = item_idx;
+	} else {
+		if (insert_new_item(tbl, pkt, start_time, prev_idx,
+				frag_offset, is_last_frag)
+			== INVALID_ARRAY_INDEX)
+			return -1;
+	}
+
+	return 0;
+}
+
+static int
+gro_vxlan6_udp6_merge_items(struct gro_vxlan6_udp6_tbl *tbl,
+			   uint32_t start_idx)
+{
+	uint16_t frag_offset;
+	uint8_t is_last_frag;
+	int16_t ip_dl;
+	struct rte_mbuf *pkt;
+	int cmp;
+	uint32_t item_idx;
+	uint16_t hdr_len;
+
+	item_idx = tbl->items[start_idx].inner_item.next_pkt_idx;
+	while (item_idx != INVALID_ARRAY_INDEX) {
+		pkt = tbl->items[item_idx].inner_item.firstseg;
+		hdr_len = pkt->outer_l2_len + pkt->outer_l3_len + pkt->l2_len +
+			pkt->l3_len;
+		ip_dl = pkt->pkt_len - hdr_len;
+		frag_offset = tbl->items[item_idx].inner_item.frag_offset;
+		is_last_frag = tbl->items[item_idx].inner_item.is_last_frag;
+		cmp = udp6_check_vxlan6_neighbor(&(tbl->items[start_idx]),
+					frag_offset, ip_dl);
+		if (cmp) {
+			if (merge_two_vxlan6_udp6_packets(
+					&(tbl->items[start_idx]),
+					pkt, cmp, frag_offset,
+					is_last_frag)) {
+				item_idx = delete_item(tbl, item_idx,
+							INVALID_ARRAY_INDEX);
+				tbl->items[start_idx].inner_item.next_pkt_idx
+					= item_idx;
+			} else {
+				return 0;
+			}
+		} else {
+			return 0;
+		}
+	}
+
+	return 0;
+}
+
+uint16_t
+gro_vxlan6_udp6_tbl_timeout_flush(struct gro_vxlan6_udp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out)
+{
+	uint16_t k = 0;
+	uint32_t i, j;
+	uint32_t max_flow_num = tbl->max_flow_num;
+
+	for (i = 0; i < max_flow_num; i++) {
+		if (unlikely(tbl->flow_num == 0))
+			return k;
+
+		j = tbl->flows[i].start_index;
+		while (j != INVALID_ARRAY_INDEX) {
+			if (tbl->items[j].inner_item.start_time <=
+					flush_timestamp) {
+				gro_vxlan6_udp6_merge_items(tbl, j);
+				out[k++] = tbl->items[j].inner_item.firstseg;
+				if (tbl->items[j].inner_item.nb_merged > 1)
+					update_vxlan6_header(&(tbl->items[j]));
+				/*
+				 * Delete the item and get the next packet
+				 * index.
+				 */
+				j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+				tbl->flows[i].start_index = j;
+				if (j == INVALID_ARRAY_INDEX)
+					tbl->flow_num--;
+
+				if (unlikely(k == nb_out))
+					return k;
+			} else
+				/*
+				 * The left packets in the flow won't be
+				 * timeout. Go to check other flows.
+				 */
+				break;
+		}
+	}
+	return k;
+}
+
+uint32_t
+gro_vxlan6_udp6_tbl_pkt_count(void *tbl)
+{
+	struct gro_vxlan6_udp6_tbl *gro_tbl = tbl;
+
+	if (gro_tbl)
+		return gro_tbl->item_num;
+
+	return 0;
+}
diff --git a/lib/librte_gro/gro_vxlan6_udp6.h b/lib/librte_gro/gro_vxlan6_udp6.h
new file mode 100644
index 0000000..7f3a8e8
--- /dev/null
+++ b/lib/librte_gro/gro_vxlan6_udp6.h
@@ -0,0 +1,148 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2021 Inspur Corporation
+ */
+
+#ifndef _GRO_VXLAN6_UDP6_H_
+#define _GRO_VXLAN6_UDP6_H_
+
+#include "gro_udp6.h"
+
+#define GRO_VXLAN6_UDP6_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/* Header fields representing a IPv6 VxLAN flow */
+struct vxlan6_udp6_flow_key {
+	struct udp6_flow_key inner_key;
+	struct rte_vxlan_hdr vxlan_hdr;
+
+	struct rte_ether_addr outer_eth_saddr;
+	struct rte_ether_addr outer_eth_daddr;
+
+	uint8_t outer_ip_saddr[16];
+	uint8_t outer_ip_daddr[16];
+
+	/* Outer UDP ports */
+	uint16_t outer_src_port;
+	uint16_t outer_dst_port;
+
+};
+
+struct gro_vxlan6_udp6_flow {
+	struct vxlan6_udp6_flow_key key;
+	/*
+	 * The index of the first packet in the flow. INVALID_ARRAY_INDEX
+	 * indicates an empty flow.
+	 */
+	uint32_t start_index;
+};
+
+struct gro_vxlan6_udp6_item {
+	struct gro_udp6_item inner_item;
+};
+
+/*
+ * IPv6 VxLAN (with an outer IPv6 header and an inner UDP/IPv6 packet)
+ * reassembly table structure
+ */
+struct gro_vxlan6_udp6_tbl {
+	/* item array */
+	struct gro_vxlan6_udp6_item *items;
+	/* flow array */
+	struct gro_vxlan6_udp6_flow *flows;
+	/* current item number */
+	uint32_t item_num;
+	/* current flow number */
+	uint32_t flow_num;
+	/* the maximum item number */
+	uint32_t max_item_num;
+	/* the maximum flow number */
+	uint32_t max_flow_num;
+};
+
+/**
+ * This function creates a IPv6 VxLAN reassembly table for IPv6 VxLAN packets
+ * which have an outer IPv6 header and an inner UDP/IPv6 packet.
+ *
+ * @param socket_id
+ *  Socket index for allocating the table
+ * @param max_flow_num
+ *  The maximum number of flows in the table
+ * @param max_item_per_flow
+ *  The maximum number of packets per flow
+ *
+ * @return
+ *  - Return the table pointer on success.
+ *  - Return NULL on failure.
+ */
+void *gro_vxlan6_udp6_tbl_create(uint16_t socket_id,
+		uint16_t max_flow_num,
+		uint16_t max_item_per_flow);
+
+/**
+ * This function destroys a IPv6 VxLAN reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ */
+void gro_vxlan6_udp6_tbl_destroy(void *tbl);
+
+/**
+ * This function merges a IPv6 VxLAN packet which has an outer IPv6 header and
+ * an inner UDP/IPv6 packet. It does not process the packet which does not
+ * have payload.
+ *
+ * This function does not check if the packet has correct checksums and
+ * does not re-calculate checksums for the merged packet. It returns the
+ * packet if there is no available space in the table.
+ *
+ * @param pkt
+ *  Packet to reassemble
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ * @start_time
+ *  The time when the packet is inserted into the table
+ *
+ * @return
+ *  - Return a positive value if the packet is merged.
+ *  - Return zero if the packet isn't merged but stored in the table.
+ *  - Return a negative value for invalid parameters or no available
+ *    space in the table.
+ */
+int32_t gro_vxlan6_udp6_reassemble(struct rte_mbuf *pkt,
+		struct gro_vxlan6_udp6_tbl *tbl,
+		uint64_t start_time);
+
+/**
+ * This function flushes timeout packets in the IPv6 VxLAN reassembly table,
+ * and without updating checksums.
+ *
+ * @param tbl
+ *  Pointer pointing to a IPv6 VxLAN GRO table
+ * @param flush_timestamp
+ *  This function flushes packets which are inserted into the table
+ *  before or at the flush_timestamp.
+ * @param out
+ *  Pointer array used to keep flushed packets
+ * @param nb_out
+ *  The element number in 'out'. It also determines the maximum number of
+ *  packets that can be flushed finally.
+ *
+ * @return
+ *  The number of flushed packets
+ */
+uint16_t gro_vxlan6_udp6_tbl_timeout_flush(struct gro_vxlan6_udp6_tbl *tbl,
+		uint64_t flush_timestamp,
+		struct rte_mbuf **out,
+		uint16_t nb_out);
+
+/**
+ * This function returns the number of the packets in a IPv6 VxLAN
+ * reassembly table.
+ *
+ * @param tbl
+ *  Pointer pointing to the IPv6 VxLAN reassembly table
+ *
+ * @return
+ *  The number of packets in the table
+ */
+uint32_t gro_vxlan6_udp6_tbl_pkt_count(void *tbl);
+#endif
diff --git a/lib/librte_gro/meson.build b/lib/librte_gro/meson.build
index e23197d..56d015a 100644
--- a/lib/librte_gro/meson.build
+++ b/lib/librte_gro/meson.build
@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c', 'gro_udp6.c', 'gro_vxlan_udp6.c', 'gro_vxlan6_udp4.c')
+sources = files('rte_gro.c', 'gro_tcp4.c', 'gro_udp4.c', 'gro_vxlan_tcp4.c', 'gro_vxlan_udp4.c', 'gro_tcp6.c', 'gro_vxlan_tcp6.c', 'gro_vxlan6_tcp4.c', 'gro_vxlan6_tcp6.c', 'gro_udp6.c', 'gro_vxlan_udp6.c', 'gro_vxlan6_udp4.c', 'gro_vxlan6_udp6.c')
 headers = files('rte_gro.h')
 deps += ['ethdev']
diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
index 021a49e..abe3911 100644
--- a/lib/librte_gro/rte_gro.c
+++ b/lib/librte_gro/rte_gro.c
@@ -15,6 +15,7 @@
 #include "gro_vxlan6_tcp4.h"
 #include "gro_vxlan6_tcp6.h"
 #include "gro_vxlan6_udp4.h"
+#include "gro_vxlan6_udp6.h"
 #include "gro_vxlan_tcp4.h"
 #include "gro_vxlan_tcp6.h"
 #include "gro_vxlan_udp4.h"
@@ -32,7 +33,8 @@
 		gro_tcp6_tbl_create, gro_vxlan_tcp6_tbl_create,
 		gro_vxlan6_tcp4_tbl_create, gro_vxlan6_tcp6_tbl_create,
 		gro_udp6_tbl_create, gro_vxlan_udp6_tbl_create,
-		gro_vxlan6_udp4_tbl_create, NULL};
+		gro_vxlan6_udp4_tbl_create, gro_vxlan6_udp6_tbl_create,
+		NULL};
 static gro_tbl_destroy_fn tbl_destroy_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_destroy, gro_vxlan_tcp4_tbl_destroy,
 			gro_udp4_tbl_destroy, gro_vxlan_udp4_tbl_destroy,
@@ -40,7 +42,8 @@
 			gro_vxlan6_tcp4_tbl_destroy,
 			gro_vxlan6_tcp6_tbl_destroy,
 			gro_udp6_tbl_destroy, gro_vxlan_udp6_tbl_destroy,
-			gro_vxlan6_udp4_tbl_destroy, NULL};
+			gro_vxlan6_udp4_tbl_destroy,
+			gro_vxlan6_udp6_tbl_destroy, NULL};
 static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
 			gro_tcp4_tbl_pkt_count, gro_vxlan_tcp4_tbl_pkt_count,
 			gro_udp4_tbl_pkt_count, gro_vxlan_udp4_tbl_pkt_count,
@@ -48,7 +51,8 @@
 			gro_vxlan6_tcp4_tbl_pkt_count,
 			gro_vxlan6_tcp6_tbl_pkt_count,
 			gro_udp6_tbl_pkt_count, gro_vxlan_udp6_tbl_pkt_count,
-			gro_vxlan6_udp4_tbl_pkt_count, NULL};
+			gro_vxlan6_udp4_tbl_pkt_count,
+			gro_vxlan6_udp6_tbl_pkt_count, NULL};
 
 #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
 		((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
@@ -157,6 +161,19 @@
 		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
 		  RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN)))
 
+#define IS_IPV6_VXLAN_UDP6_PKT(ptype) (RTE_ETH_IS_IPV6_HDR(ptype) && \
+		((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
+		((ptype & RTE_PTYPE_TUNNEL_VXLAN) == \
+		 RTE_PTYPE_TUNNEL_VXLAN) && \
+		((ptype & RTE_PTYPE_INNER_L4_UDP) == \
+		 RTE_PTYPE_INNER_L4_UDP) && \
+		(((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6_EXT) || \
+		 ((ptype & RTE_PTYPE_INNER_L3_MASK) == \
+		  RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN)))
+
 /*
  * GRO context structure. It keeps the table structures, which are
  * used to merge packets, for different GRO types. Before using
@@ -300,6 +317,13 @@ struct gro_ctx {
 	struct gro_vxlan6_udp4_item
 		vxlan6_udp4_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{{0}} };
 
+	/* Allocate a reassembly table for IPv6 VXLAN UDP/IPv6 GRO */
+	struct gro_vxlan6_udp6_tbl vxlan6_udp6_tbl;
+	struct gro_vxlan6_udp6_flow
+		vxlan6_udp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+	struct gro_vxlan6_udp6_item
+		vxlan6_udp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{{0}} };
+
 	struct rte_mbuf *unprocess_pkts[nb_pkts];
 	uint32_t item_num;
 	int32_t ret;
@@ -308,7 +332,7 @@ struct gro_ctx {
 		do_vxlan_udp_gro = 0, do_tcp6_gro = 0, do_vxlan_tcp6_gro = 0,
 		do_vxlan6_tcp4_gro = 0, do_vxlan6_tcp6_gro = 0,
 		do_udp6_gro = 0, do_vxlan_udp6_gro = 0,
-		do_vxlan6_udp4_gro = 0;
+		do_vxlan6_udp4_gro = 0, do_vxlan6_udp6_gro = 0;
 
 	if (unlikely((param->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -320,7 +344,8 @@ struct gro_ctx {
 					RTE_GRO_IPV6_VXLAN_TCP_IPV6 |
 					RTE_GRO_UDP_IPV6 |
 					RTE_GRO_IPV4_VXLAN_UDP_IPV6 |
-					RTE_GRO_IPV6_VXLAN_UDP_IPV4)) == 0))
+					RTE_GRO_IPV6_VXLAN_UDP_IPV4 |
+					RTE_GRO_IPV6_VXLAN_UDP_IPV6)) == 0))
 		return nb_pkts;
 
 	/* Get the maximum number of packets */
@@ -471,6 +496,19 @@ struct gro_ctx {
 		do_vxlan6_udp4_gro = 1;
 	}
 
+	if (param->gro_types & RTE_GRO_IPV6_VXLAN_UDP_IPV6) {
+		for (i = 0; i < item_num; i++)
+			vxlan6_udp6_flows[i].start_index = INVALID_ARRAY_INDEX;
+
+		vxlan6_udp6_tbl.flows = vxlan6_udp6_flows;
+		vxlan6_udp6_tbl.items = vxlan6_udp6_items;
+		vxlan6_udp6_tbl.flow_num = 0;
+		vxlan6_udp6_tbl.item_num = 0;
+		vxlan6_udp6_tbl.max_flow_num = item_num;
+		vxlan6_udp6_tbl.max_item_num = item_num;
+		do_vxlan6_udp6_gro = 1;
+	}
+
 	for (i = 0; i < nb_pkts; i++) {
 		/*
 		 * The timestamp is ignored, since all packets
@@ -571,6 +609,15 @@ struct gro_ctx {
 				nb_after_gro--;
 			else if (ret < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_VXLAN_UDP6_PKT(pkts[i]->packet_type) &&
+				do_vxlan6_udp6_gro) {
+			ret = gro_vxlan6_udp6_reassemble(pkts[i],
+							&vxlan6_udp6_tbl, 0);
+			if (ret > 0)
+				/* Merge successfully */
+				nb_after_gro--;
+			else if (ret < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -633,7 +680,11 @@ struct gro_ctx {
 		if (do_vxlan6_udp4_gro) {
 			i += gro_vxlan6_udp4_tbl_timeout_flush(&vxlan6_udp4_tbl,
 					0, &pkts[i], nb_pkts - i);
+		}
 
+		if (do_vxlan6_udp6_gro) {
+			i += gro_vxlan6_udp6_tbl_timeout_flush(&vxlan6_udp6_tbl,
+					0, &pkts[i], nb_pkts - i);
 		}
 
 		/* Copy unprocessed packets */
@@ -657,13 +708,14 @@ struct gro_ctx {
 	struct gro_ctx *gro_ctx = ctx;
 	void *tcp_tbl, *udp_tbl, *vxlan_tcp_tbl, *vxlan_udp_tbl, *tcp6_tbl,
 		*vxlan_tcp6_tbl, *vxlan6_tcp4_tbl, *vxlan6_tcp6_tbl,
-		*udp6_tbl, *vxlan_udp6_tbl, *vxlan6_udp4_tbl;
+		*udp6_tbl, *vxlan_udp6_tbl, *vxlan6_udp4_tbl,
+		*vxlan6_udp6_tbl;
 	uint64_t current_time;
 	uint16_t i, unprocess_num = 0;
 	uint8_t do_tcp4_gro, do_vxlan_tcp_gro, do_udp4_gro, do_vxlan_udp_gro,
 		do_tcp6_gro, do_vxlan_tcp6_gro, do_vxlan6_tcp4_gro,
 		do_vxlan6_tcp6_gro, do_udp6_gro, do_vxlan_udp6_gro,
-		do_vxlan6_udp4_gro;
+		do_vxlan6_udp4_gro, do_vxlan6_udp6_gro;
 
 	if (unlikely((gro_ctx->gro_types & (RTE_GRO_IPV4_VXLAN_TCP_IPV4 |
 					RTE_GRO_TCP_IPV4 |
@@ -675,7 +727,8 @@ struct gro_ctx {
 					RTE_GRO_IPV6_VXLAN_TCP_IPV6 |
 					RTE_GRO_UDP_IPV6 |
 					RTE_GRO_IPV4_VXLAN_UDP_IPV6 |
-					RTE_GRO_IPV6_VXLAN_UDP_IPV4)) == 0))
+					RTE_GRO_IPV6_VXLAN_UDP_IPV4 |
+					RTE_GRO_IPV6_VXLAN_UDP_IPV6)) == 0))
 		return nb_pkts;
 
 	tcp_tbl = gro_ctx->tbls[RTE_GRO_TCP_IPV4_INDEX];
@@ -689,6 +742,7 @@ struct gro_ctx {
 	udp6_tbl = gro_ctx->tbls[RTE_GRO_UDP_IPV6_INDEX];
 	vxlan_udp6_tbl = gro_ctx->tbls[RTE_GRO_IPV4_VXLAN_UDP_IPV6_INDEX];
 	vxlan6_udp4_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_UDP_IPV4_INDEX];
+	vxlan6_udp6_tbl = gro_ctx->tbls[RTE_GRO_IPV6_VXLAN_UDP_IPV6_INDEX];
 
 	do_tcp4_gro = (gro_ctx->gro_types & RTE_GRO_TCP_IPV4) ==
 		RTE_GRO_TCP_IPV4;
@@ -712,6 +766,8 @@ struct gro_ctx {
 				== RTE_GRO_IPV4_VXLAN_UDP_IPV6;
 	do_vxlan6_udp4_gro = (gro_ctx->gro_types & RTE_GRO_IPV6_VXLAN_UDP_IPV4)
 				== RTE_GRO_IPV6_VXLAN_UDP_IPV4;
+	do_vxlan6_udp6_gro = (gro_ctx->gro_types & RTE_GRO_IPV6_VXLAN_UDP_IPV6)
+				== RTE_GRO_IPV6_VXLAN_UDP_IPV6;
 
 	current_time = rte_rdtsc();
 
@@ -771,6 +827,11 @@ struct gro_ctx {
 			if (gro_vxlan6_udp4_reassemble(pkts[i], vxlan6_udp4_tbl,
 						current_time) < 0)
 				unprocess_pkts[unprocess_num++] = pkts[i];
+		} else if (IS_IPV6_VXLAN_UDP6_PKT(pkts[i]->packet_type) &&
+				do_vxlan6_udp6_gro) {
+			if (gro_vxlan6_udp6_reassemble(pkts[i], vxlan6_udp6_tbl,
+						current_time) < 0)
+				unprocess_pkts[unprocess_num++] = pkts[i];
 		} else
 			unprocess_pkts[unprocess_num++] = pkts[i];
 	}
@@ -876,6 +937,13 @@ struct gro_ctx {
 		num += gro_vxlan6_udp4_tbl_timeout_flush(gro_ctx->tbls[
 				RTE_GRO_IPV6_VXLAN_UDP_IPV4_INDEX],
 				flush_timestamp, &out[num], left_nb_out);
+		left_nb_out = max_nb_out - num;
+	}
+
+	if ((gro_types & RTE_GRO_IPV6_VXLAN_UDP_IPV6) && left_nb_out > 0) {
+		num += gro_vxlan6_udp6_tbl_timeout_flush(gro_ctx->tbls[
+				RTE_GRO_IPV6_VXLAN_UDP_IPV6_INDEX],
+				flush_timestamp, &out[num], left_nb_out);
 	}
 
 	return num;
diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
index ddcded6..fefa2e9 100644
--- a/lib/librte_gro/rte_gro.h
+++ b/lib/librte_gro/rte_gro.h
@@ -59,6 +59,9 @@
 #define RTE_GRO_IPV6_VXLAN_UDP_IPV4_INDEX 10
 #define RTE_GRO_IPV6_VXLAN_UDP_IPV4 (1ULL << RTE_GRO_IPV6_VXLAN_UDP_IPV4_INDEX)
 /**< IPv6 VxLAN UDP/IPv4 GRO flag. */
+#define RTE_GRO_IPV6_VXLAN_UDP_IPV6_INDEX 11
+#define RTE_GRO_IPV6_VXLAN_UDP_IPV6 (1ULL << RTE_GRO_IPV6_VXLAN_UDP_IPV6_INDEX)
+/**< IPv6 VxLAN UDP/IPv6 GRO flag. */
 
 /**
  * Structure used to create GRO context objects or used to pass
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN
  2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
                   ` (7 preceding siblings ...)
  2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 8/8] gro: support IPv6 VXLAN UDP/IPv6 yang_y_yi
@ 2021-03-24 21:22 ` Thomas Monjalon
  2021-07-24  8:48   ` Thomas Monjalon
  8 siblings, 1 reply; 13+ messages in thread
From: Thomas Monjalon @ 2021-03-24 21:22 UTC (permalink / raw)
  To: yang_y_yi; +Cc: dev, jiayu.hu, konstantin.ananyev, yangyi01

21/12/2020 04:50, yang_y_yi@163.com:
> Yi Yang (8):
>   gro: support TCP/IPv6
>   gro: support IPv4 VXLAN TCP/IPv6
>   gro: support IPv6 VXLAN TCP/IPv4
>   gro: support IPv6 VXLAN TCP/IPv6
>   gro: support UDP/IPv6
>   gro: support IPv4 VXLAN UDP/IPv6
>   gro: support IPv6 VXLAN UDP/IPv4
>   gro: support IPv6 VXLAN UDP/IPv6

There was no review/activity on this thread.
Is it abandoned?



^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN
  2021-03-24 21:22 ` [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN Thomas Monjalon
@ 2021-07-24  8:48   ` Thomas Monjalon
  2021-07-25 23:53     ` Hu, Jiayu
  0 siblings, 1 reply; 13+ messages in thread
From: Thomas Monjalon @ 2021-07-24  8:48 UTC (permalink / raw)
  To: yang_y_yi, jiayu.hu; +Cc: dev, konstantin.ananyev, yangyi01

24/03/2021 22:22, Thomas Monjalon:
> 21/12/2020 04:50, yang_y_yi@163.com:
> > Yi Yang (8):
> >   gro: support TCP/IPv6
> >   gro: support IPv4 VXLAN TCP/IPv6
> >   gro: support IPv6 VXLAN TCP/IPv4
> >   gro: support IPv6 VXLAN TCP/IPv6
> >   gro: support UDP/IPv6
> >   gro: support IPv4 VXLAN UDP/IPv6
> >   gro: support IPv6 VXLAN UDP/IPv4
> >   gro: support IPv6 VXLAN UDP/IPv6
> 
> There was no review/activity on this thread.
> Is it abandoned?

No reply. Should I assume it can be dropped?



^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN
  2021-07-24  8:48   ` Thomas Monjalon
@ 2021-07-25 23:53     ` Hu, Jiayu
  2023-06-12  2:11       ` Stephen Hemminger
  0 siblings, 1 reply; 13+ messages in thread
From: Hu, Jiayu @ 2021-07-25 23:53 UTC (permalink / raw)
  To: Thomas Monjalon, yang_y_yi; +Cc: dev, Ananyev, Konstantin, yangyi01



> -----Original Message-----
> From: Thomas Monjalon <thomas@monjalon.net>
> Sent: Saturday, July 24, 2021 4:49 PM
> To: yang_y_yi@163.com; Hu, Jiayu <jiayu.hu@intel.com>
> Cc: dev@dpdk.org; Ananyev, Konstantin <konstantin.ananyev@intel.com>;
> yangyi01@inspur.com
> Subject: Re: [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6
> for VLAN and VXLAN
> 
> 24/03/2021 22:22, Thomas Monjalon:
> > 21/12/2020 04:50, yang_y_yi@163.com:
> > > Yi Yang (8):
> > >   gro: support TCP/IPv6
> > >   gro: support IPv4 VXLAN TCP/IPv6
> > >   gro: support IPv6 VXLAN TCP/IPv4
> > >   gro: support IPv6 VXLAN TCP/IPv6
> > >   gro: support UDP/IPv6
> > >   gro: support IPv4 VXLAN UDP/IPv6
> > >   gro: support IPv6 VXLAN UDP/IPv4
> > >   gro: support IPv6 VXLAN UDP/IPv6
> >
> > There was no review/activity on this thread.
> > Is it abandoned?
> 
> No reply. Should I assume it can be dropped?

No further updates from Yi on this patch set. So I guess "yes"?

Thanks,
Jiayu
> 


^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN
  2021-07-25 23:53     ` Hu, Jiayu
@ 2023-06-12  2:11       ` Stephen Hemminger
  0 siblings, 0 replies; 13+ messages in thread
From: Stephen Hemminger @ 2023-06-12  2:11 UTC (permalink / raw)
  To: Hu, Jiayu; +Cc: Thomas Monjalon, yang_y_yi, dev, Ananyev, Konstantin, yangyi01

On Sun, 25 Jul 2021 23:53:28 +0000
"Hu, Jiayu" <jiayu.hu@intel.com> wrote:

> > -----Original Message-----
> > From: Thomas Monjalon <thomas@monjalon.net>
> > Sent: Saturday, July 24, 2021 4:49 PM
> > To: yang_y_yi@163.com; Hu, Jiayu <jiayu.hu@intel.com>
> > Cc: dev@dpdk.org; Ananyev, Konstantin <konstantin.ananyev@intel.com>;
> > yangyi01@inspur.com
> > Subject: Re: [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6
> > for VLAN and VXLAN
> > 
> > 24/03/2021 22:22, Thomas Monjalon:  
> > > 21/12/2020 04:50, yang_y_yi@163.com:  
> > > > Yi Yang (8):
> > > >   gro: support TCP/IPv6
> > > >   gro: support IPv4 VXLAN TCP/IPv6
> > > >   gro: support IPv6 VXLAN TCP/IPv4
> > > >   gro: support IPv6 VXLAN TCP/IPv6
> > > >   gro: support UDP/IPv6
> > > >   gro: support IPv4 VXLAN UDP/IPv6
> > > >   gro: support IPv6 VXLAN UDP/IPv4
> > > >   gro: support IPv6 VXLAN UDP/IPv6  
> > >
> > > There was no review/activity on this thread.
> > > Is it abandoned?  
> > 
> > No reply. Should I assume it can be dropped?  
> 
> No further updates from Yi on this patch set. So I guess "yes"?

Ok, patch marked rejected.
If there is a use case, then a new maintainer can find it in archive
and work on it.

What was missing was a clear use case, documentation and testing.

^ permalink raw reply	[flat|nested] 13+ messages in thread

end of thread, other threads:[~2023-06-12  2:11 UTC | newest]

Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-12-21  3:50 [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN yang_y_yi
2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 1/8] gro: support TCP/IPv6 yang_y_yi
2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 2/8] gro: support IPv4 VXLAN TCP/IPv6 yang_y_yi
2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 3/8] gro: support IPv6 VXLAN TCP/IPv4 yang_y_yi
2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 4/8] gro: support IPv6 VXLAN TCP/IPv6 yang_y_yi
2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 5/8] gro: support UDP/IPv6 yang_y_yi
2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 6/8] gro: support IPv4 VXLAN UDP/IPv6 yang_y_yi
2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 7/8] gro: support IPv6 VXLAN UDP/IPv4 yang_y_yi
2020-12-21  3:50 ` [dpdk-dev] [PATCH v1 8/8] gro: support IPv6 VXLAN UDP/IPv6 yang_y_yi
2021-03-24 21:22 ` [dpdk-dev] [PATCH v1 0/8] gro: support TCP/IPv6 and UDP/IPv6 for VLAN and VXLAN Thomas Monjalon
2021-07-24  8:48   ` Thomas Monjalon
2021-07-25 23:53     ` Hu, Jiayu
2023-06-12  2:11       ` Stephen Hemminger

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).