[v9] gro : ipv6 changes to support GRO for TCP/ipv6
Checks
Commit Message
The patch adds GRO support for TCP/ipv6 packets. This does not
include the support for vxlan, udp ipv6 packets.
Signed-off-by: Kumara Parameshwaran <kumaraparamesh92@gmail.com>
---
v1:
* Changes to support GRO for TCP/ipv6 packets. This does not include
vxlan changes.
* The GRO is performed only for ipv6 packets that does not contain
extension headers.
* The logic for the TCP coalescing remains the same, in ipv6 header
the source address, destination address, flow label, version fields
are expected to be the same.
* Re-organised the code to reuse certain tcp functions for both ipv4 and
ipv6 flows.
v2:
* Fix comments in gro_tcp6.h header file.
v3:
* Adderess review comments to fix code duplication for v4 and v6
v4:
* Addresses review comments for v3, do not use callbacks
v5:
* Address review comments
v6:
* Fix warning and coding style issues
v7:
* Fix build compilation issue
v8:
* Use gro_tcp_internal.h for functions used for gro_tcp4 and gro_tcp6 and use
gro_tcp.h for data structures and functions used across gro_vxlan_tcp4
v9:
* Resolve merge conflict and add gro_tcp.h in proper path
.../generic_receive_offload_lib.rst | 21 +-
doc/guides/rel_notes/release_23_07.rst | 2 +
lib/gro/gro_tcp.h | 187 ++++++++++++
lib/gro/gro_tcp4.c | 177 +++---------
lib/gro/gro_tcp4.h | 170 +----------
lib/gro/gro_tcp6.c | 268 ++++++++++++++++++
lib/gro/gro_tcp6.h | 160 +++++++++++
lib/gro/gro_tcp_internal.h | 128 +++++++++
lib/gro/gro_vxlan_tcp4.c | 22 +-
lib/gro/gro_vxlan_tcp4.h | 2 +-
lib/gro/meson.build | 1 +
lib/gro/rte_gro.c | 83 +++++-
lib/gro/rte_gro.h | 3 +
13 files changed, 883 insertions(+), 341 deletions(-)
create mode 100644 lib/gro/gro_tcp.h
create mode 100644 lib/gro/gro_tcp6.c
create mode 100644 lib/gro/gro_tcp6.h
create mode 100644 lib/gro/gro_tcp_internal.h
Comments
On Thu, Jun 15, 2023 at 11:50 AM Kumara Parameshwaran <
kumaraparamesh92@gmail.com> wrote:
> The patch adds GRO support for TCP/ipv6 packets. This does not
> include the support for vxlan, udp ipv6 packets.
>
> Signed-off-by: Kumara Parameshwaran <kumaraparamesh92@gmail.com>
> ---
> v1:
> * Changes to support GRO for TCP/ipv6 packets. This does not
> include
> vxlan changes.
> * The GRO is performed only for ipv6 packets that does not contain
> extension headers.
> * The logic for the TCP coalescing remains the same, in ipv6
> header
> the source address, destination address, flow label, version
> fields
> are expected to be the same.
> * Re-organised the code to reuse certain tcp functions for both
> ipv4 and
> ipv6 flows.
> v2:
> * Fix comments in gro_tcp6.h header file.
>
> v3:
> * Adderess review comments to fix code duplication for v4 and v6
>
> v4:
> * Addresses review comments for v3, do not use callbacks
>
> v5:
> * Address review comments
>
> v6:
> * Fix warning and coding style issues
>
> v7:
> * Fix build compilation issue
>
> v8:
> * Use gro_tcp_internal.h for functions used for gro_tcp4 and
> gro_tcp6 and use
> gro_tcp.h for data structures and functions used across
> gro_vxlan_tcp4
>
> v9:
> * Resolve merge conflict and add gro_tcp.h in proper path
>
> .../generic_receive_offload_lib.rst | 21 +-
> doc/guides/rel_notes/release_23_07.rst | 2 +
> lib/gro/gro_tcp.h | 187 ++++++++++++
> lib/gro/gro_tcp4.c | 177 +++---------
> lib/gro/gro_tcp4.h | 170 +----------
> lib/gro/gro_tcp6.c | 268 ++++++++++++++++++
> lib/gro/gro_tcp6.h | 160 +++++++++++
> lib/gro/gro_tcp_internal.h | 128 +++++++++
> lib/gro/gro_vxlan_tcp4.c | 22 +-
> lib/gro/gro_vxlan_tcp4.h | 2 +-
> lib/gro/meson.build | 1 +
> lib/gro/rte_gro.c | 83 +++++-
> lib/gro/rte_gro.h | 3 +
> 13 files changed, 883 insertions(+), 341 deletions(-)
> create mode 100644 lib/gro/gro_tcp.h
> create mode 100644 lib/gro/gro_tcp6.c
> create mode 100644 lib/gro/gro_tcp6.h
> create mode 100644 lib/gro/gro_tcp_internal.h
>
> diff --git a/doc/guides/prog_guide/generic_receive_offload_lib.rst
> b/doc/guides/prog_guide/generic_receive_offload_lib.rst
> index 98a5d29bbc..533cda7f5c 100644
> --- a/doc/guides/prog_guide/generic_receive_offload_lib.rst
> +++ b/doc/guides/prog_guide/generic_receive_offload_lib.rst
> @@ -138,20 +138,21 @@ The key-based algorithm has two characters:
>
> Key-based Reassembly Algorithm
>
> -TCP/IPv4 GRO
> -------------
> +TCP-IPv4/IPv6 GRO
> +-----------------
>
> -The table structure used by TCP/IPv4 GRO contains two arrays: flow array
> +The table structure used by TCP-IPv4/IPv6 GRO contains two arrays: flow
> array
> and item array. The flow array keeps flow information, and the item array
> -keeps packet information.
> +keeps packet information. The flow array is different for IPv4 and IPv6
> while
> +the item array is the same.
>
> -Header fields used to define a TCP/IPv4 flow include:
> +Header fields used to define a TCP-IPv4/IPv6 flow include:
>
> -- source and destination: Ethernet and IP address, TCP port
> +- common tcp key fields : Ethernet address, TCP port, TCP acknowledge
> number
> +- version specific IP address
> +- IPv6 flow label for IPv6 flow
>
> -- TCP acknowledge number
> -
> -TCP/IPv4 packets whose FIN, SYN, RST, URG, PSH, ECE or CWR bit is set
> +TCP packets whose FIN, SYN, RST, URG, PSH, ECE or CWR bit is set
> won't be processed.
>
> Header fields deciding if two packets are neighbors include:
> @@ -159,7 +160,7 @@ Header fields deciding if two packets are neighbors
> include:
> - TCP sequence number
>
> - IPv4 ID. The IPv4 ID fields of the packets, whose DF bit is 0, should
> - be increased by 1.
> + be increased by 1. This is applicable only for IPv4
>
> VxLAN GRO
> ---------
> diff --git a/doc/guides/rel_notes/release_23_07.rst
> b/doc/guides/rel_notes/release_23_07.rst
> index 027ae7bd2d..7124cf45c7 100644
> --- a/doc/guides/rel_notes/release_23_07.rst
> +++ b/doc/guides/rel_notes/release_23_07.rst
> @@ -170,6 +170,8 @@ New Features
>
> See :doc:`../prog_guide/pdcp_lib` for more information.
>
> +* **Added support for TCP/IPv6 GRO.**
> + * Enhanced the existing library to support GRO for TCP packets over
> IPv6 network.
>
> Removed Items
> -------------
> diff --git a/lib/gro/gro_tcp.h b/lib/gro/gro_tcp.h
> new file mode 100644
> index 0000000000..7c7a6a3008
> --- /dev/null
> +++ b/lib/gro/gro_tcp.h
> @@ -0,0 +1,187 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2023 Intel Corporation
> + */
> +#ifndef _GRO_TCP_H_
> +#define _GRO_TCP_H_
> +
> +#define INVALID_ARRAY_INDEX 0xffffffffUL
> +
> +#include <rte_tcp.h>
> +
> +/*
> + * The max length of a IPv4 packet, which includes the length of the L3
> + * header, the L4 header and the data payload.
> + */
> +#define MAX_IP_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))
> +
> +struct cmn_tcp_key {
> + struct rte_ether_addr eth_saddr;
> + struct rte_ether_addr eth_daddr;
> + uint32_t recv_ack;
> + uint16_t src_port;
> + uint16_t dst_port;
> +};
> +
> +#define ASSIGN_COMMON_TCP_KEY(k1, k2) \
> + do {\
> + rte_ether_addr_copy(&(k1->eth_saddr), &(k2->eth_saddr)); \
> + rte_ether_addr_copy(&(k1->eth_daddr), &(k2->eth_daddr)); \
> + k2->recv_ack = k1->recv_ack; \
> + k2->src_port = k1->src_port; \
> + k2->dst_port = k1->dst_port; \
> + } while (0)
> +
> +struct gro_tcp_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;
> + union {
> + /* IPv4 ID of the packet */
> + uint16_t ip_id;
> + /* Unused field for IPv6 */
> + uint16_t unused;
> + } l3;
> + /* the number of merged packets */
> + uint16_t nb_merged;
> + /* Indicate if IPv4 ID can be ignored */
> + uint8_t is_atomic;
> +};
> +
> +/*
> + * Merge two TCP 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_tcp_packets(struct gro_tcp_item *item,
> + struct rte_mbuf *pkt,
> + int cmp,
> + uint32_t sent_seq,
> + uint16_t ip_id,
> + 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 IPv4 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_IP_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);
> + /* update IP ID to the larger value */
> + item->l3.ip_id = ip_id;
> + } 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->l3.ip_id = ip_id;
> + }
> + 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 packets are neighbors.
> + */
> +static inline int
> +check_seq_option(struct gro_tcp_item *item,
> + struct rte_tcp_hdr *tcph,
> + uint32_t sent_seq,
> + uint16_t ip_id,
> + uint16_t tcp_hl,
> + uint16_t tcp_dl,
> + uint16_t l2_offset,
> + uint8_t is_atomic)
> +{
> + struct rte_mbuf *pkt_orig = item->firstseg;
> + char *iph_orig;
> + struct rte_tcp_hdr *tcph_orig;
> + uint16_t len, tcp_hl_orig;
> +
> + iph_orig = (char *)(rte_pktmbuf_mtod(pkt_orig, char *) +
> + l2_offset + pkt_orig->l2_len);
> + tcph_orig = (struct rte_tcp_hdr *)(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;
> +
> + /* Don't merge packets whose DF bits are different */
> + if (unlikely(item->is_atomic ^ is_atomic))
> + 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) && (is_atomic ||
> + (ip_id == item->l3.ip_id + 1)))
> + /* append the new packet */
> + return 1;
> + else if ((sent_seq + tcp_dl == item->sent_seq) && (is_atomic ||
> + (ip_id + item->nb_merged ==
> item->l3.ip_id)))
> + /* pre-pend the new packet */
> + return -1;
> +
> + return 0;
> +}
> +
> +static inline int
> +is_same_common_tcp_key(struct cmn_tcp_key *k1, struct cmn_tcp_key *k2)
> +{
> + return (!memcmp(k1, k2, sizeof(struct cmn_tcp_key)));
> +}
> +
> +#endif
> diff --git a/lib/gro/gro_tcp4.c b/lib/gro/gro_tcp4.c
> index 0014096e63..6645de592b 100644
> --- a/lib/gro/gro_tcp4.c
> +++ b/lib/gro/gro_tcp4.c
> @@ -7,6 +7,7 @@
> #include <rte_ethdev.h>
>
> #include "gro_tcp4.h"
> +#include "gro_tcp_internal.h"
>
> void *
> gro_tcp4_tbl_create(uint16_t socket_id,
> @@ -30,7 +31,7 @@ gro_tcp4_tbl_create(uint16_t socket_id,
> if (tbl == NULL)
> return NULL;
>
> - size = sizeof(struct gro_tcp4_item) * entries_num;
> + size = sizeof(struct gro_tcp_item) * entries_num;
> tbl->items = rte_zmalloc_socket(__func__,
> size,
> RTE_CACHE_LINE_SIZE,
> @@ -71,18 +72,6 @@ gro_tcp4_tbl_destroy(void *tbl)
> rte_free(tcp_tbl);
> }
>
> -static inline uint32_t
> -find_an_empty_item(struct gro_tcp4_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_tcp4_tbl *tbl)
> {
> @@ -95,56 +84,6 @@ find_an_empty_flow(struct gro_tcp4_tbl *tbl)
> return INVALID_ARRAY_INDEX;
> }
>
> -static inline uint32_t
> -insert_new_item(struct gro_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 (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].ip_id = ip_id;
> - tbl->items[item_idx].nb_merged = 1;
> - tbl->items[item_idx].is_atomic = is_atomic;
> - 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_tcp4_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_tcp4_tbl *tbl,
> struct tcp4_flow_key *src,
> @@ -159,13 +98,10 @@ insert_new_flow(struct gro_tcp4_tbl *tbl,
>
> 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));
> + ASSIGN_COMMON_TCP_KEY((&src->cmn_key), (&dst->cmn_key));
> +
> dst->ip_src_addr = src->ip_src_addr;
> dst->ip_dst_addr = src->ip_dst_addr;
> - 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++;
> @@ -173,21 +109,6 @@ insert_new_flow(struct gro_tcp4_tbl *tbl,
> return flow_idx;
> }
>
> -/*
> - * update the packet length for the flushed packet.
> - */
> -static inline void
> -update_header(struct gro_tcp4_item *item)
> -{
> - struct rte_ipv4_hdr *ipv4_hdr;
> - struct rte_mbuf *pkt = item->firstseg;
> -
> - ipv4_hdr = (struct rte_ipv4_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
> - pkt->l2_len);
> - ipv4_hdr->total_length = rte_cpu_to_be_16(pkt->pkt_len -
> - pkt->l2_len);
> -}
> -
> int32_t
> gro_tcp4_reassemble(struct rte_mbuf *pkt,
> struct gro_tcp4_tbl *tbl,
> @@ -202,9 +123,8 @@ gro_tcp4_reassemble(struct rte_mbuf *pkt,
> uint8_t is_atomic;
>
> struct tcp4_flow_key key;
> - uint32_t cur_idx, prev_idx, item_idx;
> + uint32_t item_idx;
> uint32_t i, max_flow_num, remaining_flow_num;
> - int cmp;
> uint8_t find;
>
> /*
> @@ -239,6 +159,14 @@ gro_tcp4_reassemble(struct rte_mbuf *pkt,
> if (tcp_dl <= 0)
> return -1;
>
> + rte_ether_addr_copy(&(eth_hdr->src_addr),
> &(key.cmn_key.eth_saddr));
> + rte_ether_addr_copy(&(eth_hdr->dst_addr),
> &(key.cmn_key.eth_daddr));
> + key.ip_src_addr = ipv4_hdr->src_addr;
> + key.ip_dst_addr = ipv4_hdr->dst_addr;
> + key.cmn_key.src_port = tcp_hdr->src_port;
> + key.cmn_key.dst_port = tcp_hdr->dst_port;
> + key.cmn_key.recv_ack = tcp_hdr->recv_ack;
> +
> /*
> * Save IPv4 ID for the packet whose DF bit is 0. For the packet
> * whose DF bit is 1, IPv4 ID is ignored.
> @@ -246,15 +174,6 @@ gro_tcp4_reassemble(struct rte_mbuf *pkt,
> 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->src_addr), &(key.eth_saddr));
> - rte_ether_addr_copy(&(eth_hdr->dst_addr), &(key.eth_daddr));
> - key.ip_src_addr = ipv4_hdr->src_addr;
> - key.ip_dst_addr = ipv4_hdr->dst_addr;
> - 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;
> @@ -270,63 +189,44 @@ gro_tcp4_reassemble(struct rte_mbuf *pkt,
> }
> }
>
> - /*
> - * 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, ip_id,
> - is_atomic);
> + sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
> + item_idx = insert_new_tcp_item(pkt, tbl->items,
> &tbl->item_num,
> + tbl->max_item_num,
> 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) {
> + INVALID_ARRAY_INDEX) {
> /*
> * Fail to insert a new flow, so delete the
> * stored packet.
> - */
> - delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
> + */
> + delete_tcp_item(tbl->items, item_idx,
> &tbl->item_num, 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(&(tbl->items[cur_idx]), tcp_hdr,
> - sent_seq, ip_id, pkt->l4_len, tcp_dl, 0,
> - is_atomic);
> - if (cmp) {
> - if (merge_two_tcp4_packets(&(tbl->items[cur_idx]),
> - pkt, cmp, sent_seq, ip_id,
> 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, cur_idx,
> - sent_seq, ip_id,
> is_atomic) ==
> - 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);
> + return process_tcp_item(pkt, tcp_hdr, tcp_dl, tbl->items,
> tbl->flows[i].start_index,
> + &tbl->item_num,
> tbl->max_item_num,
> + ip_id, is_atomic,
> start_time);
> +}
>
> - /* Fail to find a neighbor, so store 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;
> +/*
> + * update the packet length for the flushed packet.
> + */
> +static inline void
> +update_header(struct gro_tcp_item *item)
> +{
> + struct rte_ipv4_hdr *ipv4_hdr;
> + struct rte_mbuf *pkt = item->firstseg;
>
> - return 0;
> + ipv4_hdr = (struct rte_ipv4_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
> + pkt->l2_len);
> + ipv4_hdr->total_length = rte_cpu_to_be_16(pkt->pkt_len -
> + pkt->l2_len);
> }
>
> uint16_t
> @@ -353,7 +253,8 @@ gro_tcp4_tbl_timeout_flush(struct gro_tcp4_tbl *tbl,
> * Delete the packet and get the next
> * packet in the flow.
> */
> - j = delete_item(tbl, j,
> INVALID_ARRAY_INDEX);
> + j = delete_tcp_item(tbl->items, j,
> + &tbl->item_num,
> INVALID_ARRAY_INDEX);
> tbl->flows[i].start_index = j;
> if (j == INVALID_ARRAY_INDEX)
> tbl->flow_num--;
> diff --git a/lib/gro/gro_tcp4.h b/lib/gro/gro_tcp4.h
> index 212f97a042..245e5da486 100644
> --- a/lib/gro/gro_tcp4.h
> +++ b/lib/gro/gro_tcp4.h
> @@ -5,32 +5,15 @@
> #ifndef _GRO_TCP4_H_
> #define _GRO_TCP4_H_
>
> -#include <rte_tcp.h>
> +#include "gro_tcp.h"
>
> -#define INVALID_ARRAY_INDEX 0xffffffffUL
> #define GRO_TCP4_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
>
> -/*
> - * The max length of a IPv4 packet, which includes the length of the L3
> - * header, the L4 header and the data payload.
> - */
> -#define MAX_IPV4_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/IPv4 flow */
> +/* Header fields representing common fields in TCP flow */
> struct tcp4_flow_key {
> - struct rte_ether_addr eth_saddr;
> - struct rte_ether_addr eth_daddr;
> + struct cmn_tcp_key cmn_key;
> uint32_t ip_src_addr;
> uint32_t ip_dst_addr;
> -
> - uint32_t recv_ack;
> - uint16_t src_port;
> - uint16_t dst_port;
> };
>
> struct gro_tcp4_flow {
> @@ -42,42 +25,12 @@ struct gro_tcp4_flow {
> uint32_t start_index;
> };
>
> -struct gro_tcp4_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;
> - /* IPv4 ID of the packet */
> - uint16_t ip_id;
> - /* the number of merged packets */
> - uint16_t nb_merged;
> - /* Indicate if IPv4 ID can be ignored */
> - uint8_t is_atomic;
> -};
> -
> /*
> * TCP/IPv4 reassembly table structure.
> */
> struct gro_tcp4_tbl {
> /* item array */
> - struct gro_tcp4_item *items;
> + struct gro_tcp_item *items;
> /* flow array */
> struct gro_tcp4_flow *flows;
> /* current item number */
> @@ -186,120 +139,9 @@ uint32_t gro_tcp4_tbl_pkt_count(void *tbl);
> static inline int
> is_same_tcp4_flow(struct tcp4_flow_key k1, struct tcp4_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) &&
> - (k1.ip_src_addr == k2.ip_src_addr) &&
> + return ((k1.ip_src_addr == k2.ip_src_addr) &&
> (k1.ip_dst_addr == k2.ip_dst_addr) &&
> - (k1.recv_ack == k2.recv_ack) &&
> - (k1.src_port == k2.src_port) &&
> - (k1.dst_port == k2.dst_port));
> + is_same_common_tcp_key(&k1.cmn_key, &k2.cmn_key));
> }
>
> -/*
> - * Merge two TCP/IPv4 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_tcp4_packets(struct gro_tcp4_item *item,
> - struct rte_mbuf *pkt,
> - int cmp,
> - uint32_t sent_seq,
> - uint16_t ip_id,
> - 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 IPv4 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_IPV4_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);
> - /* update IP ID to the larger value */
> - item->ip_id = ip_id;
> - } 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->ip_id = ip_id;
> - }
> - 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/IPv4 packets are neighbors.
> - */
> -static inline int
> -check_seq_option(struct gro_tcp4_item *item,
> - struct rte_tcp_hdr *tcph,
> - uint32_t sent_seq,
> - uint16_t ip_id,
> - uint16_t tcp_hl,
> - uint16_t tcp_dl,
> - uint16_t l2_offset,
> - uint8_t is_atomic)
> -{
> - struct rte_mbuf *pkt_orig = item->firstseg;
> - struct rte_ipv4_hdr *iph_orig;
> - struct rte_tcp_hdr *tcph_orig;
> - uint16_t len, tcp_hl_orig;
> -
> - iph_orig = (struct rte_ipv4_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;
> -
> - /* Don't merge packets whose DF bits are different */
> - if (unlikely(item->is_atomic ^ is_atomic))
> - 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) && (is_atomic ||
> - (ip_id == item->ip_id + 1)))
> - /* append the new packet */
> - return 1;
> - else if ((sent_seq + tcp_dl == item->sent_seq) && (is_atomic ||
> - (ip_id + item->nb_merged == item->ip_id)))
> - /* pre-pend the new packet */
> - return -1;
> -
> - return 0;
> -}
> #endif
> diff --git a/lib/gro/gro_tcp6.c b/lib/gro/gro_tcp6.c
> new file mode 100644
> index 0000000000..5aa39801e1
> --- /dev/null
> +++ b/lib/gro/gro_tcp6.c
> @@ -0,0 +1,268 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2023 Intel Corporation
> + */
> +
> +#include <rte_malloc.h>
> +#include <rte_mbuf.h>
> +#include <rte_ethdev.h>
> +
> +#include "gro_tcp6.h"
> +#include "gro_tcp_internal.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_tcp_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_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_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);
> +
> + ASSIGN_COMMON_TCP_KEY((&src->cmn_key), (&dst->cmn_key));
> + memcpy(&dst->src_addr[0], &src->src_addr[0],
> sizeof(dst->src_addr));
> + memcpy(&dst->dst_addr[0], &src->dst_addr[0],
> sizeof(dst->dst_addr));
> + dst->vtc_flow = src->vtc_flow;
> +
> + 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_tcp_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;
> + int32_t tcp_dl;
> + uint16_t ip_tlen;
> + struct tcp6_flow_key key;
> + uint32_t i, max_flow_num, remaining_flow_num;
> + uint32_t sent_seq;
> + struct rte_tcp_hdr *tcp_hdr;
> + uint8_t find;
> + uint32_t item_idx;
> + /*
> + * 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 = rte_pktmbuf_mtod_offset(pkt, struct rte_tcp_hdr *,
> pkt->l2_len + 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;
> +
> + ip_tlen = rte_be_to_cpu_16(ipv6_hdr->payload_len);
> + /*
> + * Don't process the packet whose payload length is less than or
> + * equal to 0.
> + */
> + tcp_dl = ip_tlen - pkt->l4_len;
> + if (tcp_dl <= 0)
> + return -1;
> +
> + rte_ether_addr_copy(&(eth_hdr->src_addr),
> &(key.cmn_key.eth_saddr));
> + rte_ether_addr_copy(&(eth_hdr->dst_addr),
> &(key.cmn_key.eth_daddr));
> + memcpy(&key.src_addr[0], &ipv6_hdr->src_addr,
> sizeof(key.src_addr));
> + memcpy(&key.dst_addr[0], &ipv6_hdr->dst_addr,
> sizeof(key.dst_addr));
> + key.cmn_key.src_port = tcp_hdr->src_port;
> + key.cmn_key.dst_port = tcp_hdr->dst_port;
> + key.cmn_key.recv_ack = tcp_hdr->recv_ack;
> + key.vtc_flow = ipv6_hdr->vtc_flow;
> +
> + /* 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--;
> + }
> + }
> +
> + if (find == 0) {
> + sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
> + item_idx = insert_new_tcp_item(pkt, tbl->items,
> &tbl->item_num,
> + tbl->max_item_num,
> start_time,
> + INVALID_ARRAY_INDEX,
> sent_seq, 0, true);
> + 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_tcp_item(tbl->items, item_idx,
> &tbl->item_num, INVALID_ARRAY_INDEX);
> + return -1;
> + }
> + return 0;
> + }
> +
> + return process_tcp_item(pkt, tcp_hdr, tcp_dl, tbl->items,
> tbl->flows[i].start_index,
> + &tbl->item_num,
> tbl->max_item_num,
> + 0, true, start_time);
> +}
> +
> +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_tcp_item(tbl->items, j,
> + &tbl->item_num,
> 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/gro/gro_tcp6.h b/lib/gro/gro_tcp6.h
> new file mode 100644
> index 0000000000..073122f0ec
> --- /dev/null
> +++ b/lib/gro/gro_tcp6.h
> @@ -0,0 +1,160 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2023 Intel Corporation
> + */
> +
> +#ifndef _GRO_TCP6_H_
> +#define _GRO_TCP6_H_
> +
> +#include "gro_tcp.h"
> +
> +#define GRO_TCP6_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
> +
> +/* Header fields representing a TCP/IPv6 flow */
> +struct tcp6_flow_key {
> + struct cmn_tcp_key cmn_key;
> + uint8_t src_addr[16];
> + uint8_t dst_addr[16];
> + rte_be32_t vtc_flow;
> +};
> +
> +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;
> +};
> +
> +/*
> + * TCP/IPv6 reassembly table structure.
> + */
> +struct gro_tcp6_tbl {
> + /* item array */
> + struct gro_tcp_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 doesn't process the packet,
> + * which has SYN, FIN, RST, PSH, CWR, ECE or URG set, or doesn't have
> + * payload.
> + *
> + * This function doesn't check if the packet has correct checksums and
> + * doesn't 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 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)
> +{
> + rte_be32_t vtc_flow_diff;
> +
> + if (memcmp(&k1->src_addr, &k2->src_addr, 16))
> + return 0;
> + if (memcmp(&k1->dst_addr, &k2->dst_addr, 16))
> + return 0;
> + /*
> + * IP version (4) Traffic Class (8) Flow Label (20)
> + * All fields except Traffic class should be same
> + */
> + vtc_flow_diff = (k1->vtc_flow ^ k2->vtc_flow);
> + if (vtc_flow_diff & htonl(0xF00FFFFF))
> + return 0;
> +
> + return is_same_common_tcp_key(&k1->cmn_key, &k2->cmn_key);
> +}
> +
> +#endif
> diff --git a/lib/gro/gro_tcp_internal.h b/lib/gro/gro_tcp_internal.h
> new file mode 100644
> index 0000000000..cc84abeaeb
> --- /dev/null
> +++ b/lib/gro/gro_tcp_internal.h
> @@ -0,0 +1,128 @@
> +
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2023 Intel Corporation
> + */
> +
> +#ifndef _GRO_TCP_INTERNAL_H_
> +#define _GRO_TCP_INTERNAL_H_
> +
> +static inline uint32_t
> +find_an_empty_item(struct gro_tcp_item *items,
> + uint32_t max_item_num)
> +{
> + uint32_t i;
> +
> + for (i = 0; i < max_item_num; i++)
> + if (items[i].firstseg == NULL)
> + return i;
> + return INVALID_ARRAY_INDEX;
> +}
> +
> +static inline uint32_t
> +insert_new_tcp_item(struct rte_mbuf *pkt,
> + struct gro_tcp_item *items,
> + uint32_t *item_num,
> + uint32_t max_item_num,
> + 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(items, max_item_num);
> + if (item_idx == INVALID_ARRAY_INDEX)
> + return INVALID_ARRAY_INDEX;
> +
> + items[item_idx].firstseg = pkt;
> + items[item_idx].lastseg = rte_pktmbuf_lastseg(pkt);
> + items[item_idx].start_time = start_time;
> + items[item_idx].next_pkt_idx = INVALID_ARRAY_INDEX;
> + items[item_idx].sent_seq = sent_seq;
> + items[item_idx].l3.ip_id = ip_id;
> + items[item_idx].nb_merged = 1;
> + items[item_idx].is_atomic = is_atomic;
> + (*item_num) += 1;
> +
> + /* if the previous packet exists, chain them together. */
> + if (prev_idx != INVALID_ARRAY_INDEX) {
> + items[item_idx].next_pkt_idx =
> + items[prev_idx].next_pkt_idx;
> + items[prev_idx].next_pkt_idx = item_idx;
> + }
> +
> + return item_idx;
> +}
> +
> +static inline uint32_t
> +delete_tcp_item(struct gro_tcp_item *items, uint32_t item_idx,
> + uint32_t *item_num,
> + uint32_t prev_item_idx)
> +{
> + uint32_t next_idx = items[item_idx].next_pkt_idx;
> +
> + /* NULL indicates an empty item */
> + items[item_idx].firstseg = NULL;
> + (*item_num) -= 1;
> + if (prev_item_idx != INVALID_ARRAY_INDEX)
> + items[prev_item_idx].next_pkt_idx = next_idx;
> +
> + return next_idx;
> +}
> +
> +static inline int32_t
> +process_tcp_item(struct rte_mbuf *pkt,
> + struct rte_tcp_hdr *tcp_hdr,
> + int32_t tcp_dl,
> + struct gro_tcp_item *items,
> + uint32_t item_idx,
> + uint32_t *item_num,
> + uint32_t max_item_num,
> + uint16_t ip_id,
> + uint8_t is_atomic,
> + uint64_t start_time)
> +{
> + uint32_t cur_idx;
> + uint32_t prev_idx;
> + int cmp;
> + uint32_t sent_seq;
> +
> + sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
> + /*
> + * Check all packets in the flow and try to find a neighbor for
> + * the input packet.
> + */
> + cur_idx = item_idx;
> + prev_idx = cur_idx;
> + do {
> + cmp = check_seq_option(&items[cur_idx], tcp_hdr,
> + sent_seq, ip_id, pkt->l4_len, tcp_dl, 0,
> + is_atomic);
> + if (cmp) {
> + if (merge_two_tcp_packets(&items[cur_idx],
> + pkt, cmp, sent_seq, ip_id,
> 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_tcp_item(pkt, items, item_num,
> max_item_num,
> + start_time, cur_idx,
> sent_seq, ip_id, is_atomic) ==
> + INVALID_ARRAY_INDEX)
> + return -1;
> + return 0;
> + }
> + prev_idx = cur_idx;
> + cur_idx = 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_tcp_item(pkt, items, item_num, max_item_num,
> start_time, prev_idx, sent_seq,
> + ip_id, is_atomic) == INVALID_ARRAY_INDEX)
> + return -1;
> +
> + return 0;
> +}
> +#endif
> diff --git a/lib/gro/gro_vxlan_tcp4.c b/lib/gro/gro_vxlan_tcp4.c
> index 3be4deb7c7..6ab7001922 100644
> --- a/lib/gro/gro_vxlan_tcp4.c
> +++ b/lib/gro/gro_vxlan_tcp4.c
> @@ -116,7 +116,7 @@ insert_new_item(struct gro_vxlan_tcp4_tbl *tbl,
> 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.l3.ip_id = ip_id;
> tbl->items[item_idx].inner_item.nb_merged = 1;
> tbl->items[item_idx].inner_item.is_atomic = is_atomic;
> tbl->items[item_idx].outer_ip_id = outer_ip_id;
> @@ -163,15 +163,9 @@ insert_new_flow(struct gro_vxlan_tcp4_tbl *tbl,
>
> 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));
> + ASSIGN_COMMON_TCP_KEY((&(src->inner_key.cmn_key)),
> (&(dst->inner_key.cmn_key)));
> 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;
> @@ -248,7 +242,7 @@ merge_two_vxlan_tcp4_packets(struct
> gro_vxlan_tcp4_item *item,
> uint16_t outer_ip_id,
> uint16_t ip_id)
> {
> - if (merge_two_tcp4_packets(&item->inner_item, pkt, cmp, sent_seq,
> + if (merge_two_tcp_packets(&item->inner_item, pkt, cmp, sent_seq,
> ip_id, pkt->outer_l2_len +
> pkt->outer_l3_len)) {
> /* Update the outer IPv4 ID to the large value. */
> @@ -357,13 +351,13 @@ gro_vxlan_tcp4_reassemble(struct rte_mbuf *pkt,
>
> sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
>
> - rte_ether_addr_copy(&(eth_hdr->src_addr),
> &(key.inner_key.eth_saddr));
> - rte_ether_addr_copy(&(eth_hdr->dst_addr),
> &(key.inner_key.eth_daddr));
> + rte_ether_addr_copy(&(eth_hdr->src_addr),
> &(key.inner_key.cmn_key.eth_saddr));
> + rte_ether_addr_copy(&(eth_hdr->dst_addr),
> &(key.inner_key.cmn_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.inner_key.cmn_key.recv_ack = tcp_hdr->recv_ack;
> + key.inner_key.cmn_key.src_port = tcp_hdr->src_port;
> + key.inner_key.cmn_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;
> diff --git a/lib/gro/gro_vxlan_tcp4.h b/lib/gro/gro_vxlan_tcp4.h
> index 7832942a68..662db01a88 100644
> --- a/lib/gro/gro_vxlan_tcp4.h
> +++ b/lib/gro/gro_vxlan_tcp4.h
> @@ -36,7 +36,7 @@ struct gro_vxlan_tcp4_flow {
> };
>
> struct gro_vxlan_tcp4_item {
> - struct gro_tcp4_item inner_item;
> + struct gro_tcp_item inner_item;
> /* IPv4 ID in the outer IPv4 header */
> uint16_t outer_ip_id;
> /* Indicate if outer IPv4 ID can be ignored */
> diff --git a/lib/gro/meson.build b/lib/gro/meson.build
> index e4fa2958bd..dbce05220d 100644
> --- a/lib/gro/meson.build
> +++ b/lib/gro/meson.build
> @@ -4,6 +4,7 @@
> sources = files(
> 'rte_gro.c',
> 'gro_tcp4.c',
> + 'gro_tcp6.c',
> 'gro_udp4.c',
> 'gro_vxlan_tcp4.c',
> 'gro_vxlan_udp4.c',
> diff --git a/lib/gro/rte_gro.c b/lib/gro/rte_gro.c
> index e35399fd42..d824eebd93 100644
> --- a/lib/gro/rte_gro.c
> +++ b/lib/gro/rte_gro.c
> @@ -8,6 +8,7 @@
>
> #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"
> @@ -20,14 +21,16 @@ typedef uint32_t (*gro_tbl_pkt_count_fn)(void *tbl);
>
> 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,
> + 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,
> + gro_tcp6_tbl_pkt_count,
> NULL};
>
> #define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
> @@ -35,6 +38,12 @@ static gro_tbl_pkt_count_fn
> tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
> ((ptype & RTE_PTYPE_L4_FRAG) != RTE_PTYPE_L4_FRAG) && \
> (RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
>
> +/* GRO with extension headers is not supported */
> +#define IS_IPV6_TCP_PKT(ptype) (RTE_ETH_IS_IPV6_HDR(ptype) && \
> + ((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
> + ((ptype & RTE_PTYPE_L4_FRAG) != RTE_PTYPE_L4_FRAG) && \
> + (RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
> +
> #define IS_IPV4_UDP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
> ((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
> (RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
> @@ -147,7 +156,11 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
> /* allocate a reassembly table for TCP/IPv4 GRO */
> struct gro_tcp4_tbl tcp_tbl;
> struct gro_tcp4_flow tcp_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
> - struct gro_tcp4_item tcp_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0}
> };
> + struct gro_tcp_item tcp_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
> +
> + struct gro_tcp6_tbl tcp6_tbl;
> + struct gro_tcp6_flow tcp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
> + struct gro_tcp_item tcp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0}
> };
>
> /* allocate a reassembly table for UDP/IPv4 GRO */
> struct gro_udp4_tbl udp_tbl;
> @@ -171,10 +184,10 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
> 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_TCP_IPV4 |
> RTE_GRO_TCP_IPV6 |
> RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
> RTE_GRO_UDP_IPV4)) == 0))
> return nb_pkts;
> @@ -236,6 +249,18 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
> 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++) {
> /*
> @@ -276,6 +301,14 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
> 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];
> }
> @@ -283,9 +316,17 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
> if ((nb_after_gro < nb_pkts)
> || (unprocess_num < nb_pkts)) {
> i = 0;
> + /* Copy unprocessed packets */
> + if (unprocess_num > 0) {
> + memcpy(&pkts[i], unprocess_pkts,
> + sizeof(struct rte_mbuf *) *
> + unprocess_num);
> + i = unprocess_num;
> + }
> +
> /* Flush all packets from the tables */
> if (do_vxlan_tcp_gro) {
> - i =
> gro_vxlan_tcp4_tbl_timeout_flush(&vxlan_tcp_tbl,
> + i +=
> gro_vxlan_tcp4_tbl_timeout_flush(&vxlan_tcp_tbl,
> 0, pkts, nb_pkts);
> }
>
> @@ -304,13 +345,11 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
> i += gro_udp4_tbl_timeout_flush(&udp_tbl, 0,
> &pkts[i], nb_pkts - i);
> }
> - /* Copy unprocessed packets */
> - if (unprocess_num > 0) {
> - memcpy(&pkts[i], unprocess_pkts,
> - sizeof(struct rte_mbuf *) *
> - unprocess_num);
> +
> + if (do_tcp6_gro) {
> + i += gro_tcp6_tbl_timeout_flush(&tcp6_tbl, 0,
> + &pkts[i], nb_pkts - i);
> }
> - nb_after_gro = i + unprocess_num;
> }
>
> return nb_after_gro;
> @@ -323,13 +362,13 @@ rte_gro_reassemble(struct rte_mbuf **pkts,
> {
> 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_TCP_IPV4 |
> RTE_GRO_TCP_IPV6 |
> RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
> RTE_GRO_UDP_IPV4)) == 0))
> return nb_pkts;
> @@ -338,6 +377,7 @@ rte_gro_reassemble(struct rte_mbuf **pkts,
> 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;
> @@ -347,6 +387,7 @@ rte_gro_reassemble(struct rte_mbuf **pkts,
> 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();
>
> @@ -371,6 +412,11 @@ rte_gro_reassemble(struct rte_mbuf **pkts,
> 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];
> }
> @@ -426,6 +472,15 @@ rte_gro_timeout_flush(void *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/gro/rte_gro.h b/lib/gro/rte_gro.h
> index 9f9ed4935a..c83dfd9ad1 100644
> --- a/lib/gro/rte_gro.h
> +++ b/lib/gro/rte_gro.h
> @@ -38,6 +38,9 @@ extern "C" {
> #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
> --
> 2.25.1
>
>> Hi Jiyau, In the recent patch I have moved the functions that had the
>> error in Windows to the header file and since that is used by gro_tcp6.c
>> and gro_tcp4.c, I have moved it to gro_tcp_internal.h. There are other
>> functions and ds reused by gro_vxlan_tcp4.c, I have moved them to gro_tcp.h
>> so that the functions that are not used by gro_vxlan_tcp4.c will not be
>> included.
>>
>
Hi Kumara,
> -----Original Message-----
> From: Kumara Parameshwaran <kumaraparamesh92@gmail.com>
> Sent: Thursday, June 15, 2023 2:20 PM
> To: Hu, Jiayu <jiayu.hu@intel.com>
> Cc: dev@dpdk.org; Kumara Parameshwaran
> <kumaraparamesh92@gmail.com>
> Subject: [PATCH v9] gro : ipv6 changes to support GRO for TCP/ipv6
>
> The patch adds GRO support for TCP/ipv6 packets. This does not include the
> support for vxlan, udp ipv6 packets.
>
> Signed-off-by: Kumara Parameshwaran <kumaraparamesh92@gmail.com>
> ---
> v1:
> * Changes to support GRO for TCP/ipv6 packets. This does not
> include
> vxlan changes.
> * The GRO is performed only for ipv6 packets that does not contain
> extension headers.
> * The logic for the TCP coalescing remains the same, in ipv6 header
> the source address, destination address, flow label, version fields
> are expected to be the same.
> * Re-organised the code to reuse certain tcp functions for both ipv4
> and
> ipv6 flows.
> v2:
> * Fix comments in gro_tcp6.h header file.
>
> v3:
> * Adderess review comments to fix code duplication for v4 and v6
>
> v4:
> * Addresses review comments for v3, do not use callbacks
>
> v5:
> * Address review comments
>
> v6:
> * Fix warning and coding style issues
>
> v7:
> * Fix build compilation issue
>
> v8:
> * Use gro_tcp_internal.h for functions used for gro_tcp4 and
> gro_tcp6 and use
> gro_tcp.h for data structures and functions used across
> gro_vxlan_tcp4
>
> v9:
> * Resolve merge conflict and add gro_tcp.h in proper path
>
> .../generic_receive_offload_lib.rst | 21 +-
> doc/guides/rel_notes/release_23_07.rst | 2 +
> lib/gro/gro_tcp.h | 187 ++++++++++++
> lib/gro/gro_tcp4.c | 177 +++---------
> lib/gro/gro_tcp4.h | 170 +----------
> lib/gro/gro_tcp6.c | 268 ++++++++++++++++++
> lib/gro/gro_tcp6.h | 160 +++++++++++
> lib/gro/gro_tcp_internal.h | 128 +++++++++
> lib/gro/gro_vxlan_tcp4.c | 22 +-
> lib/gro/gro_vxlan_tcp4.h | 2 +-
> lib/gro/meson.build | 1 +
> lib/gro/rte_gro.c | 83 +++++-
> lib/gro/rte_gro.h | 3 +
> 13 files changed, 883 insertions(+), 341 deletions(-) create mode 100644
> lib/gro/gro_tcp.h create mode 100644 lib/gro/gro_tcp6.c create mode
> 100644 lib/gro/gro_tcp6.h create mode 100644 lib/gro/gro_tcp_internal.h
>
Reviewed-by: Jiayu Hu <Jiayu.hu@intel.com>
Hi Kumara,
I notice you write Intel in the license header. Is it intended or mistaken?
Thanks,
Jiayu
On Thu, Jun 15, 2023 at 1:31 PM Hu, Jiayu <jiayu.hu@intel.com> wrote:
> Hi Kumara,
>
> > -----Original Message-----
> > From: Kumara Parameshwaran <kumaraparamesh92@gmail.com>
> > Sent: Thursday, June 15, 2023 2:20 PM
> > To: Hu, Jiayu <jiayu.hu@intel.com>
> > Cc: dev@dpdk.org; Kumara Parameshwaran
> > <kumaraparamesh92@gmail.com>
> > Subject: [PATCH v9] gro : ipv6 changes to support GRO for TCP/ipv6
> >
> > The patch adds GRO support for TCP/ipv6 packets. This does not include
> the
> > support for vxlan, udp ipv6 packets.
> >
> > Signed-off-by: Kumara Parameshwaran <kumaraparamesh92@gmail.com>
> > ---
> > v1:
> > * Changes to support GRO for TCP/ipv6 packets. This does not
> > include
> > vxlan changes.
> > * The GRO is performed only for ipv6 packets that does not contain
> > extension headers.
> > * The logic for the TCP coalescing remains the same, in ipv6 header
> > the source address, destination address, flow label, version
> fields
> > are expected to be the same.
> > * Re-organised the code to reuse certain tcp functions for both
> ipv4
> > and
> > ipv6 flows.
> > v2:
> > * Fix comments in gro_tcp6.h header file.
> >
> > v3:
> > * Adderess review comments to fix code duplication for v4 and v6
> >
> > v4:
> > * Addresses review comments for v3, do not use callbacks
> >
> > v5:
> > * Address review comments
> >
> > v6:
> > * Fix warning and coding style issues
> >
> > v7:
> > * Fix build compilation issue
> >
> > v8:
> > * Use gro_tcp_internal.h for functions used for gro_tcp4 and
> > gro_tcp6 and use
> > gro_tcp.h for data structures and functions used across
> > gro_vxlan_tcp4
> >
> > v9:
> > * Resolve merge conflict and add gro_tcp.h in proper path
> >
> > .../generic_receive_offload_lib.rst | 21 +-
> > doc/guides/rel_notes/release_23_07.rst | 2 +
> > lib/gro/gro_tcp.h | 187 ++++++++++++
> > lib/gro/gro_tcp4.c | 177 +++---------
> > lib/gro/gro_tcp4.h | 170 +----------
> > lib/gro/gro_tcp6.c | 268 ++++++++++++++++++
> > lib/gro/gro_tcp6.h | 160 +++++++++++
> > lib/gro/gro_tcp_internal.h | 128 +++++++++
> > lib/gro/gro_vxlan_tcp4.c | 22 +-
> > lib/gro/gro_vxlan_tcp4.h | 2 +-
> > lib/gro/meson.build | 1 +
> > lib/gro/rte_gro.c | 83 +++++-
> > lib/gro/rte_gro.h | 3 +
> > 13 files changed, 883 insertions(+), 341 deletions(-) create mode
> 100644
> > lib/gro/gro_tcp.h create mode 100644 lib/gro/gro_tcp6.c create mode
> > 100644 lib/gro/gro_tcp6.h create mode 100644 lib/gro/gro_tcp_internal.h
> >
>
> Reviewed-by: Jiayu Hu <Jiayu.hu@intel.com>
>
> Hi Kumara,
>
> I notice you write Intel in the license header. Is it intended or mistaken?
>
>>
>> Hi Jiyau, This is intentional.
>>
>> Thanks,
>> Kumara.
>>
>
> Thanks,
> Jiayu
>
15/06/2023 08:20, Kumara Parameshwaran:
> The patch adds GRO support for TCP/ipv6 packets. This does not
> include the support for vxlan, udp ipv6 packets.
>
> Signed-off-by: Kumara Parameshwaran <kumaraparamesh92@gmail.com>
> ---
> lib/gro/gro_tcp.h | 187 ++++++++++++
> lib/gro/gro_tcp4.c | 177 +++---------
> lib/gro/gro_tcp4.h | 170 +----------
> lib/gro/gro_tcp6.c | 268 ++++++++++++++++++
> lib/gro/gro_tcp6.h | 160 +++++++++++
> lib/gro/gro_tcp_internal.h | 128 +++++++++
> lib/gro/gro_vxlan_tcp4.c | 22 +-
> lib/gro/gro_vxlan_tcp4.h | 2 +-
> lib/gro/meson.build | 1 +
> lib/gro/rte_gro.c | 83 +++++-
> lib/gro/rte_gro.h | 3 +
Please would it be possible to have 2 patches:
- first would move IPv4 code in a new file (without any other change)
- second would add IPv6 support
It would allow to see more clearly what is changed for IPv6.
Thanks
@@ -138,20 +138,21 @@ The key-based algorithm has two characters:
Key-based Reassembly Algorithm
-TCP/IPv4 GRO
-------------
+TCP-IPv4/IPv6 GRO
+-----------------
-The table structure used by TCP/IPv4 GRO contains two arrays: flow array
+The table structure used by TCP-IPv4/IPv6 GRO contains two arrays: flow array
and item array. The flow array keeps flow information, and the item array
-keeps packet information.
+keeps packet information. The flow array is different for IPv4 and IPv6 while
+the item array is the same.
-Header fields used to define a TCP/IPv4 flow include:
+Header fields used to define a TCP-IPv4/IPv6 flow include:
-- source and destination: Ethernet and IP address, TCP port
+- common tcp key fields : Ethernet address, TCP port, TCP acknowledge number
+- version specific IP address
+- IPv6 flow label for IPv6 flow
-- TCP acknowledge number
-
-TCP/IPv4 packets whose FIN, SYN, RST, URG, PSH, ECE or CWR bit is set
+TCP packets whose FIN, SYN, RST, URG, PSH, ECE or CWR bit is set
won't be processed.
Header fields deciding if two packets are neighbors include:
@@ -159,7 +160,7 @@ Header fields deciding if two packets are neighbors include:
- TCP sequence number
- IPv4 ID. The IPv4 ID fields of the packets, whose DF bit is 0, should
- be increased by 1.
+ be increased by 1. This is applicable only for IPv4
VxLAN GRO
---------
@@ -170,6 +170,8 @@ New Features
See :doc:`../prog_guide/pdcp_lib` for more information.
+* **Added support for TCP/IPv6 GRO.**
+ * Enhanced the existing library to support GRO for TCP packets over IPv6 network.
Removed Items
-------------
new file mode 100644
@@ -0,0 +1,187 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2023 Intel Corporation
+ */
+#ifndef _GRO_TCP_H_
+#define _GRO_TCP_H_
+
+#define INVALID_ARRAY_INDEX 0xffffffffUL
+
+#include <rte_tcp.h>
+
+/*
+ * The max length of a IPv4 packet, which includes the length of the L3
+ * header, the L4 header and the data payload.
+ */
+#define MAX_IP_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))
+
+struct cmn_tcp_key {
+ struct rte_ether_addr eth_saddr;
+ struct rte_ether_addr eth_daddr;
+ uint32_t recv_ack;
+ uint16_t src_port;
+ uint16_t dst_port;
+};
+
+#define ASSIGN_COMMON_TCP_KEY(k1, k2) \
+ do {\
+ rte_ether_addr_copy(&(k1->eth_saddr), &(k2->eth_saddr)); \
+ rte_ether_addr_copy(&(k1->eth_daddr), &(k2->eth_daddr)); \
+ k2->recv_ack = k1->recv_ack; \
+ k2->src_port = k1->src_port; \
+ k2->dst_port = k1->dst_port; \
+ } while (0)
+
+struct gro_tcp_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;
+ union {
+ /* IPv4 ID of the packet */
+ uint16_t ip_id;
+ /* Unused field for IPv6 */
+ uint16_t unused;
+ } l3;
+ /* the number of merged packets */
+ uint16_t nb_merged;
+ /* Indicate if IPv4 ID can be ignored */
+ uint8_t is_atomic;
+};
+
+/*
+ * Merge two TCP 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_tcp_packets(struct gro_tcp_item *item,
+ struct rte_mbuf *pkt,
+ int cmp,
+ uint32_t sent_seq,
+ uint16_t ip_id,
+ 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 IPv4 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_IP_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);
+ /* update IP ID to the larger value */
+ item->l3.ip_id = ip_id;
+ } 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->l3.ip_id = ip_id;
+ }
+ 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 packets are neighbors.
+ */
+static inline int
+check_seq_option(struct gro_tcp_item *item,
+ struct rte_tcp_hdr *tcph,
+ uint32_t sent_seq,
+ uint16_t ip_id,
+ uint16_t tcp_hl,
+ uint16_t tcp_dl,
+ uint16_t l2_offset,
+ uint8_t is_atomic)
+{
+ struct rte_mbuf *pkt_orig = item->firstseg;
+ char *iph_orig;
+ struct rte_tcp_hdr *tcph_orig;
+ uint16_t len, tcp_hl_orig;
+
+ iph_orig = (char *)(rte_pktmbuf_mtod(pkt_orig, char *) +
+ l2_offset + pkt_orig->l2_len);
+ tcph_orig = (struct rte_tcp_hdr *)(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;
+
+ /* Don't merge packets whose DF bits are different */
+ if (unlikely(item->is_atomic ^ is_atomic))
+ 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) && (is_atomic ||
+ (ip_id == item->l3.ip_id + 1)))
+ /* append the new packet */
+ return 1;
+ else if ((sent_seq + tcp_dl == item->sent_seq) && (is_atomic ||
+ (ip_id + item->nb_merged == item->l3.ip_id)))
+ /* pre-pend the new packet */
+ return -1;
+
+ return 0;
+}
+
+static inline int
+is_same_common_tcp_key(struct cmn_tcp_key *k1, struct cmn_tcp_key *k2)
+{
+ return (!memcmp(k1, k2, sizeof(struct cmn_tcp_key)));
+}
+
+#endif
@@ -7,6 +7,7 @@
#include <rte_ethdev.h>
#include "gro_tcp4.h"
+#include "gro_tcp_internal.h"
void *
gro_tcp4_tbl_create(uint16_t socket_id,
@@ -30,7 +31,7 @@ gro_tcp4_tbl_create(uint16_t socket_id,
if (tbl == NULL)
return NULL;
- size = sizeof(struct gro_tcp4_item) * entries_num;
+ size = sizeof(struct gro_tcp_item) * entries_num;
tbl->items = rte_zmalloc_socket(__func__,
size,
RTE_CACHE_LINE_SIZE,
@@ -71,18 +72,6 @@ gro_tcp4_tbl_destroy(void *tbl)
rte_free(tcp_tbl);
}
-static inline uint32_t
-find_an_empty_item(struct gro_tcp4_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_tcp4_tbl *tbl)
{
@@ -95,56 +84,6 @@ find_an_empty_flow(struct gro_tcp4_tbl *tbl)
return INVALID_ARRAY_INDEX;
}
-static inline uint32_t
-insert_new_item(struct gro_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 (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].ip_id = ip_id;
- tbl->items[item_idx].nb_merged = 1;
- tbl->items[item_idx].is_atomic = is_atomic;
- 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_tcp4_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_tcp4_tbl *tbl,
struct tcp4_flow_key *src,
@@ -159,13 +98,10 @@ insert_new_flow(struct gro_tcp4_tbl *tbl,
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));
+ ASSIGN_COMMON_TCP_KEY((&src->cmn_key), (&dst->cmn_key));
+
dst->ip_src_addr = src->ip_src_addr;
dst->ip_dst_addr = src->ip_dst_addr;
- 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++;
@@ -173,21 +109,6 @@ insert_new_flow(struct gro_tcp4_tbl *tbl,
return flow_idx;
}
-/*
- * update the packet length for the flushed packet.
- */
-static inline void
-update_header(struct gro_tcp4_item *item)
-{
- struct rte_ipv4_hdr *ipv4_hdr;
- struct rte_mbuf *pkt = item->firstseg;
-
- ipv4_hdr = (struct rte_ipv4_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
- pkt->l2_len);
- ipv4_hdr->total_length = rte_cpu_to_be_16(pkt->pkt_len -
- pkt->l2_len);
-}
-
int32_t
gro_tcp4_reassemble(struct rte_mbuf *pkt,
struct gro_tcp4_tbl *tbl,
@@ -202,9 +123,8 @@ gro_tcp4_reassemble(struct rte_mbuf *pkt,
uint8_t is_atomic;
struct tcp4_flow_key key;
- uint32_t cur_idx, prev_idx, item_idx;
+ uint32_t item_idx;
uint32_t i, max_flow_num, remaining_flow_num;
- int cmp;
uint8_t find;
/*
@@ -239,6 +159,14 @@ gro_tcp4_reassemble(struct rte_mbuf *pkt,
if (tcp_dl <= 0)
return -1;
+ rte_ether_addr_copy(&(eth_hdr->src_addr), &(key.cmn_key.eth_saddr));
+ rte_ether_addr_copy(&(eth_hdr->dst_addr), &(key.cmn_key.eth_daddr));
+ key.ip_src_addr = ipv4_hdr->src_addr;
+ key.ip_dst_addr = ipv4_hdr->dst_addr;
+ key.cmn_key.src_port = tcp_hdr->src_port;
+ key.cmn_key.dst_port = tcp_hdr->dst_port;
+ key.cmn_key.recv_ack = tcp_hdr->recv_ack;
+
/*
* Save IPv4 ID for the packet whose DF bit is 0. For the packet
* whose DF bit is 1, IPv4 ID is ignored.
@@ -246,15 +174,6 @@ gro_tcp4_reassemble(struct rte_mbuf *pkt,
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->src_addr), &(key.eth_saddr));
- rte_ether_addr_copy(&(eth_hdr->dst_addr), &(key.eth_daddr));
- key.ip_src_addr = ipv4_hdr->src_addr;
- key.ip_dst_addr = ipv4_hdr->dst_addr;
- 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;
@@ -270,63 +189,44 @@ gro_tcp4_reassemble(struct rte_mbuf *pkt,
}
}
- /*
- * 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, ip_id,
- is_atomic);
+ sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
+ item_idx = insert_new_tcp_item(pkt, tbl->items, &tbl->item_num,
+ tbl->max_item_num, 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) {
+ INVALID_ARRAY_INDEX) {
/*
* Fail to insert a new flow, so delete the
* stored packet.
- */
- delete_item(tbl, item_idx, INVALID_ARRAY_INDEX);
+ */
+ delete_tcp_item(tbl->items, item_idx, &tbl->item_num, 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(&(tbl->items[cur_idx]), tcp_hdr,
- sent_seq, ip_id, pkt->l4_len, tcp_dl, 0,
- is_atomic);
- if (cmp) {
- if (merge_two_tcp4_packets(&(tbl->items[cur_idx]),
- pkt, cmp, sent_seq, ip_id, 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, cur_idx,
- sent_seq, ip_id, is_atomic) ==
- 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);
+ return process_tcp_item(pkt, tcp_hdr, tcp_dl, tbl->items, tbl->flows[i].start_index,
+ &tbl->item_num, tbl->max_item_num,
+ ip_id, is_atomic, start_time);
+}
- /* Fail to find a neighbor, so store 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;
+/*
+ * update the packet length for the flushed packet.
+ */
+static inline void
+update_header(struct gro_tcp_item *item)
+{
+ struct rte_ipv4_hdr *ipv4_hdr;
+ struct rte_mbuf *pkt = item->firstseg;
- return 0;
+ ipv4_hdr = (struct rte_ipv4_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+ pkt->l2_len);
+ ipv4_hdr->total_length = rte_cpu_to_be_16(pkt->pkt_len -
+ pkt->l2_len);
}
uint16_t
@@ -353,7 +253,8 @@ gro_tcp4_tbl_timeout_flush(struct gro_tcp4_tbl *tbl,
* Delete the packet and get the next
* packet in the flow.
*/
- j = delete_item(tbl, j, INVALID_ARRAY_INDEX);
+ j = delete_tcp_item(tbl->items, j,
+ &tbl->item_num, INVALID_ARRAY_INDEX);
tbl->flows[i].start_index = j;
if (j == INVALID_ARRAY_INDEX)
tbl->flow_num--;
@@ -5,32 +5,15 @@
#ifndef _GRO_TCP4_H_
#define _GRO_TCP4_H_
-#include <rte_tcp.h>
+#include "gro_tcp.h"
-#define INVALID_ARRAY_INDEX 0xffffffffUL
#define GRO_TCP4_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
-/*
- * The max length of a IPv4 packet, which includes the length of the L3
- * header, the L4 header and the data payload.
- */
-#define MAX_IPV4_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/IPv4 flow */
+/* Header fields representing common fields in TCP flow */
struct tcp4_flow_key {
- struct rte_ether_addr eth_saddr;
- struct rte_ether_addr eth_daddr;
+ struct cmn_tcp_key cmn_key;
uint32_t ip_src_addr;
uint32_t ip_dst_addr;
-
- uint32_t recv_ack;
- uint16_t src_port;
- uint16_t dst_port;
};
struct gro_tcp4_flow {
@@ -42,42 +25,12 @@ struct gro_tcp4_flow {
uint32_t start_index;
};
-struct gro_tcp4_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;
- /* IPv4 ID of the packet */
- uint16_t ip_id;
- /* the number of merged packets */
- uint16_t nb_merged;
- /* Indicate if IPv4 ID can be ignored */
- uint8_t is_atomic;
-};
-
/*
* TCP/IPv4 reassembly table structure.
*/
struct gro_tcp4_tbl {
/* item array */
- struct gro_tcp4_item *items;
+ struct gro_tcp_item *items;
/* flow array */
struct gro_tcp4_flow *flows;
/* current item number */
@@ -186,120 +139,9 @@ uint32_t gro_tcp4_tbl_pkt_count(void *tbl);
static inline int
is_same_tcp4_flow(struct tcp4_flow_key k1, struct tcp4_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) &&
- (k1.ip_src_addr == k2.ip_src_addr) &&
+ return ((k1.ip_src_addr == k2.ip_src_addr) &&
(k1.ip_dst_addr == k2.ip_dst_addr) &&
- (k1.recv_ack == k2.recv_ack) &&
- (k1.src_port == k2.src_port) &&
- (k1.dst_port == k2.dst_port));
+ is_same_common_tcp_key(&k1.cmn_key, &k2.cmn_key));
}
-/*
- * Merge two TCP/IPv4 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_tcp4_packets(struct gro_tcp4_item *item,
- struct rte_mbuf *pkt,
- int cmp,
- uint32_t sent_seq,
- uint16_t ip_id,
- 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 IPv4 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_IPV4_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);
- /* update IP ID to the larger value */
- item->ip_id = ip_id;
- } 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->ip_id = ip_id;
- }
- 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/IPv4 packets are neighbors.
- */
-static inline int
-check_seq_option(struct gro_tcp4_item *item,
- struct rte_tcp_hdr *tcph,
- uint32_t sent_seq,
- uint16_t ip_id,
- uint16_t tcp_hl,
- uint16_t tcp_dl,
- uint16_t l2_offset,
- uint8_t is_atomic)
-{
- struct rte_mbuf *pkt_orig = item->firstseg;
- struct rte_ipv4_hdr *iph_orig;
- struct rte_tcp_hdr *tcph_orig;
- uint16_t len, tcp_hl_orig;
-
- iph_orig = (struct rte_ipv4_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;
-
- /* Don't merge packets whose DF bits are different */
- if (unlikely(item->is_atomic ^ is_atomic))
- 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) && (is_atomic ||
- (ip_id == item->ip_id + 1)))
- /* append the new packet */
- return 1;
- else if ((sent_seq + tcp_dl == item->sent_seq) && (is_atomic ||
- (ip_id + item->nb_merged == item->ip_id)))
- /* pre-pend the new packet */
- return -1;
-
- return 0;
-}
#endif
new file mode 100644
@@ -0,0 +1,268 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2023 Intel Corporation
+ */
+
+#include <rte_malloc.h>
+#include <rte_mbuf.h>
+#include <rte_ethdev.h>
+
+#include "gro_tcp6.h"
+#include "gro_tcp_internal.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_tcp_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_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_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);
+
+ ASSIGN_COMMON_TCP_KEY((&src->cmn_key), (&dst->cmn_key));
+ memcpy(&dst->src_addr[0], &src->src_addr[0], sizeof(dst->src_addr));
+ memcpy(&dst->dst_addr[0], &src->dst_addr[0], sizeof(dst->dst_addr));
+ dst->vtc_flow = src->vtc_flow;
+
+ 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_tcp_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;
+ int32_t tcp_dl;
+ uint16_t ip_tlen;
+ struct tcp6_flow_key key;
+ uint32_t i, max_flow_num, remaining_flow_num;
+ uint32_t sent_seq;
+ struct rte_tcp_hdr *tcp_hdr;
+ uint8_t find;
+ uint32_t item_idx;
+ /*
+ * 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 = rte_pktmbuf_mtod_offset(pkt, struct rte_tcp_hdr *, pkt->l2_len + 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;
+
+ ip_tlen = rte_be_to_cpu_16(ipv6_hdr->payload_len);
+ /*
+ * Don't process the packet whose payload length is less than or
+ * equal to 0.
+ */
+ tcp_dl = ip_tlen - pkt->l4_len;
+ if (tcp_dl <= 0)
+ return -1;
+
+ rte_ether_addr_copy(&(eth_hdr->src_addr), &(key.cmn_key.eth_saddr));
+ rte_ether_addr_copy(&(eth_hdr->dst_addr), &(key.cmn_key.eth_daddr));
+ memcpy(&key.src_addr[0], &ipv6_hdr->src_addr, sizeof(key.src_addr));
+ memcpy(&key.dst_addr[0], &ipv6_hdr->dst_addr, sizeof(key.dst_addr));
+ key.cmn_key.src_port = tcp_hdr->src_port;
+ key.cmn_key.dst_port = tcp_hdr->dst_port;
+ key.cmn_key.recv_ack = tcp_hdr->recv_ack;
+ key.vtc_flow = ipv6_hdr->vtc_flow;
+
+ /* 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--;
+ }
+ }
+
+ if (find == 0) {
+ sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
+ item_idx = insert_new_tcp_item(pkt, tbl->items, &tbl->item_num,
+ tbl->max_item_num, start_time,
+ INVALID_ARRAY_INDEX, sent_seq, 0, true);
+ 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_tcp_item(tbl->items, item_idx, &tbl->item_num, INVALID_ARRAY_INDEX);
+ return -1;
+ }
+ return 0;
+ }
+
+ return process_tcp_item(pkt, tcp_hdr, tcp_dl, tbl->items, tbl->flows[i].start_index,
+ &tbl->item_num, tbl->max_item_num,
+ 0, true, start_time);
+}
+
+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_tcp_item(tbl->items, j,
+ &tbl->item_num, 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;
+}
new file mode 100644
@@ -0,0 +1,160 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2023 Intel Corporation
+ */
+
+#ifndef _GRO_TCP6_H_
+#define _GRO_TCP6_H_
+
+#include "gro_tcp.h"
+
+#define GRO_TCP6_TBL_MAX_ITEM_NUM (1024UL * 1024UL)
+
+/* Header fields representing a TCP/IPv6 flow */
+struct tcp6_flow_key {
+ struct cmn_tcp_key cmn_key;
+ uint8_t src_addr[16];
+ uint8_t dst_addr[16];
+ rte_be32_t vtc_flow;
+};
+
+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;
+};
+
+/*
+ * TCP/IPv6 reassembly table structure.
+ */
+struct gro_tcp6_tbl {
+ /* item array */
+ struct gro_tcp_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 doesn't process the packet,
+ * which has SYN, FIN, RST, PSH, CWR, ECE or URG set, or doesn't have
+ * payload.
+ *
+ * This function doesn't check if the packet has correct checksums and
+ * doesn't 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 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)
+{
+ rte_be32_t vtc_flow_diff;
+
+ if (memcmp(&k1->src_addr, &k2->src_addr, 16))
+ return 0;
+ if (memcmp(&k1->dst_addr, &k2->dst_addr, 16))
+ return 0;
+ /*
+ * IP version (4) Traffic Class (8) Flow Label (20)
+ * All fields except Traffic class should be same
+ */
+ vtc_flow_diff = (k1->vtc_flow ^ k2->vtc_flow);
+ if (vtc_flow_diff & htonl(0xF00FFFFF))
+ return 0;
+
+ return is_same_common_tcp_key(&k1->cmn_key, &k2->cmn_key);
+}
+
+#endif
new file mode 100644
@@ -0,0 +1,128 @@
+
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2023 Intel Corporation
+ */
+
+#ifndef _GRO_TCP_INTERNAL_H_
+#define _GRO_TCP_INTERNAL_H_
+
+static inline uint32_t
+find_an_empty_item(struct gro_tcp_item *items,
+ uint32_t max_item_num)
+{
+ uint32_t i;
+
+ for (i = 0; i < max_item_num; i++)
+ if (items[i].firstseg == NULL)
+ return i;
+ return INVALID_ARRAY_INDEX;
+}
+
+static inline uint32_t
+insert_new_tcp_item(struct rte_mbuf *pkt,
+ struct gro_tcp_item *items,
+ uint32_t *item_num,
+ uint32_t max_item_num,
+ 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(items, max_item_num);
+ if (item_idx == INVALID_ARRAY_INDEX)
+ return INVALID_ARRAY_INDEX;
+
+ items[item_idx].firstseg = pkt;
+ items[item_idx].lastseg = rte_pktmbuf_lastseg(pkt);
+ items[item_idx].start_time = start_time;
+ items[item_idx].next_pkt_idx = INVALID_ARRAY_INDEX;
+ items[item_idx].sent_seq = sent_seq;
+ items[item_idx].l3.ip_id = ip_id;
+ items[item_idx].nb_merged = 1;
+ items[item_idx].is_atomic = is_atomic;
+ (*item_num) += 1;
+
+ /* if the previous packet exists, chain them together. */
+ if (prev_idx != INVALID_ARRAY_INDEX) {
+ items[item_idx].next_pkt_idx =
+ items[prev_idx].next_pkt_idx;
+ items[prev_idx].next_pkt_idx = item_idx;
+ }
+
+ return item_idx;
+}
+
+static inline uint32_t
+delete_tcp_item(struct gro_tcp_item *items, uint32_t item_idx,
+ uint32_t *item_num,
+ uint32_t prev_item_idx)
+{
+ uint32_t next_idx = items[item_idx].next_pkt_idx;
+
+ /* NULL indicates an empty item */
+ items[item_idx].firstseg = NULL;
+ (*item_num) -= 1;
+ if (prev_item_idx != INVALID_ARRAY_INDEX)
+ items[prev_item_idx].next_pkt_idx = next_idx;
+
+ return next_idx;
+}
+
+static inline int32_t
+process_tcp_item(struct rte_mbuf *pkt,
+ struct rte_tcp_hdr *tcp_hdr,
+ int32_t tcp_dl,
+ struct gro_tcp_item *items,
+ uint32_t item_idx,
+ uint32_t *item_num,
+ uint32_t max_item_num,
+ uint16_t ip_id,
+ uint8_t is_atomic,
+ uint64_t start_time)
+{
+ uint32_t cur_idx;
+ uint32_t prev_idx;
+ int cmp;
+ uint32_t sent_seq;
+
+ sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
+ /*
+ * Check all packets in the flow and try to find a neighbor for
+ * the input packet.
+ */
+ cur_idx = item_idx;
+ prev_idx = cur_idx;
+ do {
+ cmp = check_seq_option(&items[cur_idx], tcp_hdr,
+ sent_seq, ip_id, pkt->l4_len, tcp_dl, 0,
+ is_atomic);
+ if (cmp) {
+ if (merge_two_tcp_packets(&items[cur_idx],
+ pkt, cmp, sent_seq, ip_id, 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_tcp_item(pkt, items, item_num, max_item_num,
+ start_time, cur_idx, sent_seq, ip_id, is_atomic) ==
+ INVALID_ARRAY_INDEX)
+ return -1;
+ return 0;
+ }
+ prev_idx = cur_idx;
+ cur_idx = 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_tcp_item(pkt, items, item_num, max_item_num, start_time, prev_idx, sent_seq,
+ ip_id, is_atomic) == INVALID_ARRAY_INDEX)
+ return -1;
+
+ return 0;
+}
+#endif
@@ -116,7 +116,7 @@ insert_new_item(struct gro_vxlan_tcp4_tbl *tbl,
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.l3.ip_id = ip_id;
tbl->items[item_idx].inner_item.nb_merged = 1;
tbl->items[item_idx].inner_item.is_atomic = is_atomic;
tbl->items[item_idx].outer_ip_id = outer_ip_id;
@@ -163,15 +163,9 @@ insert_new_flow(struct gro_vxlan_tcp4_tbl *tbl,
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));
+ ASSIGN_COMMON_TCP_KEY((&(src->inner_key.cmn_key)), (&(dst->inner_key.cmn_key)));
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;
@@ -248,7 +242,7 @@ merge_two_vxlan_tcp4_packets(struct gro_vxlan_tcp4_item *item,
uint16_t outer_ip_id,
uint16_t ip_id)
{
- if (merge_two_tcp4_packets(&item->inner_item, pkt, cmp, sent_seq,
+ if (merge_two_tcp_packets(&item->inner_item, pkt, cmp, sent_seq,
ip_id, pkt->outer_l2_len +
pkt->outer_l3_len)) {
/* Update the outer IPv4 ID to the large value. */
@@ -357,13 +351,13 @@ gro_vxlan_tcp4_reassemble(struct rte_mbuf *pkt,
sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
- rte_ether_addr_copy(&(eth_hdr->src_addr), &(key.inner_key.eth_saddr));
- rte_ether_addr_copy(&(eth_hdr->dst_addr), &(key.inner_key.eth_daddr));
+ rte_ether_addr_copy(&(eth_hdr->src_addr), &(key.inner_key.cmn_key.eth_saddr));
+ rte_ether_addr_copy(&(eth_hdr->dst_addr), &(key.inner_key.cmn_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.inner_key.cmn_key.recv_ack = tcp_hdr->recv_ack;
+ key.inner_key.cmn_key.src_port = tcp_hdr->src_port;
+ key.inner_key.cmn_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;
@@ -36,7 +36,7 @@ struct gro_vxlan_tcp4_flow {
};
struct gro_vxlan_tcp4_item {
- struct gro_tcp4_item inner_item;
+ struct gro_tcp_item inner_item;
/* IPv4 ID in the outer IPv4 header */
uint16_t outer_ip_id;
/* Indicate if outer IPv4 ID can be ignored */
@@ -4,6 +4,7 @@
sources = files(
'rte_gro.c',
'gro_tcp4.c',
+ 'gro_tcp6.c',
'gro_udp4.c',
'gro_vxlan_tcp4.c',
'gro_vxlan_udp4.c',
@@ -8,6 +8,7 @@
#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"
@@ -20,14 +21,16 @@ typedef uint32_t (*gro_tbl_pkt_count_fn)(void *tbl);
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,
+ 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,
+ gro_tcp6_tbl_pkt_count,
NULL};
#define IS_IPV4_TCP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
@@ -35,6 +38,12 @@ static gro_tbl_pkt_count_fn tbl_pkt_count_fn[RTE_GRO_TYPE_MAX_NUM] = {
((ptype & RTE_PTYPE_L4_FRAG) != RTE_PTYPE_L4_FRAG) && \
(RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
+/* GRO with extension headers is not supported */
+#define IS_IPV6_TCP_PKT(ptype) (RTE_ETH_IS_IPV6_HDR(ptype) && \
+ ((ptype & RTE_PTYPE_L4_TCP) == RTE_PTYPE_L4_TCP) && \
+ ((ptype & RTE_PTYPE_L4_FRAG) != RTE_PTYPE_L4_FRAG) && \
+ (RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
+
#define IS_IPV4_UDP_PKT(ptype) (RTE_ETH_IS_IPV4_HDR(ptype) && \
((ptype & RTE_PTYPE_L4_UDP) == RTE_PTYPE_L4_UDP) && \
(RTE_ETH_IS_TUNNEL_PKT(ptype) == 0))
@@ -147,7 +156,11 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
/* allocate a reassembly table for TCP/IPv4 GRO */
struct gro_tcp4_tbl tcp_tbl;
struct gro_tcp4_flow tcp_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
- struct gro_tcp4_item tcp_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
+ struct gro_tcp_item tcp_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
+
+ struct gro_tcp6_tbl tcp6_tbl;
+ struct gro_tcp6_flow tcp6_flows[RTE_GRO_MAX_BURST_ITEM_NUM];
+ struct gro_tcp_item tcp6_items[RTE_GRO_MAX_BURST_ITEM_NUM] = {{0} };
/* allocate a reassembly table for UDP/IPv4 GRO */
struct gro_udp4_tbl udp_tbl;
@@ -171,10 +184,10 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
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_TCP_IPV4 | RTE_GRO_TCP_IPV6 |
RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
RTE_GRO_UDP_IPV4)) == 0))
return nb_pkts;
@@ -236,6 +249,18 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
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++) {
/*
@@ -276,6 +301,14 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
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];
}
@@ -283,9 +316,17 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
if ((nb_after_gro < nb_pkts)
|| (unprocess_num < nb_pkts)) {
i = 0;
+ /* Copy unprocessed packets */
+ if (unprocess_num > 0) {
+ memcpy(&pkts[i], unprocess_pkts,
+ sizeof(struct rte_mbuf *) *
+ unprocess_num);
+ i = unprocess_num;
+ }
+
/* Flush all packets from the tables */
if (do_vxlan_tcp_gro) {
- i = gro_vxlan_tcp4_tbl_timeout_flush(&vxlan_tcp_tbl,
+ i += gro_vxlan_tcp4_tbl_timeout_flush(&vxlan_tcp_tbl,
0, pkts, nb_pkts);
}
@@ -304,13 +345,11 @@ rte_gro_reassemble_burst(struct rte_mbuf **pkts,
i += gro_udp4_tbl_timeout_flush(&udp_tbl, 0,
&pkts[i], nb_pkts - i);
}
- /* Copy unprocessed packets */
- if (unprocess_num > 0) {
- memcpy(&pkts[i], unprocess_pkts,
- sizeof(struct rte_mbuf *) *
- unprocess_num);
+
+ if (do_tcp6_gro) {
+ i += gro_tcp6_tbl_timeout_flush(&tcp6_tbl, 0,
+ &pkts[i], nb_pkts - i);
}
- nb_after_gro = i + unprocess_num;
}
return nb_after_gro;
@@ -323,13 +362,13 @@ rte_gro_reassemble(struct rte_mbuf **pkts,
{
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_TCP_IPV4 | RTE_GRO_TCP_IPV6 |
RTE_GRO_IPV4_VXLAN_UDP_IPV4 |
RTE_GRO_UDP_IPV4)) == 0))
return nb_pkts;
@@ -338,6 +377,7 @@ rte_gro_reassemble(struct rte_mbuf **pkts,
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;
@@ -347,6 +387,7 @@ rte_gro_reassemble(struct rte_mbuf **pkts,
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();
@@ -371,6 +412,11 @@ rte_gro_reassemble(struct rte_mbuf **pkts,
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];
}
@@ -426,6 +472,15 @@ rte_gro_timeout_flush(void *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;
@@ -38,6 +38,9 @@ extern "C" {
#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