DPDK patches and discussions
 help / color / mirror / Atom feed
From: Jiayu Hu <jiayu.hu@intel.com>
To: dev@dpdk.org
Cc: mark.b.kavanagh@intel.com, konstantin.ananyev@intel.com,
	jianfeng.tan@intel.com, Jiayu Hu <jiayu.hu@intel.com>
Subject: [dpdk-dev] [PATCH 2/5] gso/lib: add TCP/IPv4 GSO support
Date: Thu, 24 Aug 2017 22:15:41 +0800	[thread overview]
Message-ID: <1503584144-63181-3-git-send-email-jiayu.hu@intel.com> (raw)
In-Reply-To: <1503584144-63181-1-git-send-email-jiayu.hu@intel.com>

This patch adds GSO support for TCP/IPv4 packets. Supported packets
may include a single VLAN tag. TCP/IPv4 GSO assumes that all input
packets have correct checksums, and doesn't update checksums for output
packets (the responsibility for this lies with the application).
Additionally, TCP/IPv4 GSO doesn't process IP fragmented packets.

TCP/IPv4 GSO uses two chained MBUFs, one direct MBUF and one indrect
MBUF, to organize an output packet. Note that we refer to these two
chained MBUFs as a two-segment MBUF. The direct MBUF stores the packet
header, while the indirect mbuf simply points to a location within the
original packet's payload. Consequently, use of the GSO library requires
multi-segment MBUF support in the TX functions of the NIC driver.

If a packet is GSOed, TCP/IPv4 GSO reduces its MBUF refcnt by 1. As a
result, when all of its GSOed segments are freed, the packet is freed
automatically.

Signed-off-by: Jiayu Hu <jiayu.hu@intel.com>
Signed-off-by: Mark Kavanagh <mark.b.kavanagh@intel.com>
---
 lib/librte_eal/common/include/rte_log.h |   1 +
 lib/librte_gso/Makefile                 |   2 +
 lib/librte_gso/gso_common.c             | 270 ++++++++++++++++++++++++++++++++
 lib/librte_gso/gso_common.h             | 120 ++++++++++++++
 lib/librte_gso/gso_tcp.c                |  82 ++++++++++
 lib/librte_gso/gso_tcp.h                |  73 +++++++++
 lib/librte_gso/rte_gso.c                |  44 +++++-
 lib/librte_gso/rte_gso.h                |   3 +
 8 files changed, 593 insertions(+), 2 deletions(-)
 create mode 100644 lib/librte_gso/gso_common.c
 create mode 100644 lib/librte_gso/gso_common.h
 create mode 100644 lib/librte_gso/gso_tcp.c
 create mode 100644 lib/librte_gso/gso_tcp.h

diff --git a/lib/librte_eal/common/include/rte_log.h b/lib/librte_eal/common/include/rte_log.h
index ec8dba7..2fa1199 100644
--- a/lib/librte_eal/common/include/rte_log.h
+++ b/lib/librte_eal/common/include/rte_log.h
@@ -87,6 +87,7 @@ extern struct rte_logs rte_logs;
 #define RTE_LOGTYPE_CRYPTODEV 17 /**< Log related to cryptodev. */
 #define RTE_LOGTYPE_EFD       18 /**< Log related to EFD. */
 #define RTE_LOGTYPE_EVENTDEV  19 /**< Log related to eventdev. */
+#define RTE_LOGTYPE_GSO       20 /**< Log related to GSO. */
 
 /* these log types can be used in an application */
 #define RTE_LOGTYPE_USER1     24 /**< User-defined log type 1. */
diff --git a/lib/librte_gso/Makefile b/lib/librte_gso/Makefile
index aeaacbc..0f8e38f 100644
--- a/lib/librte_gso/Makefile
+++ b/lib/librte_gso/Makefile
@@ -42,6 +42,8 @@ LIBABIVER := 1
 
 #source files
 SRCS-$(CONFIG_RTE_LIBRTE_GSO) += rte_gso.c
+SRCS-$(CONFIG_RTE_LIBRTE_GSO) += gso_common.c
+SRCS-$(CONFIG_RTE_LIBRTE_GSO) += gso_tcp.c
 
 # install this header file
 SYMLINK-$(CONFIG_RTE_LIBRTE_GSO)-include += rte_gso.h
diff --git a/lib/librte_gso/gso_common.c b/lib/librte_gso/gso_common.c
new file mode 100644
index 0000000..2b54fbd
--- /dev/null
+++ b/lib/librte_gso/gso_common.c
@@ -0,0 +1,270 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2017 Intel Corporation. All rights reserved.
+ *   All rights reserved.
+ *
+ *   Redistribution and use in source and binary forms, with or without
+ *   modification, are permitted provided that the following conditions
+ *   are met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in
+ *       the documentation and/or other materials provided with the
+ *       distribution.
+ *     * Neither the name of Intel Corporation nor the names of its
+ *       contributors may be used to endorse or promote products derived
+ *       from this software without specific prior written permission.
+ *
+ *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <stdbool.h>
+#include <string.h>
+
+#include <rte_malloc.h>
+
+#include <rte_ether.h>
+#include <rte_ip.h>
+#include <rte_tcp.h>
+
+#include "gso_common.h"
+
+static inline void
+hdr_segment_init(struct rte_mbuf *hdr_segment, struct rte_mbuf *pkt,
+		uint16_t pkt_hdr_offset)
+{
+	/* copy mbuf metadata */
+	hdr_segment->nb_segs = 1;
+	hdr_segment->port = pkt->port;
+	hdr_segment->ol_flags = pkt->ol_flags;
+	hdr_segment->packet_type = pkt->packet_type;
+	hdr_segment->pkt_len = pkt_hdr_offset;
+	hdr_segment->data_len = pkt_hdr_offset;
+	hdr_segment->tx_offload = pkt->tx_offload;
+	/* copy packet header */
+	rte_memcpy(rte_pktmbuf_mtod(hdr_segment, char *),
+			rte_pktmbuf_mtod(pkt, char *),
+			pkt_hdr_offset);
+}
+
+static inline void
+free_gso_segment(struct rte_mbuf **pkts, uint16_t nb_pkts)
+{
+	uint16_t i;
+
+	for (i = 0; i < nb_pkts; i++) {
+		rte_pktmbuf_detach(pkts[i]->next);
+		rte_pktmbuf_free(pkts[i]);
+		pkts[i] = NULL;
+	}
+}
+
+int
+gso_do_segment(struct rte_mbuf *pkt,
+		uint16_t pkt_hdr_offset,
+		uint16_t pyld_unit_size,
+		struct rte_mempool *direct_pool,
+		struct rte_mempool *indirect_pool,
+		struct rte_mbuf **pkts_out,
+		uint16_t nb_pkts_out)
+{
+	struct rte_mbuf *pkt_in;
+	struct rte_mbuf *hdr_segment, *pyld_segment;
+	uint32_t pkt_in_pyld_off;
+	uint16_t pkt_in_segment_len, pkt_out_segment_len;
+	uint16_t nb_segs;
+	bool pkt_in_segment_processed;
+
+	pkt_in_pyld_off = pkt->data_off + pkt_hdr_offset;
+	pkt_in = pkt;
+	nb_segs = 0;
+
+	while (pkt_in) {
+		pkt_in_segment_processed = false;
+		pkt_in_segment_len = pkt_in->data_off + pkt_in->data_len;
+
+		while (!pkt_in_segment_processed) {
+			if (unlikely(nb_segs >= nb_pkts_out)) {
+				free_gso_segment(pkts_out, nb_segs);
+				return -EINVAL;
+			}
+
+			/* allocate direct mbuf */
+			hdr_segment = rte_pktmbuf_alloc(direct_pool);
+			if (unlikely(hdr_segment == NULL)) {
+				free_gso_segment(pkts_out, nb_segs);
+				return -ENOMEM;
+			}
+
+			/* allocate indirect mbuf */
+			pyld_segment = rte_pktmbuf_alloc(indirect_pool);
+			if (unlikely(pyld_segment == NULL)) {
+				rte_pktmbuf_free(hdr_segment);
+				free_gso_segment(pkts_out, nb_segs);
+				return -ENOMEM;
+			}
+
+			/* copy packet header */
+			hdr_segment_init(hdr_segment, pkt, pkt_hdr_offset);
+
+			/* attach payload mbuf to current packet segment */
+			rte_pktmbuf_attach(pyld_segment, pkt_in);
+
+			hdr_segment->next = pyld_segment;
+			pkts_out[nb_segs++] = hdr_segment;
+
+			/* calculate payload length */
+			pkt_out_segment_len = pyld_unit_size;
+			if (pkt_in_pyld_off + pkt_out_segment_len >
+					pkt_in_segment_len) {
+				pkt_out_segment_len = pkt_in_segment_len -
+					pkt_in_pyld_off;
+			}
+
+			/* update payload segment */
+			pyld_segment->data_off = pkt_in_pyld_off;
+			pyld_segment->data_len = pkt_out_segment_len;
+
+			/* update header segment */
+			hdr_segment->pkt_len += pyld_segment->data_len;
+			hdr_segment->nb_segs++;
+
+			/* update pkt_in_pyld_off */
+			pkt_in_pyld_off += pkt_out_segment_len;
+			if (pkt_in_pyld_off == pkt_in_segment_len)
+				pkt_in_segment_processed = true;
+		}
+
+		/* 'pkt_in' may contain numerous segments */
+		pkt_in = pkt_in->next;
+		if (pkt_in != NULL)
+			pkt_in_pyld_off = pkt_in->data_off;
+	}
+	return nb_segs;
+}
+
+static inline void
+parse_ipv4(struct ipv4_hdr *ipv4_hdr, struct rte_mbuf *pkt)
+{
+	struct tcp_hdr *tcp_hdr;
+
+	switch (ipv4_hdr->next_proto_id) {
+	case IPPROTO_TCP:
+		pkt->packet_type |= RTE_PTYPE_L4_TCP;
+		pkt->l3_len = IPv4_HDR_LEN(ipv4_hdr);
+		tcp_hdr = (struct tcp_hdr *)((char *)ipv4_hdr + pkt->l3_len);
+		pkt->l4_len = TCP_HDR_LEN(tcp_hdr);
+		break;
+	}
+}
+
+static inline void
+parse_ethernet(struct ether_hdr *eth_hdr, struct rte_mbuf *pkt)
+{
+	struct ipv4_hdr *ipv4_hdr;
+	struct vlan_hdr *vlan_hdr;
+	uint16_t ethertype;
+
+	ethertype = rte_be_to_cpu_16(eth_hdr->ether_type);
+	if (ethertype == ETHER_TYPE_VLAN) {
+		vlan_hdr = (struct vlan_hdr *)(eth_hdr + 1);
+		pkt->l2_len = sizeof(struct vlan_hdr);
+		pkt->packet_type |= RTE_PTYPE_L2_ETHER_VLAN;
+		ethertype = rte_be_to_cpu_16(vlan_hdr->eth_proto);
+	}
+
+	switch (ethertype) {
+	case ETHER_TYPE_IPv4:
+		if (IS_VLAN_PKT(pkt)) {
+			pkt->packet_type |= RTE_PTYPE_L3_IPV4;
+		} else {
+			pkt->packet_type |= RTE_PTYPE_L2_ETHER;
+			pkt->packet_type |= RTE_PTYPE_L3_IPV4;
+		}
+		pkt->l2_len += sizeof(struct ether_hdr);
+		ipv4_hdr = (struct ipv4_hdr *) ((char *)eth_hdr +
+				pkt->l2_len);
+		parse_ipv4(ipv4_hdr, pkt);
+		break;
+	}
+}
+
+void
+gso_parse_packet(struct rte_mbuf *pkt)
+{
+	struct ether_hdr *eth_hdr = rte_pktmbuf_mtod(pkt, struct ether_hdr *);
+
+	pkt->packet_type = pkt->tx_offload = 0;
+	parse_ethernet(eth_hdr, pkt);
+}
+
+static inline void
+update_ipv4_header(char *base, uint16_t offset, uint16_t length, uint16_t id)
+{
+	struct ipv4_hdr *ipv4_hdr = (struct ipv4_hdr *)(base + offset);
+
+	ipv4_hdr->total_length = rte_cpu_to_be_16(length - offset);
+	ipv4_hdr->packet_id = rte_cpu_to_be_16(id);
+}
+
+static inline void
+update_tcp_header(char *base, uint16_t offset, uint32_t sent_seq,
+	uint8_t non_tail)
+{
+	struct tcp_hdr *tcp_hdr = (struct tcp_hdr *)(base + offset);
+
+	tcp_hdr->sent_seq = rte_cpu_to_be_32(sent_seq);
+	/* clean FIN and PSH for non-tail segments */
+	if (non_tail)
+		tcp_hdr->tcp_flags &= (~(TCP_HDR_PSH_MASK | TCP_HDR_FIN_MASK));
+}
+
+void
+gso_update_pkt_headers(struct rte_mbuf *pkt, uint16_t nb_segments,
+		struct rte_mbuf **out_segments)
+{
+	struct ipv4_hdr *ipv4_hdr;
+	struct tcp_hdr *tcp_hdr;
+	struct rte_mbuf *seg;
+	uint32_t sent_seq;
+	uint16_t offset, i;
+	uint16_t tail_seg_idx = nb_segments - 1, id;
+
+	switch (pkt->packet_type) {
+	case ETHER_VLAN_IPv4_TCP_PKT:
+	case ETHER_IPv4_TCP_PKT:
+		ipv4_hdr = (struct ipv4_hdr *)(rte_pktmbuf_mtod(pkt, char *) +
+				pkt->l2_len);
+		tcp_hdr = (struct tcp_hdr *)((char *)ipv4_hdr + pkt->l3_len);
+		id = rte_be_to_cpu_16(ipv4_hdr->packet_id);
+		sent_seq = rte_be_to_cpu_32(tcp_hdr->sent_seq);
+
+		for (i = 0; i < nb_segments; i++) {
+			seg = out_segments[i];
+
+			offset = seg->l2_len;
+			update_ipv4_header(rte_pktmbuf_mtod(seg, char *),
+					offset, seg->pkt_len, id);
+			id++;
+
+			offset += seg->l3_len;
+			update_tcp_header(rte_pktmbuf_mtod(seg, char *),
+					offset, sent_seq, i < tail_seg_idx);
+			sent_seq += seg->next->data_len;
+		}
+		break;
+	}
+}
diff --git a/lib/librte_gso/gso_common.h b/lib/librte_gso/gso_common.h
new file mode 100644
index 0000000..d750041
--- /dev/null
+++ b/lib/librte_gso/gso_common.h
@@ -0,0 +1,120 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2017 Intel Corporation. All rights reserved.
+ *   All rights reserved.
+ *
+ *   Redistribution and use in source and binary forms, with or without
+ *   modification, are permitted provided that the following conditions
+ *   are met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in
+ *       the documentation and/or other materials provided with the
+ *       distribution.
+ *     * Neither the name of Intel Corporation nor the names of its
+ *       contributors may be used to endorse or promote products derived
+ *       from this software without specific prior written permission.
+ *
+ *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _GSO_COMMON_H_
+#define _GSO_COMMON_H_
+
+#include <stdint.h>
+#include <rte_mbuf.h>
+
+#define IPV4_HDR_DF_SHIFT 14
+#define IPV4_HDR_DF_MASK (1 << IPV4_HDR_DF_SHIFT)
+#define IPv4_HDR_LEN(iph) ((iph->version_ihl & 0x0f) * 4)
+
+#define TCP_HDR_PSH_MASK ((uint8_t)0x08)
+#define TCP_HDR_FIN_MASK ((uint8_t)0x01)
+#define TCP_HDR_LEN(tcph) ((tcph->data_off & 0xf0) >> 2)
+
+#define ETHER_IPv4_PKT (RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4)
+/* Supported packet types */
+/* TCP/IPv4 packet. */
+#define ETHER_IPv4_TCP_PKT (ETHER_IPv4_PKT | RTE_PTYPE_L4_TCP)
+
+/* TCP/IPv4 packet with VLAN tag. */
+#define ETHER_VLAN_IPv4_TCP_PKT (RTE_PTYPE_L2_ETHER_VLAN | \
+		RTE_PTYPE_L3_IPV4 | RTE_PTYPE_L4_TCP)
+
+#define IS_VLAN_PKT(pkt) ((pkt->packet_type & RTE_PTYPE_L2_ETHER_VLAN) == \
+		RTE_PTYPE_L2_ETHER_VLAN)
+
+/**
+ * Internal function which parses a packet, setting outer_l2/l3_len and
+ * l2/l3/l4_len and packet_type.
+ *
+ * @param pkt
+ *  Packet to parse.
+ */
+void gso_parse_packet(struct rte_mbuf *pkt);
+
+/**
+ * Internal function which updates relevant packet headers, following
+ * segmentation. This is required to update, for example, the IPv4
+ * 'total_length' field, to reflect the reduced length of the now-
+ * segmented packet.
+ *
+ * @param pkt
+ *  The original packet.
+ * @param nb_segments
+ *  The number of GSO segments into which pkt was split.
+ * @param out_segements
+ *  Pointer array used for storing mbuf addresses for GSO segments.
+ */
+void gso_update_pkt_headers(struct rte_mbuf *pkt, uint16_t nb_segments,
+		struct rte_mbuf **out_segments);
+
+/**
+ * Internal function which divides the input packet into small segments.
+ * Each of the newly-created segments is organized as a two-segment mbuf,
+ * where the first segment is a standard mbuf, which stores a copy of
+ * packet header, and the second is an indirect mbuf which points to a
+ * section of data in the input packet.
+ *
+ * @param pkt
+ *  Packet to segment.
+ * @param pkt_hdr_offset
+ *  Packet header offset, measured in byte.
+ * @param pyld_unit_size
+ *  The max payload length of a GSO segment.
+ * @param direct_pool
+ *  MBUF pool used for allocating direct buffers for output segments.
+ * @param indirect_pool
+ *  MBUF pool used for allocating indirect buffers for output segments.
+ * @param pkts_out
+ *  Pointer array used to keep the mbuf addresses of output segments.
+ * @param nb_pkts_out
+ *  The max number of items that pkts_out can keep.
+ *
+ * @return
+ *  - The number of segments created in the event of success.
+ *  - If no GSO is performed, return 1.
+ *  - If available memory in mempools is insufficient, return -ENOMEM.
+ *  - -EINVAL for invalid parameters
+ */
+int gso_do_segment(struct rte_mbuf *pkt,
+		uint16_t pkt_hdr_offset,
+		uint16_t pyld_unit_size,
+		struct rte_mempool *direct_pool,
+		struct rte_mempool *indirect_pool,
+		struct rte_mbuf **pkts_out,
+		uint16_t nb_pkts_out);
+#endif
diff --git a/lib/librte_gso/gso_tcp.c b/lib/librte_gso/gso_tcp.c
new file mode 100644
index 0000000..9d5fc30
--- /dev/null
+++ b/lib/librte_gso/gso_tcp.c
@@ -0,0 +1,82 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2017 Intel Corporation. All rights reserved.
+ *   All rights reserved.
+ *
+ *   Redistribution and use in source and binary forms, with or without
+ *   modification, are permitted provided that the following conditions
+ *   are met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in
+ *       the documentation and/or other materials provided with the
+ *       distribution.
+ *     * Neither the name of Intel Corporation nor the names of its
+ *       contributors may be used to endorse or promote products derived
+ *       from this software without specific prior written permission.
+ *
+ *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+
+#include <rte_ether.h>
+#include <rte_ip.h>
+#include <rte_tcp.h>
+
+#include "gso_common.h"
+#include "gso_tcp.h"
+
+int
+gso_tcp4_segment(struct rte_mbuf *pkt,
+		uint16_t gso_size,
+		struct rte_mempool *direct_pool,
+		struct rte_mempool *indirect_pool,
+		struct rte_mbuf **pkts_out,
+		uint16_t nb_pkts_out)
+{
+	struct ether_hdr *eth_hdr;
+	struct ipv4_hdr *ipv4_hdr;
+	uint16_t tcp_dl;
+	uint16_t pyld_unit_size;
+	uint16_t hdr_offset;
+	int ret = 1;
+
+	eth_hdr = rte_pktmbuf_mtod(pkt, struct ether_hdr *);
+	ipv4_hdr = (struct ipv4_hdr *)((char *)eth_hdr + pkt->l2_len);
+
+	/* don't process fragmented packet */
+	if ((ipv4_hdr->fragment_offset &
+				rte_cpu_to_be_16(IPV4_HDR_DF_MASK)) == 0)
+		return ret;
+
+	tcp_dl = rte_be_to_cpu_16(ipv4_hdr->total_length) -
+		pkt->l3_len - pkt->l4_len;
+	/* don't process packet without data */
+	if (tcp_dl == 0)
+		return ret;
+
+	hdr_offset = pkt->l2_len + pkt->l3_len + pkt->l4_len;
+	pyld_unit_size = gso_size - hdr_offset - ETHER_CRC_LEN;
+
+	/* segment the payload */
+	ret = gso_do_segment(pkt, hdr_offset, pyld_unit_size, direct_pool,
+			indirect_pool, pkts_out, nb_pkts_out);
+
+	if (ret > 1)
+		gso_update_pkt_headers(pkt, ret, pkts_out);
+
+	return ret;
+}
diff --git a/lib/librte_gso/gso_tcp.h b/lib/librte_gso/gso_tcp.h
new file mode 100644
index 0000000..f291ccb
--- /dev/null
+++ b/lib/librte_gso/gso_tcp.h
@@ -0,0 +1,73 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright(c) 2017 Intel Corporation. All rights reserved.
+ *   All rights reserved.
+ *
+ *   Redistribution and use in source and binary forms, with or without
+ *   modification, are permitted provided that the following conditions
+ *   are met:
+ *
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in
+ *       the documentation and/or other materials provided with the
+ *       distribution.
+ *     * Neither the name of Intel Corporation nor the names of its
+ *       contributors may be used to endorse or promote products derived
+ *       from this software without specific prior written permission.
+ *
+ *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _GSO_TCP_H_
+#define _GSO_TCP_H_
+
+#include <stdint.h>
+#include <rte_mbuf.h>
+
+/**
+ * Segment an IPv4/TCP packet. This function assumes the input packet has
+ * correct checksums and doesn't update checksums for GSO segment.
+ * Furthermore, it doesn't process IP fragment packets.
+ *
+ * @param pkt
+ *  The packet mbuf to segment.
+ * @param gso_size
+ *  The max length of a GSO segment, measured in bytes.
+ * @param direct_pool
+ *  MBUF pool used for allocating direct buffers for output segments.
+ * @param indirect_pool
+ *  MBUF pool used for allocating indirect buffers for output segments.
+ * @param pkts_out
+ *  Pointer array, which is used to store mbuf addresses of GSO segments.
+ *  Caller should guarantee that 'pkts_out' is sufficiently large to store
+ *  all GSO segments.
+ * @param nb_pkts_out
+ *  The max number of items that 'pkts_out' can keep.
+ *
+ * @return
+ *   - The number of GSO segments on success.
+ *   - Return 1 if no GSO is performed.
+ *   - Return -ENOMEM if available memory in mempools is insufficient.
+ *   - Return -EINVAL for invalid parameters.
+ */
+int gso_tcp4_segment(struct rte_mbuf *pkt,
+		uint16_t gso_size,
+		struct rte_mempool *direct_pool,
+		struct rte_mempool *indirect_pool,
+		struct rte_mbuf **pkts_out,
+		uint16_t nb_pkts_out);
+
+#endif
diff --git a/lib/librte_gso/rte_gso.c b/lib/librte_gso/rte_gso.c
index b81afce..fac95f2 100644
--- a/lib/librte_gso/rte_gso.c
+++ b/lib/librte_gso/rte_gso.c
@@ -31,17 +31,57 @@
  *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  */
 
+#include <rte_log.h>
+
 #include "rte_gso.h"
+#include "gso_common.h"
+#include "gso_tcp.h"
 
 int
 rte_gso_segment(struct rte_mbuf *pkt,
 		struct rte_gso_ctx gso_ctx,
 		struct rte_mbuf **pkts_out,
-		uint16_t nb_pkts_out __rte_unused)
+		uint16_t nb_pkts_out)
 {
+	struct rte_mempool *direct_pool, *indirect_pool;
+	struct rte_mbuf *pkt_seg;
+	uint16_t nb_segments, gso_size;
+
 	if (pkt == NULL || pkts_out == NULL || gso_ctx.direct_pool ==
 			NULL || gso_ctx.indirect_pool == NULL)
 		return -EINVAL;
 
-	return 1;
+	if ((gso_ctx.gso_types & RTE_GSO_TCP_IPV4) == 0 ||
+			gso_ctx.gso_size >= pkt->pkt_len ||
+			gso_ctx.gso_size == 0)
+		return 1;
+
+	pkt_seg = pkt;
+	gso_size = gso_ctx.gso_size;
+	direct_pool = gso_ctx.direct_pool;
+	indirect_pool = gso_ctx.indirect_pool;
+
+	/* Parse packet headers to determine how to segment 'pkt' */
+	gso_parse_packet(pkt);
+
+	switch (pkt->packet_type) {
+	case ETHER_VLAN_IPv4_TCP_PKT:
+	case ETHER_IPv4_TCP_PKT:
+		nb_segments = gso_tcp4_segment(pkt, gso_size,
+				direct_pool, indirect_pool,
+				pkts_out, nb_pkts_out);
+		break;
+	default:
+		RTE_LOG(WARNING, GSO, "Unsupported packet type\n");
+		nb_segments = 1;
+	}
+
+	if (nb_segments > 1) {
+		while (pkt_seg) {
+			rte_mbuf_refcnt_update(pkt_seg, -1);
+			pkt_seg = pkt_seg->next;
+		}
+	}
+
+	return nb_segments;
 }
diff --git a/lib/librte_gso/rte_gso.h b/lib/librte_gso/rte_gso.h
index 5a8389a..77853fa 100644
--- a/lib/librte_gso/rte_gso.h
+++ b/lib/librte_gso/rte_gso.h
@@ -46,6 +46,9 @@ extern "C" {
 #include <stdint.h>
 #include <rte_mbuf.h>
 
+#define RTE_GSO_TCP_IPV4 (1ULL << 0)
+/**< GSO flag for TCP/IPv4 packets (containing optional VLAN tag) */
+
 /**
  * GSO context structure.
  */
-- 
2.7.4

  parent reply	other threads:[~2017-08-24 14:13 UTC|newest]

Thread overview: 157+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-08-24 14:15 [dpdk-dev] [PATCH 0/5] Support TCP/IPv4, VxLAN and GRE GSO in DPDK Jiayu Hu
2017-08-24 14:15 ` [dpdk-dev] [PATCH 1/5] lib: add Generic Segmentation Offload API framework Jiayu Hu
2017-08-30  1:38   ` Ananyev, Konstantin
2017-08-30  7:57     ` Jiayu Hu
2017-08-24 14:15 ` Jiayu Hu [this message]
2017-08-30  1:38   ` [dpdk-dev] [PATCH 2/5] gso/lib: add TCP/IPv4 GSO support Ananyev, Konstantin
2017-08-30  2:55     ` Jiayu Hu
2017-08-30  9:25       ` Kavanagh, Mark B
2017-08-30  9:39         ` Ananyev, Konstantin
2017-08-30  9:59           ` Ananyev, Konstantin
2017-08-30 13:27             ` Kavanagh, Mark B
2017-08-30  9:03     ` Jiayu Hu
2017-09-04  3:31     ` Jiayu Hu
2017-09-04  9:54       ` Ananyev, Konstantin
2017-09-05  1:09         ` Hu, Jiayu
2017-09-11 13:04           ` Ananyev, Konstantin
2017-08-24 14:15 ` [dpdk-dev] [PATCH 3/5] lib/gso: add VxLAN " Jiayu Hu
2017-08-24 14:15 ` [dpdk-dev] [PATCH 4/5] lib/gso: add GRE " Jiayu Hu
2017-08-24 14:15 ` [dpdk-dev] [PATCH 5/5] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Jiayu Hu
2017-08-30  1:37 ` [dpdk-dev] [PATCH 0/5] Support TCP/IPv4, VxLAN and GRE GSO in DPDK Ananyev, Konstantin
2017-08-30  7:36   ` Jiayu Hu
2017-08-30 10:49     ` Ananyev, Konstantin
2017-08-30 13:32       ` Kavanagh, Mark B
2017-09-05  7:57 ` [dpdk-dev] [PATCH v2 " Jiayu Hu
2017-09-05  7:57   ` [dpdk-dev] [PATCH v2 1/5] gso: add Generic Segmentation Offload API framework Jiayu Hu
2017-09-05  7:57   ` [dpdk-dev] [PATCH v2 2/5] gso: add TCP/IPv4 GSO support Jiayu Hu
2017-09-05  7:57   ` [dpdk-dev] [PATCH v2 3/5] gso: add VxLAN " Jiayu Hu
2017-09-05  7:57   ` [dpdk-dev] [PATCH v2 4/5] gso: add GRE " Jiayu Hu
2017-09-05  7:57   ` [dpdk-dev] [PATCH v2 5/5] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Jiayu Hu
2017-09-12  2:43   ` [dpdk-dev] [PATCH v3 0/5] Support TCP/IPv4, VxLAN and GRE GSO in DPDK Jiayu Hu
2017-09-12  2:43     ` [dpdk-dev] [PATCH v3 1/5] gso: add Generic Segmentation Offload API framework Jiayu Hu
2017-09-12 10:36       ` Ananyev, Konstantin
2017-09-13  2:11         ` Jiayu Hu
2017-09-14 18:33       ` Ferruh Yigit
2017-09-15  1:12         ` Hu, Jiayu
2017-09-12  2:43     ` [dpdk-dev] [PATCH v3 2/5] gso: add TCP/IPv4 GSO support Jiayu Hu
2017-09-12 11:17       ` Ananyev, Konstantin
2017-09-13  2:48         ` Jiayu Hu
2017-09-13  9:38           ` Ananyev, Konstantin
2017-09-13 10:23             ` Hu, Jiayu
2017-09-13 14:52             ` Kavanagh, Mark B
2017-09-13 15:13               ` Ananyev, Konstantin
2017-09-14  0:59                 ` Hu, Jiayu
2017-09-14  8:35                   ` Kavanagh, Mark B
2017-09-14  8:39                     ` Ananyev, Konstantin
2017-09-14  9:00                       ` Kavanagh, Mark B
2017-09-14  9:10                         ` Ananyev, Konstantin
2017-09-14  9:35                           ` Kavanagh, Mark B
2017-09-12 14:17       ` Ananyev, Konstantin
2017-09-13 10:44         ` Jiayu Hu
2017-09-13 22:10           ` Ananyev, Konstantin
2017-09-14  6:07             ` Jiayu Hu
2017-09-14  8:47               ` Ananyev, Konstantin
2017-09-14  9:29                 ` Hu, Jiayu
2017-09-14  9:35                   ` Ananyev, Konstantin
2017-09-14 10:01                     ` Hu, Jiayu
2017-09-14 15:42                       ` Kavanagh, Mark B
2017-09-14 18:38                         ` Ananyev, Konstantin
2017-09-15  7:54                           ` Hu, Jiayu
2017-09-15  8:15                             ` Ananyev, Konstantin
2017-09-15  8:17                             ` Ananyev, Konstantin
2017-09-15  8:38                               ` Hu, Jiayu
2017-09-14  8:51               ` Kavanagh, Mark B
2017-09-14  9:45                 ` Hu, Jiayu
2017-09-12  2:43     ` [dpdk-dev] [PATCH v3 3/5] gso: add VxLAN " Jiayu Hu
2017-09-12  2:43     ` [dpdk-dev] [PATCH v3 4/5] gso: add GRE " Jiayu Hu
2017-09-12  2:43     ` [dpdk-dev] [PATCH v3 5/5] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Jiayu Hu
2017-09-14 18:33       ` Ferruh Yigit
2017-09-15  1:13         ` Hu, Jiayu
2017-09-19  7:32     ` [dpdk-dev] [PATCH v4 0/5] Support TCP/IPv4, VxLAN and GRE GSO in DPDK Jiayu Hu
2017-09-19  7:32       ` [dpdk-dev] [PATCH v4 1/5] gso: add Generic Segmentation Offload API framework Jiayu Hu
2017-09-19  7:32       ` [dpdk-dev] [PATCH v4 2/5] gso: add TCP/IPv4 GSO support Jiayu Hu
2017-09-20  7:03         ` Yao, Lei A
2017-09-19  7:32       ` [dpdk-dev] [PATCH v4 3/5] gso: add VxLAN " Jiayu Hu
2017-09-20  3:11         ` Tan, Jianfeng
2017-09-20  3:17           ` Hu, Jiayu
2017-09-19  7:32       ` [dpdk-dev] [PATCH v4 4/5] gso: add GRE " Jiayu Hu
2017-09-20  2:53         ` Tan, Jianfeng
2017-09-20  6:01           ` Hu, Jiayu
2017-09-19  7:32       ` [dpdk-dev] [PATCH v4 5/5] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Jiayu Hu
2017-09-28 22:13       ` [dpdk-dev] [PATCH v5 0/6] Support TCP/IPv4, VxLAN and GRE GSO in DPDK Mark Kavanagh
2017-10-02 16:45         ` [dpdk-dev] [PATCH v6 0/6] Support TCP/IPv4, VxLAN, " Mark Kavanagh
2017-10-02 16:45           ` [dpdk-dev] [PATCH v6 1/6] gso: add Generic Segmentation Offload API framework Mark Kavanagh
2017-10-04 13:11             ` Ananyev, Konstantin
2017-10-04 13:21               ` Kavanagh, Mark B
2017-10-02 16:45           ` [dpdk-dev] [PATCH v6 2/6] gso: add TCP/IPv4 GSO support Mark Kavanagh
2017-10-04 13:32             ` Ananyev, Konstantin
2017-10-04 14:30               ` Kavanagh, Mark B
2017-10-04 14:49                 ` Ananyev, Konstantin
2017-10-04 14:59                   ` Kavanagh, Mark B
2017-10-04 13:35             ` Ananyev, Konstantin
2017-10-04 14:22               ` Kavanagh, Mark B
2017-10-02 16:45           ` [dpdk-dev] [PATCH v6 3/6] gso: add VxLAN " Mark Kavanagh
2017-10-04 14:12             ` Ananyev, Konstantin
2017-10-04 14:35               ` Kavanagh, Mark B
2017-10-04 16:13               ` Kavanagh, Mark B
2017-10-04 16:17                 ` Ananyev, Konstantin
2017-10-02 16:45           ` [dpdk-dev] [PATCH v6 4/6] gso: add GRE " Mark Kavanagh
2017-10-04 14:15             ` Ananyev, Konstantin
2017-10-04 14:36               ` Kavanagh, Mark B
2017-10-02 16:45           ` [dpdk-dev] [PATCH v6 5/6] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Mark Kavanagh
2017-10-04 15:08             ` Ananyev, Konstantin
2017-10-04 16:23               ` Kavanagh, Mark B
2017-10-04 16:26                 ` Ananyev, Konstantin
2017-10-04 16:51                   ` Kavanagh, Mark B
2017-10-02 16:45           ` [dpdk-dev] [PATCH v6 6/6] doc: add GSO programmer's guide Mark Kavanagh
2017-10-04 13:51             ` Mcnamara, John
2017-10-05 11:02           ` [dpdk-dev] [PATCH v7 0/6] Support TCP/IPv4, VxLAN, and GRE GSO in DPDK Mark Kavanagh
2017-10-05 11:02             ` [dpdk-dev] [PATCH v7 1/6] gso: add Generic Segmentation Offload API framework Mark Kavanagh
2017-10-05 11:02             ` [dpdk-dev] [PATCH v7 2/6] gso: add TCP/IPv4 GSO support Mark Kavanagh
2017-10-05 11:02             ` [dpdk-dev] [PATCH v7 3/6] gso: add VxLAN " Mark Kavanagh
2017-10-05 11:02             ` [dpdk-dev] [PATCH v7 4/6] gso: add GRE " Mark Kavanagh
2017-10-05 11:02             ` [dpdk-dev] [PATCH v7 5/6] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Mark Kavanagh
2017-10-05 11:02             ` [dpdk-dev] [PATCH v7 6/6] doc: add GSO programmer's guide Mark Kavanagh
2017-10-05 13:22             ` [dpdk-dev] [PATCH v7 0/6] Support TCP/IPv4, VxLAN, and GRE GSO in DPDK Ananyev, Konstantin
2017-10-05 14:39               ` Kavanagh, Mark B
2017-10-05 15:43             ` [dpdk-dev] [PATCH v8 " Mark Kavanagh
2017-10-05 17:12               ` Ananyev, Konstantin
2017-10-05 20:16                 ` Kavanagh, Mark B
2017-10-05 20:36               ` [dpdk-dev] [PATCH v9 " Mark Kavanagh
2017-10-05 22:24                 ` Ananyev, Konstantin
2017-10-06  8:24                   ` [dpdk-dev] FW: " Kavanagh, Mark B
2017-10-06 10:35                   ` [dpdk-dev] " Kavanagh, Mark B
2017-10-06 23:32                 ` Ferruh Yigit
2017-10-06 23:34                   ` Ferruh Yigit
2017-10-07 14:56                 ` [dpdk-dev] [PATCH v10 " Jiayu Hu
2017-10-07 14:56                   ` [dpdk-dev] [PATCH v10 1/6] gso: add Generic Segmentation Offload API framework Jiayu Hu
2017-10-07 14:56                   ` [dpdk-dev] [PATCH v10 2/6] gso: add TCP/IPv4 GSO support Jiayu Hu
2017-10-07 14:56                   ` [dpdk-dev] [PATCH v10 3/6] gso: add VxLAN " Jiayu Hu
2017-10-07 14:56                   ` [dpdk-dev] [PATCH v10 4/6] gso: add GRE " Jiayu Hu
2017-10-07 14:56                   ` [dpdk-dev] [PATCH v10 5/6] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Jiayu Hu
2017-10-07 14:56                   ` [dpdk-dev] [PATCH v10 6/6] doc: add GSO programmer's guide Jiayu Hu
2017-10-08  3:40                   ` [dpdk-dev] [PATCH v10 0/6] Support TCP/IPv4, VxLAN, and GRE GSO in DPDK Ferruh Yigit
2017-10-05 20:36               ` [dpdk-dev] [PATCH v9 1/6] gso: add Generic Segmentation Offload API framework Mark Kavanagh
2017-10-05 20:36               ` [dpdk-dev] [PATCH v9 2/6] gso: add TCP/IPv4 GSO support Mark Kavanagh
2017-10-05 20:36               ` [dpdk-dev] [PATCH v9 3/6] gso: add VxLAN " Mark Kavanagh
2017-10-05 20:36               ` [dpdk-dev] [PATCH v9 4/6] gso: add GRE " Mark Kavanagh
2017-10-05 20:36               ` [dpdk-dev] [PATCH v9 5/6] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Mark Kavanagh
2017-10-05 20:36               ` [dpdk-dev] [PATCH v9 6/6] doc: add GSO programmer's guide Mark Kavanagh
2017-10-06 13:34                 ` Mcnamara, John
2017-10-06 13:41                   ` Kavanagh, Mark B
2017-10-05 15:43             ` [dpdk-dev] [PATCH v8 1/6] gso: add Generic Segmentation Offload API framework Mark Kavanagh
2017-10-05 15:44             ` [dpdk-dev] [PATCH v8 2/6] gso: add TCP/IPv4 GSO support Mark Kavanagh
2017-10-05 15:44             ` [dpdk-dev] [PATCH v8 3/6] gso: add VxLAN " Mark Kavanagh
2017-10-05 15:44             ` [dpdk-dev] [PATCH v8 4/6] gso: add GRE " Mark Kavanagh
2017-10-05 15:44             ` [dpdk-dev] [PATCH v8 5/6] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Mark Kavanagh
2017-10-05 15:44             ` [dpdk-dev] [PATCH v8 6/6] doc: add GSO programmer's guide Mark Kavanagh
2017-10-05 17:57               ` Mcnamara, John
2017-09-28 22:13       ` [dpdk-dev] [PATCH v5 1/6] gso: add Generic Segmentation Offload API framework Mark Kavanagh
2017-09-28 22:13       ` [dpdk-dev] [PATCH v5 2/6] gso: add TCP/IPv4 GSO support Mark Kavanagh
2017-09-29  3:12         ` Jiayu Hu
2017-09-29  9:05           ` Kavanagh, Mark B
2017-09-28 22:13       ` [dpdk-dev] [PATCH v5 3/6] gso: add VxLAN " Mark Kavanagh
2017-09-28 22:13       ` [dpdk-dev] [PATCH v5 4/6] gso: add GRE " Mark Kavanagh
2017-09-28 22:13       ` [dpdk-dev] [PATCH v5 5/6] app/testpmd: enable TCP/IPv4, VxLAN and GRE GSO Mark Kavanagh
2017-09-28 22:13       ` [dpdk-dev] [PATCH v5 6/6] doc: add GSO programmer's guide Mark Kavanagh
2017-09-28 22:18       ` [dpdk-dev] [PATCH v5 2/6] gso: add TCP/IPv4 GSO support Mark Kavanagh

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1503584144-63181-3-git-send-email-jiayu.hu@intel.com \
    --to=jiayu.hu@intel.com \
    --cc=dev@dpdk.org \
    --cc=jianfeng.tan@intel.com \
    --cc=konstantin.ananyev@intel.com \
    --cc=mark.b.kavanagh@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).