[PATCH v3 1/3] node/kernel_tx: support packet transmit to kernel

[Vamsi Attunuru]

Patch adds a node to transmit the packets to kernel over
a raw socket.

Signed-off-by: Vamsi Attunuru <vattunuru at marvell.com>
---
 doc/guides/prog_guide/graph_lib.rst |   9 ++
 lib/node/kernel_tx.c                | 122 ++++++++++++++++++++++++++++
 lib/node/kernel_tx_priv.h           |  16 ++++
 lib/node/meson.build                |   1 +
 4 files changed, 148 insertions(+)

diff --git a/doc/guides/prog_guide/graph_lib.rst b/doc/guides/prog_guide/graph_lib.rst
index 1cfdc86433..fa22b014f3 100644
--- a/doc/guides/prog_guide/graph_lib.rst
+++ b/doc/guides/prog_guide/graph_lib.rst
@@ -392,3 +392,12 @@ null
 ~~~~
 This node ignores the set of objects passed to it and reports that all are
 processed.
+
+kernel_tx
+~~~~~~~~~
+This node is an exit node that forwards the packets to kernel. It will be used
+to forward any control plane traffic to kernel stack from DPDK. It uses a raw
+socket interface to transmit the packets, it uses the packet's destination
+IP address in sockaddr_in address structure and ``sendto`` function to send
+data on the raw socket. Aftering sending the burst of packets to kernel,
+this node free up the packet buffers.
diff --git a/lib/node/kernel_tx.c b/lib/node/kernel_tx.c
new file mode 100644
index 0000000000..27d1808c71
--- /dev/null
+++ b/lib/node/kernel_tx.c
@@ -0,0 +1,122 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2023 Marvell International Ltd.
+ */
+
+#include <fcntl.h>
+#include <sys/ioctl.h>
+#include <sys/socket.h>
+#include <unistd.h>
+
+#include <rte_debug.h>
+#include <rte_ethdev.h>
+#include <rte_graph.h>
+#include <rte_graph_worker.h>
+#include <rte_ip.h>
+
+#include "kernel_tx_priv.h"
+#include "node_private.h"
+
+static __rte_always_inline void
+kernel_tx_process_mbuf(struct rte_node *node, struct rte_mbuf **mbufs, uint16_t cnt)
+{
+	kernel_tx_node_ctx_t *ctx = (kernel_tx_node_ctx_t *)node->ctx;
+	struct sockaddr_in sin = {0};
+	struct rte_ipv4_hdr *ip4;
+	size_t len;
+	char *buf;
+	int i;
+
+	for (i = 0; i < cnt; i++) {
+		ip4 = rte_pktmbuf_mtod(mbufs[i], struct rte_ipv4_hdr *);
+		len = rte_pktmbuf_data_len(mbufs[i]);
+		buf = (char *)ip4;
+
+		sin.sin_family = AF_INET;
+		sin.sin_port = 0;
+		sin.sin_addr.s_addr = ip4->dst_addr;
+
+		if (sendto(ctx->sock, buf, len, 0, (struct sockaddr *)&sin, sizeof(sin)) < 0)
+			node_err("kernel_tx", "Unable to send packets: %s\n", strerror(errno));
+	}
+}
+
+static uint16_t
+kernel_tx_node_process(struct rte_graph *graph __rte_unused, struct rte_node *node, void **objs,
+			 uint16_t nb_objs)
+{
+	struct rte_mbuf **pkts = (struct rte_mbuf **)objs;
+	uint16_t obj_left = nb_objs;
+
+#define PREFETCH_CNT 4
+
+	while (obj_left >= 12) {
+		/* Prefetch next-next mbufs */
+		rte_prefetch0(pkts[8]);
+		rte_prefetch0(pkts[9]);
+		rte_prefetch0(pkts[10]);
+		rte_prefetch0(pkts[11]);
+
+		/* Prefetch next mbuf data */
+		rte_prefetch0(rte_pktmbuf_mtod_offset(pkts[4], void *, pkts[4]->l2_len));
+		rte_prefetch0(rte_pktmbuf_mtod_offset(pkts[5], void *, pkts[5]->l2_len));
+		rte_prefetch0(rte_pktmbuf_mtod_offset(pkts[6], void *, pkts[6]->l2_len));
+		rte_prefetch0(rte_pktmbuf_mtod_offset(pkts[7], void *, pkts[7]->l2_len));
+
+		kernel_tx_process_mbuf(node, pkts, PREFETCH_CNT);
+
+		obj_left -= PREFETCH_CNT;
+		pkts += PREFETCH_CNT;
+	}
+
+	while (obj_left > 0) {
+		kernel_tx_process_mbuf(node, pkts, 1);
+
+		obj_left--;
+		pkts++;
+	}
+
+	rte_pktmbuf_free_bulk((struct rte_mbuf **)objs, nb_objs);
+
+	return nb_objs;
+}
+
+static int
+kernel_tx_node_init(const struct rte_graph *graph __rte_unused, struct rte_node *node)
+{
+	kernel_tx_node_ctx_t *ctx = (kernel_tx_node_ctx_t *)node->ctx;
+
+	ctx->sock = socket(AF_INET, SOCK_RAW, IPPROTO_RAW);
+	if (ctx->sock < 0)
+		node_err("kernel_tx", "Unable to open RAW socket\n");
+
+	return 0;
+}
+
+static void
+kernel_tx_node_fini(const struct rte_graph *graph __rte_unused, struct rte_node *node)
+{
+	kernel_tx_node_ctx_t *ctx = (kernel_tx_node_ctx_t *)node->ctx;
+
+	if (ctx->sock >= 0) {
+		close(ctx->sock);
+		ctx->sock = -1;
+	}
+}
+
+static struct rte_node_register kernel_tx_node_base = {
+	.process = kernel_tx_node_process,
+	.name = "kernel_tx",
+
+	.init = kernel_tx_node_init,
+	.fini = kernel_tx_node_fini,
+
+	.nb_edges = 0,
+};
+
+struct rte_node_register *
+kernel_tx_node_get(void)
+{
+	return &kernel_tx_node_base;
+}
+
+RTE_NODE_REGISTER(kernel_tx_node_base);
diff --git a/lib/node/kernel_tx_priv.h b/lib/node/kernel_tx_priv.h
new file mode 100644
index 0000000000..6fd8a4f2af
--- /dev/null
+++ b/lib/node/kernel_tx_priv.h
@@ -0,0 +1,16 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(C) 2023 Marvell International Ltd.
+ */
+
+#ifndef __KERNEL_TX_PRIV_H__
+#define __KERNEL_TX_PRIV_H__
+
+/* kernel_tx node context structure. */
+typedef struct kernel_tx_node_ctx {
+	int sock;
+} kernel_tx_node_ctx_t;
+
+/* Get the pointer to kernel_tx node register structure */
+struct rte_node_register *kernel_tx_node_get(void);
+
+#endif /* __KERNEL_TX_PRIV_H__ */
diff --git a/lib/node/meson.build b/lib/node/meson.build
index dbdf673c86..0520be23ff 100644
--- a/lib/node/meson.build
+++ b/lib/node/meson.build
@@ -13,6 +13,7 @@ sources = files(
         'ethdev_tx.c',
         'ip4_lookup.c',
         'ip4_rewrite.c',
+        'kernel_tx.c',
         'log.c',
         'null.c',
         'pkt_cls.c',
-- 
2.25.1