[mirror_ubuntu-bionic-kernel.git] / samples / bpf / xdp_router_ipv4_kern.c

/* Copyright (C) 2017 Cavium, Inc.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of version 2 of the GNU General Public License
 * as published by the Free Software Foundation.
 */
#define KBUILD_MODNAME "foo"
#include <uapi/linux/bpf.h>
#include <linux/in.h>
#include <linux/if_ether.h>
#include <linux/if_packet.h>
#include <linux/if_vlan.h>
#include <linux/ip.h>
#include <linux/ipv6.h>
#include "bpf_helpers.h"
#include <linux/slab.h>
#include <net/ip_fib.h>

struct trie_value {
	__u8 prefix[4];
	__be64 value;
	int ifindex;
	int metric;
	__be32 gw;
};

/* Key for lpm_trie*/
union key_4 {
	u32 b32[2];
	u8 b8[8];
};

struct arp_entry {
	__be64 mac;
	__be32 dst;
};

struct direct_map {
	struct arp_entry arp;
	int ifindex;
	__be64 mac;
};

/* Map for trie implementation*/
struct bpf_map_def SEC("maps") lpm_map = {
	.type = BPF_MAP_TYPE_LPM_TRIE,
	.key_size = 8,
	.value_size = sizeof(struct trie_value),
	.max_entries = 50,
	.map_flags = BPF_F_NO_PREALLOC,
};

/* Map for counter*/
struct bpf_map_def SEC("maps") rxcnt = {
	.type = BPF_MAP_TYPE_PERCPU_ARRAY,
	.key_size = sizeof(u32),
	.value_size = sizeof(u64),
	.max_entries = 256,
};

/* Map for ARP table*/
struct bpf_map_def SEC("maps") arp_table = {
	.type = BPF_MAP_TYPE_HASH,
	.key_size = sizeof(__be32),
	.value_size = sizeof(__be64),
	.max_entries = 50,
};

/* Map to keep the exact match entries in the route table*/
struct bpf_map_def SEC("maps") exact_match = {
	.type = BPF_MAP_TYPE_HASH,
	.key_size = sizeof(__be32),
	.value_size = sizeof(struct direct_map),
	.max_entries = 50,
};

struct bpf_map_def SEC("maps") tx_port = {
	.type = BPF_MAP_TYPE_DEVMAP,
	.key_size = sizeof(int),
	.value_size = sizeof(int),
	.max_entries = 100,
};

/* Function to set source and destination mac of the packet */
static inline void set_src_dst_mac(void *data, void *src, void *dst)
{
	unsigned short *source = src;
	unsigned short *dest  = dst;
	unsigned short *p = data;

	__builtin_memcpy(p, dest, 6);
	__builtin_memcpy(p + 3, source, 6);
}

/* Parse IPV4 packet to get SRC, DST IP and protocol */
static inline int parse_ipv4(void *data, u64 nh_off, void *data_end,
			     __be32 *src, __be32 *dest)
{
	struct iphdr *iph = data + nh_off;

	if (iph + 1 > data_end)
		return 0;
	*src = iph->saddr;
	*dest = iph->daddr;
	return iph->protocol;
}

SEC("xdp_router_ipv4")
int xdp_router_ipv4_prog(struct xdp_md *ctx)
{
	void *data_end = (void *)(long)ctx->data_end;
	__be64 *dest_mac = NULL, *src_mac = NULL;
	void *data = (void *)(long)ctx->data;
	struct trie_value *prefix_value;
	int rc = XDP_DROP, forward_to;
	struct ethhdr *eth = data;
	union key_4 key4;
	long *value;
	u16 h_proto;
	u32 ipproto;
	u64 nh_off;

	nh_off = sizeof(*eth);
	if (data + nh_off > data_end)
		return rc;

	h_proto = eth->h_proto;

	if (h_proto == htons(ETH_P_8021Q) || h_proto == htons(ETH_P_8021AD)) {
		struct vlan_hdr *vhdr;

		vhdr = data + nh_off;
		nh_off += sizeof(struct vlan_hdr);
		if (data + nh_off > data_end)
			return rc;
		h_proto = vhdr->h_vlan_encapsulated_proto;
	}
	if (h_proto == htons(ETH_P_ARP)) {
		return XDP_PASS;
	} else if (h_proto == htons(ETH_P_IP)) {
		struct direct_map *direct_entry;
		__be32 src_ip = 0, dest_ip = 0;

		ipproto = parse_ipv4(data, nh_off, data_end, &src_ip, &dest_ip);
		direct_entry = bpf_map_lookup_elem(&exact_match, &dest_ip);
		/* Check for exact match, this would give a faster lookup*/
		if (direct_entry && direct_entry->mac && direct_entry->arp.mac) {
			src_mac = &direct_entry->mac;
			dest_mac = &direct_entry->arp.mac;
			forward_to = direct_entry->ifindex;
		} else {
			/* Look up in the trie for lpm*/
			key4.b32[0] = 32;
			key4.b8[4] = dest_ip & 0xff;
			key4.b8[5] = (dest_ip >> 8) & 0xff;
			key4.b8[6] = (dest_ip >> 16) & 0xff;
			key4.b8[7] = (dest_ip >> 24) & 0xff;
			prefix_value = bpf_map_lookup_elem(&lpm_map, &key4);
			if (!prefix_value)
				return XDP_DROP;
			src_mac = &prefix_value->value;
			if (!src_mac)
				return XDP_DROP;
			dest_mac = bpf_map_lookup_elem(&arp_table, &dest_ip);
			if (!dest_mac) {
				if (!prefix_value->gw)
					return XDP_DROP;
				dest_ip = prefix_value->gw;
				dest_mac = bpf_map_lookup_elem(&arp_table, &dest_ip);
			}
			forward_to = prefix_value->ifindex;
		}
	} else {
		ipproto = 0;
	}
	if (src_mac && dest_mac) {
		set_src_dst_mac(data, src_mac, dest_mac);
		value = bpf_map_lookup_elem(&rxcnt, &ipproto);
		if (value)
			*value += 1;
		return  bpf_redirect_map(&tx_port, forward_to, 0);
	}
	return rc;
}

char _license[] SEC("license") = "GPL";
Commit	Line	Data
3e29cd0e CJ	1	/* Copyright (C) 2017 Cavium, Inc.
	2	*
	3	* This program is free software; you can redistribute it and/or modify it
	4	* under the terms of version 2 of the GNU General Public License
	5	* as published by the Free Software Foundation.
	6	*/
	7	#define KBUILD_MODNAME "foo"
	8	#include <uapi/linux/bpf.h>
	9	#include <linux/in.h>
	10	#include <linux/if_ether.h>
	11	#include <linux/if_packet.h>
	12	#include <linux/if_vlan.h>
	13	#include <linux/ip.h>
	14	#include <linux/ipv6.h>
	15	#include "bpf_helpers.h"
	16	#include <linux/slab.h>
	17	#include <net/ip_fib.h>
	18
	19	struct trie_value {
	20	__u8 prefix[4];
	21	__be64 value;
	22	int ifindex;
	23	int metric;
	24	__be32 gw;
	25	};
	26
	27	/* Key for lpm_trie*/
	28	union key_4 {
	29	u32 b32[2];
	30	u8 b8[8];
	31	};
	32
	33	struct arp_entry {
	34	__be64 mac;
	35	__be32 dst;
	36	};
	37
	38	struct direct_map {
	39	struct arp_entry arp;
	40	int ifindex;
	41	__be64 mac;
	42	};
	43
	44	/* Map for trie implementation*/
	45	struct bpf_map_def SEC("maps") lpm_map = {
	46	.type = BPF_MAP_TYPE_LPM_TRIE,
	47	.key_size = 8,
	48	.value_size = sizeof(struct trie_value),
	49	.max_entries = 50,
	50	.map_flags = BPF_F_NO_PREALLOC,
	51	};
	52
	53	/* Map for counter*/
	54	struct bpf_map_def SEC("maps") rxcnt = {
	55	.type = BPF_MAP_TYPE_PERCPU_ARRAY,
	56	.key_size = sizeof(u32),
	57	.value_size = sizeof(u64),
	58	.max_entries = 256,
	59	};
	60
	61	/* Map for ARP table*/
	62	struct bpf_map_def SEC("maps") arp_table = {
	63	.type = BPF_MAP_TYPE_HASH,
	64	.key_size = sizeof(__be32),
65	.value_size = sizeof(__be64),
66	.max_entries = 50,
67	};
68
69	/* Map to keep the exact match entries in the route table*/
70	struct bpf_map_def SEC("maps") exact_match = {
71	.type = BPF_MAP_TYPE_HASH,
72	.key_size = sizeof(__be32),
73	.value_size = sizeof(struct direct_map),
74	.max_entries = 50,
75	};
76
77	struct bpf_map_def SEC("maps") tx_port = {
78	.type = BPF_MAP_TYPE_DEVMAP,
79	.key_size = sizeof(int),
80	.value_size = sizeof(int),
81	.max_entries = 100,
82	};
83
84	/* Function to set source and destination mac of the packet */
85	static inline void set_src_dst_mac(void data, void src, void *dst)
86	{
87	unsigned short *source = src;
88	unsigned short *dest = dst;
89	unsigned short *p = data;
90
91	__builtin_memcpy(p, dest, 6);
92	__builtin_memcpy(p + 3, source, 6);
93	}
94
95	/* Parse IPV4 packet to get SRC, DST IP and protocol */
96	static inline int parse_ipv4(void data, u64 nh_off, void data_end,
97	__be32 src, __be32 dest)
98	{
99	struct iphdr *iph = data + nh_off;
100
101	if (iph + 1 > data_end)
102	return 0;
103	*src = iph->saddr;
104	*dest = iph->daddr;
105	return iph->protocol;
106	}
107
108	SEC("xdp_router_ipv4")
109	int xdp_router_ipv4_prog(struct xdp_md *ctx)
110	{
111	void data_end = (void )(long)ctx->data_end;
112	__be64 dest_mac = NULL, src_mac = NULL;
113	void data = (void )(long)ctx->data;
114	struct trie_value *prefix_value;
115	int rc = XDP_DROP, forward_to;
116	struct ethhdr *eth = data;
117	union key_4 key4;
118	long *value;
119	u16 h_proto;
120	u32 ipproto;
121	u64 nh_off;
122
123	nh_off = sizeof(*eth);
124	if (data + nh_off > data_end)
125	return rc;
126
127	h_proto = eth->h_proto;
128
129	if (h_proto == htons(ETH_P_8021Q) \|\| h_proto == htons(ETH_P_8021AD)) {
130	struct vlan_hdr *vhdr;
131
132	vhdr = data + nh_off;
133	nh_off += sizeof(struct vlan_hdr);
134	if (data + nh_off > data_end)
135	return rc;
136	h_proto = vhdr->h_vlan_encapsulated_proto;
137	}
138	if (h_proto == htons(ETH_P_ARP)) {
139	return XDP_PASS;
140	} else if (h_proto == htons(ETH_P_IP)) {
141	struct direct_map *direct_entry;
142	__be32 src_ip = 0, dest_ip = 0;
143
144	ipproto = parse_ipv4(data, nh_off, data_end, &src_ip, &dest_ip);
145	direct_entry = bpf_map_lookup_elem(&exact_match, &dest_ip);
146	/* Check for exact match, this would give a faster lookup*/
147	if (direct_entry && direct_entry->mac && direct_entry->arp.mac) {
148	src_mac = &direct_entry->mac;
149	dest_mac = &direct_entry->arp.mac;
150	forward_to = direct_entry->ifindex;
151	} else {
152	/* Look up in the trie for lpm*/
153	key4.b32[0] = 32;
154	key4.b8[4] = dest_ip & 0xff;
155	key4.b8[5] = (dest_ip >> 8) & 0xff;
156	key4.b8[6] = (dest_ip >> 16) & 0xff;
157	key4.b8[7] = (dest_ip >> 24) & 0xff;
158	prefix_value = bpf_map_lookup_elem(&lpm_map, &key4);
159	if (!prefix_value)
160	return XDP_DROP;
161	src_mac = &prefix_value->value;
162	if (!src_mac)
163	return XDP_DROP;
164	dest_mac = bpf_map_lookup_elem(&arp_table, &dest_ip);
165	if (!dest_mac) {
166	if (!prefix_value->gw)
167	return XDP_DROP;
168	dest_ip = prefix_value->gw;
169	dest_mac = bpf_map_lookup_elem(&arp_table, &dest_ip);
170	}
171	forward_to = prefix_value->ifindex;
172	}
173	} else {
174	ipproto = 0;
175	}
176	if (src_mac && dest_mac) {
177	set_src_dst_mac(data, src_mac, dest_mac);
178	value = bpf_map_lookup_elem(&rxcnt, &ipproto);
179	if (value)
180	*value += 1;
181	return bpf_redirect_map(&tx_port, forward_to, 0);
182	}
183	return rc;
184	}
185
186	char _license[] SEC("license") = "GPL";