summaryrefslogtreecommitdiff
path: root/ring_rx.c
blob: 7a3d21df3b6115c3b8cbf301a545a6915bca9404 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
/*
 * netsniff-ng - the packet sniffing beast
 * Copyright 2009, 2010 Daniel Borkmann.
 * Subject to the GPL, version 2.
 */

#include <inttypes.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <sys/mman.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <arpa/inet.h>
#include <linux/if_ether.h>

#include "xmalloc.h"
#include "die.h"
#include "ring_rx.h"
#include "built_in.h"

void destroy_rx_ring(int sock, struct ring *ring)
{
	int ret;
	bool v3 = get_sockopt_tpacket(sock) == TPACKET_V3;

	munmap(ring->mm_space, ring->mm_len);
	ring->mm_len = 0;

	xfree(ring->frames);

	/* In general, this is freed during close(2) anyway. */
	if (v3)
		return;

	fmemset(&ring->layout, 0, sizeof(ring->layout));
	ret = setsockopt(sock, SOL_PACKET, PACKET_RX_RING, &ring->layout,
			 sizeof(ring->layout));
	if (unlikely(ret))
		panic("Cannot destroy the RX_RING: %s!\n", strerror(errno));
}

static void setup_rx_ring_layout(int sock, struct ring *ring, size_t size,
				 bool jumbo_support, bool v3)
{
	fmemset(&ring->layout, 0, sizeof(ring->layout));

	ring->layout.tp_block_size = (jumbo_support ?
				      RUNTIME_PAGE_SIZE << 4 :
				      RUNTIME_PAGE_SIZE << 2);

	ring->layout.tp_frame_size = (jumbo_support ?
				      TPACKET_ALIGNMENT << 12 :
				      TPACKET_ALIGNMENT << 7);

	ring->layout.tp_block_nr = size / ring->layout.tp_block_size;
	ring->layout.tp_frame_nr = ring->layout.tp_block_size /
				   ring->layout.tp_frame_size *
				   ring->layout.tp_block_nr;
	if (v3) {
		/* Pass out, if this will ever change and we do crap on it! */
		build_bug_on(offsetof(struct tpacket_req, tp_frame_nr) !=
			     offsetof(struct tpacket_req3, tp_frame_nr) &&
			     sizeof(struct tpacket_req) !=
			     offsetof(struct tpacket_req3, tp_retire_blk_tov));

		ring->layout3.tp_retire_blk_tov = 100; /* 0: let kernel decide */
		ring->layout3.tp_sizeof_priv = 0;
		ring->layout3.tp_feature_req_word = 0;

		set_sockopt_tpacket_v3(sock);
	} else {
		set_sockopt_tpacket_v2(sock);
	}

	ring_verify_layout(ring);
}

static void create_rx_ring(int sock, struct ring *ring, bool verbose)
{
	int ret;
	bool v3 = get_sockopt_tpacket(sock) == TPACKET_V3;

retry:
	ret = setsockopt(sock, SOL_PACKET, PACKET_RX_RING, &ring->raw,
			 v3 ? sizeof(ring->layout3) : sizeof(ring->layout));

	if (errno == ENOMEM && ring->layout.tp_block_nr > 1) {
		ring->layout.tp_block_nr >>= 1;
		ring->layout.tp_frame_nr = ring->layout.tp_block_size / 
					   ring->layout.tp_frame_size * 
					   ring->layout.tp_block_nr;
		goto retry;
	}
	if (ret < 0)
		panic("Cannot allocate RX_RING!\n");

	ring->mm_len = (size_t) ring->layout.tp_block_size * ring->layout.tp_block_nr;

	if (verbose) {
		if (!v3) {
			printf("RX,V2: %.2Lf MiB, %u Frames, each %u Byte allocated\n",
			       (long double) ring->mm_len / (1 << 20),
			       ring->layout.tp_frame_nr, ring->layout.tp_frame_size);
		} else {
			printf("RX,V3: %.2Lf MiB, %u Blocks, each %u Byte allocated\n",
			       (long double) ring->mm_len / (1 << 20),
			       ring->layout.tp_block_nr, ring->layout.tp_block_size);
		}
	}
}

static void alloc_rx_ring_frames(int sock, struct ring *ring)
{
	int num;
	size_t size;
	bool v3 = get_sockopt_tpacket(sock) == TPACKET_V3;

	if (v3) {
		num = ring->layout3.tp_block_nr;
		size = ring->layout3.tp_block_size;
	} else {
		num = ring->layout.tp_frame_nr;
		size = ring->layout.tp_frame_size;
	}

	alloc_ring_frames_generic(ring, num, size);
}

void ring_rx_setup(struct ring *ring, int sock, size_t size, int ifindex,
		   struct pollfd *poll, bool v3, bool jumbo_support,
		   bool verbose)
{
	fmemset(ring, 0, sizeof(*ring));
	setup_rx_ring_layout(sock, ring, size, jumbo_support, v3);
	create_rx_ring(sock, ring, verbose);
	mmap_ring_generic(sock, ring);
	alloc_rx_ring_frames(sock, ring);
	bind_ring_generic(sock, ring, ifindex, false);
	prepare_polling(sock, poll);
}

void sock_rx_net_stats(int sock, unsigned long seen)
{
	int ret;
	bool v3 = get_sockopt_tpacket(sock) == TPACKET_V3;
	union {
		struct tpacket_stats	k2;
		struct tpacket_stats_v3 k3;
	} stats;
	socklen_t slen = v3 ? sizeof(stats.k3) : sizeof(stats.k2);

	memset(&stats, 0, sizeof(stats));
	ret = getsockopt(sock, SOL_PACKET, PACKET_STATISTICS, &stats, &slen);
	if (ret > -1) {
		uint64_t packets = stats.k3.tp_packets;
		uint64_t drops = stats.k3.tp_drops;

		printf("\r%12"PRIu64"  packets incoming (%"PRIu64" unread on exit)\n",
		       v3 ? (uint64_t)seen : packets, v3 ? packets - seen : 0);
		printf("\r%12"PRIu64"  packets passed filter\n", packets - drops);
		printf("\r%12"PRIu64"  packets failed filter (out of space)\n", drops);
		if (stats.k3.tp_packets > 0)
			printf("\r%12.4lf%% packet droprate\n",
			       (1.0 * drops / packets) * 100.0);
	}
}