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
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
//! Vectors of raw packet data.
//! All of these types are Connection-level, meaning they are delivered on
//! connection termination.
//!
//! There are two types of packet lists: those containing Mbufs (`Zc` prefix)
//! and those containing raw bytes (as vectors).
//!
//! For the former: Mbufs are shared across these lists via DPDK reference counting,
//! so requesting lists of packets does not require copying. However, it may
//! introduce additional mempool requirements, as Mbufs must be kept in memory
//! for the duration of the connection. This is often particularly infeasible for
//! UDP connections, which must stay in memory until a timeout is reached.
//! In such cases, users may wish to use the non-`Zc` variants.
//!
//! For TCP connections, the non-`Zc` variants wait to clone data until the
//! first few packets have passed or until the packet data is requested.
//! After the first `PKTS_START_CLONE` packets, it is likely that some traffic has
//! been filtered out by the framework (e.g., TLS handshake has been parsed).
//! This is a middle ground between memory usage and compute performance.
//!
//! For UDP connections, this is not feasible without unacceptable mempool utilization;
//! many UDP connections are short-lived, and UDP connections are not "closed" until
//! a timeout period has passed.

use crate::PacketList;
use retina_core::{protocols::packet::tcp::TCP_PROTOCOL, L4Pdu, Mbuf};

/// Pasic raw packet bytes.
#[derive(Debug)]
pub struct PktData {
    pub data: Vec<u8>,
}

impl PktData {
    pub fn new(mbuf: &Mbuf) -> Self {
        Self {
            data: mbuf.data().to_vec(),
        }
    }
}

/// Number of Mbufs to cache before starting to clone data for
/// TCP connections only. If PKTS_START_CLONE is not reached, the
/// data is converted to Vec<u8> on first access.
const PKTS_START_CLONE: usize = 5;

pub trait PktStream {
    fn in_mbufs_own(&mut self) -> Vec<Mbuf>;
    fn in_mbufs_ref(&mut self) -> &mut Vec<Mbuf>;
    fn out_packets(&mut self) -> &mut Vec<PktData>;

    fn drain_mbufs(&mut self) {
        let mut in_mbufs = self.in_mbufs_own();
        for mbuf in in_mbufs.drain(..) {
            self.out_packets().push(PktData::new(&mbuf));
        }
    }

    fn packets(&mut self) -> &Vec<PktData> {
        if self.in_mbufs_ref().is_empty() {
            return self.out_packets();
        }
        self.drain_mbufs();
        self.out_packets()
    }

    fn push(&mut self, pdu: &L4Pdu) {
        if pdu.ctxt.proto == TCP_PROTOCOL && self.in_mbufs_ref().len() < PKTS_START_CLONE {
            self.in_mbufs_ref().push(Mbuf::new_ref(&pdu.mbuf));
            return;
        } else if !self.in_mbufs_ref().is_empty() {
            self.drain_mbufs();
        }
        self.out_packets().push(PktData::new(pdu.mbuf_ref()));
    }
}

/// For a connection, the bidirectional stream of packets
/// in the order received by the framework.
#[derive(Debug)]
pub struct BidirPktStream {
    /// The raw packet data.
    pub packets: Vec<PktData>,
    /// The first few packets are stored as Mbufs
    /// before data copies begin.
    mbufs: Vec<Mbuf>,
}

impl PktStream for BidirPktStream {
    fn in_mbufs_own(&mut self) -> Vec<Mbuf> {
        std::mem::take(&mut self.mbufs)
    }

    fn in_mbufs_ref(&mut self) -> &mut Vec<Mbuf> {
        &mut self.mbufs
    }

    fn out_packets(&mut self) -> &mut Vec<PktData> {
        &mut self.packets
    }
}

impl PacketList for BidirPktStream {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
            mbufs: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if !reassembled {
            self.push(pdu);
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
        self.mbufs.clear();
    }
}

/// For a connection, an originator's (unidirectional) stream of packets
/// in the order received by the framework. For TCP streams, the
/// "originator" is the endpoint that sends the first SYN. For UDP,
/// it is the endpoint which sends the first-seen packet.
pub struct OrigPktStream {
    /// The raw packet data.
    pub packets: Vec<PktData>,
    /// The first few packets are stored as Mbufs
    /// before data copies begin.
    mbufs: Vec<Mbuf>,
}

impl PktStream for OrigPktStream {
    fn in_mbufs_own(&mut self) -> Vec<Mbuf> {
        std::mem::take(&mut self.mbufs)
    }

    fn in_mbufs_ref(&mut self) -> &mut Vec<Mbuf> {
        &mut self.mbufs
    }

    fn out_packets(&mut self) -> &mut Vec<PktData> {
        &mut self.packets
    }
}

impl PacketList for OrigPktStream {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
            mbufs: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if pdu.dir && !reassembled {
            self.push(pdu);
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
        self.mbufs.clear();
    }
}

/// For a connection, a responder's (unidirectional) stream of packets
/// in the order received by the framework. For TCP streams, the
/// "responder" is the endpoint that receives the first SYN and responds
/// with a SYN/ACK. For UDP, it is the endpoint which does not send the
/// first packet.
pub struct RespPktStream {
    /// The raw packet data.
    pub packets: Vec<PktData>,
    /// The first few packets are stored as Mbufs
    /// before data copies begin.
    mbufs: Vec<Mbuf>,
}

impl PktStream for RespPktStream {
    fn in_mbufs_own(&mut self) -> Vec<Mbuf> {
        std::mem::take(&mut self.mbufs)
    }

    fn in_mbufs_ref(&mut self) -> &mut Vec<Mbuf> {
        &mut self.mbufs
    }

    fn out_packets(&mut self) -> &mut Vec<PktData> {
        &mut self.packets
    }
}

impl PacketList for RespPktStream {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
            mbufs: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if !pdu.dir && !reassembled {
            self.push(pdu);
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
        self.mbufs.clear();
    }
}

/// For a connection, an originator's (unidirectional) stream of packets
/// in reassembled order. This should be used for TCP only.
pub struct OrigPktsReassembled {
    /// The raw packet data.
    pub packets: Vec<PktData>,
    /// The first few packets are stored as Mbufs
    /// before data copies begin.
    mbufs: Vec<Mbuf>,
}

impl PktStream for OrigPktsReassembled {
    fn in_mbufs_own(&mut self) -> Vec<Mbuf> {
        std::mem::take(&mut self.mbufs)
    }

    fn in_mbufs_ref(&mut self) -> &mut Vec<Mbuf> {
        &mut self.mbufs
    }

    fn out_packets(&mut self) -> &mut Vec<PktData> {
        &mut self.packets
    }
}

impl PacketList for OrigPktsReassembled {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
            mbufs: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if pdu.dir && reassembled {
            self.push(pdu);
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
        self.mbufs.clear();
    }
}

/// For a connection, a responder's (unidirectional) stream of packets
/// in reassembled order. This should be used for TCP only.
pub struct RespPktsReassembled {
    /// The raw packet data.
    pub packets: Vec<PktData>,
    /// The first few packets are stored as Mbufs
    /// before data copies begin.
    mbufs: Vec<Mbuf>,
}

impl PktStream for RespPktsReassembled {
    fn in_mbufs_own(&mut self) -> Vec<Mbuf> {
        std::mem::take(&mut self.mbufs)
    }

    fn in_mbufs_ref(&mut self) -> &mut Vec<Mbuf> {
        &mut self.mbufs
    }

    fn out_packets(&mut self) -> &mut Vec<PktData> {
        &mut self.packets
    }
}

impl PacketList for RespPktsReassembled {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
            mbufs: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if !pdu.dir && reassembled {
            self.push(pdu);
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
        self.mbufs.clear();
    }
}

/// For a connection, the bidirectional stream of packets
/// in the order received by the framework.
pub struct BidirZcPktStream {
    pub packets: Vec<Mbuf>,
}

impl PacketList for BidirZcPktStream {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if !reassembled {
            self.packets.push(Mbuf::new_ref(&pdu.mbuf));
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
    }
}

/// For a connection, an originator's (unidirectional) stream of packets
/// in the order received by the framework. For TCP streams, the
/// "originator" is the endpoint that sends the first SYN. For UDP,
/// it is the endpoint which sends the first-seen packet.
pub struct OrigZcPktStream {
    pub packets: Vec<Mbuf>,
}

impl PacketList for OrigZcPktStream {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if !reassembled && pdu.dir {
            self.packets.push(Mbuf::new_ref(&pdu.mbuf));
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
    }
}

/// For a connection, a responder's (unidirectional) stream of packets
/// in the order received by the framework. For TCP streams, the
/// "responder" is the endpoint that receives the first SYN and responds
/// with a SYN/ACK. For UDP, it is the endpoint which does not send the
/// first packet.
pub struct RespZcPktStream {
    pub packets: Vec<Mbuf>,
}

impl PacketList for RespZcPktStream {
    fn new(_first_pkt: &L4Pdu) -> Self {
        // TODO figure out good default capacity
        Self {
            packets: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if !reassembled && !pdu.dir {
            self.packets.push(Mbuf::new_ref(&pdu.mbuf));
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
    }
}

/// For a connection, an originator's (unidirectional) stream of packets
/// in reassembled order. This should be used for TCP only.
pub struct OrigZcPktsReassembled {
    pub packets: Vec<Mbuf>,
}

impl PacketList for OrigZcPktsReassembled {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if reassembled && pdu.dir {
            self.packets.push(Mbuf::new_ref(&pdu.mbuf));
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
    }
}

/// For a connection, a responder's (unidirectional) stream of packets
/// in reassembled order. This should be used for TCP only.
pub struct RespZcPktsReassembled {
    pub packets: Vec<Mbuf>,
}

impl PacketList for RespZcPktsReassembled {
    fn new(_first_pkt: &L4Pdu) -> Self {
        Self {
            packets: Vec::new(),
        }
    }

    fn track_packet(&mut self, pdu: &L4Pdu, reassembled: bool) {
        if reassembled && !pdu.dir {
            self.packets.push(Mbuf::new_ref(&pdu.mbuf));
        }
    }

    fn clear(&mut self) {
        self.packets.clear();
    }
}