summaryrefslogtreecommitdiffstats
path: root/VNFs/DPPD-PROX/handle_qinq_decap4.c
blob: d74e622e5100549d2a78da03aabbf3d3cbe5c597 (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
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
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
/*
// Copyright (c) 2010-2017 Intel Corporation
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
*/

#include <rte_byteorder.h>
#include <rte_cycles.h>
#include <rte_table_hash.h>
#include <rte_lpm.h>
#include <rte_version.h>

#include "prox_lua.h"
#include "prox_lua_types.h"
#include "handle_qinq_decap4.h"
#include "handle_qinq_encap4.h"
#include "stats.h"
#include "tx_pkt.h"
#include "defines.h"
#include "handle_routing.h"
#include "prox_assert.h"
#include "task_init.h"
#include "quit.h"
#include "pkt_prototypes.h"
#include "task_base.h"
#include "task_init.h"
#include "bng_pkts.h"
#include "prox_cksum.h"
#include "expire_cpe.h"
#include "prox_port_cfg.h"
#include "prefetch.h"
#include "prox_cfg.h"
#include "lconf.h"
#include "prox_cfg.h"
#include "prox_shared.h"
#include "prox_compat.h"

struct task_qinq_decap4 {
	struct task_base        base;
	struct rte_table_hash   *cpe_table;
	struct rte_table_hash   *qinq_gre_table;
	struct qinq_gre_data    *qinq_gre_data;
	struct next_hop         *next_hops;
	struct rte_lpm          *ipv4_lpm;
	uint32_t                local_ipv4;
	uint16_t                qinq_tag;
	uint8_t                 runtime_flags;
	int                     offload_crc;
	uint64_t                keys[64];
	uint64_t                src_mac[PROX_MAX_PORTS];
	struct rte_mbuf*        fake_packets[64];
	struct expire_cpe       expire_cpe;
	uint64_t                cpe_timeout;
	uint8_t                 mapping[PROX_MAX_PORTS];
};

static uint8_t handle_qinq_decap4(struct task_qinq_decap4 *task, struct rte_mbuf *mbuf, struct qinq_gre_data* entry);
/* Convert IPv4 packets to GRE and optionally store QinQ Tags */
static void arp_update(struct task_base *tbase, struct rte_mbuf **mbufs, uint16_t n_pkts);
static void arp_msg(struct task_base *tbase, void **data, uint16_t n_msgs);

static void init_task_qinq_decap4(struct task_base *tbase, struct task_args *targ)
{
	struct task_qinq_decap4 *task = (struct task_qinq_decap4 *)tbase;
	const int socket_id = rte_lcore_to_socket_id(targ->lconf->id);
	struct lpm4 *lpm;

	task->cpe_table = targ->cpe_table;
	task->cpe_timeout = msec_to_tsc(targ->cpe_table_timeout_ms);

	PROX_PANIC(!strcmp(targ->route_table, ""), "route table not specified\n");
	lpm = prox_sh_find_socket(socket_id, targ->route_table);
	if (!lpm) {
		int ret = lua_to_lpm4(prox_lua(), GLOBAL, targ->route_table, socket_id, &lpm);
		PROX_PANIC(ret, "Failed to load IPv4 LPM:\n%s\n", get_lua_to_errors());
		prox_sh_add_socket(socket_id, targ->route_table, lpm);
	}
	task->ipv4_lpm = lpm->rte_lpm;
	task->next_hops = lpm->next_hops;

	task->qinq_tag = targ->qinq_tag;
	task->local_ipv4 = targ->local_ipv4;
	task->runtime_flags = targ->runtime_flags;
	if (strcmp(targ->task_init->sub_mode_str, "pe"))
		PROX_PANIC(targ->qinq_gre_table == NULL, "can't set up qinq gre\n");

	task->qinq_gre_table = targ->qinq_gre_table;

	if (targ->cpe_table_timeout_ms) {
		targ->lconf->period_func = check_expire_cpe;
		task->expire_cpe.cpe_table = task->cpe_table;
		targ->lconf->period_data = &task->expire_cpe;
		targ->lconf->period_timeout = msec_to_tsc(500) / NUM_VCPES;
	}

	for (uint32_t i = 0; i < 64; ++i) {
		task->fake_packets[i] = (struct rte_mbuf*)((uint8_t*)&task->keys[i] - sizeof (struct rte_mbuf));
	}
	if (task->runtime_flags & TASK_ROUTING) {
		if (targ->nb_txrings) {
			struct task_args *dtarg;
			struct core_task ct;

			for (uint32_t i = 0; i < targ->nb_txrings; ++i) {
				ct = targ->core_task_set[0].core_task[i];
				dtarg = core_targ_get(ct.core, ct.task);
				dtarg = find_reachable_task_sending_to_port(dtarg);

				PROX_PANIC(dtarg == NULL, "Error finding destination port through other tasks for outgoing ring %u\n", i);
				task->src_mac[i] = *(uint64_t*)&prox_port_cfg[dtarg->tx_port_queue[0].port].eth_addr;
			}
		}
		else {
			for (uint32_t i = 0; i < targ->nb_txports; ++i) {
				task->src_mac[i] = *(uint64_t*)&prox_port_cfg[targ->tx_port_queue[i].port].eth_addr;
			}
		}
	}

	if (targ->runtime_flags & TASK_CTRL_HANDLE_ARP) {
		targ->lconf->ctrl_func_p[targ->task] = arp_update;
	}

	/* Copy the mapping from a sibling task which is configured
	   with mode encap4. The mapping is constant, so it is faster
	   to apply it when entries are added (least common case)
	   instead of re-applying it for every packet (most common
	   case). */

	for (uint8_t task_id = 0; task_id < targ->lconf->n_tasks_all; ++task_id) {
		enum task_mode smode = targ->lconf->targs[task_id].mode;
		if (QINQ_ENCAP4 == smode) {
			for (uint8_t i = 0; i < PROX_MAX_PORTS; ++i) {
				task->mapping[i] = targ->lconf->targs[task_id].mapping[i];
			}
		}
	}

	struct prox_port_cfg *port = find_reachable_port(targ);
	if (port) {
		task->offload_crc = port->requested_tx_offload & (DEV_TX_OFFLOAD_IPV4_CKSUM | DEV_TX_OFFLOAD_UDP_CKSUM);
	}

	// By default, calling this function 1K times per second => 64K ARP per second max
	// If 4 interfaces sending to here, = ~0.1% of workload.
	// If receiving more ARP, they will be dropped, or will dramatically slow down LB if in "no drop" mode.
	targ->lconf->ctrl_timeout = freq_to_tsc(targ->ctrl_freq);
	targ->lconf->ctrl_func_m[targ->task] = arp_msg;
}

static void early_init_table(struct task_args *targ)
{
	if (!targ->qinq_gre_table && !targ->cpe_table) {
		init_qinq_gre_table(targ, get_qinq_gre_map(targ));
		init_cpe4_table(targ);
	}
}

static inline void extract_key_bulk(struct rte_mbuf **mbufs, uint16_t n_pkts, struct task_qinq_decap4 *task)
{
	for (uint16_t j = 0; j < n_pkts; ++j) {
		extract_key_cpe(mbufs[j], &task->keys[j]);
	}
}

__attribute__((cold)) static void handle_error(struct rte_mbuf *mbuf)
{
        struct cpe_pkt *packet = rte_pktmbuf_mtod(mbuf, struct cpe_pkt *);
#ifdef USE_QINQ
        uint64_t key = (*(uint64_t*)(((uint8_t *)packet) + 12)) & 0xFF0FFFFFFF0FFFFF;
	uint32_t svlan = packet->qinq_hdr.svlan.vlan_tci;
	uint32_t cvlan = packet->qinq_hdr.cvlan.vlan_tci;

	svlan = rte_be_to_cpu_16(svlan & 0xFF0F);
	cvlan = rte_be_to_cpu_16(cvlan & 0xFF0F);
#if RTE_VERSION >= RTE_VERSION_NUM(20,11,0,0)
	plogx_err("Can't convert key %016lx qinq %d|%d (%x|%x) to gre_id, rss=%x flags=%lx, status_err_len=%x, L2Tag=%d type=%d\n",
		  key, svlan, cvlan, svlan, cvlan, mbuf->hash.rss, mbuf->ol_flags, mbuf->dynfield1[0], mbuf->vlan_tci_outer, mbuf->packet_type);
#else
#if RTE_VERSION >= RTE_VERSION_NUM(2,1,0,0)
	plogx_err("Can't convert key %016lx qinq %d|%d (%x|%x) to gre_id, rss=%x flags=%lx, status_err_len=%lx, L2Tag=%d type=%d\n",
		  key, svlan, cvlan, svlan, cvlan, mbuf->hash.rss, mbuf->ol_flags, mbuf->udata64, mbuf->vlan_tci_outer, mbuf->packet_type);
#else
#if RTE_VERSION >= RTE_VERSION_NUM(1,8,0,0)
	plogx_err("Can't convert key %016lx qinq %d|%d (%x|%x) to gre_id, rss=%x flags=%lx, status_err_len=%lx, L2Tag=%d type=%d\n",
		  key, svlan, cvlan, svlan, cvlan, mbuf->hash.rss, mbuf->ol_flags, mbuf->udata64, mbuf->reserved, mbuf->packet_type);
#else
	plogx_err("Can't convert key %016lx qinq %d|%d (%x|%x) to gre_id, flags=%x, L2Tag=%d\n",
		  key, svlan, cvlan, svlan, cvlan, mbuf->ol_flags, mbuf->reserved);
#endif
#endif
#endif
#else
	plogx_err("Can't convert ip %x to gre_id\n", rte_bswap32(packet->ipv4_hdr.src_addr));
#endif
}

static int add_cpe_entry(struct rte_table_hash *hash, struct cpe_key *key, struct cpe_data *data)
{
	void* entry_in_hash;
	int ret, key_found = 0;

	ret = prox_rte_table_key8_add(hash, key, data, &key_found, &entry_in_hash);
	if (unlikely(ret)) {
		plogx_err("Failed to add key: ip %x, gre %x\n", key->ip, key->gre_id);
		return 1;
	}
	return 0;
}

static void extract_key_data_arp(struct rte_mbuf* mbuf, struct cpe_key* key, struct cpe_data* data, const struct qinq_gre_data* entry, uint64_t cpe_timeout, uint8_t* mapping)
{
	const struct cpe_packet_arp *packet = rte_pktmbuf_mtod(mbuf, const struct cpe_packet_arp *);
	uint32_t svlan = packet->qinq_hdr.svlan.vlan_tci & 0xFF0F;
	uint32_t cvlan = packet->qinq_hdr.cvlan.vlan_tci & 0xFF0F;
	uint8_t port_id;
	key->ip = packet->arp.data.spa;
	key->gre_id = entry->gre_id;

	data->mac_port_8bytes = *((const uint64_t *)(&packet->qinq_hdr.s_addr));
	data->qinq_svlan = svlan;
	data->qinq_cvlan = cvlan;
#if RTE_VERSION >= RTE_VERSION_NUM(1,8,0,0)
	port_id = mbuf->port;

#else
	port_id = mbuf->pkt.in_port;
#endif
	uint8_t mapped = mapping[port_id];
	data->mac_port.out_idx = mapping[port_id];

	if (unlikely(mapped == 255)) {
		/* This error only occurs if the system is configured incorrectly */
		plog_warn("Failed adding packet: unknown mapping for port %d", port_id);
		data->mac_port.out_idx = 0;
	}

	data->user = entry->user;
	data->tsc = rte_rdtsc() + cpe_timeout;
}

void arp_msg_to_str(char *str, struct arp_msg *msg)
{
	sprintf(str, "%u %u %u %u %u.%u.%u.%u %x:%x:%x:%x:%x:%x %u\n",
		msg->data.mac_port.out_idx, msg->key.gre_id, msg->data.qinq_svlan, msg->data.qinq_cvlan,
		msg->key.ip_bytes[0], msg->key.ip_bytes[1], msg->key.ip_bytes[2], msg->key.ip_bytes[3],
		msg->data.mac_port_b[0], msg->data.mac_port_b[1], msg->data.mac_port_b[2],
		msg->data.mac_port_b[3], msg->data.mac_port_b[4], msg->data.mac_port_b[5], msg->data.user);
}

int str_to_arp_msg(struct arp_msg *msg, const char *str)
{
	uint32_t ip[4],	interface, gre_id, svlan, cvlan, mac[6], user;

	int ret = sscanf(str, "%u %u %u %u %u.%u.%u.%u %x:%x:%x:%x:%x:%x %u",
			 &interface, &gre_id, &svlan, &cvlan,
			 ip, ip + 1, ip + 2, ip + 3,
			 mac, mac + 1, mac + 2, mac + 3, mac + 4, mac + 5, &user);

	for (uint8_t i = 0; i < 4; ++i)
		msg->key.ip_bytes[i] = ip[i];
	msg->key.gre_id = gre_id;

	for (uint8_t i = 0; i < 4; ++i)
		msg->data.mac_port_b[i] = mac[i];
	msg->data.qinq_svlan = svlan;
	msg->data.qinq_cvlan = cvlan;
	msg->data.user = user;
	msg->data.mac_port.out_idx = interface;

	return ret != 15;
}

void arp_update_from_msg(struct rte_table_hash * cpe_table, struct arp_msg **msgs, uint16_t n_msgs, uint64_t cpe_timeout)
{
	int ret, key_found = 0;
	void* entry_in_hash;

	for (uint16_t i = 0; i < n_msgs; ++i) {
		msgs[i]->data.tsc = rte_rdtsc() + cpe_timeout;
		ret = prox_rte_table_key8_add(cpe_table, &msgs[i]->key, &msgs[i]->data, &key_found, &entry_in_hash);
		if (unlikely(ret)) {
			plogx_err("Failed to add key %x, gre %x\n", msgs[i]->key.ip, msgs[i]->key.gre_id);
		}
	}
}

static void arp_msg(struct task_base *tbase, void **data, uint16_t n_msgs)
{
	struct task_qinq_decap4 *task = (struct task_qinq_decap4 *)tbase;
	struct arp_msg **msgs = (struct arp_msg **)data;

	arp_update_from_msg(task->cpe_table, msgs, n_msgs, task->cpe_timeout);
}

static void arp_update(struct task_base *tbase, struct rte_mbuf **mbufs, uint16_t n_pkts)
{
	struct task_qinq_decap4 *task = (struct task_qinq_decap4 *)tbase;

	prefetch_pkts(mbufs, n_pkts);
	extract_key_bulk(mbufs, n_pkts, task);

	uint64_t pkts_mask = RTE_LEN2MASK(n_pkts, uint64_t);
	uint64_t lookup_hit_mask = 0;
	struct qinq_gre_data* entries[64];
	prox_rte_table_key8_lookup(task->qinq_gre_table, task->fake_packets, pkts_mask, &lookup_hit_mask, (void**)entries);

	TASK_STATS_ADD_RX(&task->base.aux->stats, n_pkts);
	for (uint16_t j = 0; j < n_pkts; ++j) {
		if (unlikely(!((lookup_hit_mask >> j) & 0x1))) {
			handle_error(mbufs[j]);
			rte_pktmbuf_free(mbufs[j]);
			continue;
		}

		struct cpe_key key;
		struct cpe_data data;

		extract_key_data_arp(mbufs[j], &key, &data, entries[j], task->cpe_timeout, task->mapping);

		void* entry_in_hash;
		int ret, key_found = 0;

		ret = prox_rte_table_key8_add(task->cpe_table, &key, &data, &key_found, &entry_in_hash);

		if (unlikely(ret)) {
			plogx_err("Failed to add key %x, gre %x\n", key.ip, key.gre_id);
			TASK_STATS_ADD_DROP_DISCARD(&task->base.aux->stats, 1);
		}

		/* should do ARP reply */
		TASK_STATS_ADD_DROP_HANDLED(&task->base.aux->stats, 1);
		rte_pktmbuf_free(mbufs[j]);
	}
}

static int handle_qinq_decap4_bulk(struct task_base *tbase, struct rte_mbuf **mbufs, uint16_t n_pkts)
{
	struct task_qinq_decap4 *task = (struct task_qinq_decap4 *)tbase;
	uint64_t pkts_mask = RTE_LEN2MASK(n_pkts, uint64_t);
	struct qinq_gre_data* entries[64];
	uint8_t out[MAX_PKT_BURST];
	uint64_t lookup_hit_mask;
	prefetch_pkts(mbufs, n_pkts);

	// Prefetch headroom, as we will prepend mbuf and write to this cache line
	for (uint16_t j = 0; j < n_pkts; ++j) {
		PREFETCH0((rte_pktmbuf_mtod(mbufs[j], char*)-1));
	}

	extract_key_bulk(mbufs, n_pkts, task);
	prox_rte_table_key8_lookup(task->qinq_gre_table, task->fake_packets, pkts_mask, &lookup_hit_mask, (void**)entries);

	if (likely(lookup_hit_mask == pkts_mask)) {
		for (uint16_t j = 0; j < n_pkts; ++j) {
			out[j] = handle_qinq_decap4(task, mbufs[j], entries[j]);
		}
	}
	else {
		for (uint16_t j = 0; j < n_pkts; ++j) {
			if (unlikely(!((lookup_hit_mask >> j) & 0x1))) {
				// This might fail as the packet has not the expected QinQ or it's not an IPv4 packet
	                        handle_error(mbufs[j]);
				out[j] = OUT_DISCARD;
				continue;
			}
			out[j] = handle_qinq_decap4(task, mbufs[j], entries[j]);
		}
	}

	return task->base.tx_pkt(&task->base, mbufs, n_pkts, out);
}

/* add gre header */
static inline void gre_encap(struct task_qinq_decap4 *task, uint32_t src_ipv4, struct rte_mbuf *mbuf, uint32_t gre_id)
{
#ifdef USE_QINQ
	prox_rte_ipv4_hdr *pip = (prox_rte_ipv4_hdr *)(1 + rte_pktmbuf_mtod(mbuf, struct qinq_hdr *));
#else
	prox_rte_ipv4_hdr *pip = (prox_rte_ipv4_hdr *)(1 + rte_pktmbuf_mtod(mbuf, prox_rte_ether_hdr *));
#endif
	uint16_t ip_len = rte_be_to_cpu_16(pip->total_length);
	uint16_t padlen = rte_pktmbuf_pkt_len(mbuf) - 20 - ip_len - sizeof(struct qinq_hdr);

	if (padlen) {
		rte_pktmbuf_trim(mbuf, padlen);
	}

	PROX_PANIC(rte_pktmbuf_data_len(mbuf) - padlen + 20 > PROX_RTE_ETHER_MAX_LEN,
	           "Would need to fragment packet new size = %u - not implemented\n",
	           rte_pktmbuf_data_len(mbuf) - padlen + 20);

#ifdef USE_QINQ
	/* prepend only 20 bytes instead of 28, 8 bytes are present from the QinQ */
	prox_rte_ether_hdr *peth = (prox_rte_ether_hdr *)rte_pktmbuf_prepend(mbuf, 20);
#else
	prox_rte_ether_hdr *peth = (prox_rte_ether_hdr *)rte_pktmbuf_prepend(mbuf, 28);
#endif

	PROX_ASSERT(peth);
	PREFETCH0(peth);
	if (task->runtime_flags & TASK_TX_CRC) {
		/* calculate IP CRC here to avoid problems with -O3 flag with gcc */
#ifdef MPLS_ROUTING
		prox_ip_cksum(mbuf, pip, sizeof(prox_rte_ether_hdr) + sizeof(struct mpls_hdr), sizeof(prox_rte_ipv4_hdr), task->offload_crc);
#else
		prox_ip_cksum(mbuf, pip, sizeof(prox_rte_ether_hdr), sizeof(prox_rte_ipv4_hdr), task->offload_crc);
#endif
	}

	/* new IP header */
	prox_rte_ipv4_hdr *p_tunnel_ip = (prox_rte_ipv4_hdr *)(peth + 1);
	rte_memcpy(p_tunnel_ip, &tunnel_ip_proto, sizeof(prox_rte_ipv4_hdr));
	ip_len += sizeof(prox_rte_ipv4_hdr) + sizeof(struct gre_hdr);
	p_tunnel_ip->total_length = rte_cpu_to_be_16(ip_len);
	p_tunnel_ip->src_addr = src_ipv4;

	/* Add GRE Header values */
	struct gre_hdr *pgre = (struct gre_hdr *)(p_tunnel_ip + 1);

	rte_memcpy(pgre, &gre_hdr_proto, sizeof(struct gre_hdr));
	pgre->gre_id = gre_id;
	peth->ether_type = ETYPE_IPv4;
}

static inline uint16_t calc_padlen(const struct rte_mbuf *mbuf, const uint16_t ip_len)
{
	return rte_pktmbuf_pkt_len(mbuf) - DOWNSTREAM_DELTA - ip_len - offsetof(struct cpe_pkt, ipv4_hdr);
}

static inline uint8_t gre_encap_route(uint32_t src_ipv4, struct rte_mbuf *mbuf, uint32_t gre_id, struct task_qinq_decap4 *task)
{
	PROX_PANIC(rte_pktmbuf_data_len(mbuf) + DOWNSTREAM_DELTA  > PROX_RTE_ETHER_MAX_LEN,
	           "Would need to fragment packet new size = %u - not implemented\n",
	           rte_pktmbuf_data_len(mbuf) + DOWNSTREAM_DELTA);

	struct core_net_pkt_m *packet = (struct core_net_pkt_m *)rte_pktmbuf_prepend(mbuf, DOWNSTREAM_DELTA);
	PROX_ASSERT(packet);
	PREFETCH0(packet);

	prox_rte_ipv4_hdr *pip = &((struct cpe_pkt_delta *)packet)->pkt.ipv4_hdr;
	uint16_t ip_len = rte_be_to_cpu_16(pip->total_length);

	/* returns 0 on success, returns -ENOENT of failure (or -EINVAL if first or last parameter is NULL) */
#if RTE_VERSION >= RTE_VERSION_NUM(16,4,0,1)
	uint32_t next_hop_index;
#else
	uint8_t next_hop_index;
#endif
	if (unlikely(rte_lpm_lookup(task->ipv4_lpm, rte_bswap32(pip->dst_addr), &next_hop_index) != 0)) {
		plog_warn("lpm_lookup failed for ip %x: rc = %d\n", rte_bswap32(pip->dst_addr), -ENOENT);
		return ROUTE_ERR;
	}
	PREFETCH0(&task->next_hops[next_hop_index]);

	/* calculate outer IP CRC here to avoid problems with -O3 flag with gcc */
	const uint16_t padlen = calc_padlen(mbuf, ip_len);
	if (padlen) {
		rte_pktmbuf_trim(mbuf, padlen);
	}
	const uint8_t port_id = task->next_hops[next_hop_index].mac_port.out_idx;

	*((uint64_t *)(&packet->ether_hdr.d_addr)) = task->next_hops[next_hop_index].mac_port_8bytes;
	*((uint64_t *)(&packet->ether_hdr.s_addr)) = task->src_mac[task->next_hops[next_hop_index].mac_port.out_idx];

#ifdef MPLS_ROUTING
	packet->mpls_bytes = task->next_hops[next_hop_index].mpls | 0x00010000; // Set BoS to 1
	packet->ether_hdr.ether_type = ETYPE_MPLSU;
#else
	packet->ether_hdr.ether_type = ETYPE_IPv4;
#endif

	/* New IP header */
	rte_memcpy(&packet->tunnel_ip_hdr, &tunnel_ip_proto, sizeof(prox_rte_ipv4_hdr));
	ip_len += sizeof(prox_rte_ipv4_hdr) + sizeof(struct gre_hdr);
	packet->tunnel_ip_hdr.total_length = rte_cpu_to_be_16(ip_len);
	packet->tunnel_ip_hdr.src_addr = src_ipv4;
	packet->tunnel_ip_hdr.dst_addr = task->next_hops[next_hop_index].ip_dst;
	if (task->runtime_flags & TASK_TX_CRC) {
#ifdef MPLS_ROUTING
		prox_ip_cksum(mbuf, (void *)&(packet->tunnel_ip_hdr), sizeof(prox_rte_ether_hdr) + sizeof(struct mpls_hdr), sizeof(prox_rte_ipv4_hdr), task->offload_crc);
#else
		prox_ip_cksum(mbuf, (void *)&(packet->tunnel_ip_hdr), sizeof(prox_rte_ether_hdr), sizeof(prox_rte_ipv4_hdr), task->offload_crc);
#endif
	}

	/* Add GRE Header values */
	rte_memcpy(&packet->gre_hdr, &gre_hdr_proto, sizeof(struct gre_hdr));
	packet->gre_hdr.gre_id = rte_be_to_cpu_32(gre_id);

	return port_id;
}

static void extract_key_data(struct rte_mbuf* mbuf, struct cpe_key* key, struct cpe_data* data, const struct qinq_gre_data* entry, uint64_t cpe_timeout, uint8_t *mapping)
{
	struct cpe_pkt *packet = rte_pktmbuf_mtod(mbuf, struct cpe_pkt *);
	uint8_t port_id;

#ifndef USE_QINQ
        const uint32_t tmp = rte_bswap32(packet->ipv4_hdr.src_addr) & 0x00FFFFFF;
	const uint32_t svlan = rte_bswap16(tmp >> 12);
	const uint32_t cvlan = rte_bswap16(tmp & 0x0FFF);
#endif

#ifdef USE_QINQ
	key->ip = packet->ipv4_hdr.src_addr;
#else
	key->ip = 0;
#endif
	key->gre_id = entry->gre_id;

#ifdef USE_QINQ
	data->mac_port_8bytes = *((const uint64_t *)(&packet->qinq_hdr.s_addr));
	data->qinq_svlan      = packet->qinq_hdr.svlan.vlan_tci & 0xFF0F;
	data->qinq_cvlan      = packet->qinq_hdr.cvlan.vlan_tci & 0xFF0F;
#else
	data->mac_port_8bytes = *((const uint64_t *)(&packet->ether_hdr.s_addr));
	data->qinq_svlan      = svlan;
	data->qinq_cvlan      = cvlan;
#endif

#if RTE_VERSION >= RTE_VERSION_NUM(1,8,0,0)
	port_id = mbuf->port;

#else
	port_id = mbuf->pkt.in_port;
#endif
	uint8_t mapped = mapping[port_id];
	data->mac_port.out_idx = mapped;

	if (unlikely(mapped == 255)) {
		/* This error only occurs if the system is configured incorrectly */
		plog_warn("Failed adding packet: unknown mapping for port %d", port_id);
		data->mac_port.out_idx = 0;
	}
	else {
		data->mac_port.out_idx = mapped;
	}

	data->user             = entry->user;
	data->tsc              = rte_rdtsc() + cpe_timeout;
}

static uint8_t handle_qinq_decap4(struct task_qinq_decap4 *task, struct rte_mbuf *mbuf, struct qinq_gre_data* entry)
{
	if (!(task->runtime_flags & (TASK_CTRL_HANDLE_ARP|TASK_FP_HANDLE_ARP))) {
		// We learn CPE MAC addresses on every packets
		struct cpe_key key;
		struct cpe_data data;
		extract_key_data(mbuf, &key, &data, entry, task->cpe_timeout, task->mapping);
		//plogx_err("Adding key ip=%x/gre_id=%x data (svlan|cvlan)=%x|%x, rss=%x, gre_id=%x\n", key.ip, key.gre_id, data.qinq_svlan,data.qinq_cvlan, mbuf->hash.rss, entry->gre_id);

		if (add_cpe_entry(task->cpe_table, &key, &data)) {
			plog_warn("Failed to add ARP entry\n");
			return OUT_DISCARD;
		}
	}
	if (task->runtime_flags & TASK_FP_HANDLE_ARP) {
		// We learn CPE MAC addresses on ARP packets in Fast Path
#if RTE_VERSION >= RTE_VERSION_NUM(1,8,0,0)
		if (mbuf->packet_type == 0xB) {
			struct cpe_key key;
			struct cpe_data data;
			extract_key_data_arp(mbuf, &key, &data, entry, task->cpe_timeout, task->mapping);

			if (add_cpe_entry(task->cpe_table, &key, &data)) {
				plog_warn("Failed to add ARP entry\n");
				return OUT_DISCARD;
			}
			return OUT_HANDLED;
		} else
#endif
		{
#ifdef USE_QINQ
			struct cpe_pkt *packet = rte_pktmbuf_mtod(mbuf, struct cpe_pkt*);
			if (packet->qinq_hdr.svlan.eth_proto == task->qinq_tag &&
			    packet->qinq_hdr.ether_type == ETYPE_ARP) {
				struct cpe_key key;
				struct cpe_data data;
				extract_key_data_arp(mbuf, &key, &data, entry, task->cpe_timeout, task->mapping);

				if (add_cpe_entry(task->cpe_table, &key, &data)) {
					plog_warn("Failed to add ARP entry\n");
					return OUT_DISCARD;
				}
				return OUT_HANDLED;
			}
#endif
		}
	}
	if (task->runtime_flags & TASK_ROUTING) {
		uint8_t tx_portid;
		tx_portid = gre_encap_route(task->local_ipv4, mbuf, entry->gre_id, task);

		return tx_portid == ROUTE_ERR? OUT_DISCARD : tx_portid;
	}
	else {
		gre_encap(task, task->local_ipv4, mbuf, entry->gre_id);
		return 0;
	}
}

static void flow_iter_next(struct flow_iter *iter, struct task_args *targ)
{
	do {
		iter->idx++;
	} while (iter->idx < (int)get_qinq_gre_map(targ)->count &&
		 get_qinq_gre_map(targ)->entries[iter->idx].gre_id % targ->nb_slave_threads != targ->worker_thread_id);
}

static void flow_iter_beg(struct flow_iter *iter, struct task_args *targ)
{
	iter->idx = -1;
	flow_iter_next(iter, targ);
}

static int flow_iter_is_end(struct flow_iter *iter, struct task_args *targ)
{
	return iter->idx == (int)get_qinq_gre_map(targ)->count;
}

static uint16_t flow_iter_get_svlan(struct flow_iter *iter, struct task_args *targ)
{
	return get_qinq_gre_map(targ)->entries[iter->idx].svlan;
}

static uint16_t flow_iter_get_cvlan(struct flow_iter *iter, struct task_args *targ)
{
	return get_qinq_gre_map(targ)->entries[iter->idx].cvlan;
}

static struct task_init task_init_qinq_decapv4_table = {
	.mode = QINQ_DECAP4,
	.mode_str = "qinqdecapv4",
	.early_init = early_init_table,
	.init = init_task_qinq_decap4,
	.handle = handle_qinq_decap4_bulk,
	.flag_features = TASK_FEATURE_ROUTING,
	.flow_iter = {
		.beg       = flow_iter_beg,
		.is_end    = flow_iter_is_end,
		.next      = flow_iter_next,
		.get_svlan = flow_iter_get_svlan,
		.get_cvlan = flow_iter_get_cvlan,
	},
	.size = sizeof(struct task_qinq_decap4)
};

__attribute__((constructor)) static void reg_task_qinq_decap4(void)
{
	reg_task(&task_init_qinq_decapv4_table);
}