DPDK logo

Elixir Cross Referencer

   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
 663
 664
 665
 666
 667
 668
 669
 670
 671
 672
 673
 674
 675
 676
 677
 678
 679
 680
 681
 682
 683
 684
 685
 686
 687
 688
 689
 690
 691
 692
 693
 694
 695
 696
 697
 698
 699
 700
 701
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
/*-
 *   BSD LICENSE
 *
 *   Copyright 2015 6WIND S.A.
 *   Copyright 2015 Mellanox.
 *
 *   Redistribution and use in source and binary forms, with or without
 *   modification, are permitted provided that the following conditions
 *   are met:
 *
 *     * Redistributions of source code must retain the above copyright
 *       notice, this list of conditions and the following disclaimer.
 *     * Redistributions in binary form must reproduce the above copyright
 *       notice, this list of conditions and the following disclaimer in
 *       the documentation and/or other materials provided with the
 *       distribution.
 *     * Neither the name of 6WIND S.A. nor the names of its
 *       contributors may be used to endorse or promote products derived
 *       from this software without specific prior written permission.
 *
 *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
 *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
 *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
 *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
 *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#include <stddef.h>
#include <assert.h>
#include <errno.h>
#include <string.h>
#include <stdint.h>
#include <fcntl.h>
#include <sys/queue.h>

/* Verbs header. */
/* ISO C doesn't support unnamed structs/unions, disabling -pedantic. */
#ifdef PEDANTIC
#pragma GCC diagnostic ignored "-Wpedantic"
#endif
#include <infiniband/verbs.h>
#include <infiniband/mlx5dv.h>
#ifdef PEDANTIC
#pragma GCC diagnostic error "-Wpedantic"
#endif

#include <rte_mbuf.h>
#include <rte_malloc.h>
#include <rte_ethdev.h>
#include <rte_common.h>
#include <rte_interrupts.h>
#include <rte_debug.h>
#include <rte_io.h>

#include "mlx5.h"
#include "mlx5_rxtx.h"
#include "mlx5_utils.h"
#include "mlx5_autoconf.h"
#include "mlx5_defs.h"

/* Default RSS hash key also used for ConnectX-3. */
uint8_t rss_hash_default_key[] = {
	0x2c, 0xc6, 0x81, 0xd1,
	0x5b, 0xdb, 0xf4, 0xf7,
	0xfc, 0xa2, 0x83, 0x19,
	0xdb, 0x1a, 0x3e, 0x94,
	0x6b, 0x9e, 0x38, 0xd9,
	0x2c, 0x9c, 0x03, 0xd1,
	0xad, 0x99, 0x44, 0xa7,
	0xd9, 0x56, 0x3d, 0x59,
	0x06, 0x3c, 0x25, 0xf3,
	0xfc, 0x1f, 0xdc, 0x2a,
};

/* Length of the default RSS hash key. */
const size_t rss_hash_default_key_len = sizeof(rss_hash_default_key);

/**
 * Allocate RX queue elements.
 *
 * @param rxq_ctrl
 *   Pointer to RX queue structure.
 *
 * @return
 *   0 on success, errno value on failure.
 */
int
rxq_alloc_elts(struct mlx5_rxq_ctrl *rxq_ctrl)
{
	const unsigned int sges_n = 1 << rxq_ctrl->rxq.sges_n;
	unsigned int elts_n = 1 << rxq_ctrl->rxq.elts_n;
	unsigned int i;
	int ret = 0;

	/* Iterate on segments. */
	for (i = 0; (i != elts_n); ++i) {
		struct rte_mbuf *buf;

		buf = rte_pktmbuf_alloc(rxq_ctrl->rxq.mp);
		if (buf == NULL) {
			ERROR("%p: empty mbuf pool", (void *)rxq_ctrl);
			ret = ENOMEM;
			goto error;
		}
		/* Headroom is reserved by rte_pktmbuf_alloc(). */
		assert(DATA_OFF(buf) == RTE_PKTMBUF_HEADROOM);
		/* Buffer is supposed to be empty. */
		assert(rte_pktmbuf_data_len(buf) == 0);
		assert(rte_pktmbuf_pkt_len(buf) == 0);
		assert(!buf->next);
		/* Only the first segment keeps headroom. */
		if (i % sges_n)
			SET_DATA_OFF(buf, 0);
		PORT(buf) = rxq_ctrl->rxq.port_id;
		DATA_LEN(buf) = rte_pktmbuf_tailroom(buf);
		PKT_LEN(buf) = DATA_LEN(buf);
		NB_SEGS(buf) = 1;
		(*rxq_ctrl->rxq.elts)[i] = buf;
	}
	/* If Rx vector is activated. */
	if (rxq_check_vec_support(&rxq_ctrl->rxq) > 0) {
		struct mlx5_rxq_data *rxq = &rxq_ctrl->rxq;
		struct rte_mbuf *mbuf_init = &rxq->fake_mbuf;
		int j;

		/* Initialize default rearm_data for vPMD. */
		mbuf_init->data_off = RTE_PKTMBUF_HEADROOM;
		rte_mbuf_refcnt_set(mbuf_init, 1);
		mbuf_init->nb_segs = 1;
		mbuf_init->port = rxq->port_id;
		/*
		 * prevent compiler reordering:
		 * rearm_data covers previous fields.
		 */
		rte_compiler_barrier();
		rxq->mbuf_initializer =
			*(uint64_t *)&mbuf_init->rearm_data;
		/* Padding with a fake mbuf for vectorized Rx. */
		for (j = 0; j < MLX5_VPMD_DESCS_PER_LOOP; ++j)
			(*rxq->elts)[elts_n + j] = &rxq->fake_mbuf;
	}
	DEBUG("%p: allocated and configured %u segments (max %u packets)",
	      (void *)rxq_ctrl, elts_n, elts_n / (1 << rxq_ctrl->rxq.sges_n));
	assert(ret == 0);
	return 0;
error:
	elts_n = i;
	for (i = 0; (i != elts_n); ++i) {
		if ((*rxq_ctrl->rxq.elts)[i] != NULL)
			rte_pktmbuf_free_seg((*rxq_ctrl->rxq.elts)[i]);
		(*rxq_ctrl->rxq.elts)[i] = NULL;
	}
	DEBUG("%p: failed, freed everything", (void *)rxq_ctrl);
	assert(ret > 0);
	return ret;
}

/**
 * Free RX queue elements.
 *
 * @param rxq_ctrl
 *   Pointer to RX queue structure.
 */
static void
rxq_free_elts(struct mlx5_rxq_ctrl *rxq_ctrl)
{
	struct mlx5_rxq_data *rxq = &rxq_ctrl->rxq;
	const uint16_t q_n = (1 << rxq->elts_n);
	const uint16_t q_mask = q_n - 1;
	uint16_t used = q_n - (rxq->rq_ci - rxq->rq_pi);
	uint16_t i;

	DEBUG("%p: freeing WRs", (void *)rxq_ctrl);
	if (rxq->elts == NULL)
		return;
	/**
	 * Some mbuf in the Ring belongs to the application.  They cannot be
	 * freed.
	 */
	if (rxq_check_vec_support(rxq) > 0) {
		for (i = 0; i < used; ++i)
			(*rxq->elts)[(rxq->rq_ci + i) & q_mask] = NULL;
		rxq->rq_pi = rxq->rq_ci;
	}
	for (i = 0; (i != (1u << rxq->elts_n)); ++i) {
		if ((*rxq->elts)[i] != NULL)
			rte_pktmbuf_free_seg((*rxq->elts)[i]);
		(*rxq->elts)[i] = NULL;
	}
}

/**
 * Clean up a RX queue.
 *
 * Destroy objects, free allocated memory and reset the structure for reuse.
 *
 * @param rxq_ctrl
 *   Pointer to RX queue structure.
 */
void
mlx5_rxq_cleanup(struct mlx5_rxq_ctrl *rxq_ctrl)
{
	DEBUG("cleaning up %p", (void *)rxq_ctrl);
	if (rxq_ctrl->ibv)
		mlx5_priv_rxq_ibv_release(rxq_ctrl->priv, rxq_ctrl->ibv);
	memset(rxq_ctrl, 0, sizeof(*rxq_ctrl));
}

/**
 *
 * @param dev
 *   Pointer to Ethernet device structure.
 * @param idx
 *   RX queue index.
 * @param desc
 *   Number of descriptors to configure in queue.
 * @param socket
 *   NUMA socket on which memory must be allocated.
 * @param[in] conf
 *   Thresholds parameters.
 * @param mp
 *   Memory pool for buffer allocations.
 *
 * @return
 *   0 on success, negative errno value on failure.
 */
int
mlx5_rx_queue_setup(struct rte_eth_dev *dev, uint16_t idx, uint16_t desc,
		    unsigned int socket, const struct rte_eth_rxconf *conf,
		    struct rte_mempool *mp)
{
	struct priv *priv = dev->data->dev_private;
	struct mlx5_rxq_data *rxq = (*priv->rxqs)[idx];
	struct mlx5_rxq_ctrl *rxq_ctrl =
		container_of(rxq, struct mlx5_rxq_ctrl, rxq);
	int ret = 0;

	(void)conf;
	if (mlx5_is_secondary())
		return -E_RTE_SECONDARY;
	priv_lock(priv);
	if (!rte_is_power_of_2(desc)) {
		desc = 1 << log2above(desc);
		WARN("%p: increased number of descriptors in RX queue %u"
		     " to the next power of two (%d)",
		     (void *)dev, idx, desc);
	}
	DEBUG("%p: configuring queue %u for %u descriptors",
	      (void *)dev, idx, desc);
	if (idx >= priv->rxqs_n) {
		ERROR("%p: queue index out of range (%u >= %u)",
		      (void *)dev, idx, priv->rxqs_n);
		priv_unlock(priv);
		return -EOVERFLOW;
	}
	if (!mlx5_priv_rxq_releasable(priv, idx)) {
		ret = EBUSY;
		ERROR("%p: unable to release queue index %u",
		      (void *)dev, idx);
		goto out;
	}
	mlx5_priv_rxq_release(priv, idx);
	rxq_ctrl = mlx5_priv_rxq_new(priv, idx, desc, socket, mp);
	if (!rxq_ctrl) {
		ERROR("%p: unable to allocate queue index %u",
		      (void *)dev, idx);
		ret = ENOMEM;
		goto out;
	}
	DEBUG("%p: adding RX queue %p to list",
	      (void *)dev, (void *)rxq_ctrl);
	(*priv->rxqs)[idx] = &rxq_ctrl->rxq;
out:
	priv_unlock(priv);
	return -ret;
}

/**
 * DPDK callback to release a RX queue.
 *
 * @param dpdk_rxq
 *   Generic RX queue pointer.
 */
void
mlx5_rx_queue_release(void *dpdk_rxq)
{
	struct mlx5_rxq_data *rxq = (struct mlx5_rxq_data *)dpdk_rxq;
	struct mlx5_rxq_ctrl *rxq_ctrl;
	struct priv *priv;

	if (mlx5_is_secondary())
		return;

	if (rxq == NULL)
		return;
	rxq_ctrl = container_of(rxq, struct mlx5_rxq_ctrl, rxq);
	priv = rxq_ctrl->priv;
	priv_lock(priv);
	if (!mlx5_priv_rxq_releasable(priv, rxq_ctrl->rxq.stats.idx))
		rte_panic("Rx queue %p is still used by a flow and cannot be"
			  " removed\n", (void *)rxq_ctrl);
	mlx5_priv_rxq_release(priv, rxq_ctrl->rxq.stats.idx);
	priv_unlock(priv);
}

/**
 * Allocate queue vector and fill epoll fd list for Rx interrupts.
 *
 * @param priv
 *   Pointer to private structure.
 *
 * @return
 *   0 on success, negative on failure.
 */
int
priv_rx_intr_vec_enable(struct priv *priv)
{
	unsigned int i;
	unsigned int rxqs_n = priv->rxqs_n;
	unsigned int n = RTE_MIN(rxqs_n, (uint32_t)RTE_MAX_RXTX_INTR_VEC_ID);
	unsigned int count = 0;
	struct rte_intr_handle *intr_handle = priv->dev->intr_handle;

	assert(!mlx5_is_secondary());
	if (!priv->dev->data->dev_conf.intr_conf.rxq)
		return 0;
	priv_rx_intr_vec_disable(priv);
	intr_handle->intr_vec = malloc(n * sizeof(intr_handle->intr_vec[0]));
	if (intr_handle->intr_vec == NULL) {
		ERROR("failed to allocate memory for interrupt vector,"
		      " Rx interrupts will not be supported");
		return -ENOMEM;
	}
	intr_handle->type = RTE_INTR_HANDLE_EXT;
	for (i = 0; i != n; ++i) {
		/* This rxq ibv must not be released in this function. */
		struct mlx5_rxq_ibv *rxq_ibv = mlx5_priv_rxq_ibv_get(priv, i);
		int fd;
		int flags;
		int rc;

		/* Skip queues that cannot request interrupts. */
		if (!rxq_ibv || !rxq_ibv->channel) {
			/* Use invalid intr_vec[] index to disable entry. */
			intr_handle->intr_vec[i] =
				RTE_INTR_VEC_RXTX_OFFSET +
				RTE_MAX_RXTX_INTR_VEC_ID;
			continue;
		}
		if (count >= RTE_MAX_RXTX_INTR_VEC_ID) {
			ERROR("too many Rx queues for interrupt vector size"
			      " (%d), Rx interrupts cannot be enabled",
			      RTE_MAX_RXTX_INTR_VEC_ID);
			priv_rx_intr_vec_disable(priv);
			return -1;
		}
		fd = rxq_ibv->channel->fd;
		flags = fcntl(fd, F_GETFL);
		rc = fcntl(fd, F_SETFL, flags | O_NONBLOCK);
		if (rc < 0) {
			ERROR("failed to make Rx interrupt file descriptor"
			      " %d non-blocking for queue index %d", fd, i);
			priv_rx_intr_vec_disable(priv);
			return -1;
		}
		intr_handle->intr_vec[i] = RTE_INTR_VEC_RXTX_OFFSET + count;
		intr_handle->efds[count] = fd;
		count++;
	}
	if (!count)
		priv_rx_intr_vec_disable(priv);
	else
		intr_handle->nb_efd = count;
	return 0;
}

/**
 * Clean up Rx interrupts handler.
 *
 * @param priv
 *   Pointer to private structure.
 */
void
priv_rx_intr_vec_disable(struct priv *priv)
{
	struct rte_intr_handle *intr_handle = priv->dev->intr_handle;
	unsigned int i;
	unsigned int rxqs_n = priv->rxqs_n;
	unsigned int n = RTE_MIN(rxqs_n, (uint32_t)RTE_MAX_RXTX_INTR_VEC_ID);

	if (!priv->dev->data->dev_conf.intr_conf.rxq)
		return;
	if (!intr_handle->intr_vec)
		goto free;
	for (i = 0; i != n; ++i) {
		struct mlx5_rxq_ctrl *rxq_ctrl;
		struct mlx5_rxq_data *rxq_data;

		if (intr_handle->intr_vec[i] == RTE_INTR_VEC_RXTX_OFFSET +
		    RTE_MAX_RXTX_INTR_VEC_ID)
			continue;
		/**
		 * Need to access directly the queue to release the reference
		 * kept in priv_rx_intr_vec_enable().
		 */
		rxq_data = (*priv->rxqs)[i];
		rxq_ctrl = container_of(rxq_data, struct mlx5_rxq_ctrl, rxq);
		mlx5_priv_rxq_ibv_release(priv, rxq_ctrl->ibv);
	}
free:
	rte_intr_free_epoll_fd(intr_handle);
	if (intr_handle->intr_vec)
		free(intr_handle->intr_vec);
	intr_handle->nb_efd = 0;
	intr_handle->intr_vec = NULL;
}

/**
 *  MLX5 CQ notification .
 *
 *  @param rxq
 *     Pointer to receive queue structure.
 *  @param sq_n_rxq
 *     Sequence number per receive queue .
 */
static inline void
mlx5_arm_cq(struct mlx5_rxq_data *rxq, int sq_n_rxq)
{
	int sq_n = 0;
	uint32_t doorbell_hi;
	uint64_t doorbell;
	void *cq_db_reg = (char *)rxq->cq_uar + MLX5_CQ_DOORBELL;

	sq_n = sq_n_rxq & MLX5_CQ_SQN_MASK;
	doorbell_hi = sq_n << MLX5_CQ_SQN_OFFSET | (rxq->cq_ci & MLX5_CI_MASK);
	doorbell = (uint64_t)doorbell_hi << 32;
	doorbell |=  rxq->cqn;
	rxq->cq_db[MLX5_CQ_ARM_DB] = rte_cpu_to_be_32(doorbell_hi);
	rte_wmb();
	rte_write64(rte_cpu_to_be_64(doorbell), cq_db_reg);
}

/**
 * DPDK callback for Rx queue interrupt enable.
 *
 * @param dev
 *   Pointer to Ethernet device structure.
 * @param rx_queue_id
 *   Rx queue number.
 *
 * @return
 *   0 on success, negative on failure.
 */
int
mlx5_rx_intr_enable(struct rte_eth_dev *dev, uint16_t rx_queue_id)
{
	struct priv *priv = mlx5_get_priv(dev);
	struct mlx5_rxq_data *rxq_data;
	struct mlx5_rxq_ctrl *rxq_ctrl;
	int ret = 0;

	priv_lock(priv);
	rxq_data = (*priv->rxqs)[rx_queue_id];
	if (!rxq_data) {
		ret = EINVAL;
		goto exit;
	}
	rxq_ctrl = container_of(rxq_data, struct mlx5_rxq_ctrl, rxq);
	if (rxq_ctrl->irq) {
		struct mlx5_rxq_ibv *rxq_ibv;

		rxq_ibv = mlx5_priv_rxq_ibv_get(priv, rx_queue_id);
		if (!rxq_ibv) {
			ret = EINVAL;
			goto exit;
		}
		mlx5_arm_cq(rxq_data, rxq_data->cq_arm_sn);
		mlx5_priv_rxq_ibv_release(priv, rxq_ibv);
	}
exit:
	priv_unlock(priv);
	if (ret)
		WARN("unable to arm interrupt on rx queue %d", rx_queue_id);
	return -ret;
}

/**
 * DPDK callback for Rx queue interrupt disable.
 *
 * @param dev
 *   Pointer to Ethernet device structure.
 * @param rx_queue_id
 *   Rx queue number.
 *
 * @return
 *   0 on success, negative on failure.
 */
int
mlx5_rx_intr_disable(struct rte_eth_dev *dev, uint16_t rx_queue_id)
{
	struct priv *priv = mlx5_get_priv(dev);
	struct mlx5_rxq_data *rxq_data;
	struct mlx5_rxq_ctrl *rxq_ctrl;
	struct mlx5_rxq_ibv *rxq_ibv = NULL;
	struct ibv_cq *ev_cq;
	void *ev_ctx;
	int ret = 0;

	priv_lock(priv);
	rxq_data = (*priv->rxqs)[rx_queue_id];
	if (!rxq_data) {
		ret = EINVAL;
		goto exit;
	}
	rxq_ctrl = container_of(rxq_data, struct mlx5_rxq_ctrl, rxq);
	if (!rxq_ctrl->irq)
		goto exit;
	rxq_ibv = mlx5_priv_rxq_ibv_get(priv, rx_queue_id);
	if (!rxq_ibv) {
		ret = EINVAL;
		goto exit;
	}
	ret = ibv_get_cq_event(rxq_ibv->channel, &ev_cq, &ev_ctx);
	if (ret || ev_cq != rxq_ibv->cq) {
		ret = EINVAL;
		goto exit;
	}
	rxq_data->cq_arm_sn++;
	ibv_ack_cq_events(rxq_ibv->cq, 1);
exit:
	if (rxq_ibv)
		mlx5_priv_rxq_ibv_release(priv, rxq_ibv);
	priv_unlock(priv);
	if (ret)
		WARN("unable to disable interrupt on rx queue %d",
		     rx_queue_id);
	return -ret;
}

/**
 * Create the Rx queue Verbs object.
 *
 * @param priv
 *   Pointer to private structure.
 * @param idx
 *   Queue index in DPDK Rx queue array
 *
 * @return
 *   The Verbs object initialised if it can be created.
 */
struct mlx5_rxq_ibv*
mlx5_priv_rxq_ibv_new(struct priv *priv, uint16_t idx)
{
	struct mlx5_rxq_data *rxq_data = (*priv->rxqs)[idx];
	struct mlx5_rxq_ctrl *rxq_ctrl =
		container_of(rxq_data, struct mlx5_rxq_ctrl, rxq);
	struct ibv_wq_attr mod;
	union {
		struct {
			struct ibv_cq_init_attr_ex ibv;
			struct mlx5dv_cq_init_attr mlx5;
		} cq;
		struct ibv_wq_init_attr wq;
		struct ibv_cq_ex cq_attr;
	} attr;
	unsigned int cqe_n = (1 << rxq_data->elts_n) - 1;
	struct mlx5_rxq_ibv *tmpl;
	struct mlx5dv_cq cq_info;
	struct mlx5dv_rwq rwq;
	unsigned int i;
	int ret = 0;
	struct mlx5dv_obj obj;

	assert(rxq_data);
	assert(!rxq_ctrl->ibv);
	tmpl = rte_calloc_socket(__func__, 1, sizeof(*tmpl), 0,
				 rxq_ctrl->socket);
	if (!tmpl) {
		ERROR("%p: cannot allocate verbs resources",
		       (void *)rxq_ctrl);
		goto error;
	}
	tmpl->rxq_ctrl = rxq_ctrl;
	/* Use the entire RX mempool as the memory region. */
	tmpl->mr = priv_mr_get(priv, rxq_data->mp);
	if (!tmpl->mr) {
		tmpl->mr = priv_mr_new(priv, rxq_data->mp);
		if (!tmpl->mr) {
			ERROR("%p: MR creation failure", (void *)rxq_ctrl);
			goto error;
		}
	}
	if (rxq_ctrl->irq) {
		tmpl->channel = ibv_create_comp_channel(priv->ctx);
		if (!tmpl->channel) {
			ERROR("%p: Comp Channel creation failure",
			      (void *)rxq_ctrl);
			goto error;
		}
	}
	attr.cq.ibv = (struct ibv_cq_init_attr_ex){
		.cqe = cqe_n,
		.channel = tmpl->channel,
		.comp_mask = 0,
	};
	attr.cq.mlx5 = (struct mlx5dv_cq_init_attr){
		.comp_mask = 0,
	};
	if (priv->cqe_comp && !rxq_data->hw_timestamp) {
		attr.cq.mlx5.comp_mask |=
			MLX5DV_CQ_INIT_ATTR_MASK_COMPRESSED_CQE;
		attr.cq.mlx5.cqe_comp_res_format = MLX5DV_CQE_RES_FORMAT_HASH;
		/*
		 * For vectorized Rx, it must not be doubled in order to
		 * make cq_ci and rq_ci aligned.
		 */
		if (rxq_check_vec_support(rxq_data) < 0)
			attr.cq.ibv.cqe *= 2;
	} else if (priv->cqe_comp && rxq_data->hw_timestamp) {
		DEBUG("Rx CQE compression is disabled for HW timestamp");
	}
	tmpl->cq = ibv_cq_ex_to_cq(mlx5dv_create_cq(priv->ctx, &attr.cq.ibv,
						    &attr.cq.mlx5));
	if (tmpl->cq == NULL) {
		ERROR("%p: CQ creation failure", (void *)rxq_ctrl);
		goto error;
	}
	DEBUG("priv->device_attr.max_qp_wr is %d",
	      priv->device_attr.orig_attr.max_qp_wr);
	DEBUG("priv->device_attr.max_sge is %d",
	      priv->device_attr.orig_attr.max_sge);
	attr.wq = (struct ibv_wq_init_attr){
		.wq_context = NULL, /* Could be useful in the future. */
		.wq_type = IBV_WQT_RQ,
		/* Max number of outstanding WRs. */
		.max_wr = (1 << rxq_data->elts_n) >> rxq_data->sges_n,
		/* Max number of scatter/gather elements in a WR. */
		.max_sge = 1 << rxq_data->sges_n,
		.pd = priv->pd,
		.cq = tmpl->cq,
		.comp_mask =
			IBV_WQ_FLAGS_CVLAN_STRIPPING |
			0,
		.create_flags = (rxq_data->vlan_strip ?
				 IBV_WQ_FLAGS_CVLAN_STRIPPING :
				 0),
	};
	/* By default, FCS (CRC) is stripped by hardware. */
	if (rxq_data->crc_present) {
		attr.wq.create_flags |= IBV_WQ_FLAGS_SCATTER_FCS;
		attr.wq.comp_mask |= IBV_WQ_INIT_ATTR_FLAGS;
	}
#ifdef HAVE_IBV_WQ_FLAG_RX_END_PADDING
	if (priv->hw_padding) {
		attr.wq.create_flags |= IBV_WQ_FLAG_RX_END_PADDING;
		attr.wq.comp_mask |= IBV_WQ_INIT_ATTR_FLAGS;
	}
#endif
	tmpl->wq = ibv_create_wq(priv->ctx, &attr.wq);
	if (tmpl->wq == NULL) {
		ERROR("%p: WQ creation failure", (void *)rxq_ctrl);
		goto error;
	}
	/*
	 * Make sure number of WRs*SGEs match expectations since a queue
	 * cannot allocate more than "desc" buffers.
	 */
	if (((int)attr.wq.max_wr !=
	     ((1 << rxq_data->elts_n) >> rxq_data->sges_n)) ||
	    ((int)attr.wq.max_sge != (1 << rxq_data->sges_n))) {
		ERROR("%p: requested %u*%u but got %u*%u WRs*SGEs",
		      (void *)rxq_ctrl,
		      ((1 << rxq_data->elts_n) >> rxq_data->sges_n),
		      (1 << rxq_data->sges_n),
		      attr.wq.max_wr, attr.wq.max_sge);
		goto error;
	}
	/* Change queue state to ready. */
	mod = (struct ibv_wq_attr){
		.attr_mask = IBV_WQ_ATTR_STATE,
		.wq_state = IBV_WQS_RDY,
	};
	ret = ibv_modify_wq(tmpl->wq, &mod);
	if (ret) {
		ERROR("%p: WQ state to IBV_WQS_RDY failed",
		      (void *)rxq_ctrl);
		goto error;
	}
	obj.cq.in = tmpl->cq;
	obj.cq.out = &cq_info;
	obj.rwq.in = tmpl->wq;
	obj.rwq.out = &rwq;
	ret = mlx5dv_init_obj(&obj, MLX5DV_OBJ_CQ | MLX5DV_OBJ_RWQ);
	if (ret != 0)
		goto error;
	if (cq_info.cqe_size != RTE_CACHE_LINE_SIZE) {
		ERROR("Wrong MLX5_CQE_SIZE environment variable value: "
		      "it should be set to %u", RTE_CACHE_LINE_SIZE);
		goto error;
	}
	/* Fill the rings. */
	rxq_data->wqes = (volatile struct mlx5_wqe_data_seg (*)[])
		(uintptr_t)rwq.buf;
	for (i = 0; (i != (unsigned int)(1 << rxq_data->elts_n)); ++i) {
		struct rte_mbuf *buf = (*rxq_data->elts)[i];
		volatile struct mlx5_wqe_data_seg *scat = &(*rxq_data->wqes)[i];

		/* scat->addr must be able to store a pointer. */
		assert(sizeof(scat->addr) >= sizeof(uintptr_t));
		*scat = (struct mlx5_wqe_data_seg){
			.addr = rte_cpu_to_be_64(rte_pktmbuf_mtod(buf,
								  uintptr_t)),
			.byte_count = rte_cpu_to_be_32(DATA_LEN(buf)),
			.lkey = tmpl->mr->lkey,
		};
	}
	rxq_data->rq_db = rwq.dbrec;
	rxq_data->cqe_n = log2above(cq_info.cqe_cnt);
	rxq_data->cq_ci = 0;
	rxq_data->rq_ci = 0;
	rxq_data->rq_pi = 0;
	rxq_data->zip = (struct rxq_zip){
		.ai = 0,
	};
	rxq_data->cq_db = cq_info.dbrec;
	rxq_data->cqes = (volatile struct mlx5_cqe (*)[])(uintptr_t)cq_info.buf;
	rxq_data->cq_uar = cq_info.cq_uar;
	rxq_data->cqn = cq_info.cqn;
	rxq_data->cq_arm_sn = 0;
	/* Update doorbell counter. */
	rxq_data->rq_ci = (1 << rxq_data->elts_n) >> rxq_data->sges_n;
	rte_wmb();
	*rxq_data->rq_db = rte_cpu_to_be_32(rxq_data->rq_ci);
	DEBUG("%p: rxq updated with %p", (void *)rxq_ctrl, (void *)&tmpl);
	rte_atomic32_inc(&tmpl->refcnt);
	DEBUG("%p: Verbs Rx queue %p: refcnt %d", (void *)priv,
	      (void *)tmpl, rte_atomic32_read(&tmpl->refcnt));
	LIST_INSERT_HEAD(&priv->rxqsibv, tmpl, next);
	return tmpl;
error:
	if (tmpl->wq)
		claim_zero(ibv_destroy_wq(tmpl->wq));
	if (tmpl->cq)
		claim_zero(ibv_destroy_cq(tmpl->cq));
	if (tmpl->channel)
		claim_zero(ibv_destroy_comp_channel(tmpl->channel));
	if (tmpl->mr)
		priv_mr_release(priv, tmpl->mr);
	return NULL;
}

/**
 * Get an Rx queue Verbs object.
 *
 * @param priv
 *   Pointer to private structure.
 * @param idx
 *   Queue index in DPDK Rx queue array
 *
 * @return
 *   The Verbs object if it exists.
 */
struct mlx5_rxq_ibv*
mlx5_priv_rxq_ibv_get(struct priv *priv, uint16_t idx)
{
	struct mlx5_rxq_data *rxq_data = (*priv->rxqs)[idx];
	struct mlx5_rxq_ctrl *rxq_ctrl;

	if (idx >= priv->rxqs_n)
		return NULL;
	if (!rxq_data)
		return NULL;
	rxq_ctrl = container_of(rxq_data, struct mlx5_rxq_ctrl, rxq);
	if (rxq_ctrl->ibv) {
		priv_mr_get(priv, rxq_data->mp);
		rte_atomic32_inc(&rxq_ctrl->ibv->refcnt);
		DEBUG("%p: Verbs Rx queue %p: refcnt %d", (void *)priv,
		      (void *)rxq_ctrl->ibv,
		      rte_atomic32_read(&rxq_ctrl->ibv->refcnt));
	}
	return rxq_ctrl->ibv;
}

/**
 * Release an Rx verbs queue object.
 *
 * @param priv
 *   Pointer to private structure.
 * @param rxq_ibv
 *   Verbs Rx queue object.
 *
 * @return
 *   0 on success, errno value on failure.
 */
int
mlx5_priv_rxq_ibv_release(struct priv *priv, struct mlx5_rxq_ibv *rxq_ibv)
{
	int ret;

	assert(rxq_ibv);
	assert(rxq_ibv->wq);
	assert(rxq_ibv->cq);
	assert(rxq_ibv->mr);
	ret = priv_mr_release(priv, rxq_ibv->mr);
	if (!ret)
		rxq_ibv->mr = NULL;
	DEBUG("%p: Verbs Rx queue %p: refcnt %d", (void *)priv,
	      (void *)rxq_ibv, rte_atomic32_read(&rxq_ibv->refcnt));
	if (rte_atomic32_dec_and_test(&rxq_ibv->refcnt)) {
		rxq_free_elts(rxq_ibv->rxq_ctrl);
		claim_zero(ibv_destroy_wq(rxq_ibv->wq));
		claim_zero(ibv_destroy_cq(rxq_ibv->cq));
		if (rxq_ibv->channel)
			claim_zero(ibv_destroy_comp_channel(rxq_ibv->channel));
		LIST_REMOVE(rxq_ibv, next);
		rte_free(rxq_ibv);
		return 0;
	}
	return EBUSY;
}

/**
 * Verify the Verbs Rx queue list is empty
 *
 * @param priv
 *  Pointer to private structure.
 *
 * @return the number of object not released.
 */
int
mlx5_priv_rxq_ibv_verify(struct priv *priv)
{
	int ret = 0;
	struct mlx5_rxq_ibv *rxq_ibv;

	LIST_FOREACH(rxq_ibv, &priv->rxqsibv, next) {
		DEBUG("%p: Verbs Rx queue %p still referenced", (void *)priv,
		      (void *)rxq_ibv);
		++ret;
	}
	return ret;
}

/**
 * Return true if a single reference exists on the object.
 *
 * @param priv
 *   Pointer to private structure.
 * @param rxq_ibv
 *   Verbs Rx queue object.
 */
int
mlx5_priv_rxq_ibv_releasable(struct priv *priv, struct mlx5_rxq_ibv *rxq_ibv)
{
	(void)priv;
	assert(rxq_ibv);
	return (rte_atomic32_read(&rxq_ibv->refcnt) == 1);
}

/**
 * Create a DPDK Rx queue.
 *
 * @param priv
 *   Pointer to private structure.
 * @param idx
 *   TX queue index.
 * @param desc
 *   Number of descriptors to configure in queue.
 * @param socket
 *   NUMA socket on which memory must be allocated.
 *
 * @return
 *   A DPDK queue object on success.
 */
struct mlx5_rxq_ctrl*
mlx5_priv_rxq_new(struct priv *priv, uint16_t idx, uint16_t desc,
		  unsigned int socket, struct rte_mempool *mp)
{
	struct rte_eth_dev *dev = priv->dev;
	struct mlx5_rxq_ctrl *tmpl;
	const uint16_t desc_n =
		desc + priv->rx_vec_en * MLX5_VPMD_DESCS_PER_LOOP;
	unsigned int mb_len = rte_pktmbuf_data_room_size(mp);

	tmpl = rte_calloc_socket("RXQ", 1,
				 sizeof(*tmpl) +
				 desc_n * sizeof(struct rte_mbuf *),
				 0, socket);
	if (!tmpl)
		return NULL;
	tmpl->socket = socket;
	if (priv->dev->data->dev_conf.intr_conf.rxq)
		tmpl->irq = 1;
	/* Enable scattered packets support for this queue if necessary. */
	assert(mb_len >= RTE_PKTMBUF_HEADROOM);
	if (dev->data->dev_conf.rxmode.max_rx_pkt_len <=
	    (mb_len - RTE_PKTMBUF_HEADROOM)) {
		tmpl->rxq.sges_n = 0;
	} else if (dev->data->dev_conf.rxmode.enable_scatter) {
		unsigned int size =
			RTE_PKTMBUF_HEADROOM +
			dev->data->dev_conf.rxmode.max_rx_pkt_len;
		unsigned int sges_n;

		/*
		 * Determine the number of SGEs needed for a full packet
		 * and round it to the next power of two.
		 */
		sges_n = log2above((size / mb_len) + !!(size % mb_len));
		tmpl->rxq.sges_n = sges_n;
		/* Make sure rxq.sges_n did not overflow. */
		size = mb_len * (1 << tmpl->rxq.sges_n);
		size -= RTE_PKTMBUF_HEADROOM;
		if (size < dev->data->dev_conf.rxmode.max_rx_pkt_len) {
			ERROR("%p: too many SGEs (%u) needed to handle"
			      " requested maximum packet size %u",
			      (void *)dev,
			      1 << sges_n,
			      dev->data->dev_conf.rxmode.max_rx_pkt_len);
			goto error;
		}
	} else {
		WARN("%p: the requested maximum Rx packet size (%u) is"
		     " larger than a single mbuf (%u) and scattered"
		     " mode has not been requested",
		     (void *)dev,
		     dev->data->dev_conf.rxmode.max_rx_pkt_len,
		     mb_len - RTE_PKTMBUF_HEADROOM);
	}
	DEBUG("%p: maximum number of segments per packet: %u",
	      (void *)dev, 1 << tmpl->rxq.sges_n);
	if (desc % (1 << tmpl->rxq.sges_n)) {
		ERROR("%p: number of RX queue descriptors (%u) is not a"
		      " multiple of SGEs per packet (%u)",
		      (void *)dev,
		      desc,
		      1 << tmpl->rxq.sges_n);
		goto error;
	}
	/* Toggle RX checksum offload if hardware supports it. */
	if (priv->hw_csum)
		tmpl->rxq.csum = !!dev->data->dev_conf.rxmode.hw_ip_checksum;
	if (priv->hw_csum_l2tun)
		tmpl->rxq.csum_l2tun =
			!!dev->data->dev_conf.rxmode.hw_ip_checksum;
	tmpl->rxq.hw_timestamp =
			!!dev->data->dev_conf.rxmode.hw_timestamp;
	/* Configure VLAN stripping. */
	tmpl->rxq.vlan_strip = (priv->hw_vlan_strip &&
			       !!dev->data->dev_conf.rxmode.hw_vlan_strip);
	/* By default, FCS (CRC) is stripped by hardware. */
	if (dev->data->dev_conf.rxmode.hw_strip_crc) {
		tmpl->rxq.crc_present = 0;
	} else if (priv->hw_fcs_strip) {
		tmpl->rxq.crc_present = 1;
	} else {
		WARN("%p: CRC stripping has been disabled but will still"
		     " be performed by hardware, make sure MLNX_OFED and"
		     " firmware are up to date",
		     (void *)dev);
		tmpl->rxq.crc_present = 0;
	}
	DEBUG("%p: CRC stripping is %s, %u bytes will be subtracted from"
	      " incoming frames to hide it",
	      (void *)dev,
	      tmpl->rxq.crc_present ? "disabled" : "enabled",
	      tmpl->rxq.crc_present << 2);
	/* Save port ID. */
	tmpl->rxq.rss_hash = priv->rxqs_n > 1;
	tmpl->rxq.port_id = dev->data->port_id;
	tmpl->priv = priv;
	tmpl->rxq.mp = mp;
	tmpl->rxq.stats.idx = idx;
	tmpl->rxq.elts_n = log2above(desc);
	tmpl->rxq.elts =
		(struct rte_mbuf *(*)[1 << tmpl->rxq.elts_n])(tmpl + 1);
	rte_atomic32_inc(&tmpl->refcnt);
	DEBUG("%p: Rx queue %p: refcnt %d", (void *)priv,
	      (void *)tmpl, rte_atomic32_read(&tmpl->refcnt));
	LIST_INSERT_HEAD(&priv->rxqsctrl, tmpl, next);
	return tmpl;
error:
	rte_free(tmpl);
	return NULL;
}

/**
 * Get a Rx queue.
 *
 * @param priv
 *   Pointer to private structure.
 * @param idx
 *   TX queue index.
 *
 * @return
 *   A pointer to the queue if it exists.
 */
struct mlx5_rxq_ctrl*
mlx5_priv_rxq_get(struct priv *priv, uint16_t idx)
{
	struct mlx5_rxq_ctrl *rxq_ctrl = NULL;

	if ((*priv->rxqs)[idx]) {
		rxq_ctrl = container_of((*priv->rxqs)[idx],
					struct mlx5_rxq_ctrl,
					rxq);

		mlx5_priv_rxq_ibv_get(priv, idx);
		rte_atomic32_inc(&rxq_ctrl->refcnt);
		DEBUG("%p: Rx queue %p: refcnt %d", (void *)priv,
		      (void *)rxq_ctrl, rte_atomic32_read(&rxq_ctrl->refcnt));
	}
	return rxq_ctrl;
}

/**
 * Release a Rx queue.
 *
 * @param priv
 *   Pointer to private structure.
 * @param idx
 *   TX queue index.
 *
 * @return
 *   0 on success, errno value on failure.
 */
int
mlx5_priv_rxq_release(struct priv *priv, uint16_t idx)
{
	struct mlx5_rxq_ctrl *rxq_ctrl;

	if (!(*priv->rxqs)[idx])
		return 0;
	rxq_ctrl = container_of((*priv->rxqs)[idx], struct mlx5_rxq_ctrl, rxq);
	assert(rxq_ctrl->priv);
	if (rxq_ctrl->ibv) {
		int ret;

		ret = mlx5_priv_rxq_ibv_release(rxq_ctrl->priv, rxq_ctrl->ibv);
		if (!ret)
			rxq_ctrl->ibv = NULL;
	}
	DEBUG("%p: Rx queue %p: refcnt %d", (void *)priv,
	      (void *)rxq_ctrl, rte_atomic32_read(&rxq_ctrl->refcnt));
	if (rte_atomic32_dec_and_test(&rxq_ctrl->refcnt)) {
		LIST_REMOVE(rxq_ctrl, next);
		rte_free(rxq_ctrl);
		(*priv->rxqs)[idx] = NULL;
		return 0;
	}
	return EBUSY;
}

/**
 * Verify if the queue can be released.
 *
 * @param priv
 *   Pointer to private structure.
 * @param idx
 *   TX queue index.
 *
 * @return
 *   1 if the queue can be released.
 */
int
mlx5_priv_rxq_releasable(struct priv *priv, uint16_t idx)
{
	struct mlx5_rxq_ctrl *rxq_ctrl;

	if (!(*priv->rxqs)[idx])
		return -1;
	rxq_ctrl = container_of((*priv->rxqs)[idx], struct mlx5_rxq_ctrl, rxq);
	return (rte_atomic32_read(&rxq_ctrl->refcnt) == 1);
}

/**
 * Verify the Rx Queue list is empty
 *
 * @param priv
 *  Pointer to private structure.
 *
 * @return the number of object not released.
 */
int
mlx5_priv_rxq_verify(struct priv *priv)
{
	struct mlx5_rxq_ctrl *rxq_ctrl;
	int ret = 0;

	LIST_FOREACH(rxq_ctrl, &priv->rxqsctrl, next) {
		DEBUG("%p: Rx Queue %p still referenced", (void *)priv,
		      (void *)rxq_ctrl);
		++ret;
	}
	return ret;
}

/**
 * Create an indirection table.
 *
 * @param priv
 *   Pointer to private structure.
 * @param queues
 *   Queues entering in the indirection table.
 * @param queues_n
 *   Number of queues in the array.
 *
 * @return
 *   A new indirection table.
 */
struct mlx5_ind_table_ibv*
mlx5_priv_ind_table_ibv_new(struct priv *priv, uint16_t queues[],
			    uint16_t queues_n)
{
	struct mlx5_ind_table_ibv *ind_tbl;
	const unsigned int wq_n = rte_is_power_of_2(queues_n) ?
		log2above(queues_n) :
		log2above(priv->ind_table_max_size);
	struct ibv_wq *wq[1 << wq_n];
	unsigned int i;
	unsigned int j;

	ind_tbl = rte_calloc(__func__, 1, sizeof(*ind_tbl) +
			     queues_n * sizeof(uint16_t), 0);
	if (!ind_tbl)
		return NULL;
	for (i = 0; i != queues_n; ++i) {
		struct mlx5_rxq_ctrl *rxq =
			mlx5_priv_rxq_get(priv, queues[i]);

		if (!rxq)
			goto error;
		wq[i] = rxq->ibv->wq;
		ind_tbl->queues[i] = queues[i];
	}
	ind_tbl->queues_n = queues_n;
	/* Finalise indirection table. */
	for (j = 0; i != (unsigned int)(1 << wq_n); ++i, ++j)
		wq[i] = wq[j];
	ind_tbl->ind_table = ibv_create_rwq_ind_table(
		priv->ctx,
		&(struct ibv_rwq_ind_table_init_attr){
			.log_ind_tbl_size = wq_n,
			.ind_tbl = wq,
			.comp_mask = 0,
		});
	if (!ind_tbl->ind_table)
		goto error;
	rte_atomic32_inc(&ind_tbl->refcnt);
	LIST_INSERT_HEAD(&priv->ind_tbls, ind_tbl, next);
	DEBUG("%p: Indirection table %p: refcnt %d", (void *)priv,
	      (void *)ind_tbl, rte_atomic32_read(&ind_tbl->refcnt));
	return ind_tbl;
error:
	rte_free(ind_tbl);
	DEBUG("%p cannot create indirection table", (void *)priv);
	return NULL;
}

/**
 * Get an indirection table.
 *
 * @param priv
 *   Pointer to private structure.
 * @param queues
 *   Queues entering in the indirection table.
 * @param queues_n
 *   Number of queues in the array.
 *
 * @return
 *   An indirection table if found.
 */
struct mlx5_ind_table_ibv*
mlx5_priv_ind_table_ibv_get(struct priv *priv, uint16_t queues[],
			    uint16_t queues_n)
{
	struct mlx5_ind_table_ibv *ind_tbl;

	LIST_FOREACH(ind_tbl, &priv->ind_tbls, next) {
		if ((ind_tbl->queues_n == queues_n) &&
		    (memcmp(ind_tbl->queues, queues,
			    ind_tbl->queues_n * sizeof(ind_tbl->queues[0]))
		     == 0))
			break;
	}
	if (ind_tbl) {
		unsigned int i;

		rte_atomic32_inc(&ind_tbl->refcnt);
		DEBUG("%p: Indirection table %p: refcnt %d", (void *)priv,
		      (void *)ind_tbl, rte_atomic32_read(&ind_tbl->refcnt));
		for (i = 0; i != ind_tbl->queues_n; ++i)
			mlx5_priv_rxq_get(priv, ind_tbl->queues[i]);
	}
	return ind_tbl;
}

/**
 * Release an indirection table.
 *
 * @param priv
 *   Pointer to private structure.
 * @param ind_table
 *   Indirection table to release.
 *
 * @return
 *   0 on success, errno value on failure.
 */
int
mlx5_priv_ind_table_ibv_release(struct priv *priv,
				struct mlx5_ind_table_ibv *ind_tbl)
{
	unsigned int i;

	DEBUG("%p: Indirection table %p: refcnt %d", (void *)priv,
	      (void *)ind_tbl, rte_atomic32_read(&ind_tbl->refcnt));
	if (rte_atomic32_dec_and_test(&ind_tbl->refcnt))
		claim_zero(ibv_destroy_rwq_ind_table(ind_tbl->ind_table));
	for (i = 0; i != ind_tbl->queues_n; ++i)
		claim_nonzero(mlx5_priv_rxq_release(priv, ind_tbl->queues[i]));
	if (!rte_atomic32_read(&ind_tbl->refcnt)) {
		LIST_REMOVE(ind_tbl, next);
		rte_free(ind_tbl);
		return 0;
	}
	return EBUSY;
}

/**
 * Verify the Rx Queue list is empty
 *
 * @param priv
 *  Pointer to private structure.
 *
 * @return the number of object not released.
 */
int
mlx5_priv_ind_table_ibv_verify(struct priv *priv)
{
	struct mlx5_ind_table_ibv *ind_tbl;
	int ret = 0;

	LIST_FOREACH(ind_tbl, &priv->ind_tbls, next) {
		DEBUG("%p: Verbs indirection table %p still referenced",
		      (void *)priv, (void *)ind_tbl);
		++ret;
	}
	return ret;
}

/**
 * Create an Rx Hash queue.
 *
 * @param priv
 *   Pointer to private structure.
 * @param rss_key
 *   RSS key for the Rx hash queue.
 * @param rss_key_len
 *   RSS key length.
 * @param hash_fields
 *   Verbs protocol hash field to make the RSS on.
 * @param queues
 *   Queues entering in hash queue. In case of empty hash_fields only the
 *   first queue index will be taken for the indirection table.
 * @param queues_n
 *   Number of queues.
 *
 * @return
 *   An hash Rx queue on success.
 */
struct mlx5_hrxq*
mlx5_priv_hrxq_new(struct priv *priv, uint8_t *rss_key, uint8_t rss_key_len,
		   uint64_t hash_fields, uint16_t queues[], uint16_t queues_n)
{
	struct mlx5_hrxq *hrxq;
	struct mlx5_ind_table_ibv *ind_tbl;
	struct ibv_qp *qp;

	queues_n = hash_fields ? queues_n : 1;
	ind_tbl = mlx5_priv_ind_table_ibv_get(priv, queues, queues_n);
	if (!ind_tbl)
		ind_tbl = mlx5_priv_ind_table_ibv_new(priv, queues, queues_n);
	if (!ind_tbl)
		return NULL;
	qp = ibv_create_qp_ex(
		priv->ctx,
		&(struct ibv_qp_init_attr_ex){
			.qp_type = IBV_QPT_RAW_PACKET,
			.comp_mask =
				IBV_QP_INIT_ATTR_PD |
				IBV_QP_INIT_ATTR_IND_TABLE |
				IBV_QP_INIT_ATTR_RX_HASH,
			.rx_hash_conf = (struct ibv_rx_hash_conf){
				.rx_hash_function = IBV_RX_HASH_FUNC_TOEPLITZ,
				.rx_hash_key_len = rss_key_len,
				.rx_hash_key = rss_key,
				.rx_hash_fields_mask = hash_fields,
			},
			.rwq_ind_tbl = ind_tbl->ind_table,
			.pd = priv->pd,
		});
	if (!qp)
		goto error;
	hrxq = rte_calloc(__func__, 1, sizeof(*hrxq) + rss_key_len, 0);
	if (!hrxq)
		goto error;
	hrxq->ind_table = ind_tbl;
	hrxq->qp = qp;
	hrxq->rss_key_len = rss_key_len;
	hrxq->hash_fields = hash_fields;
	memcpy(hrxq->rss_key, rss_key, rss_key_len);
	rte_atomic32_inc(&hrxq->refcnt);
	LIST_INSERT_HEAD(&priv->hrxqs, hrxq, next);
	DEBUG("%p: Hash Rx queue %p: refcnt %d", (void *)priv,
	      (void *)hrxq, rte_atomic32_read(&hrxq->refcnt));
	return hrxq;
error:
	mlx5_priv_ind_table_ibv_release(priv, ind_tbl);
	if (qp)
		claim_zero(ibv_destroy_qp(qp));
	return NULL;
}

/**
 * Get an Rx Hash queue.
 *
 * @param priv
 *   Pointer to private structure.
 * @param rss_conf
 *   RSS configuration for the Rx hash queue.
 * @param queues
 *   Queues entering in hash queue. In case of empty hash_fields only the
 *   first queue index will be taken for the indirection table.
 * @param queues_n
 *   Number of queues.
 *
 * @return
 *   An hash Rx queue on success.
 */
struct mlx5_hrxq*
mlx5_priv_hrxq_get(struct priv *priv, uint8_t *rss_key, uint8_t rss_key_len,
		   uint64_t hash_fields, uint16_t queues[], uint16_t queues_n)
{
	struct mlx5_hrxq *hrxq;

	queues_n = hash_fields ? queues_n : 1;
	LIST_FOREACH(hrxq, &priv->hrxqs, next) {
		struct mlx5_ind_table_ibv *ind_tbl;

		if (hrxq->rss_key_len != rss_key_len)
			continue;
		if (memcmp(hrxq->rss_key, rss_key, rss_key_len))
			continue;
		if (hrxq->hash_fields != hash_fields)
			continue;
		ind_tbl = mlx5_priv_ind_table_ibv_get(priv, queues, queues_n);
		if (!ind_tbl)
			continue;
		if (ind_tbl != hrxq->ind_table) {
			mlx5_priv_ind_table_ibv_release(priv, ind_tbl);
			continue;
		}
		rte_atomic32_inc(&hrxq->refcnt);
		DEBUG("%p: Hash Rx queue %p: refcnt %d", (void *)priv,
		      (void *)hrxq, rte_atomic32_read(&hrxq->refcnt));
		return hrxq;
	}
	return NULL;
}

/**
 * Release the hash Rx queue.
 *
 * @param priv
 *   Pointer to private structure.
 * @param hrxq
 *   Pointer to Hash Rx queue to release.
 *
 * @return
 *   0 on success, errno value on failure.
 */
int
mlx5_priv_hrxq_release(struct priv *priv, struct mlx5_hrxq *hrxq)
{
	DEBUG("%p: Hash Rx queue %p: refcnt %d", (void *)priv,
	      (void *)hrxq, rte_atomic32_read(&hrxq->refcnt));
	if (rte_atomic32_dec_and_test(&hrxq->refcnt)) {
		claim_zero(ibv_destroy_qp(hrxq->qp));
		mlx5_priv_ind_table_ibv_release(priv, hrxq->ind_table);
		LIST_REMOVE(hrxq, next);
		rte_free(hrxq);
		return 0;
	}
	claim_nonzero(mlx5_priv_ind_table_ibv_release(priv, hrxq->ind_table));
	return EBUSY;
}

/**
 * Verify the Rx Queue list is empty
 *
 * @param priv
 *  Pointer to private structure.
 *
 * @return the number of object not released.
 */
int
mlx5_priv_hrxq_ibv_verify(struct priv *priv)
{
	struct mlx5_hrxq *hrxq;
	int ret = 0;

	LIST_FOREACH(hrxq, &priv->hrxqs, next) {
		DEBUG("%p: Verbs Hash Rx queue %p still referenced",
		      (void *)priv, (void *)hrxq);
		++ret;
	}
	return ret;
}