Loading...
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 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 | /* QLogic qede NIC Driver
* Copyright (c) 2015-2017 QLogic Corporation
*
* This software is available to you under a choice of one of two
* licenses. You may choose to be licensed under the terms of the GNU
* General Public License (GPL) Version 2, available from the file
* COPYING in the main directory of this source tree, or the
* OpenIB.org BSD license below:
*
* 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.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/skbuff.h>
#include <linux/bpf_trace.h>
#include <net/udp_tunnel.h>
#include <linux/ip.h>
#include <net/ipv6.h>
#include <net/tcp.h>
#include <linux/if_ether.h>
#include <linux/if_vlan.h>
#include <net/ip6_checksum.h>
#include "qede_ptp.h"
#include <linux/qed/qed_if.h>
#include "qede.h"
/*********************************
* Content also used by slowpath *
*********************************/
int qede_alloc_rx_buffer(struct qede_rx_queue *rxq, bool allow_lazy)
{
struct sw_rx_data *sw_rx_data;
struct eth_rx_bd *rx_bd;
dma_addr_t mapping;
struct page *data;
/* In case lazy-allocation is allowed, postpone allocation until the
* end of the NAPI run. We'd still need to make sure the Rx ring has
* sufficient buffers to guarantee an additional Rx interrupt.
*/
if (allow_lazy && likely(rxq->filled_buffers > 12)) {
rxq->filled_buffers--;
return 0;
}
data = alloc_pages(GFP_ATOMIC, 0);
if (unlikely(!data))
return -ENOMEM;
/* Map the entire page as it would be used
* for multiple RX buffer segment size mapping.
*/
mapping = dma_map_page(rxq->dev, data, 0,
PAGE_SIZE, rxq->data_direction);
if (unlikely(dma_mapping_error(rxq->dev, mapping))) {
__free_page(data);
return -ENOMEM;
}
sw_rx_data = &rxq->sw_rx_ring[rxq->sw_rx_prod & NUM_RX_BDS_MAX];
sw_rx_data->page_offset = 0;
sw_rx_data->data = data;
sw_rx_data->mapping = mapping;
/* Advance PROD and get BD pointer */
rx_bd = (struct eth_rx_bd *)qed_chain_produce(&rxq->rx_bd_ring);
WARN_ON(!rx_bd);
rx_bd->addr.hi = cpu_to_le32(upper_32_bits(mapping));
rx_bd->addr.lo = cpu_to_le32(lower_32_bits(mapping) +
rxq->rx_headroom);
rxq->sw_rx_prod++;
rxq->filled_buffers++;
return 0;
}
/* Unmap the data and free skb */
int qede_free_tx_pkt(struct qede_dev *edev, struct qede_tx_queue *txq, int *len)
{
u16 idx = txq->sw_tx_cons & NUM_TX_BDS_MAX;
struct sk_buff *skb = txq->sw_tx_ring.skbs[idx].skb;
struct eth_tx_1st_bd *first_bd;
struct eth_tx_bd *tx_data_bd;
int bds_consumed = 0;
int nbds;
bool data_split = txq->sw_tx_ring.skbs[idx].flags & QEDE_TSO_SPLIT_BD;
int i, split_bd_len = 0;
if (unlikely(!skb)) {
DP_ERR(edev,
"skb is null for txq idx=%d txq->sw_tx_cons=%d txq->sw_tx_prod=%d\n",
idx, txq->sw_tx_cons, txq->sw_tx_prod);
return -1;
}
*len = skb->len;
first_bd = (struct eth_tx_1st_bd *)qed_chain_consume(&txq->tx_pbl);
bds_consumed++;
nbds = first_bd->data.nbds;
if (data_split) {
struct eth_tx_bd *split = (struct eth_tx_bd *)
qed_chain_consume(&txq->tx_pbl);
split_bd_len = BD_UNMAP_LEN(split);
bds_consumed++;
}
dma_unmap_single(&edev->pdev->dev, BD_UNMAP_ADDR(first_bd),
BD_UNMAP_LEN(first_bd) + split_bd_len, DMA_TO_DEVICE);
/* Unmap the data of the skb frags */
for (i = 0; i < skb_shinfo(skb)->nr_frags; i++, bds_consumed++) {
tx_data_bd = (struct eth_tx_bd *)
qed_chain_consume(&txq->tx_pbl);
dma_unmap_page(&edev->pdev->dev, BD_UNMAP_ADDR(tx_data_bd),
BD_UNMAP_LEN(tx_data_bd), DMA_TO_DEVICE);
}
while (bds_consumed++ < nbds)
qed_chain_consume(&txq->tx_pbl);
/* Free skb */
dev_kfree_skb_any(skb);
txq->sw_tx_ring.skbs[idx].skb = NULL;
txq->sw_tx_ring.skbs[idx].flags = 0;
return 0;
}
/* Unmap the data and free skb when mapping failed during start_xmit */
static void qede_free_failed_tx_pkt(struct qede_tx_queue *txq,
struct eth_tx_1st_bd *first_bd,
int nbd, bool data_split)
{
u16 idx = txq->sw_tx_prod & NUM_TX_BDS_MAX;
struct sk_buff *skb = txq->sw_tx_ring.skbs[idx].skb;
struct eth_tx_bd *tx_data_bd;
int i, split_bd_len = 0;
/* Return prod to its position before this skb was handled */
qed_chain_set_prod(&txq->tx_pbl,
le16_to_cpu(txq->tx_db.data.bd_prod), first_bd);
first_bd = (struct eth_tx_1st_bd *)qed_chain_produce(&txq->tx_pbl);
if (data_split) {
struct eth_tx_bd *split = (struct eth_tx_bd *)
qed_chain_produce(&txq->tx_pbl);
split_bd_len = BD_UNMAP_LEN(split);
nbd--;
}
dma_unmap_single(txq->dev, BD_UNMAP_ADDR(first_bd),
BD_UNMAP_LEN(first_bd) + split_bd_len, DMA_TO_DEVICE);
/* Unmap the data of the skb frags */
for (i = 0; i < nbd; i++) {
tx_data_bd = (struct eth_tx_bd *)
qed_chain_produce(&txq->tx_pbl);
if (tx_data_bd->nbytes)
dma_unmap_page(txq->dev,
BD_UNMAP_ADDR(tx_data_bd),
BD_UNMAP_LEN(tx_data_bd), DMA_TO_DEVICE);
}
/* Return again prod to its position before this skb was handled */
qed_chain_set_prod(&txq->tx_pbl,
le16_to_cpu(txq->tx_db.data.bd_prod), first_bd);
/* Free skb */
dev_kfree_skb_any(skb);
txq->sw_tx_ring.skbs[idx].skb = NULL;
txq->sw_tx_ring.skbs[idx].flags = 0;
}
static u32 qede_xmit_type(struct sk_buff *skb, int *ipv6_ext)
{
u32 rc = XMIT_L4_CSUM;
__be16 l3_proto;
if (skb->ip_summed != CHECKSUM_PARTIAL)
return XMIT_PLAIN;
l3_proto = vlan_get_protocol(skb);
if (l3_proto == htons(ETH_P_IPV6) &&
(ipv6_hdr(skb)->nexthdr == NEXTHDR_IPV6))
*ipv6_ext = 1;
if (skb->encapsulation) {
rc |= XMIT_ENC;
if (skb_is_gso(skb)) {
unsigned short gso_type = skb_shinfo(skb)->gso_type;
if ((gso_type & SKB_GSO_UDP_TUNNEL_CSUM) ||
(gso_type & SKB_GSO_GRE_CSUM))
rc |= XMIT_ENC_GSO_L4_CSUM;
rc |= XMIT_LSO;
return rc;
}
}
if (skb_is_gso(skb))
rc |= XMIT_LSO;
return rc;
}
static void qede_set_params_for_ipv6_ext(struct sk_buff *skb,
struct eth_tx_2nd_bd *second_bd,
struct eth_tx_3rd_bd *third_bd)
{
u8 l4_proto;
u16 bd2_bits1 = 0, bd2_bits2 = 0;
bd2_bits1 |= (1 << ETH_TX_DATA_2ND_BD_IPV6_EXT_SHIFT);
bd2_bits2 |= ((((u8 *)skb_transport_header(skb) - skb->data) >> 1) &
ETH_TX_DATA_2ND_BD_L4_HDR_START_OFFSET_W_MASK)
<< ETH_TX_DATA_2ND_BD_L4_HDR_START_OFFSET_W_SHIFT;
bd2_bits1 |= (ETH_L4_PSEUDO_CSUM_CORRECT_LENGTH <<
ETH_TX_DATA_2ND_BD_L4_PSEUDO_CSUM_MODE_SHIFT);
if (vlan_get_protocol(skb) == htons(ETH_P_IPV6))
l4_proto = ipv6_hdr(skb)->nexthdr;
else
l4_proto = ip_hdr(skb)->protocol;
if (l4_proto == IPPROTO_UDP)
bd2_bits1 |= 1 << ETH_TX_DATA_2ND_BD_L4_UDP_SHIFT;
if (third_bd)
third_bd->data.bitfields |=
cpu_to_le16(((tcp_hdrlen(skb) / 4) &
ETH_TX_DATA_3RD_BD_TCP_HDR_LEN_DW_MASK) <<
ETH_TX_DATA_3RD_BD_TCP_HDR_LEN_DW_SHIFT);
second_bd->data.bitfields1 = cpu_to_le16(bd2_bits1);
second_bd->data.bitfields2 = cpu_to_le16(bd2_bits2);
}
static int map_frag_to_bd(struct qede_tx_queue *txq,
skb_frag_t *frag, struct eth_tx_bd *bd)
{
dma_addr_t mapping;
/* Map skb non-linear frag data for DMA */
mapping = skb_frag_dma_map(txq->dev, frag, 0,
skb_frag_size(frag), DMA_TO_DEVICE);
if (unlikely(dma_mapping_error(txq->dev, mapping)))
return -ENOMEM;
/* Setup the data pointer of the frag data */
BD_SET_UNMAP_ADDR_LEN(bd, mapping, skb_frag_size(frag));
return 0;
}
static u16 qede_get_skb_hlen(struct sk_buff *skb, bool is_encap_pkt)
{
if (is_encap_pkt)
return (skb_inner_transport_header(skb) +
inner_tcp_hdrlen(skb) - skb->data);
else
return (skb_transport_header(skb) +
tcp_hdrlen(skb) - skb->data);
}
/* +2 for 1st BD for headers and 2nd BD for headlen (if required) */
#if ((MAX_SKB_FRAGS + 2) > ETH_TX_MAX_BDS_PER_NON_LSO_PACKET)
static bool qede_pkt_req_lin(struct sk_buff *skb, u8 xmit_type)
{
int allowed_frags = ETH_TX_MAX_BDS_PER_NON_LSO_PACKET - 1;
if (xmit_type & XMIT_LSO) {
int hlen;
hlen = qede_get_skb_hlen(skb, xmit_type & XMIT_ENC);
/* linear payload would require its own BD */
if (skb_headlen(skb) > hlen)
allowed_frags--;
}
return (skb_shinfo(skb)->nr_frags > allowed_frags);
}
#endif
static inline void qede_update_tx_producer(struct qede_tx_queue *txq)
{
/* wmb makes sure that the BDs data is updated before updating the
* producer, otherwise FW may read old data from the BDs.
*/
wmb();
barrier();
writel(txq->tx_db.raw, txq->doorbell_addr);
/* mmiowb is needed to synchronize doorbell writes from more than one
* processor. It guarantees that the write arrives to the device before
* the queue lock is released and another start_xmit is called (possibly
* on another CPU). Without this barrier, the next doorbell can bypass
* this doorbell. This is applicable to IA64/Altix systems.
*/
mmiowb();
}
static int qede_xdp_xmit(struct qede_dev *edev, struct qede_fastpath *fp,
struct sw_rx_data *metadata, u16 padding, u16 length)
{
struct qede_tx_queue *txq = fp->xdp_tx;
u16 idx = txq->sw_tx_prod & NUM_TX_BDS_MAX;
struct eth_tx_1st_bd *first_bd;
if (!qed_chain_get_elem_left(&txq->tx_pbl)) {
txq->stopped_cnt++;
return -ENOMEM;
}
first_bd = (struct eth_tx_1st_bd *)qed_chain_produce(&txq->tx_pbl);
memset(first_bd, 0, sizeof(*first_bd));
first_bd->data.bd_flags.bitfields =
BIT(ETH_TX_1ST_BD_FLAGS_START_BD_SHIFT);
first_bd->data.bitfields |=
(length & ETH_TX_DATA_1ST_BD_PKT_LEN_MASK) <<
ETH_TX_DATA_1ST_BD_PKT_LEN_SHIFT;
first_bd->data.nbds = 1;
/* We can safely ignore the offset, as it's 0 for XDP */
BD_SET_UNMAP_ADDR_LEN(first_bd, metadata->mapping + padding, length);
/* Synchronize the buffer back to device, as program [probably]
* has changed it.
*/
dma_sync_single_for_device(&edev->pdev->dev,
metadata->mapping + padding,
length, PCI_DMA_TODEVICE);
txq->sw_tx_ring.xdp[idx].page = metadata->data;
txq->sw_tx_ring.xdp[idx].mapping = metadata->mapping;
txq->sw_tx_prod++;
/* Mark the fastpath for future XDP doorbell */
fp->xdp_xmit = 1;
return 0;
}
int qede_txq_has_work(struct qede_tx_queue *txq)
{
u16 hw_bd_cons;
/* Tell compiler that consumer and producer can change */
barrier();
hw_bd_cons = le16_to_cpu(*txq->hw_cons_ptr);
if (qed_chain_get_cons_idx(&txq->tx_pbl) == hw_bd_cons + 1)
return 0;
return hw_bd_cons != qed_chain_get_cons_idx(&txq->tx_pbl);
}
static void qede_xdp_tx_int(struct qede_dev *edev, struct qede_tx_queue *txq)
{
u16 hw_bd_cons, idx;
hw_bd_cons = le16_to_cpu(*txq->hw_cons_ptr);
barrier();
while (hw_bd_cons != qed_chain_get_cons_idx(&txq->tx_pbl)) {
qed_chain_consume(&txq->tx_pbl);
idx = txq->sw_tx_cons & NUM_TX_BDS_MAX;
dma_unmap_page(&edev->pdev->dev,
txq->sw_tx_ring.xdp[idx].mapping,
PAGE_SIZE, DMA_BIDIRECTIONAL);
__free_page(txq->sw_tx_ring.xdp[idx].page);
txq->sw_tx_cons++;
txq->xmit_pkts++;
}
}
static int qede_tx_int(struct qede_dev *edev, struct qede_tx_queue *txq)
{
struct netdev_queue *netdev_txq;
u16 hw_bd_cons;
unsigned int pkts_compl = 0, bytes_compl = 0;
int rc;
netdev_txq = netdev_get_tx_queue(edev->ndev, txq->index);
hw_bd_cons = le16_to_cpu(*txq->hw_cons_ptr);
barrier();
while (hw_bd_cons != qed_chain_get_cons_idx(&txq->tx_pbl)) {
int len = 0;
rc = qede_free_tx_pkt(edev, txq, &len);
if (rc) {
DP_NOTICE(edev, "hw_bd_cons = %d, chain_cons=%d\n",
hw_bd_cons,
qed_chain_get_cons_idx(&txq->tx_pbl));
break;
}
bytes_compl += len;
pkts_compl++;
txq->sw_tx_cons++;
txq->xmit_pkts++;
}
netdev_tx_completed_queue(netdev_txq, pkts_compl, bytes_compl);
/* Need to make the tx_bd_cons update visible to start_xmit()
* before checking for netif_tx_queue_stopped(). Without the
* memory barrier, there is a small possibility that
* start_xmit() will miss it and cause the queue to be stopped
* forever.
* On the other hand we need an rmb() here to ensure the proper
* ordering of bit testing in the following
* netif_tx_queue_stopped(txq) call.
*/
smp_mb();
if (unlikely(netif_tx_queue_stopped(netdev_txq))) {
/* Taking tx_lock is needed to prevent reenabling the queue
* while it's empty. This could have happen if rx_action() gets
* suspended in qede_tx_int() after the condition before
* netif_tx_wake_queue(), while tx_action (qede_start_xmit()):
*
* stops the queue->sees fresh tx_bd_cons->releases the queue->
* sends some packets consuming the whole queue again->
* stops the queue
*/
__netif_tx_lock(netdev_txq, smp_processor_id());
if ((netif_tx_queue_stopped(netdev_txq)) &&
(edev->state == QEDE_STATE_OPEN) &&
(qed_chain_get_elem_left(&txq->tx_pbl)
>= (MAX_SKB_FRAGS + 1))) {
netif_tx_wake_queue(netdev_txq);
DP_VERBOSE(edev, NETIF_MSG_TX_DONE,
"Wake queue was called\n");
}
__netif_tx_unlock(netdev_txq);
}
return 0;
}
bool qede_has_rx_work(struct qede_rx_queue *rxq)
{
u16 hw_comp_cons, sw_comp_cons;
/* Tell compiler that status block fields can change */
barrier();
hw_comp_cons = le16_to_cpu(*rxq->hw_cons_ptr);
sw_comp_cons = qed_chain_get_cons_idx(&rxq->rx_comp_ring);
return hw_comp_cons != sw_comp_cons;
}
static inline void qede_rx_bd_ring_consume(struct qede_rx_queue *rxq)
{
qed_chain_consume(&rxq->rx_bd_ring);
rxq->sw_rx_cons++;
}
/* This function reuses the buffer(from an offset) from
* consumer index to producer index in the bd ring
*/
static inline void qede_reuse_page(struct qede_rx_queue *rxq,
struct sw_rx_data *curr_cons)
{
struct eth_rx_bd *rx_bd_prod = qed_chain_produce(&rxq->rx_bd_ring);
struct sw_rx_data *curr_prod;
dma_addr_t new_mapping;
curr_prod = &rxq->sw_rx_ring[rxq->sw_rx_prod & NUM_RX_BDS_MAX];
*curr_prod = *curr_cons;
new_mapping = curr_prod->mapping + curr_prod->page_offset;
rx_bd_prod->addr.hi = cpu_to_le32(upper_32_bits(new_mapping));
rx_bd_prod->addr.lo = cpu_to_le32(lower_32_bits(new_mapping) +
rxq->rx_headroom);
rxq->sw_rx_prod++;
curr_cons->data = NULL;
}
/* In case of allocation failures reuse buffers
* from consumer index to produce buffers for firmware
*/
void qede_recycle_rx_bd_ring(struct qede_rx_queue *rxq, u8 count)
{
struct sw_rx_data *curr_cons;
for (; count > 0; count--) {
curr_cons = &rxq->sw_rx_ring[rxq->sw_rx_cons & NUM_RX_BDS_MAX];
qede_reuse_page(rxq, curr_cons);
qede_rx_bd_ring_consume(rxq);
}
}
static inline int qede_realloc_rx_buffer(struct qede_rx_queue *rxq,
struct sw_rx_data *curr_cons)
{
/* Move to the next segment in the page */
curr_cons->page_offset += rxq->rx_buf_seg_size;
if (curr_cons->page_offset == PAGE_SIZE) {
if (unlikely(qede_alloc_rx_buffer(rxq, true))) {
/* Since we failed to allocate new buffer
* current buffer can be used again.
*/
curr_cons->page_offset -= rxq->rx_buf_seg_size;
return -ENOMEM;
}
dma_unmap_page(rxq->dev, curr_cons->mapping,
PAGE_SIZE, rxq->data_direction);
} else {
/* Increment refcount of the page as we don't want
* network stack to take the ownership of the page
* which can be recycled multiple times by the driver.
*/
page_ref_inc(curr_cons->data);
qede_reuse_page(rxq, curr_cons);
}
return 0;
}
void qede_update_rx_prod(struct qede_dev *edev, struct qede_rx_queue *rxq)
{
u16 bd_prod = qed_chain_get_prod_idx(&rxq->rx_bd_ring);
u16 cqe_prod = qed_chain_get_prod_idx(&rxq->rx_comp_ring);
struct eth_rx_prod_data rx_prods = {0};
/* Update producers */
rx_prods.bd_prod = cpu_to_le16(bd_prod);
rx_prods.cqe_prod = cpu_to_le16(cqe_prod);
/* Make sure that the BD and SGE data is updated before updating the
* producers since FW might read the BD/SGE right after the producer
* is updated.
*/
wmb();
internal_ram_wr(rxq->hw_rxq_prod_addr, sizeof(rx_prods),
(u32 *)&rx_prods);
/* mmiowb is needed to synchronize doorbell writes from more than one
* processor. It guarantees that the write arrives to the device before
* the napi lock is released and another qede_poll is called (possibly
* on another CPU). Without this barrier, the next doorbell can bypass
* this doorbell. This is applicable to IA64/Altix systems.
*/
mmiowb();
}
static void qede_get_rxhash(struct sk_buff *skb, u8 bitfields, __le32 rss_hash)
{
enum pkt_hash_types hash_type = PKT_HASH_TYPE_NONE;
enum rss_hash_type htype;
u32 hash = 0;
htype = GET_FIELD(bitfields, ETH_FAST_PATH_RX_REG_CQE_RSS_HASH_TYPE);
if (htype) {
hash_type = ((htype == RSS_HASH_TYPE_IPV4) ||
(htype == RSS_HASH_TYPE_IPV6)) ?
PKT_HASH_TYPE_L3 : PKT_HASH_TYPE_L4;
hash = le32_to_cpu(rss_hash);
}
skb_set_hash(skb, hash, hash_type);
}
static void qede_set_skb_csum(struct sk_buff *skb, u8 csum_flag)
{
skb_checksum_none_assert(skb);
if (csum_flag & QEDE_CSUM_UNNECESSARY)
skb->ip_summed = CHECKSUM_UNNECESSARY;
if (csum_flag & QEDE_TUNN_CSUM_UNNECESSARY) {
skb->csum_level = 1;
skb->encapsulation = 1;
}
}
static inline void qede_skb_receive(struct qede_dev *edev,
struct qede_fastpath *fp,
struct qede_rx_queue *rxq,
struct sk_buff *skb, u16 vlan_tag)
{
if (vlan_tag)
__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vlan_tag);
napi_gro_receive(&fp->napi, skb);
}
static void qede_set_gro_params(struct qede_dev *edev,
struct sk_buff *skb,
struct eth_fast_path_rx_tpa_start_cqe *cqe)
{
u16 parsing_flags = le16_to_cpu(cqe->pars_flags.flags);
if (((parsing_flags >> PARSING_AND_ERR_FLAGS_L3TYPE_SHIFT) &
PARSING_AND_ERR_FLAGS_L3TYPE_MASK) == 2)
skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6;
else
skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
skb_shinfo(skb)->gso_size = __le16_to_cpu(cqe->len_on_first_bd) -
cqe->header_len;
}
static int qede_fill_frag_skb(struct qede_dev *edev,
struct qede_rx_queue *rxq,
u8 tpa_agg_index, u16 len_on_bd)
{
struct sw_rx_data *current_bd = &rxq->sw_rx_ring[rxq->sw_rx_cons &
NUM_RX_BDS_MAX];
struct qede_agg_info *tpa_info = &rxq->tpa_info[tpa_agg_index];
struct sk_buff *skb = tpa_info->skb;
if (unlikely(tpa_info->state != QEDE_AGG_STATE_START))
goto out;
/* Add one frag and update the appropriate fields in the skb */
skb_fill_page_desc(skb, tpa_info->frag_id++,
current_bd->data, current_bd->page_offset,
len_on_bd);
if (unlikely(qede_realloc_rx_buffer(rxq, current_bd))) {
/* Incr page ref count to reuse on allocation failure
* so that it doesn't get freed while freeing SKB.
*/
page_ref_inc(current_bd->data);
goto out;
}
qed_chain_consume(&rxq->rx_bd_ring);
rxq->sw_rx_cons++;
skb->data_len += len_on_bd;
skb->truesize += rxq->rx_buf_seg_size;
skb->len += len_on_bd;
return 0;
out:
tpa_info->state = QEDE_AGG_STATE_ERROR;
qede_recycle_rx_bd_ring(rxq, 1);
return -ENOMEM;
}
static bool qede_tunn_exist(u16 flag)
{
return !!(flag & (PARSING_AND_ERR_FLAGS_TUNNELEXIST_MASK <<
PARSING_AND_ERR_FLAGS_TUNNELEXIST_SHIFT));
}
static u8 qede_check_tunn_csum(u16 flag)
{
u16 csum_flag = 0;
u8 tcsum = 0;
if (flag & (PARSING_AND_ERR_FLAGS_TUNNELL4CHKSMWASCALCULATED_MASK <<
PARSING_AND_ERR_FLAGS_TUNNELL4CHKSMWASCALCULATED_SHIFT))
csum_flag |= PARSING_AND_ERR_FLAGS_TUNNELL4CHKSMERROR_MASK <<
PARSING_AND_ERR_FLAGS_TUNNELL4CHKSMERROR_SHIFT;
if (flag & (PARSING_AND_ERR_FLAGS_L4CHKSMWASCALCULATED_MASK <<
PARSING_AND_ERR_FLAGS_L4CHKSMWASCALCULATED_SHIFT)) {
csum_flag |= PARSING_AND_ERR_FLAGS_L4CHKSMERROR_MASK <<
PARSING_AND_ERR_FLAGS_L4CHKSMERROR_SHIFT;
tcsum = QEDE_TUNN_CSUM_UNNECESSARY;
}
csum_flag |= PARSING_AND_ERR_FLAGS_TUNNELIPHDRERROR_MASK <<
PARSING_AND_ERR_FLAGS_TUNNELIPHDRERROR_SHIFT |
PARSING_AND_ERR_FLAGS_IPHDRERROR_MASK <<
PARSING_AND_ERR_FLAGS_IPHDRERROR_SHIFT;
if (csum_flag & flag)
return QEDE_CSUM_ERROR;
return QEDE_CSUM_UNNECESSARY | tcsum;
}
static void qede_tpa_start(struct qede_dev *edev,
struct qede_rx_queue *rxq,
struct eth_fast_path_rx_tpa_start_cqe *cqe)
{
struct qede_agg_info *tpa_info = &rxq->tpa_info[cqe->tpa_agg_index];
struct eth_rx_bd *rx_bd_cons = qed_chain_consume(&rxq->rx_bd_ring);
struct eth_rx_bd *rx_bd_prod = qed_chain_produce(&rxq->rx_bd_ring);
struct sw_rx_data *replace_buf = &tpa_info->buffer;
dma_addr_t mapping = tpa_info->buffer_mapping;
struct sw_rx_data *sw_rx_data_cons;
struct sw_rx_data *sw_rx_data_prod;
sw_rx_data_cons = &rxq->sw_rx_ring[rxq->sw_rx_cons & NUM_RX_BDS_MAX];
sw_rx_data_prod = &rxq->sw_rx_ring[rxq->sw_rx_prod & NUM_RX_BDS_MAX];
/* Use pre-allocated replacement buffer - we can't release the agg.
* start until its over and we don't want to risk allocation failing
* here, so re-allocate when aggregation will be over.
*/
sw_rx_data_prod->mapping = replace_buf->mapping;
sw_rx_data_prod->data = replace_buf->data;
rx_bd_prod->addr.hi = cpu_to_le32(upper_32_bits(mapping));
rx_bd_prod->addr.lo = cpu_to_le32(lower_32_bits(mapping));
sw_rx_data_prod->page_offset = replace_buf->page_offset;
rxq->sw_rx_prod++;
/* move partial skb from cons to pool (don't unmap yet)
* save mapping, incase we drop the packet later on.
*/
tpa_info->buffer = *sw_rx_data_cons;
mapping = HILO_U64(le32_to_cpu(rx_bd_cons->addr.hi),
le32_to_cpu(rx_bd_cons->addr.lo));
tpa_info->buffer_mapping = mapping;
rxq->sw_rx_cons++;
/* set tpa state to start only if we are able to allocate skb
* for this aggregation, otherwise mark as error and aggregation will
* be dropped
*/
tpa_info->skb = netdev_alloc_skb(edev->ndev,
le16_to_cpu(cqe->len_on_first_bd));
if (unlikely(!tpa_info->skb)) {
DP_NOTICE(edev, "Failed to allocate SKB for gro\n");
tpa_info->state = QEDE_AGG_STATE_ERROR;
goto cons_buf;
}
/* Start filling in the aggregation info */
skb_put(tpa_info->skb, le16_to_cpu(cqe->len_on_first_bd));
tpa_info->frag_id = 0;
tpa_info->state = QEDE_AGG_STATE_START;
/* Store some information from first CQE */
tpa_info->start_cqe_placement_offset = cqe->placement_offset;
tpa_info->start_cqe_bd_len = le16_to_cpu(cqe->len_on_first_bd);
if ((le16_to_cpu(cqe->pars_flags.flags) >>
PARSING_AND_ERR_FLAGS_TAG8021QEXIST_SHIFT) &
PARSING_AND_ERR_FLAGS_TAG8021QEXIST_MASK)
tpa_info->vlan_tag = le16_to_cpu(cqe->vlan_tag);
else
tpa_info->vlan_tag = 0;
qede_get_rxhash(tpa_info->skb, cqe->bitfields, cqe->rss_hash);
/* This is needed in order to enable forwarding support */
qede_set_gro_params(edev, tpa_info->skb, cqe);
cons_buf: /* We still need to handle bd_len_list to consume buffers */
if (likely(cqe->ext_bd_len_list[0]))
qede_fill_frag_skb(edev, rxq, cqe->tpa_agg_index,
le16_to_cpu(cqe->ext_bd_len_list[0]));
if (unlikely(cqe->ext_bd_len_list[1])) {
DP_ERR(edev,
"Unlikely - got a TPA aggregation with more than one ext_bd_len_list entry in the TPA start\n");
tpa_info->state = QEDE_AGG_STATE_ERROR;
}
}
#ifdef CONFIG_INET
static void qede_gro_ip_csum(struct sk_buff *skb)
{
const struct iphdr *iph = ip_hdr(skb);
struct tcphdr *th;
skb_set_transport_header(skb, sizeof(struct iphdr));
th = tcp_hdr(skb);
th->check = ~tcp_v4_check(skb->len - skb_transport_offset(skb),
iph->saddr, iph->daddr, 0);
tcp_gro_complete(skb);
}
static void qede_gro_ipv6_csum(struct sk_buff *skb)
{
struct ipv6hdr *iph = ipv6_hdr(skb);
struct tcphdr *th;
skb_set_transport_header(skb, sizeof(struct ipv6hdr));
th = tcp_hdr(skb);
th->check = ~tcp_v6_check(skb->len - skb_transport_offset(skb),
&iph->saddr, &iph->daddr, 0);
tcp_gro_complete(skb);
}
#endif
static void qede_gro_receive(struct qede_dev *edev,
struct qede_fastpath *fp,
struct sk_buff *skb,
u16 vlan_tag)
{
/* FW can send a single MTU sized packet from gro flow
* due to aggregation timeout/last segment etc. which
* is not expected to be a gro packet. If a skb has zero
* frags then simply push it in the stack as non gso skb.
*/
if (unlikely(!skb->data_len)) {
skb_shinfo(skb)->gso_type = 0;
skb_shinfo(skb)->gso_size = 0;
goto send_skb;
}
#ifdef CONFIG_INET
if (skb_shinfo(skb)->gso_size) {
skb_reset_network_header(skb);
switch (skb->protocol) {
case htons(ETH_P_IP):
qede_gro_ip_csum(skb);
break;
case htons(ETH_P_IPV6):
qede_gro_ipv6_csum(skb);
break;
default:
DP_ERR(edev,
"Error: FW GRO supports only IPv4/IPv6, not 0x%04x\n",
ntohs(skb->protocol));
}
}
#endif
send_skb:
skb_record_rx_queue(skb, fp->rxq->rxq_id);
qede_skb_receive(edev, fp, fp->rxq, skb, vlan_tag);
}
static inline void qede_tpa_cont(struct qede_dev *edev,
struct qede_rx_queue *rxq,
struct eth_fast_path_rx_tpa_cont_cqe *cqe)
{
int i;
for (i = 0; cqe->len_list[i]; i++)
qede_fill_frag_skb(edev, rxq, cqe->tpa_agg_index,
le16_to_cpu(cqe->len_list[i]));
if (unlikely(i > 1))
DP_ERR(edev,
"Strange - TPA cont with more than a single len_list entry\n");
}
static int qede_tpa_end(struct qede_dev *edev,
struct qede_fastpath *fp,
struct eth_fast_path_rx_tpa_end_cqe *cqe)
{
struct qede_rx_queue *rxq = fp->rxq;
struct qede_agg_info *tpa_info;
struct sk_buff *skb;
int i;
tpa_info = &rxq->tpa_info[cqe->tpa_agg_index];
skb = tpa_info->skb;
for (i = 0; cqe->len_list[i]; i++)
qede_fill_frag_skb(edev, rxq, cqe->tpa_agg_index,
le16_to_cpu(cqe->len_list[i]));
if (unlikely(i > 1))
DP_ERR(edev,
"Strange - TPA emd with more than a single len_list entry\n");
if (unlikely(tpa_info->state != QEDE_AGG_STATE_START))
goto err;
/* Sanity */
if (unlikely(cqe->num_of_bds != tpa_info->frag_id + 1))
DP_ERR(edev,
"Strange - TPA had %02x BDs, but SKB has only %d frags\n",
cqe->num_of_bds, tpa_info->frag_id);
if (unlikely(skb->len != le16_to_cpu(cqe->total_packet_len)))
DP_ERR(edev,
"Strange - total packet len [cqe] is %4x but SKB has len %04x\n",
le16_to_cpu(cqe->total_packet_len), skb->len);
memcpy(skb->data,
page_address(tpa_info->buffer.data) +
tpa_info->start_cqe_placement_offset +
tpa_info->buffer.page_offset, tpa_info->start_cqe_bd_len);
/* Finalize the SKB */
skb->protocol = eth_type_trans(skb, edev->ndev);
skb->ip_summed = CHECKSUM_UNNECESSARY;
/* tcp_gro_complete() will copy NAPI_GRO_CB(skb)->count
* to skb_shinfo(skb)->gso_segs
*/
NAPI_GRO_CB(skb)->count = le16_to_cpu(cqe->num_of_coalesced_segs);
qede_gro_receive(edev, fp, skb, tpa_info->vlan_tag);
tpa_info->state = QEDE_AGG_STATE_NONE;
return 1;
err:
tpa_info->state = QEDE_AGG_STATE_NONE;
dev_kfree_skb_any(tpa_info->skb);
tpa_info->skb = NULL;
return 0;
}
static u8 qede_check_notunn_csum(u16 flag)
{
u16 csum_flag = 0;
u8 csum = 0;
if (flag & (PARSING_AND_ERR_FLAGS_L4CHKSMWASCALCULATED_MASK <<
PARSING_AND_ERR_FLAGS_L4CHKSMWASCALCULATED_SHIFT)) {
csum_flag |= PARSING_AND_ERR_FLAGS_L4CHKSMERROR_MASK <<
PARSING_AND_ERR_FLAGS_L4CHKSMERROR_SHIFT;
csum = QEDE_CSUM_UNNECESSARY;
}
csum_flag |= PARSING_AND_ERR_FLAGS_IPHDRERROR_MASK <<
PARSING_AND_ERR_FLAGS_IPHDRERROR_SHIFT;
if (csum_flag & flag)
return QEDE_CSUM_ERROR;
return csum;
}
static u8 qede_check_csum(u16 flag)
{
if (!qede_tunn_exist(flag))
return qede_check_notunn_csum(flag);
else
return qede_check_tunn_csum(flag);
}
static bool qede_pkt_is_ip_fragmented(struct eth_fast_path_rx_reg_cqe *cqe,
u16 flag)
{
u8 tun_pars_flg = cqe->tunnel_pars_flags.flags;
if ((tun_pars_flg & (ETH_TUNNEL_PARSING_FLAGS_IPV4_FRAGMENT_MASK <<
ETH_TUNNEL_PARSING_FLAGS_IPV4_FRAGMENT_SHIFT)) ||
(flag & (PARSING_AND_ERR_FLAGS_IPV4FRAG_MASK <<
PARSING_AND_ERR_FLAGS_IPV4FRAG_SHIFT)))
return true;
return false;
}
/* Return true iff packet is to be passed to stack */
static bool qede_rx_xdp(struct qede_dev *edev,
struct qede_fastpath *fp,
struct qede_rx_queue *rxq,
struct bpf_prog *prog,
struct sw_rx_data *bd,
struct eth_fast_path_rx_reg_cqe *cqe,
u16 *data_offset, u16 *len)
{
struct xdp_buff xdp;
enum xdp_action act;
xdp.data_hard_start = page_address(bd->data);
xdp.data = xdp.data_hard_start + *data_offset;
xdp.data_end = xdp.data + *len;
/* Queues always have a full reset currently, so for the time
* being until there's atomic program replace just mark read
* side for map helpers.
*/
rcu_read_lock();
act = bpf_prog_run_xdp(prog, &xdp);
rcu_read_unlock();
/* Recalculate, as XDP might have changed the headers */
*data_offset = xdp.data - xdp.data_hard_start;
*len = xdp.data_end - xdp.data;
if (act == XDP_PASS)
return true;
/* Count number of packets not to be passed to stack */
rxq->xdp_no_pass++;
switch (act) {
case XDP_TX:
/* We need the replacement buffer before transmit. */
if (qede_alloc_rx_buffer(rxq, true)) {
qede_recycle_rx_bd_ring(rxq, 1);
trace_xdp_exception(edev->ndev, prog, act);
return false;
}
/* Now if there's a transmission problem, we'd still have to
* throw current buffer, as replacement was already allocated.
*/
if (qede_xdp_xmit(edev, fp, bd, *data_offset, *len)) {
dma_unmap_page(rxq->dev, bd->mapping,
PAGE_SIZE, DMA_BIDIRECTIONAL);
__free_page(bd->data);
trace_xdp_exception(edev->ndev, prog, act);
}
/* Regardless, we've consumed an Rx BD */
qede_rx_bd_ring_consume(rxq);
return false;
default:
bpf_warn_invalid_xdp_action(act);
case XDP_ABORTED:
trace_xdp_exception(edev->ndev, prog, act);
case XDP_DROP:
qede_recycle_rx_bd_ring(rxq, cqe->bd_num);
}
return false;
}
static struct sk_buff *qede_rx_allocate_skb(struct qede_dev *edev,
struct qede_rx_queue *rxq,
struct sw_rx_data *bd, u16 len,
u16 pad)
{
unsigned int offset = bd->page_offset + pad;
struct skb_frag_struct *frag;
struct page *page = bd->data;
unsigned int pull_len;
struct sk_buff *skb;
unsigned char *va;
/* Allocate a new SKB with a sufficient large header len */
skb = netdev_alloc_skb(edev->ndev, QEDE_RX_HDR_SIZE);
if (unlikely(!skb))
return NULL;
/* Copy data into SKB - if it's small, we can simply copy it and
* re-use the already allcoated & mapped memory.
*/
if (len + pad <= edev->rx_copybreak) {
memcpy(skb_put(skb, len),
page_address(page) + offset, len);
qede_reuse_page(rxq, bd);
goto out;
}
frag = &skb_shinfo(skb)->frags[0];
skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags,
page, offset, len, rxq->rx_buf_seg_size);
va = skb_frag_address(frag);
pull_len = eth_get_headlen(va, QEDE_RX_HDR_SIZE);
/* Align the pull_len to optimize memcpy */
memcpy(skb->data, va, ALIGN(pull_len, sizeof(long)));
/* Correct the skb & frag sizes offset after the pull */
skb_frag_size_sub(frag, pull_len);
frag->page_offset += pull_len;
skb->data_len -= pull_len;
skb->tail += pull_len;
if (unlikely(qede_realloc_rx_buffer(rxq, bd))) {
/* Incr page ref count to reuse on allocation failure so
* that it doesn't get freed while freeing SKB [as its
* already mapped there].
*/
page_ref_inc(page);
dev_kfree_skb_any(skb);
return NULL;
}
out:
/* We've consumed the first BD and prepared an SKB */
qede_rx_bd_ring_consume(rxq);
return skb;
}
static int qede_rx_build_jumbo(struct qede_dev *edev,
struct qede_rx_queue *rxq,
struct sk_buff *skb,
struct eth_fast_path_rx_reg_cqe *cqe,
u16 first_bd_len)
{
u16 pkt_len = le16_to_cpu(cqe->pkt_len);
struct sw_rx_data *bd;
u16 bd_cons_idx;
u8 num_frags;
pkt_len -= first_bd_len;
/* We've already used one BD for the SKB. Now take care of the rest */
for (num_frags = cqe->bd_num - 1; num_frags > 0; num_frags--) {
u16 cur_size = pkt_len > rxq->rx_buf_size ? rxq->rx_buf_size :
pkt_len;
if (unlikely(!cur_size)) {
DP_ERR(edev,
"Still got %d BDs for mapping jumbo, but length became 0\n",
num_frags);
goto out;
}
/* We need a replacement buffer for each BD */
if (unlikely(qede_alloc_rx_buffer(rxq, true)))
goto out;
/* Now that we've allocated the replacement buffer,
* we can safely consume the next BD and map it to the SKB.
*/
bd_cons_idx = rxq->sw_rx_cons & NUM_RX_BDS_MAX;
bd = &rxq->sw_rx_ring[bd_cons_idx];
qede_rx_bd_ring_consume(rxq);
dma_unmap_page(rxq->dev, bd->mapping,
PAGE_SIZE, DMA_FROM_DEVICE);
skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags++,
bd->data, 0, cur_size);
skb->truesize += PAGE_SIZE;
skb->data_len += cur_size;
skb->len += cur_size;
pkt_len -= cur_size;
}
if (unlikely(pkt_len))
DP_ERR(edev,
"Mapped all BDs of jumbo, but still have %d bytes\n",
pkt_len);
out:
return num_frags;
}
static int qede_rx_process_tpa_cqe(struct qede_dev *edev,
struct qede_fastpath *fp,
struct qede_rx_queue *rxq,
union eth_rx_cqe *cqe,
enum eth_rx_cqe_type type)
{
switch (type) {
case ETH_RX_CQE_TYPE_TPA_START:
qede_tpa_start(edev, rxq, &cqe->fast_path_tpa_start);
return 0;
case ETH_RX_CQE_TYPE_TPA_CONT:
qede_tpa_cont(edev, rxq, &cqe->fast_path_tpa_cont);
return 0;
case ETH_RX_CQE_TYPE_TPA_END:
return qede_tpa_end(edev, fp, &cqe->fast_path_tpa_end);
default:
return 0;
}
}
static int qede_rx_process_cqe(struct qede_dev *edev,
struct qede_fastpath *fp,
struct qede_rx_queue *rxq)
{
struct bpf_prog *xdp_prog = READ_ONCE(rxq->xdp_prog);
struct eth_fast_path_rx_reg_cqe *fp_cqe;
u16 len, pad, bd_cons_idx, parse_flag;
enum eth_rx_cqe_type cqe_type;
union eth_rx_cqe *cqe;
struct sw_rx_data *bd;
struct sk_buff *skb;
__le16 flags;
u8 csum_flag;
/* Get the CQE from the completion ring */
cqe = (union eth_rx_cqe *)qed_chain_consume(&rxq->rx_comp_ring);
cqe_type = cqe->fast_path_regular.type;
/* Process an unlikely slowpath event */
if (unlikely(cqe_type == ETH_RX_CQE_TYPE_SLOW_PATH)) {
struct eth_slow_path_rx_cqe *sp_cqe;
sp_cqe = (struct eth_slow_path_rx_cqe *)cqe;
edev->ops->eth_cqe_completion(edev->cdev, fp->id, sp_cqe);
return 0;
}
/* Handle TPA cqes */
if (cqe_type != ETH_RX_CQE_TYPE_REGULAR)
return qede_rx_process_tpa_cqe(edev, fp, rxq, cqe, cqe_type);
/* Get the data from the SW ring; Consume it only after it's evident
* we wouldn't recycle it.
*/
bd_cons_idx = rxq->sw_rx_cons & NUM_RX_BDS_MAX;
bd = &rxq->sw_rx_ring[bd_cons_idx];
fp_cqe = &cqe->fast_path_regular;
len = le16_to_cpu(fp_cqe->len_on_first_bd);
pad = fp_cqe->placement_offset + rxq->rx_headroom;
/* Run eBPF program if one is attached */
if (xdp_prog)
if (!qede_rx_xdp(edev, fp, rxq, xdp_prog, bd, fp_cqe,
&pad, &len))
return 0;
/* If this is an error packet then drop it */
flags = cqe->fast_path_regular.pars_flags.flags;
parse_flag = le16_to_cpu(flags);
csum_flag = qede_check_csum(parse_flag);
if (unlikely(csum_flag == QEDE_CSUM_ERROR)) {
if (qede_pkt_is_ip_fragmented(fp_cqe, parse_flag)) {
rxq->rx_ip_frags++;
} else {
DP_NOTICE(edev,
"CQE has error, flags = %x, dropping incoming packet\n",
parse_flag);
rxq->rx_hw_errors++;
qede_recycle_rx_bd_ring(rxq, fp_cqe->bd_num);
return 0;
}
}
/* Basic validation passed; Need to prepare an SKB. This would also
* guarantee to finally consume the first BD upon success.
*/
skb = qede_rx_allocate_skb(edev, rxq, bd, len, pad);
if (!skb) {
rxq->rx_alloc_errors++;
qede_recycle_rx_bd_ring(rxq, fp_cqe->bd_num);
return 0;
}
/* In case of Jumbo packet, several PAGE_SIZEd buffers will be pointed
* by a single cqe.
*/
if (fp_cqe->bd_num > 1) {
u16 unmapped_frags = qede_rx_build_jumbo(edev, rxq, skb,
fp_cqe, len);
if (unlikely(unmapped_frags > 0)) {
qede_recycle_rx_bd_ring(rxq, unmapped_frags);
dev_kfree_skb_any(skb);
return 0;
}
}
/* The SKB contains all the data. Now prepare meta-magic */
skb->protocol = eth_type_trans(skb, edev->ndev);
qede_get_rxhash(skb, fp_cqe->bitfields, fp_cqe->rss_hash);
qede_set_skb_csum(skb, csum_flag);
skb_record_rx_queue(skb, rxq->rxq_id);
qede_ptp_record_rx_ts(edev, cqe, skb);
/* SKB is prepared - pass it to stack */
qede_skb_receive(edev, fp, rxq, skb, le16_to_cpu(fp_cqe->vlan_tag));
return 1;
}
static int qede_rx_int(struct qede_fastpath *fp, int budget)
{
struct qede_rx_queue *rxq = fp->rxq;
struct qede_dev *edev = fp->edev;
int work_done = 0, rcv_pkts = 0;
u16 hw_comp_cons, sw_comp_cons;
hw_comp_cons = le16_to_cpu(*rxq->hw_cons_ptr);
sw_comp_cons = qed_chain_get_cons_idx(&rxq->rx_comp_ring);
/* Memory barrier to prevent the CPU from doing speculative reads of CQE
* / BD in the while-loop before reading hw_comp_cons. If the CQE is
* read before it is written by FW, then FW writes CQE and SB, and then
* the CPU reads the hw_comp_cons, it will use an old CQE.
*/
rmb();
/* Loop to complete all indicated BDs */
while ((sw_comp_cons != hw_comp_cons) && (work_done < budget)) {
rcv_pkts += qede_rx_process_cqe(edev, fp, rxq);
qed_chain_recycle_consumed(&rxq->rx_comp_ring);
sw_comp_cons = qed_chain_get_cons_idx(&rxq->rx_comp_ring);
work_done++;
}
rxq->rcv_pkts += rcv_pkts;
/* Allocate replacement buffers */
while (rxq->num_rx_buffers - rxq->filled_buffers)
if (qede_alloc_rx_buffer(rxq, false))
break;
/* Update producers */
qede_update_rx_prod(edev, rxq);
return work_done;
}
static bool qede_poll_is_more_work(struct qede_fastpath *fp)
{
qed_sb_update_sb_idx(fp->sb_info);
/* *_has_*_work() reads the status block, thus we need to ensure that
* status block indices have been actually read (qed_sb_update_sb_idx)
* prior to this check (*_has_*_work) so that we won't write the
* "newer" value of the status block to HW (if there was a DMA right
* after qede_has_rx_work and if there is no rmb, the memory reading
* (qed_sb_update_sb_idx) may be postponed to right before *_ack_sb).
* In this case there will never be another interrupt until there is
* another update of the status block, while there is still unhandled
* work.
*/
rmb();
if (likely(fp->type & QEDE_FASTPATH_RX))
if (qede_has_rx_work(fp->rxq))
return true;
if (fp->type & QEDE_FASTPATH_XDP)
if (qede_txq_has_work(fp->xdp_tx))
return true;
if (likely(fp->type & QEDE_FASTPATH_TX))
if (qede_txq_has_work(fp->txq))
return true;
return false;
}
/*********************
* NDO & API related *
*********************/
int qede_poll(struct napi_struct *napi, int budget)
{
struct qede_fastpath *fp = container_of(napi, struct qede_fastpath,
napi);
struct qede_dev *edev = fp->edev;
int rx_work_done = 0;
if (likely(fp->type & QEDE_FASTPATH_TX) && qede_txq_has_work(fp->txq))
qede_tx_int(edev, fp->txq);
if ((fp->type & QEDE_FASTPATH_XDP) && qede_txq_has_work(fp->xdp_tx))
qede_xdp_tx_int(edev, fp->xdp_tx);
rx_work_done = (likely(fp->type & QEDE_FASTPATH_RX) &&
qede_has_rx_work(fp->rxq)) ?
qede_rx_int(fp, budget) : 0;
if (rx_work_done < budget) {
if (!qede_poll_is_more_work(fp)) {
napi_complete_done(napi, rx_work_done);
/* Update and reenable interrupts */
qed_sb_ack(fp->sb_info, IGU_INT_ENABLE, 1);
} else {
rx_work_done = budget;
}
}
if (fp->xdp_xmit) {
u16 xdp_prod = qed_chain_get_prod_idx(&fp->xdp_tx->tx_pbl);
fp->xdp_xmit = 0;
fp->xdp_tx->tx_db.data.bd_prod = cpu_to_le16(xdp_prod);
qede_update_tx_producer(fp->xdp_tx);
}
return rx_work_done;
}
irqreturn_t qede_msix_fp_int(int irq, void *fp_cookie)
{
struct qede_fastpath *fp = fp_cookie;
qed_sb_ack(fp->sb_info, IGU_INT_DISABLE, 0 /*do not update*/);
napi_schedule_irqoff(&fp->napi);
return IRQ_HANDLED;
}
/* Main transmit function */
netdev_tx_t qede_start_xmit(struct sk_buff *skb, struct net_device *ndev)
{
struct qede_dev *edev = netdev_priv(ndev);
struct netdev_queue *netdev_txq;
struct qede_tx_queue *txq;
struct eth_tx_1st_bd *first_bd;
struct eth_tx_2nd_bd *second_bd = NULL;
struct eth_tx_3rd_bd *third_bd = NULL;
struct eth_tx_bd *tx_data_bd = NULL;
u16 txq_index;
u8 nbd = 0;
dma_addr_t mapping;
int rc, frag_idx = 0, ipv6_ext = 0;
u8 xmit_type;
u16 idx;
u16 hlen;
bool data_split = false;
/* Get tx-queue context and netdev index */
txq_index = skb_get_queue_mapping(skb);
WARN_ON(txq_index >= QEDE_TSS_COUNT(edev));
txq = edev->fp_array[edev->fp_num_rx + txq_index].txq;
netdev_txq = netdev_get_tx_queue(ndev, txq_index);
WARN_ON(qed_chain_get_elem_left(&txq->tx_pbl) < (MAX_SKB_FRAGS + 1));
xmit_type = qede_xmit_type(skb, &ipv6_ext);
#if ((MAX_SKB_FRAGS + 2) > ETH_TX_MAX_BDS_PER_NON_LSO_PACKET)
if (qede_pkt_req_lin(skb, xmit_type)) {
if (skb_linearize(skb)) {
DP_NOTICE(edev,
"SKB linearization failed - silently dropping this SKB\n");
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
}
}
#endif
/* Fill the entry in the SW ring and the BDs in the FW ring */
idx = txq->sw_tx_prod & NUM_TX_BDS_MAX;
txq->sw_tx_ring.skbs[idx].skb = skb;
first_bd = (struct eth_tx_1st_bd *)
qed_chain_produce(&txq->tx_pbl);
memset(first_bd, 0, sizeof(*first_bd));
first_bd->data.bd_flags.bitfields =
1 << ETH_TX_1ST_BD_FLAGS_START_BD_SHIFT;
if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP))
qede_ptp_tx_ts(edev, skb);
/* Map skb linear data for DMA and set in the first BD */
mapping = dma_map_single(txq->dev, skb->data,
skb_headlen(skb), DMA_TO_DEVICE);
if (unlikely(dma_mapping_error(txq->dev, mapping))) {
DP_NOTICE(edev, "SKB mapping failed\n");
qede_free_failed_tx_pkt(txq, first_bd, 0, false);
qede_update_tx_producer(txq);
return NETDEV_TX_OK;
}
nbd++;
BD_SET_UNMAP_ADDR_LEN(first_bd, mapping, skb_headlen(skb));
/* In case there is IPv6 with extension headers or LSO we need 2nd and
* 3rd BDs.
*/
if (unlikely((xmit_type & XMIT_LSO) | ipv6_ext)) {
second_bd = (struct eth_tx_2nd_bd *)
qed_chain_produce(&txq->tx_pbl);
memset(second_bd, 0, sizeof(*second_bd));
nbd++;
third_bd = (struct eth_tx_3rd_bd *)
qed_chain_produce(&txq->tx_pbl);
memset(third_bd, 0, sizeof(*third_bd));
nbd++;
/* We need to fill in additional data in second_bd... */
tx_data_bd = (struct eth_tx_bd *)second_bd;
}
if (skb_vlan_tag_present(skb)) {
first_bd->data.vlan = cpu_to_le16(skb_vlan_tag_get(skb));
first_bd->data.bd_flags.bitfields |=
1 << ETH_TX_1ST_BD_FLAGS_VLAN_INSERTION_SHIFT;
}
/* Fill the parsing flags & params according to the requested offload */
if (xmit_type & XMIT_L4_CSUM) {
/* We don't re-calculate IP checksum as it is already done by
* the upper stack
*/
first_bd->data.bd_flags.bitfields |=
1 << ETH_TX_1ST_BD_FLAGS_L4_CSUM_SHIFT;
if (xmit_type & XMIT_ENC) {
first_bd->data.bd_flags.bitfields |=
1 << ETH_TX_1ST_BD_FLAGS_IP_CSUM_SHIFT;
first_bd->data.bitfields |=
1 << ETH_TX_DATA_1ST_BD_TUNN_FLAG_SHIFT;
}
/* Legacy FW had flipped behavior in regard to this bit -
* I.e., needed to set to prevent FW from touching encapsulated
* packets when it didn't need to.
*/
if (unlikely(txq->is_legacy))
first_bd->data.bitfields ^=
1 << ETH_TX_DATA_1ST_BD_TUNN_FLAG_SHIFT;
/* If the packet is IPv6 with extension header, indicate that
* to FW and pass few params, since the device cracker doesn't
* support parsing IPv6 with extension header/s.
*/
if (unlikely(ipv6_ext))
qede_set_params_for_ipv6_ext(skb, second_bd, third_bd);
}
if (xmit_type & XMIT_LSO) {
first_bd->data.bd_flags.bitfields |=
(1 << ETH_TX_1ST_BD_FLAGS_LSO_SHIFT);
third_bd->data.lso_mss =
cpu_to_le16(skb_shinfo(skb)->gso_size);
if (unlikely(xmit_type & XMIT_ENC)) {
first_bd->data.bd_flags.bitfields |=
1 << ETH_TX_1ST_BD_FLAGS_TUNN_IP_CSUM_SHIFT;
if (xmit_type & XMIT_ENC_GSO_L4_CSUM) {
u8 tmp = ETH_TX_1ST_BD_FLAGS_TUNN_L4_CSUM_SHIFT;
first_bd->data.bd_flags.bitfields |= 1 << tmp;
}
hlen = qede_get_skb_hlen(skb, true);
} else {
first_bd->data.bd_flags.bitfields |=
1 << ETH_TX_1ST_BD_FLAGS_IP_CSUM_SHIFT;
hlen = qede_get_skb_hlen(skb, false);
}
/* @@@TBD - if will not be removed need to check */
third_bd->data.bitfields |=
cpu_to_le16(1 << ETH_TX_DATA_3RD_BD_HDR_NBD_SHIFT);
/* Make life easier for FW guys who can't deal with header and
* data on same BD. If we need to split, use the second bd...
*/
if (unlikely(skb_headlen(skb) > hlen)) {
DP_VERBOSE(edev, NETIF_MSG_TX_QUEUED,
"TSO split header size is %d (%x:%x)\n",
first_bd->nbytes, first_bd->addr.hi,
first_bd->addr.lo);
mapping = HILO_U64(le32_to_cpu(first_bd->addr.hi),
le32_to_cpu(first_bd->addr.lo)) +
hlen;
BD_SET_UNMAP_ADDR_LEN(tx_data_bd, mapping,
le16_to_cpu(first_bd->nbytes) -
hlen);
/* this marks the BD as one that has no
* individual mapping
*/
txq->sw_tx_ring.skbs[idx].flags |= QEDE_TSO_SPLIT_BD;
first_bd->nbytes = cpu_to_le16(hlen);
tx_data_bd = (struct eth_tx_bd *)third_bd;
data_split = true;
}
} else {
first_bd->data.bitfields |=
(skb->len & ETH_TX_DATA_1ST_BD_PKT_LEN_MASK) <<
ETH_TX_DATA_1ST_BD_PKT_LEN_SHIFT;
}
/* Handle fragmented skb */
/* special handle for frags inside 2nd and 3rd bds.. */
while (tx_data_bd && frag_idx < skb_shinfo(skb)->nr_frags) {
rc = map_frag_to_bd(txq,
&skb_shinfo(skb)->frags[frag_idx],
tx_data_bd);
if (rc) {
qede_free_failed_tx_pkt(txq, first_bd, nbd, data_split);
qede_update_tx_producer(txq);
return NETDEV_TX_OK;
}
if (tx_data_bd == (struct eth_tx_bd *)second_bd)
tx_data_bd = (struct eth_tx_bd *)third_bd;
else
tx_data_bd = NULL;
frag_idx++;
}
/* map last frags into 4th, 5th .... */
for (; frag_idx < skb_shinfo(skb)->nr_frags; frag_idx++, nbd++) {
tx_data_bd = (struct eth_tx_bd *)
qed_chain_produce(&txq->tx_pbl);
memset(tx_data_bd, 0, sizeof(*tx_data_bd));
rc = map_frag_to_bd(txq,
&skb_shinfo(skb)->frags[frag_idx],
tx_data_bd);
if (rc) {
qede_free_failed_tx_pkt(txq, first_bd, nbd, data_split);
qede_update_tx_producer(txq);
return NETDEV_TX_OK;
}
}
/* update the first BD with the actual num BDs */
first_bd->data.nbds = nbd;
netdev_tx_sent_queue(netdev_txq, skb->len);
skb_tx_timestamp(skb);
/* Advance packet producer only before sending the packet since mapping
* of pages may fail.
*/
txq->sw_tx_prod++;
/* 'next page' entries are counted in the producer value */
txq->tx_db.data.bd_prod =
cpu_to_le16(qed_chain_get_prod_idx(&txq->tx_pbl));
if (!skb->xmit_more || netif_xmit_stopped(netdev_txq))
qede_update_tx_producer(txq);
if (unlikely(qed_chain_get_elem_left(&txq->tx_pbl)
< (MAX_SKB_FRAGS + 1))) {
if (skb->xmit_more)
qede_update_tx_producer(txq);
netif_tx_stop_queue(netdev_txq);
txq->stopped_cnt++;
DP_VERBOSE(edev, NETIF_MSG_TX_QUEUED,
"Stop queue was called\n");
/* paired memory barrier is in qede_tx_int(), we have to keep
* ordering of set_bit() in netif_tx_stop_queue() and read of
* fp->bd_tx_cons
*/
smp_mb();
if ((qed_chain_get_elem_left(&txq->tx_pbl) >=
(MAX_SKB_FRAGS + 1)) &&
(edev->state == QEDE_STATE_OPEN)) {
netif_tx_wake_queue(netdev_txq);
DP_VERBOSE(edev, NETIF_MSG_TX_QUEUED,
"Wake queue was called\n");
}
}
return NETDEV_TX_OK;
}
/* 8B udp header + 8B base tunnel header + 32B option length */
#define QEDE_MAX_TUN_HDR_LEN 48
netdev_features_t qede_features_check(struct sk_buff *skb,
struct net_device *dev,
netdev_features_t features)
{
if (skb->encapsulation) {
u8 l4_proto = 0;
switch (vlan_get_protocol(skb)) {
case htons(ETH_P_IP):
l4_proto = ip_hdr(skb)->protocol;
break;
case htons(ETH_P_IPV6):
l4_proto = ipv6_hdr(skb)->nexthdr;
break;
default:
return features;
}
/* Disable offloads for geneve tunnels, as HW can't parse
* the geneve header which has option length greater than 32b
* and disable offloads for the ports which are not offloaded.
*/
if (l4_proto == IPPROTO_UDP) {
struct qede_dev *edev = netdev_priv(dev);
u16 hdrlen, vxln_port, gnv_port;
hdrlen = QEDE_MAX_TUN_HDR_LEN;
vxln_port = edev->vxlan_dst_port;
gnv_port = edev->geneve_dst_port;
if ((skb_inner_mac_header(skb) -
skb_transport_header(skb)) > hdrlen ||
(ntohs(udp_hdr(skb)->dest) != vxln_port &&
ntohs(udp_hdr(skb)->dest) != gnv_port))
return features & ~(NETIF_F_CSUM_MASK |
NETIF_F_GSO_MASK);
}
}
return features;
}
|