summaryrefslogtreecommitdiffstats
path: root/drivers/net/cxgb4vf/cxgb4vf_main.c
blob: 24808ac9976e00f6942b93372040670df4800012 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
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
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
/*
 * This file is part of the Chelsio T4 PCI-E SR-IOV Virtual Function Ethernet
 * driver for Linux.
 *
 * Copyright (c) 2009-2010 Chelsio Communications, Inc. All rights reserved.
 *
 * 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/version.h>
#include <linux/module.h>
#include <linux/moduleparam.h>
#include <linux/init.h>
#include <linux/pci.h>
#include <linux/dma-mapping.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/debugfs.h>
#include <linux/ethtool.h>

#include "t4vf_common.h"
#include "t4vf_defs.h"

#include "../cxgb4/t4_regs.h"
#include "../cxgb4/t4_msg.h"

/*
 * Generic information about the driver.
 */
#define DRV_VERSION "1.0.0"
#define DRV_DESC "Chelsio T4 Virtual Function (VF) Network Driver"

/*
 * Module Parameters.
 * ==================
 */

/*
 * Default ethtool "message level" for adapters.
 */
#define DFLT_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_PROBE | NETIF_MSG_LINK | \
			 NETIF_MSG_TIMER | NETIF_MSG_IFDOWN | NETIF_MSG_IFUP |\
			 NETIF_MSG_RX_ERR | NETIF_MSG_TX_ERR)

static int dflt_msg_enable = DFLT_MSG_ENABLE;

module_param(dflt_msg_enable, int, 0644);
MODULE_PARM_DESC(dflt_msg_enable,
		 "default adapter ethtool message level bitmap");

/*
 * The driver uses the best interrupt scheme available on a platform in the
 * order MSI-X then MSI.  This parameter determines which of these schemes the
 * driver may consider as follows:
 *
 *     msi = 2: choose from among MSI-X and MSI
 *     msi = 1: only consider MSI interrupts
 *
 * Note that unlike the Physical Function driver, this Virtual Function driver
 * does _not_ support legacy INTx interrupts (this limitation is mandated by
 * the PCI-E SR-IOV standard).
 */
#define MSI_MSIX	2
#define MSI_MSI		1
#define MSI_DEFAULT	MSI_MSIX

static int msi = MSI_DEFAULT;

module_param(msi, int, 0644);
MODULE_PARM_DESC(msi, "whether to use MSI-X or MSI");

/*
 * Fundamental constants.
 * ======================
 */

enum {
	MAX_TXQ_ENTRIES		= 16384,
	MAX_RSPQ_ENTRIES	= 16384,
	MAX_RX_BUFFERS		= 16384,

	MIN_TXQ_ENTRIES		= 32,
	MIN_RSPQ_ENTRIES	= 128,
	MIN_FL_ENTRIES		= 16,

	/*
	 * For purposes of manipulating the Free List size we need to
	 * recognize that Free Lists are actually Egress Queues (the host
	 * produces free buffers which the hardware consumes), Egress Queues
	 * indices are all in units of Egress Context Units bytes, and free
	 * list entries are 64-bit PCI DMA addresses.  And since the state of
	 * the Producer Index == the Consumer Index implies an EMPTY list, we
	 * always have at least one Egress Unit's worth of Free List entries
	 * unused.  See sge.c for more details ...
	 */
	EQ_UNIT = SGE_EQ_IDXSIZE,
	FL_PER_EQ_UNIT = EQ_UNIT / sizeof(__be64),
	MIN_FL_RESID = FL_PER_EQ_UNIT,
};

/*
 * Global driver state.
 * ====================
 */

static struct dentry *cxgb4vf_debugfs_root;

/*
 * OS "Callback" functions.
 * ========================
 */

/*
 * The link status has changed on the indicated "port" (Virtual Interface).
 */
void t4vf_os_link_changed(struct adapter *adapter, int pidx, int link_ok)
{
	struct net_device *dev = adapter->port[pidx];

	/*
	 * If the port is disabled or the current recorded "link up"
	 * status matches the new status, just return.
	 */
	if (!netif_running(dev) || link_ok == netif_carrier_ok(dev))
		return;

	/*
	 * Tell the OS that the link status has changed and print a short
	 * informative message on the console about the event.
	 */
	if (link_ok) {
		const char *s;
		const char *fc;
		const struct port_info *pi = netdev_priv(dev);

		netif_carrier_on(dev);

		switch (pi->link_cfg.speed) {
		case SPEED_10000:
			s = "10Gbps";
			break;

		case SPEED_1000:
			s = "1000Mbps";
			break;

		case SPEED_100:
			s = "100Mbps";
			break;

		default:
			s = "unknown";
			break;
		}

		switch (pi->link_cfg.fc) {
		case PAUSE_RX:
			fc = "RX";
			break;

		case PAUSE_TX:
			fc = "TX";
			break;

		case PAUSE_RX|PAUSE_TX:
			fc = "RX/TX";
			break;

		default:
			fc = "no";
			break;
		}

		printk(KERN_INFO "%s: link up, %s, full-duplex, %s PAUSE\n",
		       dev->name, s, fc);
	} else {
		netif_carrier_off(dev);
		printk(KERN_INFO "%s: link down\n", dev->name);
	}
}

/*
 * Net device operations.
 * ======================
 */

/*
 * Record our new VLAN Group and enable/disable hardware VLAN Tag extraction
 * based on whether the specified VLAN Group pointer is NULL or not.
 */
static void cxgb4vf_vlan_rx_register(struct net_device *dev,
				     struct vlan_group *grp)
{
	struct port_info *pi = netdev_priv(dev);

	pi->vlan_grp = grp;
	t4vf_set_rxmode(pi->adapter, pi->viid, -1, -1, -1, -1, grp != NULL, 0);
}

/*
 * Perform the MAC and PHY actions needed to enable a "port" (Virtual
 * Interface).
 */
static int link_start(struct net_device *dev)
{
	int ret;
	struct port_info *pi = netdev_priv(dev);

	/*
	 * We do not set address filters and promiscuity here, the stack does
	 * that step explicitly.
	 */
	ret = t4vf_set_rxmode(pi->adapter, pi->viid, dev->mtu, -1, -1, -1, -1,
			      true);
	if (ret == 0) {
		ret = t4vf_change_mac(pi->adapter, pi->viid,
				      pi->xact_addr_filt, dev->dev_addr, true);
		if (ret >= 0) {
			pi->xact_addr_filt = ret;
			ret = 0;
		}
	}

	/*
	 * We don't need to actually "start the link" itself since the
	 * firmware will do that for us when the first Virtual Interface
	 * is enabled on a port.
	 */
	if (ret == 0)
		ret = t4vf_enable_vi(pi->adapter, pi->viid, true, true);
	return ret;
}

/*
 * Name the MSI-X interrupts.
 */
static void name_msix_vecs(struct adapter *adapter)
{
	int namelen = sizeof(adapter->msix_info[0].desc) - 1;
	int pidx;

	/*
	 * Firmware events.
	 */
	snprintf(adapter->msix_info[MSIX_FW].desc, namelen,
		 "%s-FWeventq", adapter->name);
	adapter->msix_info[MSIX_FW].desc[namelen] = 0;

	/*
	 * Ethernet queues.
	 */
	for_each_port(adapter, pidx) {
		struct net_device *dev = adapter->port[pidx];
		const struct port_info *pi = netdev_priv(dev);
		int qs, msi;

		for (qs = 0, msi = MSIX_NIQFLINT;
		     qs < pi->nqsets;
		     qs++, msi++) {
			snprintf(adapter->msix_info[msi].desc, namelen,
				 "%s-%d", dev->name, qs);
			adapter->msix_info[msi].desc[namelen] = 0;
		}
	}
}

/*
 * Request all of our MSI-X resources.
 */
static int request_msix_queue_irqs(struct adapter *adapter)
{
	struct sge *s = &adapter->sge;
	int rxq, msi, err;

	/*
	 * Firmware events.
	 */
	err = request_irq(adapter->msix_info[MSIX_FW].vec, t4vf_sge_intr_msix,
			  0, adapter->msix_info[MSIX_FW].desc, &s->fw_evtq);
	if (err)
		return err;

	/*
	 * Ethernet queues.
	 */
	msi = MSIX_NIQFLINT;
	for_each_ethrxq(s, rxq) {
		err = request_irq(adapter->msix_info[msi].vec,
				  t4vf_sge_intr_msix, 0,
				  adapter->msix_info[msi].desc,
				  &s->ethrxq[rxq].rspq);
		if (err)
			goto err_free_irqs;
		msi++;
	}
	return 0;

err_free_irqs:
	while (--rxq >= 0)
		free_irq(adapter->msix_info[--msi].vec, &s->ethrxq[rxq].rspq);
	free_irq(adapter->msix_info[MSIX_FW].vec, &s->fw_evtq);
	return err;
}

/*
 * Free our MSI-X resources.
 */
static void free_msix_queue_irqs(struct adapter *adapter)
{
	struct sge *s = &adapter->sge;
	int rxq, msi;

	free_irq(adapter->msix_info[MSIX_FW].vec, &s->fw_evtq);
	msi = MSIX_NIQFLINT;
	for_each_ethrxq(s, rxq)
		free_irq(adapter->msix_info[msi++].vec,
			 &s->ethrxq[rxq].rspq);
}

/*
 * Turn on NAPI and start up interrupts on a response queue.
 */
static void qenable(struct sge_rspq *rspq)
{
	napi_enable(&rspq->napi);

	/*
	 * 0-increment the Going To Sleep register to start the timer and
	 * enable interrupts.
	 */
	t4_write_reg(rspq->adapter, T4VF_SGE_BASE_ADDR + SGE_VF_GTS,
		     CIDXINC(0) |
		     SEINTARM(rspq->intr_params) |
		     INGRESSQID(rspq->cntxt_id));
}

/*
 * Enable NAPI scheduling and interrupt generation for all Receive Queues.
 */
static void enable_rx(struct adapter *adapter)
{
	int rxq;
	struct sge *s = &adapter->sge;

	for_each_ethrxq(s, rxq)
		qenable(&s->ethrxq[rxq].rspq);
	qenable(&s->fw_evtq);

	/*
	 * The interrupt queue doesn't use NAPI so we do the 0-increment of
	 * its Going To Sleep register here to get it started.
	 */
	if (adapter->flags & USING_MSI)
		t4_write_reg(adapter, T4VF_SGE_BASE_ADDR + SGE_VF_GTS,
			     CIDXINC(0) |
			     SEINTARM(s->intrq.intr_params) |
			     INGRESSQID(s->intrq.cntxt_id));

}

/*
 * Wait until all NAPI handlers are descheduled.
 */
static void quiesce_rx(struct adapter *adapter)
{
	struct sge *s = &adapter->sge;
	int rxq;

	for_each_ethrxq(s, rxq)
		napi_disable(&s->ethrxq[rxq].rspq.napi);
	napi_disable(&s->fw_evtq.napi);
}

/*
 * Response queue handler for the firmware event queue.
 */
static int fwevtq_handler(struct sge_rspq *rspq, const __be64 *rsp,
			  const struct pkt_gl *gl)
{
	/*
	 * Extract response opcode and get pointer to CPL message body.
	 */
	struct adapter *adapter = rspq->adapter;
	u8 opcode = ((const struct rss_header *)rsp)->opcode;
	void *cpl = (void *)(rsp + 1);

	switch (opcode) {
	case CPL_FW6_MSG: {
		/*
		 * We've received an asynchronous message from the firmware.
		 */
		const struct cpl_fw6_msg *fw_msg = cpl;
		if (fw_msg->type == FW6_TYPE_CMD_RPL)
			t4vf_handle_fw_rpl(adapter, fw_msg->data);
		break;
	}

	case CPL_SGE_EGR_UPDATE: {
		/*
		 * We've received an Egress Queue Status Update message.  We
		 * get these, if the SGE is configured to send these when the
		 * firmware passes certain points in processing our TX
		 * Ethernet Queue or if we make an explicit request for one.
		 * We use these updates to determine when we may need to
		 * restart a TX Ethernet Queue which was stopped for lack of
		 * free TX Queue Descriptors ...
		 */
		const struct cpl_sge_egr_update *p = (void *)cpl;
		unsigned int qid = EGR_QID(be32_to_cpu(p->opcode_qid));
		struct sge *s = &adapter->sge;
		struct sge_txq *tq;
		struct sge_eth_txq *txq;
		unsigned int eq_idx;

		/*
		 * Perform sanity checking on the Queue ID to make sure it
		 * really refers to one of our TX Ethernet Egress Queues which
		 * is active and matches the queue's ID.  None of these error
		 * conditions should ever happen so we may want to either make
		 * them fatal and/or conditionalized under DEBUG.
		 */
		eq_idx = EQ_IDX(s, qid);
		if (unlikely(eq_idx >= MAX_EGRQ)) {
			dev_err(adapter->pdev_dev,
				"Egress Update QID %d out of range\n", qid);
			break;
		}
		tq = s->egr_map[eq_idx];
		if (unlikely(tq == NULL)) {
			dev_err(adapter->pdev_dev,
				"Egress Update QID %d TXQ=NULL\n", qid);
			break;
		}
		txq = container_of(tq, struct sge_eth_txq, q);
		if (unlikely(tq->abs_id != qid)) {
			dev_err(adapter->pdev_dev,
				"Egress Update QID %d refers to TXQ %d\n",
				qid, tq->abs_id);
			break;
		}

		/*
		 * Restart a stopped TX Queue which has less than half of its
		 * TX ring in use ...
		 */
		txq->q.restarts++;
		netif_tx_wake_queue(txq->txq);
		break;
	}

	default:
		dev_err(adapter->pdev_dev,
			"unexpected CPL %#x on FW event queue\n", opcode);
	}

	return 0;
}

/*
 * Allocate SGE TX/RX response queues.  Determine how many sets of SGE queues
 * to use and initializes them.  We support multiple "Queue Sets" per port if
 * we have MSI-X, otherwise just one queue set per port.
 */
static int setup_sge_queues(struct adapter *adapter)
{
	struct sge *s = &adapter->sge;
	int err, pidx, msix;

	/*
	 * Clear "Queue Set" Free List Starving and TX Queue Mapping Error
	 * state.
	 */
	bitmap_zero(s->starving_fl, MAX_EGRQ);

	/*
	 * If we're using MSI interrupt mode we need to set up a "forwarded
	 * interrupt" queue which we'll set up with our MSI vector.  The rest
	 * of the ingress queues will be set up to forward their interrupts to
	 * this queue ...  This must be first since t4vf_sge_alloc_rxq() uses
	 * the intrq's queue ID as the interrupt forwarding queue for the
	 * subsequent calls ...
	 */
	if (adapter->flags & USING_MSI) {
		err = t4vf_sge_alloc_rxq(adapter, &s->intrq, false,
					 adapter->port[0], 0, NULL, NULL);
		if (err)
			goto err_free_queues;
	}

	/*
	 * Allocate our ingress queue for asynchronous firmware messages.
	 */
	err = t4vf_sge_alloc_rxq(adapter, &s->fw_evtq, true, adapter->port[0],
				 MSIX_FW, NULL, fwevtq_handler);
	if (err)
		goto err_free_queues;

	/*
	 * Allocate each "port"'s initial Queue Sets.  These can be changed
	 * later on ... up to the point where any interface on the adapter is
	 * brought up at which point lots of things get nailed down
	 * permanently ...
	 */
	msix = MSIX_NIQFLINT;
	for_each_port(adapter, pidx) {
		struct net_device *dev = adapter->port[pidx];
		struct port_info *pi = netdev_priv(dev);
		struct sge_eth_rxq *rxq = &s->ethrxq[pi->first_qset];
		struct sge_eth_txq *txq = &s->ethtxq[pi->first_qset];
		int qs;

		for (qs = 0; qs < pi->nqsets; qs++, rxq++, txq++) {
			err = t4vf_sge_alloc_rxq(adapter, &rxq->rspq, false,
						 dev, msix++,
						 &rxq->fl, t4vf_ethrx_handler);
			if (err)
				goto err_free_queues;

			err = t4vf_sge_alloc_eth_txq(adapter, txq, dev,
					     netdev_get_tx_queue(dev, qs),
					     s->fw_evtq.cntxt_id);
			if (err)
				goto err_free_queues;

			rxq->rspq.idx = qs;
			memset(&rxq->stats, 0, sizeof(rxq->stats));
		}
	}

	/*
	 * Create the reverse mappings for the queues.
	 */
	s->egr_base = s->ethtxq[0].q.abs_id - s->ethtxq[0].q.cntxt_id;
	s->ingr_base = s->ethrxq[0].rspq.abs_id - s->ethrxq[0].rspq.cntxt_id;
	IQ_MAP(s, s->fw_evtq.abs_id) = &s->fw_evtq;
	for_each_port(adapter, pidx) {
		struct net_device *dev = adapter->port[pidx];
		struct port_info *pi = netdev_priv(dev);
		struct sge_eth_rxq *rxq = &s->ethrxq[pi->first_qset];
		struct sge_eth_txq *txq = &s->ethtxq[pi->first_qset];
		int qs;

		for (qs = 0; qs < pi->nqsets; qs++, rxq++, txq++) {
			IQ_MAP(s, rxq->rspq.abs_id) = &rxq->rspq;
			EQ_MAP(s, txq->q.abs_id) = &txq->q;

			/*
			 * The FW_IQ_CMD doesn't return the Absolute Queue IDs
			 * for Free Lists but since all of the Egress Queues
			 * (including Free Lists) have Relative Queue IDs
			 * which are computed as Absolute - Base Queue ID, we
			 * can synthesize the Absolute Queue IDs for the Free
			 * Lists.  This is useful for debugging purposes when
			 * we want to dump Queue Contexts via the PF Driver.
			 */
			rxq->fl.abs_id = rxq->fl.cntxt_id + s->egr_base;
			EQ_MAP(s, rxq->fl.abs_id) = &rxq->fl;
		}
	}
	return 0;

err_free_queues:
	t4vf_free_sge_resources(adapter);
	return err;
}

/*
 * Set up Receive Side Scaling (RSS) to distribute packets to multiple receive
 * queues.  We configure the RSS CPU lookup table to distribute to the number
 * of HW receive queues, and the response queue lookup table to narrow that
 * down to the response queues actually configured for each "port" (Virtual
 * Interface).  We always configure the RSS mapping for all ports since the
 * mapping table has plenty of entries.
 */
static int setup_rss(struct adapter *adapter)
{
	int pidx;

	for_each_port(adapter, pidx) {
		struct port_info *pi = adap2pinfo(adapter, pidx);
		struct sge_eth_rxq *rxq = &adapter->sge.ethrxq[pi->first_qset];
		u16 rss[MAX_PORT_QSETS];
		int qs, err;

		for (qs = 0; qs < pi->nqsets; qs++)
			rss[qs] = rxq[qs].rspq.abs_id;

		err = t4vf_config_rss_range(adapter, pi->viid,
					    0, pi->rss_size, rss, pi->nqsets);
		if (err)
			return err;

		/*
		 * Perform Global RSS Mode-specific initialization.
		 */
		switch (adapter->params.rss.mode) {
		case FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL:
			/*
			 * If Tunnel All Lookup isn't specified in the global
			 * RSS Configuration, then we need to specify a
			 * default Ingress Queue for any ingress packets which
			 * aren't hashed.  We'll use our first ingress queue
			 * ...
			 */
			if (!adapter->params.rss.u.basicvirtual.tnlalllookup) {
				union rss_vi_config config;
				err = t4vf_read_rss_vi_config(adapter,
							      pi->viid,
							      &config);
				if (err)
					return err;
				config.basicvirtual.defaultq =
					rxq[0].rspq.abs_id;
				err = t4vf_write_rss_vi_config(adapter,
							       pi->viid,
							       &config);
				if (err)
					return err;
			}
			break;
		}
	}

	return 0;
}

/*
 * Bring the adapter up.  Called whenever we go from no "ports" open to having
 * one open.  This function performs the actions necessary to make an adapter
 * operational, such as completing the initialization of HW modules, and
 * enabling interrupts.  Must be called with the rtnl lock held.  (Note that
 * this is called "cxgb_up" in the PF Driver.)
 */
static int adapter_up(struct adapter *adapter)
{
	int err;

	/*
	 * If this is the first time we've been called, perform basic
	 * adapter setup.  Once we've done this, many of our adapter
	 * parameters can no longer be changed ...
	 */
	if ((adapter->flags & FULL_INIT_DONE) == 0) {
		err = setup_sge_queues(adapter);
		if (err)
			return err;
		err = setup_rss(adapter);
		if (err) {
			t4vf_free_sge_resources(adapter);
			return err;
		}

		if (adapter->flags & USING_MSIX)
			name_msix_vecs(adapter);
		adapter->flags |= FULL_INIT_DONE;
	}

	/*
	 * Acquire our interrupt resources.  We only support MSI-X and MSI.
	 */
	BUG_ON((adapter->flags & (USING_MSIX|USING_MSI)) == 0);
	if (adapter->flags & USING_MSIX)
		err = request_msix_queue_irqs(adapter);
	else
		err = request_irq(adapter->pdev->irq,
				  t4vf_intr_handler(adapter), 0,
				  adapter->name, adapter);
	if (err) {
		dev_err(adapter->pdev_dev, "request_irq failed, err %d\n",
			err);
		return err;
	}

	/*
	 * Enable NAPI ingress processing and return success.
	 */
	enable_rx(adapter);
	t4vf_sge_start(adapter);
	return 0;
}

/*
 * Bring the adapter down.  Called whenever the last "port" (Virtual
 * Interface) closed.  (Note that this routine is called "cxgb_down" in the PF
 * Driver.)
 */
static void adapter_down(struct adapter *adapter)
{
	/*
	 * Free interrupt resources.
	 */
	if (adapter->flags & USING_MSIX)
		free_msix_queue_irqs(adapter);
	else
		free_irq(adapter->pdev->irq, adapter);

	/*
	 * Wait for NAPI handlers to finish.
	 */
	quiesce_rx(adapter);
}

/*
 * Start up a net device.
 */
static int cxgb4vf_open(struct net_device *dev)
{
	int err;
	struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;

	/*
	 * If this is the first interface that we're opening on the "adapter",
	 * bring the "adapter" up now.
	 */
	if (adapter->open_device_map == 0) {
		err = adapter_up(adapter);
		if (err)
			return err;
	}

	/*
	 * Note that this interface is up and start everything up ...
	 */
	netif_set_real_num_tx_queues(dev, pi->nqsets);
	err = netif_set_real_num_rx_queues(dev, pi->nqsets);
	if (err)
		return err;
	set_bit(pi->port_id, &adapter->open_device_map);
	link_start(dev);
	netif_tx_start_all_queues(dev);
	return 0;
}

/*
 * Shut down a net device.  This routine is called "cxgb_close" in the PF
 * Driver ...
 */
static int cxgb4vf_stop(struct net_device *dev)
{
	int ret;
	struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;

	netif_tx_stop_all_queues(dev);
	netif_carrier_off(dev);
	ret = t4vf_enable_vi(adapter, pi->viid, false, false);
	pi->link_cfg.link_ok = 0;

	clear_bit(pi->port_id, &adapter->open_device_map);
	if (adapter->open_device_map == 0)
		adapter_down(adapter);
	return 0;
}

/*
 * Translate our basic statistics into the standard "ifconfig" statistics.
 */
static struct net_device_stats *cxgb4vf_get_stats(struct net_device *dev)
{
	struct t4vf_port_stats stats;
	struct port_info *pi = netdev2pinfo(dev);
	struct adapter *adapter = pi->adapter;
	struct net_device_stats *ns = &dev->stats;
	int err;

	spin_lock(&adapter->stats_lock);
	err = t4vf_get_port_stats(adapter, pi->pidx, &stats);
	spin_unlock(&adapter->stats_lock);

	memset(ns, 0, sizeof(*ns));
	if (err)
		return ns;

	ns->tx_bytes = (stats.tx_bcast_bytes + stats.tx_mcast_bytes +
			stats.tx_ucast_bytes + stats.tx_offload_bytes);
	ns->tx_packets = (stats.tx_bcast_frames + stats.tx_mcast_frames +
			  stats.tx_ucast_frames + stats.tx_offload_frames);
	ns->rx_bytes = (stats.rx_bcast_bytes + stats.rx_mcast_bytes +
			stats.rx_ucast_bytes);
	ns->rx_packets = (stats.rx_bcast_frames + stats.rx_mcast_frames +
			  stats.rx_ucast_frames);
	ns->multicast = stats.rx_mcast_frames;
	ns->tx_errors = stats.tx_drop_frames;
	ns->rx_errors = stats.rx_err_frames;

	return ns;
}

/*
 * Collect up to maxaddrs worth of a netdevice's unicast addresses into an
 * array of addrss pointers and return the number collected.
 */
static inline int collect_netdev_uc_list_addrs(const struct net_device *dev,
					       const u8 **addr,
					       unsigned int maxaddrs)
{
	unsigned int naddr = 0;
	const struct netdev_hw_addr *ha;

	for_each_dev_addr(dev, ha) {
		addr[naddr++] = ha->addr;
		if (naddr >= maxaddrs)
			break;
	}
	return naddr;
}

/*
 * Collect up to maxaddrs worth of a netdevice's multicast addresses into an
 * array of addrss pointers and return the number collected.
 */
static inline int collect_netdev_mc_list_addrs(const struct net_device *dev,
					       const u8 **addr,
					       unsigned int maxaddrs)
{
	unsigned int naddr = 0;
	const struct netdev_hw_addr *ha;

	netdev_for_each_mc_addr(ha, dev) {
		addr[naddr++] = ha->addr;
		if (naddr >= maxaddrs)
			break;
	}
	return naddr;
}

/*
 * Configure the exact and hash address filters to handle a port's multicast
 * and secondary unicast MAC addresses.
 */
static int set_addr_filters(const struct net_device *dev, bool sleep)
{
	u64 mhash = 0;
	u64 uhash = 0;
	bool free = true;
	u16 filt_idx[7];
	const u8 *addr[7];
	int ret, naddr = 0;
	const struct port_info *pi = netdev_priv(dev);

	/* first do the secondary unicast addresses */
	naddr = collect_netdev_uc_list_addrs(dev, addr, ARRAY_SIZE(addr));
	if (naddr > 0) {
		ret = t4vf_alloc_mac_filt(pi->adapter, pi->viid, free,
					  naddr, addr, filt_idx, &uhash, sleep);
		if (ret < 0)
			return ret;

		free = false;
	}

	/* next set up the multicast addresses */
	naddr = collect_netdev_mc_list_addrs(dev, addr, ARRAY_SIZE(addr));
	if (naddr > 0) {
		ret = t4vf_alloc_mac_filt(pi->adapter, pi->viid, free,
					  naddr, addr, filt_idx, &mhash, sleep);
		if (ret < 0)
			return ret;
	}

	return t4vf_set_addr_hash(pi->adapter, pi->viid, uhash != 0,
				  uhash | mhash, sleep);
}

/*
 * Set RX properties of a port, such as promiscruity, address filters, and MTU.
 * If @mtu is -1 it is left unchanged.
 */
static int set_rxmode(struct net_device *dev, int mtu, bool sleep_ok)
{
	int ret;
	struct port_info *pi = netdev_priv(dev);

	ret = set_addr_filters(dev, sleep_ok);
	if (ret == 0)
		ret = t4vf_set_rxmode(pi->adapter, pi->viid, -1,
				      (dev->flags & IFF_PROMISC) != 0,
				      (dev->flags & IFF_ALLMULTI) != 0,
				      1, -1, sleep_ok);
	return ret;
}

/*
 * Set the current receive modes on the device.
 */
static void cxgb4vf_set_rxmode(struct net_device *dev)
{
	/* unfortunately we can't return errors to the stack */
	set_rxmode(dev, -1, false);
}

/*
 * Find the entry in the interrupt holdoff timer value array which comes
 * closest to the specified interrupt holdoff value.
 */
static int closest_timer(const struct sge *s, int us)
{
	int i, timer_idx = 0, min_delta = INT_MAX;

	for (i = 0; i < ARRAY_SIZE(s->timer_val); i++) {
		int delta = us - s->timer_val[i];
		if (delta < 0)
			delta = -delta;
		if (delta < min_delta) {
			min_delta = delta;
			timer_idx = i;
		}
	}
	return timer_idx;
}

static int closest_thres(const struct sge *s, int thres)
{
	int i, delta, pktcnt_idx = 0, min_delta = INT_MAX;

	for (i = 0; i < ARRAY_SIZE(s->counter_val); i++) {
		delta = thres - s->counter_val[i];
		if (delta < 0)
			delta = -delta;
		if (delta < min_delta) {
			min_delta = delta;
			pktcnt_idx = i;
		}
	}
	return pktcnt_idx;
}

/*
 * Return a queue's interrupt hold-off time in us.  0 means no timer.
 */
static unsigned int qtimer_val(const struct adapter *adapter,
			       const struct sge_rspq *rspq)
{
	unsigned int timer_idx = QINTR_TIMER_IDX_GET(rspq->intr_params);

	return timer_idx < SGE_NTIMERS
		? adapter->sge.timer_val[timer_idx]
		: 0;
}

/**
 *	set_rxq_intr_params - set a queue's interrupt holdoff parameters
 *	@adapter: the adapter
 *	@rspq: the RX response queue
 *	@us: the hold-off time in us, or 0 to disable timer
 *	@cnt: the hold-off packet count, or 0 to disable counter
 *
 *	Sets an RX response queue's interrupt hold-off time and packet count.
 *	At least one of the two needs to be enabled for the queue to generate
 *	interrupts.
 */
static int set_rxq_intr_params(struct adapter *adapter, struct sge_rspq *rspq,
			       unsigned int us, unsigned int cnt)
{
	unsigned int timer_idx;

	/*
	 * If both the interrupt holdoff timer and count are specified as
	 * zero, default to a holdoff count of 1 ...
	 */
	if ((us | cnt) == 0)
		cnt = 1;

	/*
	 * If an interrupt holdoff count has been specified, then find the
	 * closest configured holdoff count and use that.  If the response
	 * queue has already been created, then update its queue context
	 * parameters ...
	 */
	if (cnt) {
		int err;
		u32 v, pktcnt_idx;

		pktcnt_idx = closest_thres(&adapter->sge, cnt);
		if (rspq->desc && rspq->pktcnt_idx != pktcnt_idx) {
			v = FW_PARAMS_MNEM(FW_PARAMS_MNEM_DMAQ) |
			    FW_PARAMS_PARAM_X(
					FW_PARAMS_PARAM_DMAQ_IQ_INTCNTTHRESH) |
			    FW_PARAMS_PARAM_YZ(rspq->cntxt_id);
			err = t4vf_set_params(adapter, 1, &v, &pktcnt_idx);
			if (err)
				return err;
		}
		rspq->pktcnt_idx = pktcnt_idx;
	}

	/*
	 * Compute the closest holdoff timer index from the supplied holdoff
	 * timer value.
	 */
	timer_idx = (us == 0
		     ? SGE_TIMER_RSTRT_CNTR
		     : closest_timer(&adapter->sge, us));

	/*
	 * Update the response queue's interrupt coalescing parameters and
	 * return success.
	 */
	rspq->intr_params = (QINTR_TIMER_IDX(timer_idx) |
			     (cnt > 0 ? QINTR_CNT_EN : 0));
	return 0;
}

/*
 * Return a version number to identify the type of adapter.  The scheme is:
 * - bits 0..9: chip version
 * - bits 10..15: chip revision
 */
static inline unsigned int mk_adap_vers(const struct adapter *adapter)
{
	/*
	 * Chip version 4, revision 0x3f (cxgb4vf).
	 */
	return 4 | (0x3f << 10);
}

/*
 * Execute the specified ioctl command.
 */
static int cxgb4vf_do_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
{
	int ret = 0;

	switch (cmd) {
	    /*
	     * The VF Driver doesn't have access to any of the other
	     * common Ethernet device ioctl()'s (like reading/writing
	     * PHY registers, etc.
	     */

	default:
		ret = -EOPNOTSUPP;
		break;
	}
	return ret;
}

/*
 * Change the device's MTU.
 */
static int cxgb4vf_change_mtu(struct net_device *dev, int new_mtu)
{
	int ret;
	struct port_info *pi = netdev_priv(dev);

	/* accommodate SACK */
	if (new_mtu < 81)
		return -EINVAL;

	ret = t4vf_set_rxmode(pi->adapter, pi->viid, new_mtu,
			      -1, -1, -1, -1, true);
	if (!ret)
		dev->mtu = new_mtu;
	return ret;
}

/*
 * Change the devices MAC address.
 */
static int cxgb4vf_set_mac_addr(struct net_device *dev, void *_addr)
{
	int ret;
	struct sockaddr *addr = _addr;
	struct port_info *pi = netdev_priv(dev);

	if (!is_valid_ether_addr(addr->sa_data))
		return -EINVAL;

	ret = t4vf_change_mac(pi->adapter, pi->viid, pi->xact_addr_filt,
			      addr->sa_data, true);
	if (ret < 0)
		return ret;

	memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
	pi->xact_addr_filt = ret;
	return 0;
}

#ifdef CONFIG_NET_POLL_CONTROLLER
/*
 * Poll all of our receive queues.  This is called outside of normal interrupt
 * context.
 */
static void cxgb4vf_poll_controller(struct net_device *dev)
{
	struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;

	if (adapter->flags & USING_MSIX) {
		struct sge_eth_rxq *rxq;
		int nqsets;

		rxq = &adapter->sge.ethrxq[pi->first_qset];
		for (nqsets = pi->nqsets; nqsets; nqsets--) {
			t4vf_sge_intr_msix(0, &rxq->rspq);
			rxq++;
		}
	} else
		t4vf_intr_handler(adapter)(0, adapter);
}
#endif

/*
 * Ethtool operations.
 * ===================
 *
 * Note that we don't support any ethtool operations which change the physical
 * state of the port to which we're linked.
 */

/*
 * Return current port link settings.
 */
static int cxgb4vf_get_settings(struct net_device *dev,
				struct ethtool_cmd *cmd)
{
	const struct port_info *pi = netdev_priv(dev);

	cmd->supported = pi->link_cfg.supported;
	cmd->advertising = pi->link_cfg.advertising;
	cmd->speed = netif_carrier_ok(dev) ? pi->link_cfg.speed : -1;
	cmd->duplex = DUPLEX_FULL;

	cmd->port = (cmd->supported & SUPPORTED_TP) ? PORT_TP : PORT_FIBRE;
	cmd->phy_address = pi->port_id;
	cmd->transceiver = XCVR_EXTERNAL;
	cmd->autoneg = pi->link_cfg.autoneg;
	cmd->maxtxpkt = 0;
	cmd->maxrxpkt = 0;
	return 0;
}

/*
 * Return our driver information.
 */
static void cxgb4vf_get_drvinfo(struct net_device *dev,
				struct ethtool_drvinfo *drvinfo)
{
	struct adapter *adapter = netdev2adap(dev);

	strcpy(drvinfo->driver, KBUILD_MODNAME);
	strcpy(drvinfo->version, DRV_VERSION);
	strcpy(drvinfo->bus_info, pci_name(to_pci_dev(dev->dev.parent)));
	snprintf(drvinfo->fw_version, sizeof(drvinfo->fw_version),
		 "%u.%u.%u.%u, TP %u.%u.%u.%u",
		 FW_HDR_FW_VER_MAJOR_GET(adapter->params.dev.fwrev),
		 FW_HDR_FW_VER_MINOR_GET(adapter->params.dev.fwrev),
		 FW_HDR_FW_VER_MICRO_GET(adapter->params.dev.fwrev),
		 FW_HDR_FW_VER_BUILD_GET(adapter->params.dev.fwrev),
		 FW_HDR_FW_VER_MAJOR_GET(adapter->params.dev.tprev),
		 FW_HDR_FW_VER_MINOR_GET(adapter->params.dev.tprev),
		 FW_HDR_FW_VER_MICRO_GET(adapter->params.dev.tprev),
		 FW_HDR_FW_VER_BUILD_GET(adapter->params.dev.tprev));
}

/*
 * Return current adapter message level.
 */
static u32 cxgb4vf_get_msglevel(struct net_device *dev)
{
	return netdev2adap(dev)->msg_enable;
}

/*
 * Set current adapter message level.
 */
static void cxgb4vf_set_msglevel(struct net_device *dev, u32 msglevel)
{
	netdev2adap(dev)->msg_enable = msglevel;
}

/*
 * Return the device's current Queue Set ring size parameters along with the
 * allowed maximum values.  Since ethtool doesn't understand the concept of
 * multi-queue devices, we just return the current values associated with the
 * first Queue Set.
 */
static void cxgb4vf_get_ringparam(struct net_device *dev,
				  struct ethtool_ringparam *rp)
{
	const struct port_info *pi = netdev_priv(dev);
	const struct sge *s = &pi->adapter->sge;

	rp->rx_max_pending = MAX_RX_BUFFERS;
	rp->rx_mini_max_pending = MAX_RSPQ_ENTRIES;
	rp->rx_jumbo_max_pending = 0;
	rp->tx_max_pending = MAX_TXQ_ENTRIES;

	rp->rx_pending = s->ethrxq[pi->first_qset].fl.size - MIN_FL_RESID;
	rp->rx_mini_pending = s->ethrxq[pi->first_qset].rspq.size;
	rp->rx_jumbo_pending = 0;
	rp->tx_pending = s->ethtxq[pi->first_qset].q.size;
}

/*
 * Set the Queue Set ring size parameters for the device.  Again, since
 * ethtool doesn't allow for the concept of multiple queues per device, we'll
 * apply these new values across all of the Queue Sets associated with the
 * device -- after vetting them of course!
 */
static int cxgb4vf_set_ringparam(struct net_device *dev,
				 struct ethtool_ringparam *rp)
{
	const struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;
	struct sge *s = &adapter->sge;
	int qs;

	if (rp->rx_pending > MAX_RX_BUFFERS ||
	    rp->rx_jumbo_pending ||
	    rp->tx_pending > MAX_TXQ_ENTRIES ||
	    rp->rx_mini_pending > MAX_RSPQ_ENTRIES ||
	    rp->rx_mini_pending < MIN_RSPQ_ENTRIES ||
	    rp->rx_pending < MIN_FL_ENTRIES ||
	    rp->tx_pending < MIN_TXQ_ENTRIES)
		return -EINVAL;

	if (adapter->flags & FULL_INIT_DONE)
		return -EBUSY;

	for (qs = pi->first_qset; qs < pi->first_qset + pi->nqsets; qs++) {
		s->ethrxq[qs].fl.size = rp->rx_pending + MIN_FL_RESID;
		s->ethrxq[qs].rspq.size = rp->rx_mini_pending;
		s->ethtxq[qs].q.size = rp->tx_pending;
	}
	return 0;
}

/*
 * Return the interrupt holdoff timer and count for the first Queue Set on the
 * device.  Our extension ioctl() (the cxgbtool interface) allows the
 * interrupt holdoff timer to be read on all of the device's Queue Sets.
 */
static int cxgb4vf_get_coalesce(struct net_device *dev,
				struct ethtool_coalesce *coalesce)
{
	const struct port_info *pi = netdev_priv(dev);
	const struct adapter *adapter = pi->adapter;
	const struct sge_rspq *rspq = &adapter->sge.ethrxq[pi->first_qset].rspq;

	coalesce->rx_coalesce_usecs = qtimer_val(adapter, rspq);
	coalesce->rx_max_coalesced_frames =
		((rspq->intr_params & QINTR_CNT_EN)
		 ? adapter->sge.counter_val[rspq->pktcnt_idx]
		 : 0);
	return 0;
}

/*
 * Set the RX interrupt holdoff timer and count for the first Queue Set on the
 * interface.  Our extension ioctl() (the cxgbtool interface) allows us to set
 * the interrupt holdoff timer on any of the device's Queue Sets.
 */
static int cxgb4vf_set_coalesce(struct net_device *dev,
				struct ethtool_coalesce *coalesce)
{
	const struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;

	return set_rxq_intr_params(adapter,
				   &adapter->sge.ethrxq[pi->first_qset].rspq,
				   coalesce->rx_coalesce_usecs,
				   coalesce->rx_max_coalesced_frames);
}

/*
 * Report current port link pause parameter settings.
 */
static void cxgb4vf_get_pauseparam(struct net_device *dev,
				   struct ethtool_pauseparam *pauseparam)
{
	struct port_info *pi = netdev_priv(dev);

	pauseparam->autoneg = (pi->link_cfg.requested_fc & PAUSE_AUTONEG) != 0;
	pauseparam->rx_pause = (pi->link_cfg.fc & PAUSE_RX) != 0;
	pauseparam->tx_pause = (pi->link_cfg.fc & PAUSE_TX) != 0;
}

/*
 * Return whether RX Checksum Offloading is currently enabled for the device.
 */
static u32 cxgb4vf_get_rx_csum(struct net_device *dev)
{
	struct port_info *pi = netdev_priv(dev);

	return (pi->rx_offload & RX_CSO) != 0;
}

/*
 * Turn RX Checksum Offloading on or off for the device.
 */
static int cxgb4vf_set_rx_csum(struct net_device *dev, u32 csum)
{
	struct port_info *pi = netdev_priv(dev);

	if (csum)
		pi->rx_offload |= RX_CSO;
	else
		pi->rx_offload &= ~RX_CSO;
	return 0;
}

/*
 * Identify the port by blinking the port's LED.
 */
static int cxgb4vf_phys_id(struct net_device *dev, u32 id)
{
	struct port_info *pi = netdev_priv(dev);

	return t4vf_identify_port(pi->adapter, pi->viid, 5);
}

/*
 * Port stats maintained per queue of the port.
 */
struct queue_port_stats {
	u64 tso;
	u64 tx_csum;
	u64 rx_csum;
	u64 vlan_ex;
	u64 vlan_ins;
};

/*
 * Strings for the ETH_SS_STATS statistics set ("ethtool -S").  Note that
 * these need to match the order of statistics returned by
 * t4vf_get_port_stats().
 */
static const char stats_strings[][ETH_GSTRING_LEN] = {
	/*
	 * These must match the layout of the t4vf_port_stats structure.
	 */
	"TxBroadcastBytes  ",
	"TxBroadcastFrames ",
	"TxMulticastBytes  ",
	"TxMulticastFrames ",
	"TxUnicastBytes    ",
	"TxUnicastFrames   ",
	"TxDroppedFrames   ",
	"TxOffloadBytes    ",
	"TxOffloadFrames   ",
	"RxBroadcastBytes  ",
	"RxBroadcastFrames ",
	"RxMulticastBytes  ",
	"RxMulticastFrames ",
	"RxUnicastBytes    ",
	"RxUnicastFrames   ",
	"RxErrorFrames     ",

	/*
	 * These are accumulated per-queue statistics and must match the
	 * order of the fields in the queue_port_stats structure.
	 */
	"TSO               ",
	"TxCsumOffload     ",
	"RxCsumGood        ",
	"VLANextractions   ",
	"VLANinsertions    ",
};

/*
 * Return the number of statistics in the specified statistics set.
 */
static int cxgb4vf_get_sset_count(struct net_device *dev, int sset)
{
	switch (sset) {
	case ETH_SS_STATS:
		return ARRAY_SIZE(stats_strings);
	default:
		return -EOPNOTSUPP;
	}
	/*NOTREACHED*/
}

/*
 * Return the strings for the specified statistics set.
 */
static void cxgb4vf_get_strings(struct net_device *dev,
				u32 sset,
				u8 *data)
{
	switch (sset) {
	case ETH_SS_STATS:
		memcpy(data, stats_strings, sizeof(stats_strings));
		break;
	}
}

/*
 * Small utility routine to accumulate queue statistics across the queues of
 * a "port".
 */
static void collect_sge_port_stats(const struct adapter *adapter,
				   const struct port_info *pi,
				   struct queue_port_stats *stats)
{
	const struct sge_eth_txq *txq = &adapter->sge.ethtxq[pi->first_qset];
	const struct sge_eth_rxq *rxq = &adapter->sge.ethrxq[pi->first_qset];
	int qs;

	memset(stats, 0, sizeof(*stats));
	for (qs = 0; qs < pi->nqsets; qs++, rxq++, txq++) {
		stats->tso += txq->tso;
		stats->tx_csum += txq->tx_cso;
		stats->rx_csum += rxq->stats.rx_cso;
		stats->vlan_ex += rxq->stats.vlan_ex;
		stats->vlan_ins += txq->vlan_ins;
	}
}

/*
 * Return the ETH_SS_STATS statistics set.
 */
static void cxgb4vf_get_ethtool_stats(struct net_device *dev,
				      struct ethtool_stats *stats,
				      u64 *data)
{
	struct port_info *pi = netdev2pinfo(dev);
	struct adapter *adapter = pi->adapter;
	int err = t4vf_get_port_stats(adapter, pi->pidx,
				      (struct t4vf_port_stats *)data);
	if (err)
		memset(data, 0, sizeof(struct t4vf_port_stats));

	data += sizeof(struct t4vf_port_stats) / sizeof(u64);
	collect_sge_port_stats(adapter, pi, (struct queue_port_stats *)data);
}

/*
 * Return the size of our register map.
 */
static int cxgb4vf_get_regs_len(struct net_device *dev)
{
	return T4VF_REGMAP_SIZE;
}

/*
 * Dump a block of registers, start to end inclusive, into a buffer.
 */
static void reg_block_dump(struct adapter *adapter, void *regbuf,
			   unsigned int start, unsigned int end)
{
	u32 *bp = regbuf + start - T4VF_REGMAP_START;

	for ( ; start <= end; start += sizeof(u32)) {
		/*
		 * Avoid reading the Mailbox Control register since that
		 * can trigger a Mailbox Ownership Arbitration cycle and
		 * interfere with communication with the firmware.
		 */
		if (start == T4VF_CIM_BASE_ADDR + CIM_VF_EXT_MAILBOX_CTRL)
			*bp++ = 0xffff;
		else
			*bp++ = t4_read_reg(adapter, start);
	}
}

/*
 * Copy our entire register map into the provided buffer.
 */
static void cxgb4vf_get_regs(struct net_device *dev,
			     struct ethtool_regs *regs,
			     void *regbuf)
{
	struct adapter *adapter = netdev2adap(dev);

	regs->version = mk_adap_vers(adapter);

	/*
	 * Fill in register buffer with our register map.
	 */
	memset(regbuf, 0, T4VF_REGMAP_SIZE);

	reg_block_dump(adapter, regbuf,
		       T4VF_SGE_BASE_ADDR + T4VF_MOD_MAP_SGE_FIRST,
		       T4VF_SGE_BASE_ADDR + T4VF_MOD_MAP_SGE_LAST);
	reg_block_dump(adapter, regbuf,
		       T4VF_MPS_BASE_ADDR + T4VF_MOD_MAP_MPS_FIRST,
		       T4VF_MPS_BASE_ADDR + T4VF_MOD_MAP_MPS_LAST);
	reg_block_dump(adapter, regbuf,
		       T4VF_PL_BASE_ADDR + T4VF_MOD_MAP_PL_FIRST,
		       T4VF_PL_BASE_ADDR + T4VF_MOD_MAP_PL_LAST);
	reg_block_dump(adapter, regbuf,
		       T4VF_CIM_BASE_ADDR + T4VF_MOD_MAP_CIM_FIRST,
		       T4VF_CIM_BASE_ADDR + T4VF_MOD_MAP_CIM_LAST);

	reg_block_dump(adapter, regbuf,
		       T4VF_MBDATA_BASE_ADDR + T4VF_MBDATA_FIRST,
		       T4VF_MBDATA_BASE_ADDR + T4VF_MBDATA_LAST);
}

/*
 * Report current Wake On LAN settings.
 */
static void cxgb4vf_get_wol(struct net_device *dev,
			    struct ethtool_wolinfo *wol)
{
	wol->supported = 0;
	wol->wolopts = 0;
	memset(&wol->sopass, 0, sizeof(wol->sopass));
}

/*
 * Set TCP Segmentation Offloading feature capabilities.
 */
static int cxgb4vf_set_tso(struct net_device *dev, u32 tso)
{
	if (tso)
		dev->features |= NETIF_F_TSO | NETIF_F_TSO6;
	else
		dev->features &= ~(NETIF_F_TSO | NETIF_F_TSO6);
	return 0;
}

static struct ethtool_ops cxgb4vf_ethtool_ops = {
	.get_settings		= cxgb4vf_get_settings,
	.get_drvinfo		= cxgb4vf_get_drvinfo,
	.get_msglevel		= cxgb4vf_get_msglevel,
	.set_msglevel		= cxgb4vf_set_msglevel,
	.get_ringparam		= cxgb4vf_get_ringparam,
	.set_ringparam		= cxgb4vf_set_ringparam,
	.get_coalesce		= cxgb4vf_get_coalesce,
	.set_coalesce		= cxgb4vf_set_coalesce,
	.get_pauseparam		= cxgb4vf_get_pauseparam,
	.get_rx_csum		= cxgb4vf_get_rx_csum,
	.set_rx_csum		= cxgb4vf_set_rx_csum,
	.set_tx_csum		= ethtool_op_set_tx_ipv6_csum,
	.set_sg			= ethtool_op_set_sg,
	.get_link		= ethtool_op_get_link,
	.get_strings		= cxgb4vf_get_strings,
	.phys_id		= cxgb4vf_phys_id,
	.get_sset_count		= cxgb4vf_get_sset_count,
	.get_ethtool_stats	= cxgb4vf_get_ethtool_stats,
	.get_regs_len		= cxgb4vf_get_regs_len,
	.get_regs		= cxgb4vf_get_regs,
	.get_wol		= cxgb4vf_get_wol,
	.set_tso		= cxgb4vf_set_tso,
};

/*
 * /sys/kernel/debug/cxgb4vf support code and data.
 * ================================================
 */

/*
 * Show SGE Queue Set information.  We display QPL Queues Sets per line.
 */
#define QPL	4

static int sge_qinfo_show(struct seq_file *seq, void *v)
{
	struct adapter *adapter = seq->private;
	int eth_entries = DIV_ROUND_UP(adapter->sge.ethqsets, QPL);
	int qs, r = (uintptr_t)v - 1;

	if (r)
		seq_putc(seq, '\n');

	#define S3(fmt_spec, s, v) \
		do {\
			seq_printf(seq, "%-12s", s); \
			for (qs = 0; qs < n; ++qs) \
				seq_printf(seq, " %16" fmt_spec, v); \
			seq_putc(seq, '\n'); \
		} while (0)
	#define S(s, v)		S3("s", s, v)
	#define T(s, v)		S3("u", s, txq[qs].v)
	#define R(s, v)		S3("u", s, rxq[qs].v)

	if (r < eth_entries) {
		const struct sge_eth_rxq *rxq = &adapter->sge.ethrxq[r * QPL];
		const struct sge_eth_txq *txq = &adapter->sge.ethtxq[r * QPL];
		int n = min(QPL, adapter->sge.ethqsets - QPL * r);

		S("QType:", "Ethernet");
		S("Interface:",
		  (rxq[qs].rspq.netdev
		   ? rxq[qs].rspq.netdev->name
		   : "N/A"));
		S3("d", "Port:",
		   (rxq[qs].rspq.netdev
		    ? ((struct port_info *)
		       netdev_priv(rxq[qs].rspq.netdev))->port_id
		    : -1));
		T("TxQ ID:", q.abs_id);
		T("TxQ size:", q.size);
		T("TxQ inuse:", q.in_use);
		T("TxQ PIdx:", q.pidx);
		T("TxQ CIdx:", q.cidx);
		R("RspQ ID:", rspq.abs_id);
		R("RspQ size:", rspq.size);
		R("RspQE size:", rspq.iqe_len);
		S3("u", "Intr delay:", qtimer_val(adapter, &rxq[qs].rspq));
		S3("u", "Intr pktcnt:",
		   adapter->sge.counter_val[rxq[qs].rspq.pktcnt_idx]);
		R("RspQ CIdx:", rspq.cidx);
		R("RspQ Gen:", rspq.gen);
		R("FL ID:", fl.abs_id);
		R("FL size:", fl.size - MIN_FL_RESID);
		R("FL avail:", fl.avail);
		R("FL PIdx:", fl.pidx);
		R("FL CIdx:", fl.cidx);
		return 0;
	}

	r -= eth_entries;
	if (r == 0) {
		const struct sge_rspq *evtq = &adapter->sge.fw_evtq;

		seq_printf(seq, "%-12s %16s\n", "QType:", "FW event queue");
		seq_printf(seq, "%-12s %16u\n", "RspQ ID:", evtq->abs_id);
		seq_printf(seq, "%-12s %16u\n", "Intr delay:",
			   qtimer_val(adapter, evtq));
		seq_printf(seq, "%-12s %16u\n", "Intr pktcnt:",
			   adapter->sge.counter_val[evtq->pktcnt_idx]);
		seq_printf(seq, "%-12s %16u\n", "RspQ Cidx:", evtq->cidx);
		seq_printf(seq, "%-12s %16u\n", "RspQ Gen:", evtq->gen);
	} else if (r == 1) {
		const struct sge_rspq *intrq = &adapter->sge.intrq;

		seq_printf(seq, "%-12s %16s\n", "QType:", "Interrupt Queue");
		seq_printf(seq, "%-12s %16u\n", "RspQ ID:", intrq->abs_id);
		seq_printf(seq, "%-12s %16u\n", "Intr delay:",
			   qtimer_val(adapter, intrq));
		seq_printf(seq, "%-12s %16u\n", "Intr pktcnt:",
			   adapter->sge.counter_val[intrq->pktcnt_idx]);
		seq_printf(seq, "%-12s %16u\n", "RspQ Cidx:", intrq->cidx);
		seq_printf(seq, "%-12s %16u\n", "RspQ Gen:", intrq->gen);
	}

	#undef R
	#undef T
	#undef S
	#undef S3

	return 0;
}

/*
 * Return the number of "entries" in our "file".  We group the multi-Queue
 * sections with QPL Queue Sets per "entry".  The sections of the output are:
 *
 *     Ethernet RX/TX Queue Sets
 *     Firmware Event Queue
 *     Forwarded Interrupt Queue (if in MSI mode)
 */
static int sge_queue_entries(const struct adapter *adapter)
{
	return DIV_ROUND_UP(adapter->sge.ethqsets, QPL) + 1 +
		((adapter->flags & USING_MSI) != 0);
}

static void *sge_queue_start(struct seq_file *seq, loff_t *pos)
{
	int entries = sge_queue_entries(seq->private);

	return *pos < entries ? (void *)((uintptr_t)*pos + 1) : NULL;
}

static void sge_queue_stop(struct seq_file *seq, void *v)
{
}

static void *sge_queue_next(struct seq_file *seq, void *v, loff_t *pos)
{
	int entries = sge_queue_entries(seq->private);

	++*pos;
	return *pos < entries ? (void *)((uintptr_t)*pos + 1) : NULL;
}

static const struct seq_operations sge_qinfo_seq_ops = {
	.start = sge_queue_start,
	.next  = sge_queue_next,
	.stop  = sge_queue_stop,
	.show  = sge_qinfo_show
};

static int sge_qinfo_open(struct inode *inode, struct file *file)
{
	int res = seq_open(file, &sge_qinfo_seq_ops);

	if (!res) {
		struct seq_file *seq = file->private_data;
		seq->private = inode->i_private;
	}
	return res;
}

static const struct file_operations sge_qinfo_debugfs_fops = {
	.owner   = THIS_MODULE,
	.open    = sge_qinfo_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release,
};

/*
 * Show SGE Queue Set statistics.  We display QPL Queues Sets per line.
 */
#define QPL	4

static int sge_qstats_show(struct seq_file *seq, void *v)
{
	struct adapter *adapter = seq->private;
	int eth_entries = DIV_ROUND_UP(adapter->sge.ethqsets, QPL);
	int qs, r = (uintptr_t)v - 1;

	if (r)
		seq_putc(seq, '\n');

	#define S3(fmt, s, v) \
		do { \
			seq_printf(seq, "%-16s", s); \
			for (qs = 0; qs < n; ++qs) \
				seq_printf(seq, " %8" fmt, v); \
			seq_putc(seq, '\n'); \
		} while (0)
	#define S(s, v)		S3("s", s, v)

	#define T3(fmt, s, v)	S3(fmt, s, txq[qs].v)
	#define T(s, v)		T3("lu", s, v)

	#define R3(fmt, s, v)	S3(fmt, s, rxq[qs].v)
	#define R(s, v)		R3("lu", s, v)

	if (r < eth_entries) {
		const struct sge_eth_rxq *rxq = &adapter->sge.ethrxq[r * QPL];
		const struct sge_eth_txq *txq = &adapter->sge.ethtxq[r * QPL];
		int n = min(QPL, adapter->sge.ethqsets - QPL * r);

		S("QType:", "Ethernet");
		S("Interface:",
		  (rxq[qs].rspq.netdev
		   ? rxq[qs].rspq.netdev->name
		   : "N/A"));
		R3("u", "RspQNullInts:", rspq.unhandled_irqs);
		R("RxPackets:", stats.pkts);
		R("RxCSO:", stats.rx_cso);
		R("VLANxtract:", stats.vlan_ex);
		R("LROmerged:", stats.lro_merged);
		R("LROpackets:", stats.lro_pkts);
		R("RxDrops:", stats.rx_drops);
		T("TSO:", tso);
		T("TxCSO:", tx_cso);
		T("VLANins:", vlan_ins);
		T("TxQFull:", q.stops);
		T("TxQRestarts:", q.restarts);
		T("TxMapErr:", mapping_err);
		R("FLAllocErr:", fl.alloc_failed);
		R("FLLrgAlcErr:", fl.large_alloc_failed);
		R("FLStarving:", fl.starving);
		return 0;
	}

	r -= eth_entries;
	if (r == 0) {
		const struct sge_rspq *evtq = &adapter->sge.fw_evtq;

		seq_printf(seq, "%-8s %16s\n", "QType:", "FW event queue");
		seq_printf(seq, "%-16s %8u\n", "RspQNullInts:",
			   evtq->unhandled_irqs);
		seq_printf(seq, "%-16s %8u\n", "RspQ CIdx:", evtq->cidx);
		seq_printf(seq, "%-16s %8u\n", "RspQ Gen:", evtq->gen);
	} else if (r == 1) {
		const struct sge_rspq *intrq = &adapter->sge.intrq;

		seq_printf(seq, "%-8s %16s\n", "QType:", "Interrupt Queue");
		seq_printf(seq, "%-16s %8u\n", "RspQNullInts:",
			   intrq->unhandled_irqs);
		seq_printf(seq, "%-16s %8u\n", "RspQ CIdx:", intrq->cidx);
		seq_printf(seq, "%-16s %8u\n", "RspQ Gen:", intrq->gen);
	}

	#undef R
	#undef T
	#undef S
	#undef R3
	#undef T3
	#undef S3

	return 0;
}

/*
 * Return the number of "entries" in our "file".  We group the multi-Queue
 * sections with QPL Queue Sets per "entry".  The sections of the output are:
 *
 *     Ethernet RX/TX Queue Sets
 *     Firmware Event Queue
 *     Forwarded Interrupt Queue (if in MSI mode)
 */
static int sge_qstats_entries(const struct adapter *adapter)
{
	return DIV_ROUND_UP(adapter->sge.ethqsets, QPL) + 1 +
		((adapter->flags & USING_MSI) != 0);
}

static void *sge_qstats_start(struct seq_file *seq, loff_t *pos)
{
	int entries = sge_qstats_entries(seq->private);

	return *pos < entries ? (void *)((uintptr_t)*pos + 1) : NULL;
}

static void sge_qstats_stop(struct seq_file *seq, void *v)
{
}

static void *sge_qstats_next(struct seq_file *seq, void *v, loff_t *pos)
{
	int entries = sge_qstats_entries(seq->private);

	(*pos)++;
	return *pos < entries ? (void *)((uintptr_t)*pos + 1) : NULL;
}

static const struct seq_operations sge_qstats_seq_ops = {
	.start = sge_qstats_start,
	.next  = sge_qstats_next,
	.stop  = sge_qstats_stop,
	.show  = sge_qstats_show
};

static int sge_qstats_open(struct inode *inode, struct file *file)
{
	int res = seq_open(file, &sge_qstats_seq_ops);

	if (res == 0) {
		struct seq_file *seq = file->private_data;
		seq->private = inode->i_private;
	}
	return res;
}

static const struct file_operations sge_qstats_proc_fops = {
	.owner   = THIS_MODULE,
	.open    = sge_qstats_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release,
};

/*
 * Show PCI-E SR-IOV Virtual Function Resource Limits.
 */
static int resources_show(struct seq_file *seq, void *v)
{
	struct adapter *adapter = seq->private;
	struct vf_resources *vfres = &adapter->params.vfres;

	#define S(desc, fmt, var) \
		seq_printf(seq, "%-60s " fmt "\n", \
			   desc " (" #var "):", vfres->var)

	S("Virtual Interfaces", "%d", nvi);
	S("Egress Queues", "%d", neq);
	S("Ethernet Control", "%d", nethctrl);
	S("Ingress Queues/w Free Lists/Interrupts", "%d", niqflint);
	S("Ingress Queues", "%d", niq);
	S("Traffic Class", "%d", tc);
	S("Port Access Rights Mask", "%#x", pmask);
	S("MAC Address Filters", "%d", nexactf);
	S("Firmware Command Read Capabilities", "%#x", r_caps);
	S("Firmware Command Write/Execute Capabilities", "%#x", wx_caps);

	#undef S

	return 0;
}

static int resources_open(struct inode *inode, struct file *file)
{
	return single_open(file, resources_show, inode->i_private);
}

static const struct file_operations resources_proc_fops = {
	.owner   = THIS_MODULE,
	.open    = resources_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = single_release,
};

/*
 * Show Virtual Interfaces.
 */
static int interfaces_show(struct seq_file *seq, void *v)
{
	if (v == SEQ_START_TOKEN) {
		seq_puts(seq, "Interface  Port   VIID\n");
	} else {
		struct adapter *adapter = seq->private;
		int pidx = (uintptr_t)v - 2;
		struct net_device *dev = adapter->port[pidx];
		struct port_info *pi = netdev_priv(dev);

		seq_printf(seq, "%9s  %4d  %#5x\n",
			   dev->name, pi->port_id, pi->viid);
	}
	return 0;
}

static inline void *interfaces_get_idx(struct adapter *adapter, loff_t pos)
{
	return pos <= adapter->params.nports
		? (void *)(uintptr_t)(pos + 1)
		: NULL;
}

static void *interfaces_start(struct seq_file *seq, loff_t *pos)
{
	return *pos
		? interfaces_get_idx(seq->private, *pos)
		: SEQ_START_TOKEN;
}

static void *interfaces_next(struct seq_file *seq, void *v, loff_t *pos)
{
	(*pos)++;
	return interfaces_get_idx(seq->private, *pos);
}

static void interfaces_stop(struct seq_file *seq, void *v)
{
}

static const struct seq_operations interfaces_seq_ops = {
	.start = interfaces_start,
	.next  = interfaces_next,
	.stop  = interfaces_stop,
	.show  = interfaces_show
};

static int interfaces_open(struct inode *inode, struct file *file)
{
	int res = seq_open(file, &interfaces_seq_ops);

	if (res == 0) {
		struct seq_file *seq = file->private_data;
		seq->private = inode->i_private;
	}
	return res;
}

static const struct file_operations interfaces_proc_fops = {
	.owner   = THIS_MODULE,
	.open    = interfaces_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release,
};

/*
 * /sys/kernel/debugfs/cxgb4vf/ files list.
 */
struct cxgb4vf_debugfs_entry {
	const char *name;		/* name of debugfs node */
	mode_t mode;			/* file system mode */
	const struct file_operations *fops;
};

static struct cxgb4vf_debugfs_entry debugfs_files[] = {
	{ "sge_qinfo",  S_IRUGO, &sge_qinfo_debugfs_fops },
	{ "sge_qstats", S_IRUGO, &sge_qstats_proc_fops },
	{ "resources",  S_IRUGO, &resources_proc_fops },
	{ "interfaces", S_IRUGO, &interfaces_proc_fops },
};

/*
 * Module and device initialization and cleanup code.
 * ==================================================
 */

/*
 * Set up out /sys/kernel/debug/cxgb4vf sub-nodes.  We assume that the
 * directory (debugfs_root) has already been set up.
 */
static int __devinit setup_debugfs(struct adapter *adapter)
{
	int i;

	BUG_ON(adapter->debugfs_root == NULL);

	/*
	 * Debugfs support is best effort.
	 */
	for (i = 0; i < ARRAY_SIZE(debugfs_files); i++)
		(void)debugfs_create_file(debugfs_files[i].name,
				  debugfs_files[i].mode,
				  adapter->debugfs_root,
				  (void *)adapter,
				  debugfs_files[i].fops);

	return 0;
}

/*
 * Tear down the /sys/kernel/debug/cxgb4vf sub-nodes created above.  We leave
 * it to our caller to tear down the directory (debugfs_root).
 */
static void __devexit cleanup_debugfs(struct adapter *adapter)
{
	BUG_ON(adapter->debugfs_root == NULL);

	/*
	 * Unlike our sister routine cleanup_proc(), we don't need to remove
	 * individual entries because a call will be made to
	 * debugfs_remove_recursive().  We just need to clean up any ancillary
	 * persistent state.
	 */
	/* nothing to do */
}

/*
 * Perform early "adapter" initialization.  This is where we discover what
 * adapter parameters we're going to be using and initialize basic adapter
 * hardware support.
 */
static int adap_init0(struct adapter *adapter)
{
	struct vf_resources *vfres = &adapter->params.vfres;
	struct sge_params *sge_params = &adapter->params.sge;
	struct sge *s = &adapter->sge;
	unsigned int ethqsets;
	int err;

	/*
	 * Wait for the device to become ready before proceeding ...
	 */
	err = t4vf_wait_dev_ready(adapter);
	if (err) {
		dev_err(adapter->pdev_dev, "device didn't become ready:"
			" err=%d\n", err);
		return err;
	}

	/*
	 * Grab basic operational parameters.  These will predominantly have
	 * been set up by the Physical Function Driver or will be hard coded
	 * into the adapter.  We just have to live with them ...  Note that
	 * we _must_ get our VPD parameters before our SGE parameters because
	 * we need to know the adapter's core clock from the VPD in order to
	 * properly decode the SGE Timer Values.
	 */
	err = t4vf_get_dev_params(adapter);
	if (err) {
		dev_err(adapter->pdev_dev, "unable to retrieve adapter"
			" device parameters: err=%d\n", err);
		return err;
	}
	err = t4vf_get_vpd_params(adapter);
	if (err) {
		dev_err(adapter->pdev_dev, "unable to retrieve adapter"
			" VPD parameters: err=%d\n", err);
		return err;
	}
	err = t4vf_get_sge_params(adapter);
	if (err) {
		dev_err(adapter->pdev_dev, "unable to retrieve adapter"
			" SGE parameters: err=%d\n", err);
		return err;
	}
	err = t4vf_get_rss_glb_config(adapter);
	if (err) {
		dev_err(adapter->pdev_dev, "unable to retrieve adapter"
			" RSS parameters: err=%d\n", err);
		return err;
	}
	if (adapter->params.rss.mode !=
	    FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL) {
		dev_err(adapter->pdev_dev, "unable to operate with global RSS"
			" mode %d\n", adapter->params.rss.mode);
		return -EINVAL;
	}
	err = t4vf_sge_init(adapter);
	if (err) {
		dev_err(adapter->pdev_dev, "unable to use adapter parameters:"
			" err=%d\n", err);
		return err;
	}

	/*
	 * Retrieve our RX interrupt holdoff timer values and counter
	 * threshold values from the SGE parameters.
	 */
	s->timer_val[0] = core_ticks_to_us(adapter,
		TIMERVALUE0_GET(sge_params->sge_timer_value_0_and_1));
	s->timer_val[1] = core_ticks_to_us(adapter,
		TIMERVALUE1_GET(sge_params->sge_timer_value_0_and_1));
	s->timer_val[2] = core_ticks_to_us(adapter,
		TIMERVALUE0_GET(sge_params->sge_timer_value_2_and_3));
	s->timer_val[3] = core_ticks_to_us(adapter,
		TIMERVALUE1_GET(sge_params->sge_timer_value_2_and_3));
	s->timer_val[4] = core_ticks_to_us(adapter,
		TIMERVALUE0_GET(sge_params->sge_timer_value_4_and_5));
	s->timer_val[5] = core_ticks_to_us(adapter,
		TIMERVALUE1_GET(sge_params->sge_timer_value_4_and_5));

	s->counter_val[0] =
		THRESHOLD_0_GET(sge_params->sge_ingress_rx_threshold);
	s->counter_val[1] =
		THRESHOLD_1_GET(sge_params->sge_ingress_rx_threshold);
	s->counter_val[2] =
		THRESHOLD_2_GET(sge_params->sge_ingress_rx_threshold);
	s->counter_val[3] =
		THRESHOLD_3_GET(sge_params->sge_ingress_rx_threshold);

	/*
	 * Grab our Virtual Interface resource allocation, extract the
	 * features that we're interested in and do a bit of sanity testing on
	 * what we discover.
	 */
	err = t4vf_get_vfres(adapter);
	if (err) {
		dev_err(adapter->pdev_dev, "unable to get virtual interface"
			" resources: err=%d\n", err);
		return err;
	}

	/*
	 * The number of "ports" which we support is equal to the number of
	 * Virtual Interfaces with which we've been provisioned.
	 */
	adapter->params.nports = vfres->nvi;
	if (adapter->params.nports > MAX_NPORTS) {
		dev_warn(adapter->pdev_dev, "only using %d of %d allowed"
			 " virtual interfaces\n", MAX_NPORTS,
			 adapter->params.nports);
		adapter->params.nports = MAX_NPORTS;
	}

	/*
	 * We need to reserve a number of the ingress queues with Free List
	 * and Interrupt capabilities for special interrupt purposes (like
	 * asynchronous firmware messages, or forwarded interrupts if we're
	 * using MSI).  The rest of the FL/Intr-capable ingress queues will be
	 * matched up one-for-one with Ethernet/Control egress queues in order
	 * to form "Queue Sets" which will be aportioned between the "ports".
	 * For each Queue Set, we'll need the ability to allocate two Egress
	 * Contexts -- one for the Ingress Queue Free List and one for the TX
	 * Ethernet Queue.
	 */
	ethqsets = vfres->niqflint - INGQ_EXTRAS;
	if (vfres->nethctrl != ethqsets) {
		dev_warn(adapter->pdev_dev, "unequal number of [available]"
			 " ingress/egress queues (%d/%d); using minimum for"
			 " number of Queue Sets\n", ethqsets, vfres->nethctrl);
		ethqsets = min(vfres->nethctrl, ethqsets);
	}
	if (vfres->neq < ethqsets*2) {
		dev_warn(adapter->pdev_dev, "Not enough Egress Contexts (%d)"
			 " to support Queue Sets (%d); reducing allowed Queue"
			 " Sets\n", vfres->neq, ethqsets);
		ethqsets = vfres->neq/2;
	}
	if (ethqsets > MAX_ETH_QSETS) {
		dev_warn(adapter->pdev_dev, "only using %d of %d allowed Queue"
			 " Sets\n", MAX_ETH_QSETS, adapter->sge.max_ethqsets);
		ethqsets = MAX_ETH_QSETS;
	}
	if (vfres->niq != 0 || vfres->neq > ethqsets*2) {
		dev_warn(adapter->pdev_dev, "unused resources niq/neq (%d/%d)"
			 " ignored\n", vfres->niq, vfres->neq - ethqsets*2);
	}
	adapter->sge.max_ethqsets = ethqsets;

	/*
	 * Check for various parameter sanity issues.  Most checks simply
	 * result in us using fewer resources than our provissioning but we
	 * do need at least  one "port" with which to work ...
	 */
	if (adapter->sge.max_ethqsets < adapter->params.nports) {
		dev_warn(adapter->pdev_dev, "only using %d of %d available"
			 " virtual interfaces (too few Queue Sets)\n",
			 adapter->sge.max_ethqsets, adapter->params.nports);
		adapter->params.nports = adapter->sge.max_ethqsets;
	}
	if (adapter->params.nports == 0) {
		dev_err(adapter->pdev_dev, "no virtual interfaces configured/"
			"usable!\n");
		return -EINVAL;
	}
	return 0;
}

static inline void init_rspq(struct sge_rspq *rspq, u8 timer_idx,
			     u8 pkt_cnt_idx, unsigned int size,
			     unsigned int iqe_size)
{
	rspq->intr_params = (QINTR_TIMER_IDX(timer_idx) |
			     (pkt_cnt_idx < SGE_NCOUNTERS ? QINTR_CNT_EN : 0));
	rspq->pktcnt_idx = (pkt_cnt_idx < SGE_NCOUNTERS
			    ? pkt_cnt_idx
			    : 0);
	rspq->iqe_len = iqe_size;
	rspq->size = size;
}

/*
 * Perform default configuration of DMA queues depending on the number and
 * type of ports we found and the number of available CPUs.  Most settings can
 * be modified by the admin via ethtool and cxgbtool prior to the adapter
 * being brought up for the first time.
 */
static void __devinit cfg_queues(struct adapter *adapter)
{
	struct sge *s = &adapter->sge;
	int q10g, n10g, qidx, pidx, qs;

	/*
	 * We should not be called till we know how many Queue Sets we can
	 * support.  In particular, this means that we need to know what kind
	 * of interrupts we'll be using ...
	 */
	BUG_ON((adapter->flags & (USING_MSIX|USING_MSI)) == 0);

	/*
	 * Count the number of 10GbE Virtual Interfaces that we have.
	 */
	n10g = 0;
	for_each_port(adapter, pidx)
		n10g += is_10g_port(&adap2pinfo(adapter, pidx)->link_cfg);

	/*
	 * We default to 1 queue per non-10G port and up to # of cores queues
	 * per 10G port.
	 */
	if (n10g == 0)
		q10g = 0;
	else {
		int n1g = (adapter->params.nports - n10g);
		q10g = (adapter->sge.max_ethqsets - n1g) / n10g;
		if (q10g > num_online_cpus())
			q10g = num_online_cpus();
	}

	/*
	 * Allocate the "Queue Sets" to the various Virtual Interfaces.
	 * The layout will be established in setup_sge_queues() when the
	 * adapter is brough up for the first time.
	 */
	qidx = 0;
	for_each_port(adapter, pidx) {
		struct port_info *pi = adap2pinfo(adapter, pidx);

		pi->first_qset = qidx;
		pi->nqsets = is_10g_port(&pi->link_cfg) ? q10g : 1;
		qidx += pi->nqsets;
	}
	s->ethqsets = qidx;

	/*
	 * Set up default Queue Set parameters ...  Start off with the
	 * shortest interrupt holdoff timer.
	 */
	for (qs = 0; qs < s->max_ethqsets; qs++) {
		struct sge_eth_rxq *rxq = &s->ethrxq[qs];
		struct sge_eth_txq *txq = &s->ethtxq[qs];

		init_rspq(&rxq->rspq, 0, 0, 1024, L1_CACHE_BYTES);
		rxq->fl.size = 72;
		txq->q.size = 1024;
	}

	/*
	 * The firmware event queue is used for link state changes and
	 * notifications of TX DMA completions.
	 */
	init_rspq(&s->fw_evtq, SGE_TIMER_RSTRT_CNTR, 0, 512,
		  L1_CACHE_BYTES);

	/*
	 * The forwarded interrupt queue is used when we're in MSI interrupt
	 * mode.  In this mode all interrupts associated with RX queues will
	 * be forwarded to a single queue which we'll associate with our MSI
	 * interrupt vector.  The messages dropped in the forwarded interrupt
	 * queue will indicate which ingress queue needs servicing ...  This
	 * queue needs to be large enough to accommodate all of the ingress
	 * queues which are forwarding their interrupt (+1 to prevent the PIDX
	 * from equalling the CIDX if every ingress queue has an outstanding
	 * interrupt).  The queue doesn't need to be any larger because no
	 * ingress queue will ever have more than one outstanding interrupt at
	 * any time ...
	 */
	init_rspq(&s->intrq, SGE_TIMER_RSTRT_CNTR, 0, MSIX_ENTRIES + 1,
		  L1_CACHE_BYTES);
}

/*
 * Reduce the number of Ethernet queues across all ports to at most n.
 * n provides at least one queue per port.
 */
static void __devinit reduce_ethqs(struct adapter *adapter, int n)
{
	int i;
	struct port_info *pi;

	/*
	 * While we have too many active Ether Queue Sets, interate across the
	 * "ports" and reduce their individual Queue Set allocations.
	 */
	BUG_ON(n < adapter->params.nports);
	while (n < adapter->sge.ethqsets)
		for_each_port(adapter, i) {
			pi = adap2pinfo(adapter, i);
			if (pi->nqsets > 1) {
				pi->nqsets--;
				adapter->sge.ethqsets--;
				if (adapter->sge.ethqsets <= n)
					break;
			}
		}

	/*
	 * Reassign the starting Queue Sets for each of the "ports" ...
	 */
	n = 0;
	for_each_port(adapter, i) {
		pi = adap2pinfo(adapter, i);
		pi->first_qset = n;
		n += pi->nqsets;
	}
}

/*
 * We need to grab enough MSI-X vectors to cover our interrupt needs.  Ideally
 * we get a separate MSI-X vector for every "Queue Set" plus any extras we
 * need.  Minimally we need one for every Virtual Interface plus those needed
 * for our "extras".  Note that this process may lower the maximum number of
 * allowed Queue Sets ...
 */
static int __devinit enable_msix(struct adapter *adapter)
{
	int i, err, want, need;
	struct msix_entry entries[MSIX_ENTRIES];
	struct sge *s = &adapter->sge;

	for (i = 0; i < MSIX_ENTRIES; ++i)
		entries[i].entry = i;

	/*
	 * We _want_ enough MSI-X interrupts to cover all of our "Queue Sets"
	 * plus those needed for our "extras" (for example, the firmware
	 * message queue).  We _need_ at least one "Queue Set" per Virtual
	 * Interface plus those needed for our "extras".  So now we get to see
	 * if the song is right ...
	 */
	want = s->max_ethqsets + MSIX_EXTRAS;
	need = adapter->params.nports + MSIX_EXTRAS;
	while ((err = pci_enable_msix(adapter->pdev, entries, want)) >= need)
		want = err;

	if (err == 0) {
		int nqsets = want - MSIX_EXTRAS;
		if (nqsets < s->max_ethqsets) {
			dev_warn(adapter->pdev_dev, "only enough MSI-X vectors"
				 " for %d Queue Sets\n", nqsets);
			s->max_ethqsets = nqsets;
			if (nqsets < s->ethqsets)
				reduce_ethqs(adapter, nqsets);
		}
		for (i = 0; i < want; ++i)
			adapter->msix_info[i].vec = entries[i].vector;
	} else if (err > 0) {
		pci_disable_msix(adapter->pdev);
		dev_info(adapter->pdev_dev, "only %d MSI-X vectors left,"
			 " not using MSI-X\n", err);
	}
	return err;
}

#ifdef HAVE_NET_DEVICE_OPS
static const struct net_device_ops cxgb4vf_netdev_ops	= {
	.ndo_open		= cxgb4vf_open,
	.ndo_stop		= cxgb4vf_stop,
	.ndo_start_xmit		= t4vf_eth_xmit,
	.ndo_get_stats		= cxgb4vf_get_stats,
	.ndo_set_rx_mode	= cxgb4vf_set_rxmode,
	.ndo_set_mac_address	= cxgb4vf_set_mac_addr,
	.ndo_validate_addr	= eth_validate_addr,
	.ndo_do_ioctl		= cxgb4vf_do_ioctl,
	.ndo_change_mtu		= cxgb4vf_change_mtu,
	.ndo_vlan_rx_register	= cxgb4vf_vlan_rx_register,
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= cxgb4vf_poll_controller,
#endif
};
#endif

/*
 * "Probe" a device: initialize a device and construct all kernel and driver
 * state needed to manage the device.  This routine is called "init_one" in
 * the PF Driver ...
 */
static int __devinit cxgb4vf_pci_probe(struct pci_dev *pdev,
				       const struct pci_device_id *ent)
{
	static int version_printed;

	int pci_using_dac;
	int err, pidx;
	unsigned int pmask;
	struct adapter *adapter;
	struct port_info *pi;
	struct net_device *netdev;

	/*
	 * Vet our module parameters.
	 */
	if (msi != MSI_MSIX && msi != MSI_MSI) {
		dev_err(&pdev->dev, "bad module parameter msi=%d; must be %d"
			" (MSI-X or MSI) or %d (MSI)\n", msi, MSI_MSIX,
			MSI_MSI);
		err = -EINVAL;
		goto err_out;
	}

	/*
	 * Print our driver banner the first time we're called to initialize a
	 * device.
	 */
	if (version_printed == 0) {
		printk(KERN_INFO "%s - version %s\n", DRV_DESC, DRV_VERSION);
		version_printed = 1;
	}


	/*
	 * Initialize generic PCI device state.
	 */
	err = pci_enable_device(pdev);
	if (err) {
		dev_err(&pdev->dev, "cannot enable PCI device\n");
		return err;
	}

	/*
	 * Reserve PCI resources for the device.  If we can't get them some
	 * other driver may have already claimed the device ...
	 */
	err = pci_request_regions(pdev, KBUILD_MODNAME);
	if (err) {
		dev_err(&pdev->dev, "cannot obtain PCI resources\n");
		goto err_disable_device;
	}

	/*
	 * Set up our DMA mask: try for 64-bit address masking first and
	 * fall back to 32-bit if we can't get 64 bits ...
	 */
	err = pci_set_dma_mask(pdev, DMA_BIT_MASK(64));
	if (err == 0) {
		err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64));
		if (err) {
			dev_err(&pdev->dev, "unable to obtain 64-bit DMA for"
				" coherent allocations\n");
			goto err_release_regions;
		}
		pci_using_dac = 1;
	} else {
		err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
		if (err != 0) {
			dev_err(&pdev->dev, "no usable DMA configuration\n");
			goto err_release_regions;
		}
		pci_using_dac = 0;
	}

	/*
	 * Enable bus mastering for the device ...
	 */
	pci_set_master(pdev);

	/*
	 * Allocate our adapter data structure and attach it to the device.
	 */
	adapter = kzalloc(sizeof(*adapter), GFP_KERNEL);
	if (!adapter) {
		err = -ENOMEM;
		goto err_release_regions;
	}
	pci_set_drvdata(pdev, adapter);
	adapter->pdev = pdev;
	adapter->pdev_dev = &pdev->dev;

	/*
	 * Initialize SMP data synchronization resources.
	 */
	spin_lock_init(&adapter->stats_lock);

	/*
	 * Map our I/O registers in BAR0.
	 */
	adapter->regs = pci_ioremap_bar(pdev, 0);
	if (!adapter->regs) {
		dev_err(&pdev->dev, "cannot map device registers\n");
		err = -ENOMEM;
		goto err_free_adapter;
	}

	/*
	 * Initialize adapter level features.
	 */
	adapter->name = pci_name(pdev);
	adapter->msg_enable = dflt_msg_enable;
	err = adap_init0(adapter);
	if (err)
		goto err_unmap_bar;

	/*
	 * Allocate our "adapter ports" and stitch everything together.
	 */
	pmask = adapter->params.vfres.pmask;
	for_each_port(adapter, pidx) {
		int port_id, viid;

		/*
		 * We simplistically allocate our virtual interfaces
		 * sequentially across the port numbers to which we have
		 * access rights.  This should be configurable in some manner
		 * ...
		 */
		if (pmask == 0)
			break;
		port_id = ffs(pmask) - 1;
		pmask &= ~(1 << port_id);
		viid = t4vf_alloc_vi(adapter, port_id);
		if (viid < 0) {
			dev_err(&pdev->dev, "cannot allocate VI for port %d:"
				" err=%d\n", port_id, viid);
			err = viid;
			goto err_free_dev;
		}

		/*
		 * Allocate our network device and stitch things together.
		 */
		netdev = alloc_etherdev_mq(sizeof(struct port_info),
					   MAX_PORT_QSETS);
		if (netdev == NULL) {
			dev_err(&pdev->dev, "cannot allocate netdev for"
				" port %d\n", port_id);
			t4vf_free_vi(adapter, viid);
			err = -ENOMEM;
			goto err_free_dev;
		}
		adapter->port[pidx] = netdev;
		SET_NETDEV_DEV(netdev, &pdev->dev);
		pi = netdev_priv(netdev);
		pi->adapter = adapter;
		pi->pidx = pidx;
		pi->port_id = port_id;
		pi->viid = viid;

		/*
		 * Initialize the starting state of our "port" and register
		 * it.
		 */
		pi->xact_addr_filt = -1;
		pi->rx_offload = RX_CSO;
		netif_carrier_off(netdev);
		netdev->irq = pdev->irq;

		netdev->features = (NETIF_F_SG | NETIF_F_TSO | NETIF_F_TSO6 |
				    NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
				    NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_RX |
				    NETIF_F_GRO);
		if (pci_using_dac)
			netdev->features |= NETIF_F_HIGHDMA;
		netdev->vlan_features =
			(netdev->features &
			 ~(NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_RX));

#ifdef HAVE_NET_DEVICE_OPS
		netdev->netdev_ops = &cxgb4vf_netdev_ops;
#else
		netdev->vlan_rx_register = cxgb4vf_vlan_rx_register;
		netdev->open = cxgb4vf_open;
		netdev->stop = cxgb4vf_stop;
		netdev->hard_start_xmit = t4vf_eth_xmit;
		netdev->get_stats = cxgb4vf_get_stats;
		netdev->set_rx_mode = cxgb4vf_set_rxmode;
		netdev->do_ioctl = cxgb4vf_do_ioctl;
		netdev->change_mtu = cxgb4vf_change_mtu;
		netdev->set_mac_address = cxgb4vf_set_mac_addr;
#ifdef CONFIG_NET_POLL_CONTROLLER
		netdev->poll_controller = cxgb4vf_poll_controller;
#endif
#endif
		SET_ETHTOOL_OPS(netdev, &cxgb4vf_ethtool_ops);

		/*
		 * Initialize the hardware/software state for the port.
		 */
		err = t4vf_port_init(adapter, pidx);
		if (err) {
			dev_err(&pdev->dev, "cannot initialize port %d\n",
				pidx);
			goto err_free_dev;
		}
	}

	/*
	 * The "card" is now ready to go.  If any errors occur during device
	 * registration we do not fail the whole "card" but rather proceed
	 * only with the ports we manage to register successfully.  However we
	 * must register at least one net device.
	 */
	for_each_port(adapter, pidx) {
		netdev = adapter->port[pidx];
		if (netdev == NULL)
			continue;

		err = register_netdev(netdev);
		if (err) {
			dev_warn(&pdev->dev, "cannot register net device %s,"
				 " skipping\n", netdev->name);
			continue;
		}

		set_bit(pidx, &adapter->registered_device_map);
	}
	if (adapter->registered_device_map == 0) {
		dev_err(&pdev->dev, "could not register any net devices\n");
		goto err_free_dev;
	}

	/*
	 * Set up our debugfs entries.
	 */
	if (cxgb4vf_debugfs_root) {
		adapter->debugfs_root =
			debugfs_create_dir(pci_name(pdev),
					   cxgb4vf_debugfs_root);
		if (adapter->debugfs_root == NULL)
			dev_warn(&pdev->dev, "could not create debugfs"
				 " directory");
		else
			setup_debugfs(adapter);
	}

	/*
	 * See what interrupts we'll be using.  If we've been configured to
	 * use MSI-X interrupts, try to enable them but fall back to using
	 * MSI interrupts if we can't enable MSI-X interrupts.  If we can't
	 * get MSI interrupts we bail with the error.
	 */
	if (msi == MSI_MSIX && enable_msix(adapter) == 0)
		adapter->flags |= USING_MSIX;
	else {
		err = pci_enable_msi(pdev);
		if (err) {
			dev_err(&pdev->dev, "Unable to allocate %s interrupts;"
				" err=%d\n",
				msi == MSI_MSIX ? "MSI-X or MSI" : "MSI", err);
			goto err_free_debugfs;
		}
		adapter->flags |= USING_MSI;
	}

	/*
	 * Now that we know how many "ports" we have and what their types are,
	 * and how many Queue Sets we can support, we can configure our queue
	 * resources.
	 */
	cfg_queues(adapter);

	/*
	 * Print a short notice on the existance and configuration of the new
	 * VF network device ...
	 */
	for_each_port(adapter, pidx) {
		dev_info(adapter->pdev_dev, "%s: Chelsio VF NIC PCIe %s\n",
			 adapter->port[pidx]->name,
			 (adapter->flags & USING_MSIX) ? "MSI-X" :
			 (adapter->flags & USING_MSI)  ? "MSI" : "");
	}

	/*
	 * Return success!
	 */
	return 0;

	/*
	 * Error recovery and exit code.  Unwind state that's been created
	 * so far and return the error.
	 */

err_free_debugfs:
	if (adapter->debugfs_root) {
		cleanup_debugfs(adapter);
		debugfs_remove_recursive(adapter->debugfs_root);
	}

err_free_dev:
	for_each_port(adapter, pidx) {
		netdev = adapter->port[pidx];
		if (netdev == NULL)
			continue;
		pi = netdev_priv(netdev);
		t4vf_free_vi(adapter, pi->viid);
		if (test_bit(pidx, &adapter->registered_device_map))
			unregister_netdev(netdev);
		free_netdev(netdev);
	}

err_unmap_bar:
	iounmap(adapter->regs);

err_free_adapter:
	kfree(adapter);
	pci_set_drvdata(pdev, NULL);

err_release_regions:
	pci_release_regions(pdev);
	pci_set_drvdata(pdev, NULL);
	pci_clear_master(pdev);

err_disable_device:
	pci_disable_device(pdev);

err_out:
	return err;
}

/*
 * "Remove" a device: tear down all kernel and driver state created in the
 * "probe" routine and quiesce the device (disable interrupts, etc.).  (Note
 * that this is called "remove_one" in the PF Driver.)
 */
static void __devexit cxgb4vf_pci_remove(struct pci_dev *pdev)
{
	struct adapter *adapter = pci_get_drvdata(pdev);

	/*
	 * Tear down driver state associated with device.
	 */
	if (adapter) {
		int pidx;

		/*
		 * Stop all of our activity.  Unregister network port,
		 * disable interrupts, etc.
		 */
		for_each_port(adapter, pidx)
			if (test_bit(pidx, &adapter->registered_device_map))
				unregister_netdev(adapter->port[pidx]);
		t4vf_sge_stop(adapter);
		if (adapter->flags & USING_MSIX) {
			pci_disable_msix(adapter->pdev);
			adapter->flags &= ~USING_MSIX;
		} else if (adapter->flags & USING_MSI) {
			pci_disable_msi(adapter->pdev);
			adapter->flags &= ~USING_MSI;
		}

		/*
		 * Tear down our debugfs entries.
		 */
		if (adapter->debugfs_root) {
			cleanup_debugfs(adapter);
			debugfs_remove_recursive(adapter->debugfs_root);
		}

		/*
		 * Free all of the various resources which we've acquired ...
		 */
		t4vf_free_sge_resources(adapter);
		for_each_port(adapter, pidx) {
			struct net_device *netdev = adapter->port[pidx];
			struct port_info *pi;

			if (netdev == NULL)
				continue;

			pi = netdev_priv(netdev);
			t4vf_free_vi(adapter, pi->viid);
			free_netdev(netdev);
		}
		iounmap(adapter->regs);
		kfree(adapter);
		pci_set_drvdata(pdev, NULL);
	}

	/*
	 * Disable the device and release its PCI resources.
	 */
	pci_disable_device(pdev);
	pci_clear_master(pdev);
	pci_release_regions(pdev);
}

/*
 * PCI Device registration data structures.
 */
#define CH_DEVICE(devid, idx) \
	{ PCI_VENDOR_ID_CHELSIO, devid, PCI_ANY_ID, PCI_ANY_ID, 0, 0, idx }

static struct pci_device_id cxgb4vf_pci_tbl[] = {
	CH_DEVICE(0xb000, 0),	/* PE10K FPGA */
	CH_DEVICE(0x4800, 0),	/* T440-dbg */
	CH_DEVICE(0x4801, 0),	/* T420-cr */
	CH_DEVICE(0x4802, 0),	/* T422-cr */
	{ 0, }
};

MODULE_DESCRIPTION(DRV_DESC);
MODULE_AUTHOR("Chelsio Communications");
MODULE_LICENSE("Dual BSD/GPL");
MODULE_VERSION(DRV_VERSION);
MODULE_DEVICE_TABLE(pci, cxgb4vf_pci_tbl);

static struct pci_driver cxgb4vf_driver = {
	.name		= KBUILD_MODNAME,
	.id_table	= cxgb4vf_pci_tbl,
	.probe		= cxgb4vf_pci_probe,
	.remove		= __devexit_p(cxgb4vf_pci_remove),
};

/*
 * Initialize global driver state.
 */
static int __init cxgb4vf_module_init(void)
{
	int ret;

	/* Debugfs support is optional, just warn if this fails */
	cxgb4vf_debugfs_root = debugfs_create_dir(KBUILD_MODNAME, NULL);
	if (!cxgb4vf_debugfs_root)
		printk(KERN_WARNING KBUILD_MODNAME ": could not create"
		       " debugfs entry, continuing\n");

	ret = pci_register_driver(&cxgb4vf_driver);
	if (ret < 0)
		debugfs_remove(cxgb4vf_debugfs_root);
	return ret;
}

/*
 * Tear down global driver state.
 */
static void __exit cxgb4vf_module_exit(void)
{
	pci_unregister_driver(&cxgb4vf_driver);
	debugfs_remove(cxgb4vf_debugfs_root);
}

module_init(cxgb4vf_module_init);
module_exit(cxgb4vf_module_exit);