aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/net/sfc/efx.c
blob: 00c23b1babcad720154fe7f73b346cddc1b3b3f9 (plain) (blame)
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
/****************************************************************************
 * Driver for Solarflare Solarstorm network controllers and boards
 * Copyright 2005-2006 Fen Systems Ltd.
 * Copyright 2005-2008 Solarflare Communications Inc.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 as published
 * by the Free Software Foundation, incorporated herein by reference.
 */

#include <linux/module.h>
#include <linux/pci.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/delay.h>
#include <linux/notifier.h>
#include <linux/ip.h>
#include <linux/tcp.h>
#include <linux/in.h>
#include <linux/crc32.h>
#include <linux/ethtool.h>
#include <linux/topology.h>
#include "net_driver.h"
#include "ethtool.h"
#include "tx.h"
#include "rx.h"
#include "efx.h"
#include "mdio_10g.h"
#include "falcon.h"

#define EFX_MAX_MTU (9 * 1024)

/* RX slow fill workqueue. If memory allocation fails in the fast path,
 * a work item is pushed onto this work queue to retry the allocation later,
 * to avoid the NIC being starved of RX buffers. Since this is a per cpu
 * workqueue, there is nothing to be gained in making it per NIC
 */
static struct workqueue_struct *refill_workqueue;

/* Reset workqueue. If any NIC has a hardware failure then a reset will be
 * queued onto this work queue. This is not a per-nic work queue, because
 * efx_reset_work() acquires the rtnl lock, so resets are naturally serialised.
 */
static struct workqueue_struct *reset_workqueue;

/**************************************************************************
 *
 * Configurable values
 *
 *************************************************************************/

/*
 * Enable large receive offload (LRO) aka soft segment reassembly (SSR)
 *
 * This sets the default for new devices.  It can be controlled later
 * using ethtool.
 */
static int lro = true;
module_param(lro, int, 0644);
MODULE_PARM_DESC(lro, "Large receive offload acceleration");

/*
 * Use separate channels for TX and RX events
 *
 * Set this to 1 to use separate channels for TX and RX. It allows us
 * to control interrupt affinity separately for TX and RX.
 *
 * This is only used in MSI-X interrupt mode
 */
static unsigned int separate_tx_channels;
module_param(separate_tx_channels, uint, 0644);
MODULE_PARM_DESC(separate_tx_channels,
		 "Use separate channels for TX and RX");

/* This is the weight assigned to each of the (per-channel) virtual
 * NAPI devices.
 */
static int napi_weight = 64;

/* This is the time (in jiffies) between invocations of the hardware
 * monitor, which checks for known hardware bugs and resets the
 * hardware and driver as necessary.
 */
unsigned int efx_monitor_interval = 1 * HZ;

/* This controls whether or not the driver will initialise devices
 * with invalid MAC addresses stored in the EEPROM or flash.  If true,
 * such devices will be initialised with a random locally-generated
 * MAC address.  This allows for loading the sfc_mtd driver to
 * reprogram the flash, even if the flash contents (including the MAC
 * address) have previously been erased.
 */
static unsigned int allow_bad_hwaddr;

/* Initial interrupt moderation settings.  They can be modified after
 * module load with ethtool.
 *
 * The default for RX should strike a balance between increasing the
 * round-trip latency and reducing overhead.
 */
static unsigned int rx_irq_mod_usec = 60;

/* Initial interrupt moderation settings.  They can be modified after
 * module load with ethtool.
 *
 * This default is chosen to ensure that a 10G link does not go idle
 * while a TX queue is stopped after it has become full.  A queue is
 * restarted when it drops below half full.  The time this takes (assuming
 * worst case 3 descriptors per packet and 1024 descriptors) is
 *   512 / 3 * 1.2 = 205 usec.
 */
static unsigned int tx_irq_mod_usec = 150;

/* This is the first interrupt mode to try out of:
 * 0 => MSI-X
 * 1 => MSI
 * 2 => legacy
 */
static unsigned int interrupt_mode;

/* This is the requested number of CPUs to use for Receive-Side Scaling (RSS),
 * i.e. the number of CPUs among which we may distribute simultaneous
 * interrupt handling.
 *
 * Cards without MSI-X will only target one CPU via legacy or MSI interrupt.
 * The default (0) means to assign an interrupt to each package (level II cache)
 */
static unsigned int rss_cpus;
module_param(rss_cpus, uint, 0444);
MODULE_PARM_DESC(rss_cpus, "Number of CPUs to use for Receive-Side Scaling");

static int phy_flash_cfg;
module_param(phy_flash_cfg, int, 0644);
MODULE_PARM_DESC(phy_flash_cfg, "Set PHYs into reflash mode initially");

static unsigned irq_adapt_low_thresh = 10000;
module_param(irq_adapt_low_thresh, uint, 0644);
MODULE_PARM_DESC(irq_adapt_low_thresh,
		 "Threshold score for reducing IRQ moderation");

static unsigned irq_adapt_high_thresh = 20000;
module_param(irq_adapt_high_thresh, uint, 0644);
MODULE_PARM_DESC(irq_adapt_high_thresh,
		 "Threshold score for increasing IRQ moderation");

/**************************************************************************
 *
 * Utility functions and prototypes
 *
 *************************************************************************/
static void efx_remove_channel(struct efx_channel *channel);
static void efx_remove_port(struct efx_nic *efx);
static void efx_fini_napi(struct efx_nic *efx);
static void efx_fini_channels(struct efx_nic *efx);

#define EFX_ASSERT_RESET_SERIALISED(efx)		\
	do {						\
		if (efx->state == STATE_RUNNING)	\
			ASSERT_RTNL();			\
	} while (0)

/**************************************************************************
 *
 * Event queue processing
 *
 *************************************************************************/

/* Process channel's event queue
 *
 * This function is responsible for processing the event queue of a
 * single channel.  The caller must guarantee that this function will
 * never be concurrently called more than once on the same channel,
 * though different channels may be being processed concurrently.
 */
static int efx_process_channel(struct efx_channel *channel, int rx_quota)
{
	struct efx_nic *efx = channel->efx;
	int rx_packets;

	if (unlikely(efx->reset_pending != RESET_TYPE_NONE ||
		     !channel->enabled))
		return 0;

	rx_packets = falcon_process_eventq(channel, rx_quota);
	if (rx_packets == 0)
		return 0;

	/* Deliver last RX packet. */
	if (channel->rx_pkt) {
		__efx_rx_packet(channel, channel->rx_pkt,
				channel->rx_pkt_csummed);
		channel->rx_pkt = NULL;
	}

	efx_rx_strategy(channel);

	efx_fast_push_rx_descriptors(&efx->rx_queue[channel->channel]);

	return rx_packets;
}

/* Mark channel as finished processing
 *
 * Note that since we will not receive further interrupts for this
 * channel before we finish processing and call the eventq_read_ack()
 * method, there is no need to use the interrupt hold-off timers.
 */
static inline void efx_channel_processed(struct efx_channel *channel)
{
	/* The interrupt handler for this channel may set work_pending
	 * as soon as we acknowledge the events we've seen.  Make sure
	 * it's cleared before then. */
	channel->work_pending = false;
	smp_wmb();

	falcon_eventq_read_ack(channel);
}

/* NAPI poll handler
 *
 * NAPI guarantees serialisation of polls of the same device, which
 * provides the guarantee required by efx_process_channel().
 */
static int efx_poll(struct napi_struct *napi, int budget)
{
	struct efx_channel *channel =
		container_of(napi, struct efx_channel, napi_str);
	int rx_packets;

	EFX_TRACE(channel->efx, "channel %d NAPI poll executing on CPU %d\n",
		  channel->channel, raw_smp_processor_id());

	rx_packets = efx_process_channel(channel, budget);

	if (rx_packets < budget) {
		struct efx_nic *efx = channel->efx;

		if (channel->used_flags & EFX_USED_BY_RX &&
		    efx->irq_rx_adaptive &&
		    unlikely(++channel->irq_count == 1000)) {
			unsigned old_irq_moderation = channel->irq_moderation;

			if (unlikely(channel->irq_mod_score <
				     irq_adapt_low_thresh)) {
				channel->irq_moderation =
					max_t(int,
					      channel->irq_moderation -
					      FALCON_IRQ_MOD_RESOLUTION,
					      FALCON_IRQ_MOD_RESOLUTION);
			} else if (unlikely(channel->irq_mod_score >
					    irq_adapt_high_thresh)) {
				channel->irq_moderation =
					min(channel->irq_moderation +
					    FALCON_IRQ_MOD_RESOLUTION,
					    efx->irq_rx_moderation);
			}

			if (channel->irq_moderation != old_irq_moderation)
				falcon_set_int_moderation(channel);

			channel->irq_count = 0;
			channel->irq_mod_score = 0;
		}

		/* There is no race here; although napi_disable() will
		 * only wait for napi_complete(), this isn't a problem
		 * since efx_channel_processed() will have no effect if
		 * interrupts have already been disabled.
		 */
		napi_complete(napi);
		efx_channel_processed(channel);
	}

	return rx_packets;
}

/* Process the eventq of the specified channel immediately on this CPU
 *
 * Disable hardware generated interrupts, wait for any existing
 * processing to finish, then directly poll (and ack ) the eventq.
 * Finally reenable NAPI and interrupts.
 *
 * Since we are touching interrupts the caller should hold the suspend lock
 */
void efx_process_channel_now(struct efx_channel *channel)
{
	struct efx_nic *efx = channel->efx;

	BUG_ON(!channel->used_flags);
	BUG_ON(!channel->enabled);

	/* Disable interrupts and wait for ISRs to complete */
	falcon_disable_interrupts(efx);
	if (efx->legacy_irq)
		synchronize_irq(efx->legacy_irq);
	if (channel->irq)
		synchronize_irq(channel->irq);

	/* Wait for any NAPI processing to complete */
	napi_disable(&channel->napi_str);

	/* Poll the channel */
	efx_process_channel(channel, efx->type->evq_size);

	/* Ack the eventq. This may cause an interrupt to be generated
	 * when they are reenabled */
	efx_channel_processed(channel);

	napi_enable(&channel->napi_str);
	falcon_enable_interrupts(efx);
}

/* Create event queue
 * Event queue memory allocations are done only once.  If the channel
 * is reset, the memory buffer will be reused; this guards against
 * errors during channel reset and also simplifies interrupt handling.
 */
static int efx_probe_eventq(struct efx_channel *channel)
{
	EFX_LOG(channel->efx, "chan %d create event queue\n", channel->channel);

	return falcon_probe_eventq(channel);
}

/* Prepare channel's event queue */
static void efx_init_eventq(struct efx_channel *channel)
{
	EFX_LOG(channel->efx, "chan %d init event queue\n", channel->channel);

	channel->eventq_read_ptr = 0;

	falcon_init_eventq(channel);
}

static void efx_fini_eventq(struct efx_channel *channel)
{
	EFX_LOG(channel->efx, "chan %d fini event queue\n", channel->channel);

	falcon_fini_eventq(channel);
}

static void efx_remove_eventq(struct efx_channel *channel)
{
	EFX_LOG(channel->efx, "chan %d remove event queue\n", channel->channel);

	falcon_remove_eventq(channel);
}

/**************************************************************************
 *
 * Channel handling
 *
 *************************************************************************/

static int efx_probe_channel(struct efx_channel *channel)
{
	struct efx_tx_queue *tx_queue;
	struct efx_rx_queue *rx_queue;
	int rc;

	EFX_LOG(channel->efx, "creating channel %d\n", channel->channel);

	rc = efx_probe_eventq(channel);
	if (rc)
		goto fail1;

	efx_for_each_channel_tx_queue(tx_queue, channel) {
		rc = efx_probe_tx_queue(tx_queue);
		if (rc)
			goto fail2;
	}

	efx_for_each_channel_rx_queue(rx_queue, channel) {
		rc = efx_probe_rx_queue(rx_queue);
		if (rc)
			goto fail3;
	}

	channel->n_rx_frm_trunc = 0;

	return 0;

 fail3:
	efx_for_each_channel_rx_queue(rx_queue, channel)
		efx_remove_rx_queue(rx_queue);
 fail2:
	efx_for_each_channel_tx_queue(tx_queue, channel)
		efx_remove_tx_queue(tx_queue);
 fail1:
	return rc;
}


static void efx_set_channel_names(struct efx_nic *efx)
{
	struct efx_channel *channel;
	const char *type = "";
	int number;

	efx_for_each_channel(channel, efx) {
		number = channel->channel;
		if (efx->n_channels > efx->n_rx_queues) {
			if (channel->channel < efx->n_rx_queues) {
				type = "-rx";
			} else {
				type = "-tx";
				number -= efx->n_rx_queues;
			}
		}
		snprintf(channel->name, sizeof(channel->name),
			 "%s%s-%d", efx->name, type, number);
	}
}

/* Channels are shutdown and reinitialised whilst the NIC is running
 * to propagate configuration changes (mtu, checksum offload), or
 * to clear hardware error conditions
 */
static void efx_init_channels(struct efx_nic *efx)
{
	struct efx_tx_queue *tx_queue;
	struct efx_rx_queue *rx_queue;
	struct efx_channel *channel;

	/* Calculate the rx buffer allocation parameters required to
	 * support the current MTU, including padding for header
	 * alignment and overruns.
	 */
	efx->rx_buffer_len = (max(EFX_PAGE_IP_ALIGN, NET_IP_ALIGN) +
			      EFX_MAX_FRAME_LEN(efx->net_dev->mtu) +
			      efx->type->rx_buffer_padding);
	efx->rx_buffer_order = get_order(efx->rx_buffer_len);

	/* Initialise the channels */
	efx_for_each_channel(channel, efx) {
		EFX_LOG(channel->efx, "init chan %d\n", channel->channel);

		efx_init_eventq(channel);

		efx_for_each_channel_tx_queue(tx_queue, channel)
			efx_init_tx_queue(tx_queue);

		/* The rx buffer allocation strategy is MTU dependent */
		efx_rx_strategy(channel);

		efx_for_each_channel_rx_queue(rx_queue, channel)
			efx_init_rx_queue(rx_queue);

		WARN_ON(channel->rx_pkt != NULL);
		efx_rx_strategy(channel);
	}
}

/* This enables event queue processing and packet transmission.
 *
 * Note that this function is not allowed to fail, since that would
 * introduce too much complexity into the suspend/resume path.
 */
static void efx_start_channel(struct efx_channel *channel)
{
	struct efx_rx_queue *rx_queue;

	EFX_LOG(channel->efx, "starting chan %d\n", channel->channel);

	if (!(channel->efx->net_dev->flags & IFF_UP))
		netif_napi_add(channel->napi_dev, &channel->napi_str,
			       efx_poll, napi_weight);

	/* The interrupt handler for this channel may set work_pending
	 * as soon as we enable it.  Make sure it's cleared before
	 * then.  Similarly, make sure it sees the enabled flag set. */
	channel->work_pending = false;
	channel->enabled = true;
	smp_wmb();

	napi_enable(&channel->napi_str);

	/* Load up RX descriptors */
	efx_for_each_channel_rx_queue(rx_queue, channel)
		efx_fast_push_rx_descriptors(rx_queue);
}

/* This disables event queue processing and packet transmission.
 * This function does not guarantee that all queue processing
 * (e.g. RX refill) is complete.
 */
static void efx_stop_channel(struct efx_channel *channel)
{
	struct efx_rx_queue *rx_queue;

	if (!channel->enabled)
		return;

	EFX_LOG(channel->efx, "stop chan %d\n", channel->channel);

	channel->enabled = false;
	napi_disable(&channel->napi_str);

	/* Ensure that any worker threads have exited or will be no-ops */
	efx_for_each_channel_rx_queue(rx_queue, channel) {
		spin_lock_bh(&rx_queue->add_lock);
		spin_unlock_bh(&rx_queue->add_lock);
	}
}

static void efx_fini_channels(struct efx_nic *efx)
{
	struct efx_channel *channel;
	struct efx_tx_queue *tx_queue;
	struct efx_rx_queue *rx_queue;
	int rc;

	EFX_ASSERT_RESET_SERIALISED(efx);
	BUG_ON(efx->port_enabled);

	rc = falcon_flush_queues(efx);
	if (rc)
		EFX_ERR(efx, "failed to flush queues\n");
	else
		EFX_LOG(efx, "successfully flushed all queues\n");

	efx_for_each_channel(channel, efx) {
		EFX_LOG(channel->efx, "shut down chan %d\n", channel->channel);

		efx_for_each_channel_rx_queue(rx_queue, channel)
			efx_fini_rx_queue(rx_queue);
		efx_for_each_channel_tx_queue(tx_queue, channel)
			efx_fini_tx_queue(tx_queue);
		efx_fini_eventq(channel);
	}
}

static void efx_remove_channel(struct efx_channel *channel)
{
	struct efx_tx_queue *tx_queue;
	struct efx_rx_queue *rx_queue;

	EFX_LOG(channel->efx, "destroy chan %d\n", channel->channel);

	efx_for_each_channel_rx_queue(rx_queue, channel)
		efx_remove_rx_queue(rx_queue);
	efx_for_each_channel_tx_queue(tx_queue, channel)
		efx_remove_tx_queue(tx_queue);
	efx_remove_eventq(channel);

	channel->used_flags = 0;
}

void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue, int delay)
{
	queue_delayed_work(refill_workqueue, &rx_queue->work, delay);
}

/**************************************************************************
 *
 * Port handling
 *
 **************************************************************************/

/* This ensures that the kernel is kept informed (via
 * netif_carrier_on/off) of the link status, and also maintains the
 * link status's stop on the port's TX queue.
 */
static void efx_link_status_changed(struct efx_nic *efx)
{
	/* SFC Bug 5356: A net_dev notifier is registered, so we must ensure
	 * that no events are triggered between unregister_netdev() and the
	 * driver unloading. A more general condition is that NETDEV_CHANGE
	 * can only be generated between NETDEV_UP and NETDEV_DOWN */
	if (!netif_running(efx->net_dev))
		return;

	if (efx->port_inhibited) {
		netif_carrier_off(efx->net_dev);
		return;
	}

	if (efx->link_up != netif_carrier_ok(efx->net_dev)) {
		efx->n_link_state_changes++;

		if (efx->link_up)
			netif_carrier_on(efx->net_dev);
		else
			netif_carrier_off(efx->net_dev);
	}

	/* Status message for kernel log */
	if (efx->link_up) {
		EFX_INFO(efx, "link up at %uMbps %s-duplex (MTU %d)%s\n",
			 efx->link_speed, efx->link_fd ? "full" : "half",
			 efx->net_dev->mtu,
			 (efx->promiscuous ? " [PROMISC]" : ""));
	} else {
		EFX_INFO(efx, "link down\n");
	}

}

static void efx_fini_port(struct efx_nic *efx);

/* This call reinitialises the MAC to pick up new PHY settings. The
 * caller must hold the mac_lock */
void __efx_reconfigure_port(struct efx_nic *efx)
{
	WARN_ON(!mutex_is_locked(&efx->mac_lock));

	EFX_LOG(efx, "reconfiguring MAC from PHY settings on CPU %d\n",
		raw_smp_processor_id());

	/* Serialise the promiscuous flag with efx_set_multicast_list. */
	if (efx_dev_registered(efx)) {
		netif_addr_lock_bh(efx->net_dev);
		netif_addr_unlock_bh(efx->net_dev);
	}

	falcon_deconfigure_mac_wrapper(efx);

	/* Reconfigure the PHY, disabling transmit in mac level loopback. */
	if (LOOPBACK_INTERNAL(efx))
		efx->phy_mode |= PHY_MODE_TX_DISABLED;
	else
		efx->phy_mode &= ~PHY_MODE_TX_DISABLED;
	efx->phy_op->reconfigure(efx);

	if (falcon_switch_mac(efx))
		goto fail;

	efx->mac_op->reconfigure(efx);

	/* Inform kernel of loss/gain of carrier */
	efx_link_status_changed(efx);
	return;

fail:
	EFX_ERR(efx, "failed to reconfigure MAC\n");
	efx->port_enabled = false;
	efx_fini_port(efx);
}

/* Reinitialise the MAC to pick up new PHY settings, even if the port is
 * disabled. */
void efx_reconfigure_port(struct efx_nic *efx)
{
	EFX_ASSERT_RESET_SERIALISED(efx);

	mutex_lock(&efx->mac_lock);
	__efx_reconfigure_port(efx);
	mutex_unlock(&efx->mac_lock);
}

/* Asynchronous efx_reconfigure_port work item. To speed up efx_flush_all()
 * we don't efx_reconfigure_port() if the port is disabled. Care is taken
 * in efx_stop_all() and efx_start_port() to prevent PHY events being lost */
static void efx_phy_work(struct work_struct *data)
{
	struct efx_nic *efx = container_of(data, struct efx_nic, phy_work);

	mutex_lock(&efx->mac_lock);
	if (efx->port_enabled)
		__efx_reconfigure_port(efx);
	mutex_unlock(&efx->mac_lock);
}

static void efx_mac_work(struct work_struct *data)
{
	struct efx_nic *efx = container_of(data, struct efx_nic, mac_work);

	mutex_lock(&efx->mac_lock);
	if (efx->port_enabled)
		efx->mac_op->irq(efx);
	mutex_unlock(&efx->mac_lock);
}

static int efx_probe_port(struct efx_nic *efx)
{
	int rc;

	EFX_LOG(efx, "create port\n");

	/* Connect up MAC/PHY operations table and read MAC address */
	rc = falcon_probe_port(efx);
	if (rc)
		goto err;

	if (phy_flash_cfg)
		efx->phy_mode = PHY_MODE_SPECIAL;

	/* Sanity check MAC address */
	if (is_valid_ether_addr(efx->mac_address)) {
		memcpy(efx->net_dev->dev_addr, efx->mac_address, ETH_ALEN);
	} else {
		EFX_ERR(efx, "invalid MAC address %pM\n",
			efx->mac_address);
		if (!allow_bad_hwaddr) {
			rc = -EINVAL;
			goto err;
		}
		random_ether_addr(efx->net_dev->dev_addr);
		EFX_INFO(efx, "using locally-generated MAC %pM\n",
			 efx->net_dev->dev_addr);
	}

	return 0;

 err:
	efx_remove_port(efx);
	return rc;
}

static int efx_init_port(struct efx_nic *efx)
{
	int rc;

	EFX_LOG(efx, "init port\n");

	rc = efx->phy_op->init(efx);
	if (rc)
		return rc;
	mutex_lock(&efx->mac_lock);
	efx->phy_op->reconfigure(efx);
	rc = falcon_switch_mac(efx);
	mutex_unlock(&efx->mac_lock);
	if (rc)
		goto fail;
	efx->mac_op->reconfigure(efx);

	efx->port_initialized = true;
	efx_stats_enable(efx);
	return 0;

fail:
	efx->phy_op->fini(efx);
	return rc;
}

/* Allow efx_reconfigure_port() to be scheduled, and close the window
 * between efx_stop_port and efx_flush_all whereby a previously scheduled
 * efx_phy_work()/efx_mac_work() may have been cancelled */
static void efx_start_port(struct efx_nic *efx)
{
	EFX_LOG(efx, "start port\n");
	BUG_ON(efx->port_enabled);

	mutex_lock(&efx->mac_lock);
	efx->port_enabled = true;
	__efx_reconfigure_port(efx);
	efx->mac_op->irq(efx);
	mutex_unlock(&efx->mac_lock);
}

/* Prevent efx_phy_work, efx_mac_work, and efx_monitor() from executing,
 * and efx_set_multicast_list() from scheduling efx_phy_work. efx_phy_work
 * and efx_mac_work may still be scheduled via NAPI processing until
 * efx_flush_all() is called */
static void efx_stop_port(struct efx_nic *efx)
{
	EFX_LOG(efx, "stop port\n");

	mutex_lock(&efx->mac_lock);
	efx->port_enabled = false;
	mutex_unlock(&efx->mac_lock);

	/* Serialise against efx_set_multicast_list() */
	if (efx_dev_registered(efx)) {
		netif_addr_lock_bh(efx->net_dev);
		netif_addr_unlock_bh(efx->net_dev);
	}
}

static void efx_fini_port(struct efx_nic *efx)
{
	EFX_LOG(efx, "shut down port\n");

	if (!efx->port_initialized)
		return;

	efx_stats_disable(efx);
	efx->phy_op->fini(efx);
	efx->port_initialized = false;

	efx->link_up = false;
	efx_link_status_changed(efx);
}

static void efx_remove_port(struct efx_nic *efx)
{
	EFX_LOG(efx, "destroying port\n");

	falcon_remove_port(efx);
}

/**************************************************************************
 *
 * NIC handling
 *
 **************************************************************************/

/* This configures the PCI device to enable I/O and DMA. */
static int efx_init_io(struct efx_nic *efx)
{
	struct pci_dev *pci_dev = efx->pci_dev;
	dma_addr_t dma_mask = efx->type->max_dma_mask;
	int rc;

	EFX_LOG(efx, "initialising I/O\n");

	rc = pci_enable_device(pci_dev);
	if (rc) {
		EFX_ERR(efx, "failed to enable PCI device\n");
		goto fail1;
	}

	pci_set_master(pci_dev);

	/* Set the PCI DMA mask.  Try all possibilities from our
	 * genuine mask down to 32 bits, because some architectures
	 * (e.g. x86_64 with iommu_sac_force set) will allow 40 bit
	 * masks event though they reject 46 bit masks.
	 */
	while (dma_mask > 0x7fffffffUL) {
		if (pci_dma_supported(pci_dev, dma_mask) &&
		    ((rc = pci_set_dma_mask(pci_dev, dma_mask)) == 0))
			break;
		dma_mask >>= 1;
	}
	if (rc) {
		EFX_ERR(efx, "could not find a suitable DMA mask\n");
		goto fail2;
	}
	EFX_LOG(efx, "using DMA mask %llx\n", (unsigned long long) dma_mask);
	rc = pci_set_consistent_dma_mask(pci_dev, dma_mask);
	if (rc) {
		/* pci_set_consistent_dma_mask() is not *allowed* to
		 * fail with a mask that pci_set_dma_mask() accepted,
		 * but just in case...
		 */
		EFX_ERR(efx, "failed to set consistent DMA mask\n");
		goto fail2;
	}

	efx->membase_phys = pci_resource_start(efx->pci_dev,
					       efx->type->mem_bar);
	rc = pci_request_region(pci_dev, efx->type->mem_bar, "sfc");
	if (rc) {
		EFX_ERR(efx, "request for memory BAR failed\n");
		rc = -EIO;
		goto fail3;
	}
	efx->membase = ioremap_nocache(efx->membase_phys,
				       efx->type->mem_map_size);
	if (!efx->membase) {
		EFX_ERR(efx, "could not map memory BAR %d at %llx+%x\n",
			efx->type->mem_bar,
			(unsigned long long)efx->membase_phys,
			efx->type->mem_map_size);
		rc = -ENOMEM;
		goto fail4;
	}
	EFX_LOG(efx, "memory BAR %u at %llx+%x (virtual %p)\n",
		efx->type->mem_bar, (unsigned long long)efx->membase_phys,
		efx->type->mem_map_size, efx->membase);

	return 0;

 fail4:
	pci_release_region(efx->pci_dev, efx->type->mem_bar);
 fail3:
	efx->membase_phys = 0;
 fail2:
	pci_disable_device(efx->pci_dev);
 fail1:
	return rc;
}

static void efx_fini_io(struct efx_nic *efx)
{
	EFX_LOG(efx, "shutting down I/O\n");

	if (efx->membase) {
		iounmap(efx->membase);
		efx->membase = NULL;
	}

	if (efx->membase_phys) {
		pci_release_region(efx->pci_dev, efx->type->mem_bar);
		efx->membase_phys = 0;
	}

	pci_disable_device(efx->pci_dev);
}

/* Get number of RX queues wanted.  Return number of online CPU
 * packages in the expectation that an IRQ balancer will spread
 * interrupts across them. */
static int efx_wanted_rx_queues(void)
{
	cpumask_var_t core_mask;
	int count;
	int cpu;

	if (!alloc_cpumask_var(&core_mask, GFP_KERNEL)) {
		printk(KERN_WARNING
		       "efx.c: allocation failure, irq balancing hobbled\n");
		return 1;
	}

	cpumask_clear(core_mask);
	count = 0;
	for_each_online_cpu(cpu) {
		if (!cpumask_test_cpu(cpu, core_mask)) {
			++count;
			cpumask_or(core_mask, core_mask,
				   topology_core_cpumask(cpu));
		}
	}

	free_cpumask_var(core_mask);
	return count;
}

/* Probe the number and type of interrupts we are able to obtain, and
 * the resulting numbers of channels and RX queues.
 */
static void efx_probe_interrupts(struct efx_nic *efx)
{
	int max_channels =
		min_t(int, efx->type->phys_addr_channels, EFX_MAX_CHANNELS);
	int rc, i;

	if (efx->interrupt_mode == EFX_INT_MODE_MSIX) {
		struct msix_entry xentries[EFX_MAX_CHANNELS];
		int wanted_ints;
		int rx_queues;

		/* We want one RX queue and interrupt per CPU package
		 * (or as specified by the rss_cpus module parameter).
		 * We will need one channel per interrupt.
		 */
		rx_queues = rss_cpus ? rss_cpus : efx_wanted_rx_queues();
		wanted_ints = rx_queues + (separate_tx_channels ? 1 : 0);
		wanted_ints = min(wanted_ints, max_channels);

		for (i = 0; i < wanted_ints; i++)
			xentries[i].entry = i;
		rc = pci_enable_msix(efx->pci_dev, xentries, wanted_ints);
		if (rc > 0) {
			EFX_ERR(efx, "WARNING: Insufficient MSI-X vectors"
				" available (%d < %d).\n", rc, wanted_ints);
			EFX_ERR(efx, "WARNING: Performance may be reduced.\n");
			EFX_BUG_ON_PARANOID(rc >= wanted_ints);
			wanted_ints = rc;
			rc = pci_enable_msix(efx->pci_dev, xentries,
					     wanted_ints);
		}

		if (rc == 0) {
			efx->n_rx_queues = min(rx_queues, wanted_ints);
			efx->n_channels = wanted_ints;
			for (i = 0; i < wanted_ints; i++)
				efx->channel[i].irq = xentries[i].vector;
		} else {
			/* Fall back to single channel MSI */
			efx->interrupt_mode = EFX_INT_MODE_MSI;
			EFX_ERR(efx, "could not enable MSI-X\n");
		}
	}

	/* Try single interrupt MSI */
	if (efx->interrupt_mode == EFX_INT_MODE_MSI) {
		efx->n_rx_queues = 1;
		efx->n_channels = 1;
		rc = pci_enable_msi(efx->pci_dev);
		if (rc == 0) {
			efx->channel[0].irq = efx->pci_dev->irq;
		} else {
			EFX_ERR(efx, "could not enable MSI\n");
			efx->interrupt_mode = EFX_INT_MODE_LEGACY;
		}
	}

	/* Assume legacy interrupts */
	if (efx->interrupt_mode == EFX_INT_MODE_LEGACY) {
		efx->n_rx_queues = 1;
		efx->n_channels = 1 + (separate_tx_channels ? 1 : 0);
		efx->legacy_irq = efx->pci_dev->irq;
	}
}

static void efx_remove_interrupts(struct efx_nic *efx)
{
	struct efx_channel *channel;

	/* Remove MSI/MSI-X interrupts */
	efx_for_each_channel(channel, efx)
		channel->irq = 0;
	pci_disable_msi(efx->pci_dev);
	pci_disable_msix(efx->pci_dev);

	/* Remove legacy interrupt */
	efx->legacy_irq = 0;
}

static void efx_set_channels(struct efx_nic *efx)
{
	struct efx_tx_queue *tx_queue;
	struct efx_rx_queue *rx_queue;

	efx_for_each_tx_queue(tx_queue, efx) {
		if (separate_tx_channels)
			tx_queue->channel = &efx->channel[efx->n_channels-1];
		else
			tx_queue->channel = &efx->channel[0];
		tx_queue->channel->used_flags |= EFX_USED_BY_TX;
	}

	efx_for_each_rx_queue(rx_queue, efx) {
		rx_queue->channel = &efx->channel[rx_queue->queue];
		rx_queue->channel->used_flags |= EFX_USED_BY_RX;
	}
}

static int efx_probe_nic(struct efx_nic *efx)
{
	int rc;

	EFX_LOG(efx, "creating NIC\n");

	/* Carry out hardware-type specific initialisation */
	rc = falcon_probe_nic(efx);
	if (rc)
		return rc;

	/* Determine the number of channels and RX queues by trying to hook
	 * in MSI-X interrupts. */
	efx_probe_interrupts(efx);

	efx_set_channels(efx);

	/* Initialise the interrupt moderation settings */
	efx_init_irq_moderation(efx, tx_irq_mod_usec, rx_irq_mod_usec, true);

	return 0;
}

static void efx_remove_nic(struct efx_nic *efx)
{
	EFX_LOG(efx, "destroying NIC\n");

	efx_remove_interrupts(efx);
	falcon_remove_nic(efx);
}

/**************************************************************************
 *
 * NIC startup/shutdown
 *
 *************************************************************************/

static int efx_probe_all(struct efx_nic *efx)
{
	struct efx_channel *channel;
	int rc;

	/* Create NIC */
	rc = efx_probe_nic(efx);
	if (rc) {
		EFX_ERR(efx, "failed to create NIC\n");
		goto fail1;
	}

	/* Create port */
	rc = efx_probe_port(efx);
	if (rc) {
		EFX_ERR(efx, "failed to create port\n");
		goto fail2;
	}

	/* Create channels */
	efx_for_each_channel(channel, efx) {
		rc = efx_probe_channel(channel);
		if (rc) {
			EFX_ERR(efx, "failed to create channel %d\n",
				channel->channel);
			goto fail3;
		}
	}
	efx_set_channel_names(efx);

	return 0;

 fail3:
	efx_for_each_channel(channel, efx)
		efx_remove_channel(channel);
	efx_remove_port(efx);
 fail2:
	efx_remove_nic(efx);
 fail1:
	return rc;
}

/* Called after previous invocation(s) of efx_stop_all, restarts the
 * port, kernel transmit queue, NAPI processing and hardware interrupts,
 * and ensures that the port is scheduled to be reconfigured.
 * This function is safe to call multiple times when the NIC is in any
 * state. */
static void efx_start_all(struct efx_nic *efx)
{
	struct efx_channel *channel;

	EFX_ASSERT_RESET_SERIALISED(efx);

	/* Check that it is appropriate to restart the interface. All
	 * of these flags are safe to read under just the rtnl lock */
	if (efx->port_enabled)
		return;
	if ((efx->state != STATE_RUNNING) && (efx->state != STATE_INIT))
		return;
	if (efx_dev_registered(efx) && !netif_running(efx->net_dev))
		return;

	/* Mark the port as enabled so port reconfigurations can start, then
	 * restart the transmit interface early so the watchdog timer stops */
	efx_start_port(efx);
	if (efx_dev_registered(efx))
		efx_wake_queue(efx);

	efx_for_each_channel(channel, efx)
		efx_start_channel(channel);

	falcon_enable_interrupts(efx);

	/* Start hardware monitor if we're in RUNNING */
	if (efx->state == STATE_RUNNING)
		queue_delayed_work(efx->workqueue, &efx->monitor_work,
				   efx_monitor_interval);
}

/* Flush all delayed work. Should only be called when no more delayed work
 * will be scheduled. This doesn't flush pending online resets (efx_reset),
 * since we're holding the rtnl_lock at this point. */
static void efx_flush_all(struct efx_nic *efx)
{
	struct efx_rx_queue *rx_queue;

	/* Make sure the hardware monitor is stopped */
	cancel_delayed_work_sync(&efx->monitor_work);

	/* Ensure that all RX slow refills are complete. */
	efx_for_each_rx_queue(rx_queue, efx)
		cancel_delayed_work_sync(&rx_queue->work);

	/* Stop scheduled port reconfigurations */
	cancel_work_sync(&efx->mac_work);
	cancel_work_sync(&efx->phy_work);

}

/* Quiesce hardware and software without bringing the link down.
 * Safe to call multiple times, when the nic and interface is in any
 * state. The caller is guaranteed to subsequently be in a position
 * to modify any hardware and software state they see fit without
 * taking locks. */
static void efx_stop_all(struct efx_nic *efx)
{
	struct efx_channel *channel;

	EFX_ASSERT_RESET_SERIALISED(efx);

	/* port_enabled can be read safely under the rtnl lock */
	if (!efx->port_enabled)
		return;

	/* Disable interrupts and wait for ISR to complete */
	falcon_disable_interrupts(efx);
	if (efx->legacy_irq)
		synchronize_irq(efx->legacy_irq);
	efx_for_each_channel(channel, efx) {
		if (channel->irq)
			synchronize_irq(channel->irq);
	}

	/* Stop all NAPI processing and synchronous rx refills */
	efx_for_each_channel(channel, efx)
		efx_stop_channel(channel);

	/* Stop all asynchronous port reconfigurations. Since all
	 * event processing has already been stopped, there is no
	 * window to loose phy events */
	efx_stop_port(efx);

	/* Flush efx_phy_work, efx_mac_work, refill_workqueue, monitor_work */
	efx_flush_all(efx);

	/* Isolate the MAC from the TX and RX engines, so that queue
	 * flushes will complete in a timely fashion. */
	falcon_drain_tx_fifo(efx);

	/* Stop the kernel transmit interface late, so the watchdog
	 * timer isn't ticking over the flush */
	if (efx_dev_registered(efx)) {
		efx_stop_queue(efx);
		netif_tx_lock_bh(efx->net_dev);
		netif_tx_unlock_bh(efx->net_dev);
	}
}

static void efx_remove_all(struct efx_nic *efx)
{
	struct efx_channel *channel;

	efx_for_each_channel(channel, efx)
		efx_remove_channel(channel);
	efx_remove_port(efx);
	efx_remove_nic(efx);
}

/* A convinience function to safely flush all the queues */
void efx_flush_queues(struct efx_nic *efx)
{
	EFX_ASSERT_RESET_SERIALISED(efx);

	efx_stop_all(efx);

	efx_fini_channels(efx);
	efx_init_channels(efx);

	efx_start_all(efx);
}

/**************************************************************************
 *
 * Interrupt moderation
 *
 **************************************************************************/

/* Set interrupt moderation parameters */
void efx_init_irq_moderation(struct efx_nic *efx, int tx_usecs, int rx_usecs,
			     bool rx_adaptive)
{
	struct efx_tx_queue *tx_queue;
	struct efx_rx_queue *rx_queue;

	EFX_ASSERT_RESET_SERIALISED(efx);

	efx_for_each_tx_queue(tx_queue, efx)
		tx_queue->channel->irq_moderation = tx_usecs;

	efx->irq_rx_adaptive = rx_adaptive;
	efx->irq_rx_moderation = rx_usecs;
	efx_for_each_rx_queue(rx_queue, efx)
		rx_queue->channel->irq_moderation = rx_usecs;
}

/**************************************************************************
 *
 * Hardware monitor
 *
 **************************************************************************/

/* Run periodically off the general workqueue. Serialised against
 * efx_reconfigure_port via the mac_lock */
static void efx_monitor(struct work_struct *data)
{
	struct efx_nic *efx = container_of(data, struct efx_nic,
					   monitor_work.work);
	int rc;

	EFX_TRACE(efx, "hardware monitor executing on CPU %d\n",
		  raw_smp_processor_id());

	/* If the mac_lock is already held then it is likely a port
	 * reconfiguration is already in place, which will likely do
	 * most of the work of check_hw() anyway. */
	if (!mutex_trylock(&efx->mac_lock))
		goto out_requeue;
	if (!efx->port_enabled)
		goto out_unlock;
	rc = efx->board_info.monitor(efx);
	if (rc) {
		EFX_ERR(efx, "Board sensor %s; shutting down PHY\n",
			(rc == -ERANGE) ? "reported fault" : "failed");
		efx->phy_mode |= PHY_MODE_LOW_POWER;
		falcon_sim_phy_event(efx);
	}
	efx->phy_op->poll(efx);
	efx->mac_op->poll(efx);

out_unlock:
	mutex_unlock(&efx->mac_lock);
out_requeue:
	queue_delayed_work(efx->workqueue, &efx->monitor_work,
			   efx_monitor_interval);
}

/**************************************************************************
 *
 * ioctls
 *
 *************************************************************************/

/* Net device ioctl
 * Context: process, rtnl_lock() held.
 */
static int efx_ioctl(struct net_device *net_dev, struct ifreq *ifr, int cmd)
{
	struct efx_nic *efx = netdev_priv(net_dev);

	EFX_ASSERT_RESET_SERIALISED(efx);

	return generic_mii_ioctl(&efx->mii, if_mii(ifr), cmd, NULL);
}

/**************************************************************************
 *
 * NAPI interface
 *
 **************************************************************************/

static int efx_init_napi(struct efx_nic *efx)
{
	struct efx_channel *channel;

	efx_for_each_channel(channel, efx) {
		channel->napi_dev = efx->net_dev;
	}
	return 0;
}

static void efx_fini_napi(struct efx_nic *efx)
{
	struct efx_channel *channel;

	efx_for_each_channel(channel, efx) {
		channel->napi_dev = NULL;
	}
}

/**************************************************************************
 *
 * Kernel netpoll interface
 *
 *************************************************************************/

#ifdef CONFIG_NET_POLL_CONTROLLER

/* Although in the common case interrupts will be disabled, this is not
 * guaranteed. However, all our work happens inside the NAPI callback,
 * so no locking is required.
 */
static void efx_netpoll(struct net_device *net_dev)
{
	struct efx_nic *efx = netdev_priv(net_dev);
	struct efx_channel *channel;

	efx_for_each_channel(channel, efx)
		efx_schedule_channel(channel);
}

#endif

/**************************************************************************
 *
 * Kernel net device interface
 *
 *************************************************************************/

/* Context: process, rtnl_lock() held. */
static int efx_net_open(struct net_device *net_dev)
{
	struct efx_nic *efx = netdev_priv(net_dev);
	EFX_ASSERT_RESET_SERIALISED(efx);

	EFX_LOG(efx, "opening device %s on CPU %d\n", net_dev->name,
		raw_smp_processor_id());

	if (efx->state == STATE_DISABLED)
		return -EIO;
	if (efx->phy_mode & PHY_MODE_SPECIAL)
		return -EBUSY;

	efx_start_all(efx);
	return 0;
}

/* Context: process, rtnl_lock() held.
 * Note that the kernel will ignore our return code; this method
 * should really be a void.
 */
static int efx_net_stop(struct net_device *net_dev)
{
	struct efx_nic *efx = netdev_priv(net_dev);

	EFX_LOG(efx, "closing %s on CPU %d\n", net_dev->name,
		raw_smp_processor_id());

	if (efx->state != STATE_DISABLED) {
		/* Stop the device and flush all the channels */
		efx_stop_all(efx);
		efx_fini_channels(efx);
		efx_init_channels(efx);
	}

	return 0;
}

void efx_stats_disable(struct efx_nic *efx)
{
	spin_lock(&efx->stats_lock);
	++efx->stats_disable_count;
	spin_unlock(&efx->stats_lock);
}

void efx_stats_enable(struct efx_nic *efx)
{
	spin_lock(&efx->stats_lock);
	--efx->stats_disable_count;
	spin_unlock(&efx->stats_lock);
}

/* Context: process, dev_base_lock or RTNL held, non-blocking. */
static struct net_device_stats *efx_net_stats(struct net_device *net_dev)
{
	struct efx_nic *efx = netdev_priv(net_dev);
	struct efx_mac_stats *mac_stats = &efx->mac_stats;
	struct net_device_stats *stats = &net_dev->stats;

	/* Update stats if possible, but do not wait if another thread
	 * is updating them or if MAC stats fetches are temporarily
	 * disabled; slightly stale stats are acceptable.
	 */
	if (!spin_trylock(&efx->stats_lock))
		return stats;
	if (!efx->stats_disable_count) {
		efx->mac_op->update_stats(efx);
		falcon_update_nic_stats(efx);
	}
	spin_unlock(&efx->stats_lock);

	stats->rx_packets = mac_stats->rx_packets;
	stats->tx_packets = mac_stats->tx_packets;
	stats->rx_bytes = mac_stats->rx_bytes;
	stats->tx_bytes = mac_stats->tx_bytes;
	stats->multicast = mac_stats->rx_multicast;
	stats->collisions = mac_stats->tx_collision;
	stats->rx_length_errors = (mac_stats->rx_gtjumbo +
				   mac_stats->rx_length_error);
	stats->rx_over_errors = efx->n_rx_nodesc_drop_cnt;
	stats->rx_crc_errors = mac_stats->rx_bad;
	stats->rx_frame_errors = mac_stats->rx_align_error;
	stats->rx_fifo_errors = mac_stats->rx_overflow;
	stats->rx_missed_errors = mac_stats->rx_missed;
	stats->tx_window_errors = mac_stats->tx_late_collision;

	stats->rx_errors = (stats->rx_length_errors +
			    stats->rx_over_errors +
			    stats->rx_crc_errors +
			    stats->rx_frame_errors +
			    stats->rx_fifo_errors +
			    stats->rx_missed_errors +
			    mac_stats->rx_symbol_error);
	stats->tx_errors = (stats->tx_window_errors +
			    mac_stats->tx_bad);

	return stats;
}

/* Context: netif_tx_lock held, BHs disabled. */
static void efx_watchdog(struct net_device *net_dev)
{
	struct efx_nic *efx = netdev_priv(net_dev);

	EFX_ERR(efx, "TX stuck with stop_count=%d port_enabled=%d:"
		" resetting channels\n",
		atomic_read(&efx->netif_stop_count), efx->port_enabled);

	efx_schedule_reset(efx, RESET_TYPE_TX_WATCHDOG);
}


/* Context: process, rtnl_lock() held. */
static int efx_change_mtu(struct net_device *net_dev, int new_mtu)
{
	struct efx_nic *efx = netdev_priv(net_dev);
	int rc = 0;

	EFX_ASSERT_RESET_SERIALISED(efx);

	if (new_mtu > EFX_MAX_MTU)
		return -EINVAL;

	efx_stop_all(efx);

	EFX_LOG(efx, "changing MTU to %d\n", new_mtu);

	efx_fini_channels(efx);
	net_dev->mtu = new_mtu;
	efx_init_channels(efx);

	efx_start_all(efx);
	return rc;
}

static int efx_set_mac_address(struct net_device *net_dev, void *data)
{
	struct efx_nic *efx = netdev_priv(net_dev);
	struct sockaddr *addr = data;
	char *new_addr = addr->sa_data;

	EFX_ASSERT_RESET_SERIALISED(efx);

	if (!is_valid_ether_addr(new_addr)) {
		EFX_ERR(efx, "invalid ethernet MAC address requested: %pM\n",
			new_addr);
		return -EINVAL;
	}

	memcpy(net_dev->dev_addr, new_addr, net_dev->addr_len);

	/* Reconfigure the MAC */
	efx_reconfigure_port(efx);

	return 0;
}

/* Context: netif_addr_lock held, BHs disabled. */
static void efx_set_multicast_list(struct net_device *net_dev)
{
	struct efx_nic *efx = netdev_priv(net_dev);
	struct dev_mc_list *mc_list = net_dev->mc_list;
	union efx_multicast_hash *mc_hash = &efx->multicast_hash;
	bool promiscuous = !!(net_dev->flags & IFF_PROMISC);
	bool changed = (efx->promiscuous != promiscuous);
	u32 crc;
	int bit;
	int i;

	efx->promiscuous = promiscuous;

	/* Build multicast hash table */
	if (promiscuous || (net_dev->flags & IFF_ALLMULTI)) {
		memset(mc_hash, 0xff, sizeof(*mc_hash));
	} else {
		memset(mc_hash, 0x00, sizeof(*mc_hash));
		for (i = 0; i < net_dev->mc_count; i++) {
			crc = ether_crc_le(ETH_ALEN, mc_list->dmi_addr);
			bit = crc & (EFX_MCAST_HASH_ENTRIES - 1);
			set_bit_le(bit, mc_hash->byte);
			mc_list = mc_list->next;
		}
	}

	if (!efx->port_enabled)
		/* Delay pushing settings until efx_start_port() */
		return;

	if (changed)
		queue_work(efx->workqueue, &efx->phy_work);

	/* Create and activate new global multicast hash table */
	falcon_set_multicast_hash(efx);
}

static const struct net_device_ops efx_netdev_ops = {
	.ndo_open		= efx_net_open,
	.ndo_stop		= efx_net_stop,
	.ndo_get_stats		= efx_net_stats,
	.ndo_tx_timeout		= efx_watchdog,
	.ndo_start_xmit		= efx_hard_start_xmit,
	.ndo_validate_addr	= eth_validate_addr,
	.ndo_do_ioctl		= efx_ioctl,
	.ndo_change_mtu		= efx_change_mtu,
	.ndo_set_mac_address	= efx_set_mac_address,
	.ndo_set_multicast_list = efx_set_multicast_list,
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller = efx_netpoll,
#endif
};

static void efx_update_name(struct efx_nic *efx)
{
	strcpy(efx->name, efx->net_dev->name);
	efx_mtd_rename(efx);
	efx_set_channel_names(efx);
}

static int efx_netdev_event(struct notifier_block *this,
			    unsigned long event, void *ptr)
{
	struct net_device *net_dev = ptr;

	if (net_dev->netdev_ops == &efx_netdev_ops &&
	    event == NETDEV_CHANGENAME)
		efx_update_name(netdev_priv(net_dev));

	return NOTIFY_DONE;
}

static struct notifier_block efx_netdev_notifier = {
	.notifier_call = efx_netdev_event,
};

static ssize_t
show_phy_type(struct device *dev, struct device_attribute *attr, char *buf)
{
	struct efx_nic *efx = pci_get_drvdata(to_pci_dev(dev));
	return sprintf(buf, "%d\n", efx->phy_type);
}
static DEVICE_ATTR(phy_type, 0644, show_phy_type, NULL);

static int efx_register_netdev(struct efx_nic *efx)
{
	struct net_device *net_dev = efx->net_dev;
	int rc;

	net_dev->watchdog_timeo = 5 * HZ;
	net_dev->irq = efx->pci_dev->irq;
	net_dev->netdev_ops = &efx_netdev_ops;
	SET_NETDEV_DEV(net_dev, &efx->pci_dev->dev);
	SET_ETHTOOL_OPS(net_dev, &efx_ethtool_ops);

	/* Always start with carrier off; PHY events will detect the link */
	netif_carrier_off(efx->net_dev);

	/* Clear MAC statistics */
	efx->mac_op->update_stats(efx);
	memset(&efx->mac_stats, 0, sizeof(efx->mac_stats));

	rc = register_netdev(net_dev);
	if (rc) {
		EFX_ERR(efx, "could not register net dev\n");
		return rc;
	}

	rtnl_lock();
	efx_update_name(efx);
	rtnl_unlock();

	rc = device_create_file(&efx->pci_dev->dev, &dev_attr_phy_type);
	if (rc) {
		EFX_ERR(efx, "failed to init net dev attributes\n");
		goto fail_registered;
	}

	return 0;

fail_registered:
	unregister_netdev(net_dev);
	return rc;
}

static void efx_unregister_netdev(struct efx_nic *efx)
{
	struct efx_tx_queue *tx_queue;

	if (!efx->net_dev)
		return;

	BUG_ON(netdev_priv(efx->net_dev) != efx);

	/* Free up any skbs still remaining. This has to happen before
	 * we try to unregister the netdev as running their destructors
	 * may be needed to get the device ref. count to 0. */
	efx_for_each_tx_queue(tx_queue, efx)
		efx_release_tx_buffers(tx_queue);

	if (efx_dev_registered(efx)) {
		strlcpy(efx->name, pci_name(efx->pci_dev), sizeof(efx->name));
		device_remove_file(&efx->pci_dev->dev, &dev_attr_phy_type);
		unregister_netdev(efx->net_dev);
	}
}

/**************************************************************************
 *
 * Device reset and suspend
 *
 **************************************************************************/

/* Tears down the entire software state and most of the hardware state
 * before reset.  */
void efx_reset_down(struct efx_nic *efx, enum reset_type method,
		    struct ethtool_cmd *ecmd)
{
	EFX_ASSERT_RESET_SERIALISED(efx);

	efx_stats_disable(efx);
	efx_stop_all(efx);
	mutex_lock(&efx->mac_lock);
	mutex_lock(&efx->spi_lock);

	efx->phy_op->get_settings(efx, ecmd);

	efx_fini_channels(efx);
	if (efx->port_initialized && method != RESET_TYPE_INVISIBLE)
		efx->phy_op->fini(efx);
}

/* This function will always ensure that the locks acquired in
 * efx_reset_down() are released. A failure return code indicates
 * that we were unable to reinitialise the hardware, and the
 * driver should be disabled. If ok is false, then the rx and tx
 * engines are not restarted, pending a RESET_DISABLE. */
int efx_reset_up(struct efx_nic *efx, enum reset_type method,
		 struct ethtool_cmd *ecmd, bool ok)
{
	int rc;

	EFX_ASSERT_RESET_SERIALISED(efx);

	rc = falcon_init_nic(efx);
	if (rc) {
		EFX_ERR(efx, "failed to initialise NIC\n");
		ok = false;
	}

	if (efx->port_initialized && method != RESET_TYPE_INVISIBLE) {
		if (ok) {
			rc = efx->phy_op->init(efx);
			if (rc)
				ok = false;
		}
		if (!ok)
			efx->port_initialized = false;
	}

	if (ok) {
		efx_init_channels(efx);

		if (efx->phy_op->set_settings(efx, ecmd))
			EFX_ERR(efx, "could not restore PHY settings\n");
	}

	mutex_unlock(&efx->spi_lock);
	mutex_unlock(&efx->mac_lock);

	if (ok) {
		efx_start_all(efx);
		efx_stats_enable(efx);
	}
	return rc;
}

/* Reset the NIC as transparently as possible. Do not reset the PHY
 * Note that the reset may fail, in which case the card will be left
 * in a most-probably-unusable state.
 *
 * This function will sleep.  You cannot reset from within an atomic
 * state; use efx_schedule_reset() instead.
 *
 * Grabs the rtnl_lock.
 */
static int efx_reset(struct efx_nic *efx)
{
	struct ethtool_cmd ecmd;
	enum reset_type method = efx->reset_pending;
	int rc = 0;

	/* Serialise with kernel interfaces */
	rtnl_lock();

	/* If we're not RUNNING then don't reset. Leave the reset_pending
	 * flag set so that efx_pci_probe_main will be retried */
	if (efx->state != STATE_RUNNING) {
		EFX_INFO(efx, "scheduled reset quenched. NIC not RUNNING\n");
		goto out_unlock;
	}

	EFX_INFO(efx, "resetting (%d)\n", method);

	efx_reset_down(efx, method, &ecmd);

	rc = falcon_reset_hw(efx, method);
	if (rc) {
		EFX_ERR(efx, "failed to reset hardware\n");
		goto out_disable;
	}

	/* Allow resets to be rescheduled. */
	efx->reset_pending = RESET_TYPE_NONE;

	/* Reinitialise bus-mastering, which may have been turned off before
	 * the reset was scheduled. This is still appropriate, even in the
	 * RESET_TYPE_DISABLE since this driver generally assumes the hardware
	 * can respond to requests. */
	pci_set_master(efx->pci_dev);

	/* Leave device stopped if necessary */
	if (method == RESET_TYPE_DISABLE) {
		efx_reset_up(efx, method, &ecmd, false);
		rc = -EIO;
	} else {
		rc = efx_reset_up(efx, method, &ecmd, true);
	}

out_disable:
	if (rc) {
		EFX_ERR(efx, "has been disabled\n");
		efx->state = STATE_DISABLED;
		dev_close(efx->net_dev);
	} else {
		EFX_LOG(efx, "reset complete\n");
	}

out_unlock:
	rtnl_unlock();
	return rc;
}

/* The worker thread exists so that code that cannot sleep can
 * schedule a reset for later.
 */
static void efx_reset_work(struct work_struct *data)
{
	struct efx_nic *nic = container_of(data, struct efx_nic, reset_work);

	efx_reset(nic);
}

void efx_schedule_reset(struct efx_nic *efx, enum reset_type type)
{
	enum reset_type method;

	if (efx->reset_pending != RESET_TYPE_NONE) {
		EFX_INFO(efx, "quenching already scheduled reset\n");
		return;
	}

	switch (type) {
	case RESET_TYPE_INVISIBLE:
	case RESET_TYPE_ALL:
	case RESET_TYPE_WORLD:
	case RESET_TYPE_DISABLE:
		method = type;
		break;
	case RESET_TYPE_RX_RECOVERY:
	case RESET_TYPE_RX_DESC_FETCH:
	case RESET_TYPE_TX_DESC_FETCH:
	case RESET_TYPE_TX_SKIP:
		method = RESET_TYPE_INVISIBLE;
		break;
	default:
		method = RESET_TYPE_ALL;
		break;
	}

	if (method != type)
		EFX_LOG(efx, "scheduling reset (%d:%d)\n", type, method);
	else
		EFX_LOG(efx, "scheduling reset (%d)\n", method);

	efx->reset_pending = method;

	queue_work(reset_workqueue, &efx->reset_work);
}

/**************************************************************************
 *
 * List of NICs we support
 *
 **************************************************************************/

/* PCI device ID table */
static struct pci_device_id efx_pci_table[] __devinitdata = {
	{PCI_DEVICE(EFX_VENDID_SFC, FALCON_A_P_DEVID),
	 .driver_data = (unsigned long) &falcon_a_nic_type},
	{PCI_DEVICE(EFX_VENDID_SFC, FALCON_B_P_DEVID),
	 .driver_data = (unsigned long) &falcon_b_nic_type},
	{0}			/* end of list */
};

/**************************************************************************
 *
 * Dummy PHY/MAC/Board operations
 *
 * Can be used for some unimplemented operations
 * Needed so all function pointers are valid and do not have to be tested
 * before use
 *
 **************************************************************************/
int efx_port_dummy_op_int(struct efx_nic *efx)
{
	return 0;
}
void efx_port_dummy_op_void(struct efx_nic *efx) {}
void efx_port_dummy_op_blink(struct efx_nic *efx, bool blink) {}

static struct efx_mac_operations efx_dummy_mac_operations = {
	.reconfigure	= efx_port_dummy_op_void,
	.poll		= efx_port_dummy_op_void,
	.irq		= efx_port_dummy_op_void,
};

static struct efx_phy_operations efx_dummy_phy_operations = {
	.init		 = efx_port_dummy_op_int,
	.reconfigure	 = efx_port_dummy_op_void,
	.poll		 = efx_port_dummy_op_void,
	.fini		 = efx_port_dummy_op_void,
	.clear_interrupt = efx_port_dummy_op_void,
};

static struct efx_board efx_dummy_board_info = {
	.init		= efx_port_dummy_op_int,
	.init_leds	= efx_port_dummy_op_void,
	.set_id_led	= efx_port_dummy_op_blink,
	.monitor	= efx_port_dummy_op_int,
	.blink		= efx_port_dummy_op_blink,
	.fini		= efx_port_dummy_op_void,
};

/**************************************************************************
 *
 * Data housekeeping
 *
 **************************************************************************/

/* This zeroes out and then fills in the invariants in a struct
 * efx_nic (including all sub-structures).
 */
static int efx_init_struct(struct efx_nic *efx, struct efx_nic_type *type,
			   struct pci_dev *pci_dev, struct net_device *net_dev)
{
	struct efx_channel *channel;
	struct efx_tx_queue *tx_queue;
	struct efx_rx_queue *rx_queue;
	int i;

	/* Initialise common structures */
	memset(efx, 0, sizeof(*efx));
	spin_lock_init(&efx->biu_lock);
	spin_lock_init(&efx->phy_lock);
	mutex_init(&efx->spi_lock);
	INIT_WORK(&efx->reset_work, efx_reset_work);
	INIT_DELAYED_WORK(&efx->monitor_work, efx_monitor);
	efx->pci_dev = pci_dev;
	efx->state = STATE_INIT;
	efx->reset_pending = RESET_TYPE_NONE;
	strlcpy(efx->name, pci_name(pci_dev), sizeof(efx->name));
	efx->board_info = efx_dummy_board_info;

	efx->net_dev = net_dev;
	efx->rx_checksum_enabled = true;
	spin_lock_init(&efx->netif_stop_lock);
	spin_lock_init(&efx->stats_lock);
	efx->stats_disable_count = 1;
	mutex_init(&efx->mac_lock);
	efx->mac_op = &efx_dummy_mac_operations;
	efx->phy_op = &efx_dummy_phy_operations;
	efx->mii.dev = net_dev;
	INIT_WORK(&efx->phy_work, efx_phy_work);
	INIT_WORK(&efx->mac_work, efx_mac_work);
	atomic_set(&efx->netif_stop_count, 1);

	for (i = 0; i < EFX_MAX_CHANNELS; i++) {
		channel = &efx->channel[i];
		channel->efx = efx;
		channel->channel = i;
		channel->work_pending = false;
	}
	for (i = 0; i < EFX_TX_QUEUE_COUNT; i++) {
		tx_queue = &efx->tx_queue[i];
		tx_queue->efx = efx;
		tx_queue->queue = i;
		tx_queue->buffer = NULL;
		tx_queue->channel = &efx->channel[0]; /* for safety */
		tx_queue->tso_headers_free = NULL;
	}
	for (i = 0; i < EFX_MAX_RX_QUEUES; i++) {
		rx_queue = &efx->rx_queue[i];
		rx_queue->efx = efx;
		rx_queue->queue = i;
		rx_queue->channel = &efx->channel[0]; /* for safety */
		rx_queue->buffer = NULL;
		spin_lock_init(&rx_queue->add_lock);
		INIT_DELAYED_WORK(&rx_queue->work, efx_rx_work);
	}

	efx->type = type;

	/* Sanity-check NIC type */
	EFX_BUG_ON_PARANOID(efx->type->txd_ring_mask &
			    (efx->type->txd_ring_mask + 1));
	EFX_BUG_ON_PARANOID(efx->type->rxd_ring_mask &
			    (efx->type->rxd_ring_mask + 1));
	EFX_BUG_ON_PARANOID(efx->type->evq_size &
			    (efx->type->evq_size - 1));
	/* As close as we can get to guaranteeing that we don't overflow */
	EFX_BUG_ON_PARANOID(efx->type->evq_size <
			    (efx->type->txd_ring_mask + 1 +
			     efx->type->rxd_ring_mask + 1));
	EFX_BUG_ON_PARANOID(efx->type->phys_addr_channels > EFX_MAX_CHANNELS);

	/* Higher numbered interrupt modes are less capable! */
	efx->interrupt_mode = max(efx->type->max_interrupt_mode,
				  interrupt_mode);

	/* Would be good to use the net_dev name, but we're too early */
	snprintf(efx->workqueue_name, sizeof(efx->workqueue_name), "sfc%s",
		 pci_name(pci_dev));
	efx->workqueue = create_singlethread_workqueue(efx->workqueue_name);
	if (!efx->workqueue)
		return -ENOMEM;

	return 0;
}

static void efx_fini_struct(struct efx_nic *efx)
{
	if (efx->workqueue) {
		destroy_workqueue(efx->workqueue);
		efx->workqueue = NULL;
	}
}

/**************************************************************************
 *
 * PCI interface
 *
 **************************************************************************/

/* Main body of final NIC shutdown code
 * This is called only at module unload (or hotplug removal).
 */
static void efx_pci_remove_main(struct efx_nic *efx)
{
	EFX_ASSERT_RESET_SERIALISED(efx);

	/* Skip everything if we never obtained a valid membase */
	if (!efx->membase)
		return;

	efx_fini_channels(efx);
	efx_fini_port(efx);

	/* Shutdown the board, then the NIC and board state */
	efx->board_info.fini(efx);
	falcon_fini_interrupt(efx);

	efx_fini_napi(efx);
	efx_remove_all(efx);
}

/* Final NIC shutdown
 * This is called only at module unload (or hotplug removal).
 */
static void efx_pci_remove(struct pci_dev *pci_dev)
{
	struct efx_nic *efx;

	efx = pci_get_drvdata(pci_dev);
	if (!efx)
		return;

	/* Mark the NIC as fini, then stop the interface */
	rtnl_lock();
	efx->state = STATE_FINI;
	dev_close(efx->net_dev);

	/* Allow any queued efx_resets() to complete */
	rtnl_unlock();

	if (efx->membase == NULL)
		goto out;

	efx_unregister_netdev(efx);

	efx_mtd_remove(efx);

	/* Wait for any scheduled resets to complete. No more will be
	 * scheduled from this point because efx_stop_all() has been
	 * called, we are no longer registered with driverlink, and
	 * the net_device's have been removed. */
	cancel_work_sync(&efx->reset_work);

	efx_pci_remove_main(efx);

out:
	efx_fini_io(efx);
	EFX_LOG(efx, "shutdown successful\n");

	pci_set_drvdata(pci_dev, NULL);
	efx_fini_struct(efx);
	free_netdev(efx->net_dev);
};

/* Main body of NIC initialisation
 * This is called at module load (or hotplug insertion, theoretically).
 */
static int efx_pci_probe_main(struct efx_nic *efx)
{
	int rc;

	/* Do start-of-day initialisation */
	rc = efx_probe_all(efx);
	if (rc)
		goto fail1;

	rc = efx_init_napi(efx);
	if (rc)
		goto fail2;

	/* Initialise the board */
	rc = efx->board_info.init(efx);
	if (rc) {
		EFX_ERR(efx, "failed to initialise board\n");
		goto fail3;
	}

	rc = falcon_init_nic(efx);
	if (rc) {
		EFX_ERR(efx, "failed to initialise NIC\n");
		goto fail4;
	}

	rc = efx_init_port(efx);
	if (rc) {
		EFX_ERR(efx, "failed to initialise port\n");
		goto fail5;
	}

	efx_init_channels(efx);

	rc = falcon_init_interrupt(efx);
	if (rc)
		goto fail6;

	return 0;

 fail6:
	efx_fini_channels(efx);
	efx_fini_port(efx);
 fail5:
 fail4:
	efx->board_info.fini(efx);
 fail3:
	efx_fini_napi(efx);
 fail2:
	efx_remove_all(efx);
 fail1:
	return rc;
}

/* NIC initialisation
 *
 * This is called at module load (or hotplug insertion,
 * theoretically).  It sets up PCI mappings, tests and resets the NIC,
 * sets up and registers the network devices with the kernel and hooks
 * the interrupt service routine.  It does not prepare the device for
 * transmission; this is left to the first time one of the network
 * interfaces is brought up (i.e. efx_net_open).
 */
static int __devinit efx_pci_probe(struct pci_dev *pci_dev,
				   const struct pci_device_id *entry)
{
	struct efx_nic_type *type = (struct efx_nic_type *) entry->driver_data;
	struct net_device *net_dev;
	struct efx_nic *efx;
	int i, rc;

	/* Allocate and initialise a struct net_device and struct efx_nic */
	net_dev = alloc_etherdev(sizeof(*efx));
	if (!net_dev)
		return -ENOMEM;
	net_dev->features |= (NETIF_F_IP_CSUM | NETIF_F_SG |
			      NETIF_F_HIGHDMA | NETIF_F_TSO);
	if (lro)
		net_dev->features |= NETIF_F_GRO;
	/* Mask for features that also apply to VLAN devices */
	net_dev->vlan_features |= (NETIF_F_ALL_CSUM | NETIF_F_SG |
				   NETIF_F_HIGHDMA | NETIF_F_TSO);
	efx = netdev_priv(net_dev);
	pci_set_drvdata(pci_dev, efx);
	rc = efx_init_struct(efx, type, pci_dev, net_dev);
	if (rc)
		goto fail1;

	EFX_INFO(efx, "Solarflare Communications NIC detected\n");

	/* Set up basic I/O (BAR mappings etc) */
	rc = efx_init_io(efx);
	if (rc)
		goto fail2;

	/* No serialisation is required with the reset path because
	 * we're in STATE_INIT. */
	for (i = 0; i < 5; i++) {
		rc = efx_pci_probe_main(efx);

		/* Serialise against efx_reset(). No more resets will be
		 * scheduled since efx_stop_all() has been called, and we
		 * have not and never have been registered with either
		 * the rtnetlink or driverlink layers. */
		cancel_work_sync(&efx->reset_work);

		if (rc == 0) {
			if (efx->reset_pending != RESET_TYPE_NONE) {
				/* If there was a scheduled reset during
				 * probe, the NIC is probably hosed anyway */
				efx_pci_remove_main(efx);
				rc = -EIO;
			} else {
				break;
			}
		}

		/* Retry if a recoverably reset event has been scheduled */
		if ((efx->reset_pending != RESET_TYPE_INVISIBLE) &&
		    (efx->reset_pending != RESET_TYPE_ALL))
			goto fail3;

		efx->reset_pending = RESET_TYPE_NONE;
	}

	if (rc) {
		EFX_ERR(efx, "Could not reset NIC\n");
		goto fail4;
	}

	/* Switch to the running state before we expose the device to
	 * the OS.  This is to ensure that the initial gathering of
	 * MAC stats succeeds. */
	efx->state = STATE_RUNNING;

	efx_mtd_probe(efx); /* allowed to fail */

	rc = efx_register_netdev(efx);
	if (rc)
		goto fail5;

	EFX_LOG(efx, "initialisation successful\n");
	return 0;

 fail5:
	efx_pci_remove_main(efx);
 fail4:
 fail3:
	efx_fini_io(efx);
 fail2:
	efx_fini_struct(efx);
 fail1:
	EFX_LOG(efx, "initialisation failed. rc=%d\n", rc);
	free_netdev(net_dev);
	return rc;
}

static struct pci_driver efx_pci_driver = {
	.name		= EFX_DRIVER_NAME,
	.id_table	= efx_pci_table,
	.probe		= efx_pci_probe,
	.remove		= efx_pci_remove,
};

/**************************************************************************
 *
 * Kernel module interface
 *
 *************************************************************************/

module_param(interrupt_mode, uint, 0444);
MODULE_PARM_DESC(interrupt_mode,
		 "Interrupt mode (0=>MSIX 1=>MSI 2=>legacy)");

static int __init efx_init_module(void)
{
	int rc;

	printk(KERN_INFO "Solarflare NET driver v" EFX_DRIVER_VERSION "\n");

	rc = register_netdevice_notifier(&efx_netdev_notifier);
	if (rc)
		goto err_notifier;

	refill_workqueue = create_workqueue("sfc_refill");
	if (!refill_workqueue) {
		rc = -ENOMEM;
		goto err_refill;
	}
	reset_workqueue = create_singlethread_workqueue("sfc_reset");
	if (!reset_workqueue) {
		rc = -ENOMEM;
		goto err_reset;
	}

	rc = pci_register_driver(&efx_pci_driver);
	if (rc < 0)
		goto err_pci;

	return 0;

 err_pci:
	destroy_workqueue(reset_workqueue);
 err_reset:
	destroy_workqueue(refill_workqueue);
 err_refill:
	unregister_netdevice_notifier(&efx_netdev_notifier);
 err_notifier:
	return rc;
}

static void __exit efx_exit_module(void)
{
	printk(KERN_INFO "Solarflare NET driver unloading\n");

	pci_unregister_driver(&efx_pci_driver);
	destroy_workqueue(reset_workqueue);
	destroy_workqueue(refill_workqueue);
	unregister_netdevice_notifier(&efx_netdev_notifier);

}

module_init(efx_init_module);
module_exit(efx_exit_module);

MODULE_AUTHOR("Michael Brown <mbrown@fensystems.co.uk> and "
	      "Solarflare Communications");
MODULE_DESCRIPTION("Solarflare Communications network driver");
MODULE_LICENSE("GPL");
MODULE_DEVICE_TABLE(pci, efx_pci_table);