aboutsummaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/sfc/mcdi.c
blob: 4b0bd8a1514dbb7035c7c0562cefc072901e585c (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
/****************************************************************************
 * Driver for Solarflare network controllers and boards
 * Copyright 2008-2013 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/delay.h>
#include <asm/cmpxchg.h>
#include "net_driver.h"
#include "nic.h"
#include "io.h"
#include "farch_regs.h"
#include "mcdi_pcol.h"
#include "phy.h"

/**************************************************************************
 *
 * Management-Controller-to-Driver Interface
 *
 **************************************************************************
 */

#define MCDI_RPC_TIMEOUT       (10 * HZ)

/* A reboot/assertion causes the MCDI status word to be set after the
 * command word is set or a REBOOT event is sent. If we notice a reboot
 * via these mechanisms then wait 250ms for the status word to be set.
 */
#define MCDI_STATUS_DELAY_US		100
#define MCDI_STATUS_DELAY_COUNT		2500
#define MCDI_STATUS_SLEEP_MS						\
	(MCDI_STATUS_DELAY_US * MCDI_STATUS_DELAY_COUNT / 1000)

#define SEQ_MASK							\
	EFX_MASK32(EFX_WIDTH(MCDI_HEADER_SEQ))

struct efx_mcdi_async_param {
	struct list_head list;
	unsigned int cmd;
	size_t inlen;
	size_t outlen;
	efx_mcdi_async_completer *complete;
	unsigned long cookie;
	/* followed by request/response buffer */
};

static void efx_mcdi_timeout_async(unsigned long context);
static int efx_mcdi_drv_attach(struct efx_nic *efx, bool driver_operating,
			       bool *was_attached_out);
static bool efx_mcdi_poll_once(struct efx_nic *efx);

static inline struct efx_mcdi_iface *efx_mcdi(struct efx_nic *efx)
{
	EFX_BUG_ON_PARANOID(!efx->mcdi);
	return &efx->mcdi->iface;
}

int efx_mcdi_init(struct efx_nic *efx)
{
	struct efx_mcdi_iface *mcdi;
	bool already_attached;
	int rc;

	efx->mcdi = kzalloc(sizeof(*efx->mcdi), GFP_KERNEL);
	if (!efx->mcdi)
		return -ENOMEM;

	mcdi = efx_mcdi(efx);
	mcdi->efx = efx;
	init_waitqueue_head(&mcdi->wq);
	spin_lock_init(&mcdi->iface_lock);
	mcdi->state = MCDI_STATE_QUIESCENT;
	mcdi->mode = MCDI_MODE_POLL;
	spin_lock_init(&mcdi->async_lock);
	INIT_LIST_HEAD(&mcdi->async_list);
	setup_timer(&mcdi->async_timer, efx_mcdi_timeout_async,
		    (unsigned long)mcdi);

	(void) efx_mcdi_poll_reboot(efx);
	mcdi->new_epoch = true;

	/* Recover from a failed assertion before probing */
	rc = efx_mcdi_handle_assertion(efx);
	if (rc)
		return rc;

	/* Let the MC (and BMC, if this is a LOM) know that the driver
	 * is loaded. We should do this before we reset the NIC.
	 */
	rc = efx_mcdi_drv_attach(efx, true, &already_attached);
	if (rc) {
		netif_err(efx, probe, efx->net_dev,
			  "Unable to register driver with MCPU\n");
		return rc;
	}
	if (already_attached)
		/* Not a fatal error */
		netif_err(efx, probe, efx->net_dev,
			  "Host already registered with MCPU\n");

	return 0;
}

void efx_mcdi_fini(struct efx_nic *efx)
{
	if (!efx->mcdi)
		return;

	BUG_ON(efx->mcdi->iface.state != MCDI_STATE_QUIESCENT);

	/* Relinquish the device (back to the BMC, if this is a LOM) */
	efx_mcdi_drv_attach(efx, false, NULL);

	kfree(efx->mcdi);
}

static void efx_mcdi_send_request(struct efx_nic *efx, unsigned cmd,
				  const efx_dword_t *inbuf, size_t inlen)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);
	efx_dword_t hdr[2];
	size_t hdr_len;
	u32 xflags, seqno;

	BUG_ON(mcdi->state == MCDI_STATE_QUIESCENT);

	/* Serialise with efx_mcdi_ev_cpl() and efx_mcdi_ev_death() */
	spin_lock_bh(&mcdi->iface_lock);
	++mcdi->seqno;
	spin_unlock_bh(&mcdi->iface_lock);

	seqno = mcdi->seqno & SEQ_MASK;
	xflags = 0;
	if (mcdi->mode == MCDI_MODE_EVENTS)
		xflags |= MCDI_HEADER_XFLAGS_EVREQ;

	if (efx->type->mcdi_max_ver == 1) {
		/* MCDI v1 */
		EFX_POPULATE_DWORD_7(hdr[0],
				     MCDI_HEADER_RESPONSE, 0,
				     MCDI_HEADER_RESYNC, 1,
				     MCDI_HEADER_CODE, cmd,
				     MCDI_HEADER_DATALEN, inlen,
				     MCDI_HEADER_SEQ, seqno,
				     MCDI_HEADER_XFLAGS, xflags,
				     MCDI_HEADER_NOT_EPOCH, !mcdi->new_epoch);
		hdr_len = 4;
	} else {
		/* MCDI v2 */
		BUG_ON(inlen > MCDI_CTL_SDU_LEN_MAX_V2);
		EFX_POPULATE_DWORD_7(hdr[0],
				     MCDI_HEADER_RESPONSE, 0,
				     MCDI_HEADER_RESYNC, 1,
				     MCDI_HEADER_CODE, MC_CMD_V2_EXTN,
				     MCDI_HEADER_DATALEN, 0,
				     MCDI_HEADER_SEQ, seqno,
				     MCDI_HEADER_XFLAGS, xflags,
				     MCDI_HEADER_NOT_EPOCH, !mcdi->new_epoch);
		EFX_POPULATE_DWORD_2(hdr[1],
				     MC_CMD_V2_EXTN_IN_EXTENDED_CMD, cmd,
				     MC_CMD_V2_EXTN_IN_ACTUAL_LEN, inlen);
		hdr_len = 8;
	}

	efx->type->mcdi_request(efx, hdr, hdr_len, inbuf, inlen);

	mcdi->new_epoch = false;
}

static int efx_mcdi_errno(unsigned int mcdi_err)
{
	switch (mcdi_err) {
	case 0:
		return 0;
#define TRANSLATE_ERROR(name)					\
	case MC_CMD_ERR_ ## name:				\
		return -name;
	TRANSLATE_ERROR(EPERM);
	TRANSLATE_ERROR(ENOENT);
	TRANSLATE_ERROR(EINTR);
	TRANSLATE_ERROR(EAGAIN);
	TRANSLATE_ERROR(EACCES);
	TRANSLATE_ERROR(EBUSY);
	TRANSLATE_ERROR(EINVAL);
	TRANSLATE_ERROR(EDEADLK);
	TRANSLATE_ERROR(ENOSYS);
	TRANSLATE_ERROR(ETIME);
	TRANSLATE_ERROR(EALREADY);
	TRANSLATE_ERROR(ENOSPC);
#undef TRANSLATE_ERROR
	case MC_CMD_ERR_ALLOC_FAIL:
		return -ENOBUFS;
	case MC_CMD_ERR_MAC_EXIST:
		return -EADDRINUSE;
	default:
		return -EPROTO;
	}
}

static void efx_mcdi_read_response_header(struct efx_nic *efx)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);
	unsigned int respseq, respcmd, error;
	efx_dword_t hdr;

	efx->type->mcdi_read_response(efx, &hdr, 0, 4);
	respseq = EFX_DWORD_FIELD(hdr, MCDI_HEADER_SEQ);
	respcmd = EFX_DWORD_FIELD(hdr, MCDI_HEADER_CODE);
	error = EFX_DWORD_FIELD(hdr, MCDI_HEADER_ERROR);

	if (respcmd != MC_CMD_V2_EXTN) {
		mcdi->resp_hdr_len = 4;
		mcdi->resp_data_len = EFX_DWORD_FIELD(hdr, MCDI_HEADER_DATALEN);
	} else {
		efx->type->mcdi_read_response(efx, &hdr, 4, 4);
		mcdi->resp_hdr_len = 8;
		mcdi->resp_data_len =
			EFX_DWORD_FIELD(hdr, MC_CMD_V2_EXTN_IN_ACTUAL_LEN);
	}

	if (error && mcdi->resp_data_len == 0) {
		netif_err(efx, hw, efx->net_dev, "MC rebooted\n");
		mcdi->resprc = -EIO;
	} else if ((respseq ^ mcdi->seqno) & SEQ_MASK) {
		netif_err(efx, hw, efx->net_dev,
			  "MC response mismatch tx seq 0x%x rx seq 0x%x\n",
			  respseq, mcdi->seqno);
		mcdi->resprc = -EIO;
	} else if (error) {
		efx->type->mcdi_read_response(efx, &hdr, mcdi->resp_hdr_len, 4);
		mcdi->resprc =
			efx_mcdi_errno(EFX_DWORD_FIELD(hdr, EFX_DWORD_0));
	} else {
		mcdi->resprc = 0;
	}
}

static bool efx_mcdi_poll_once(struct efx_nic *efx)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);

	rmb();
	if (!efx->type->mcdi_poll_response(efx))
		return false;

	spin_lock_bh(&mcdi->iface_lock);
	efx_mcdi_read_response_header(efx);
	spin_unlock_bh(&mcdi->iface_lock);

	return true;
}

static int efx_mcdi_poll(struct efx_nic *efx)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);
	unsigned long time, finish;
	unsigned int spins;
	int rc;

	/* Check for a reboot atomically with respect to efx_mcdi_copyout() */
	rc = efx_mcdi_poll_reboot(efx);
	if (rc) {
		spin_lock_bh(&mcdi->iface_lock);
		mcdi->resprc = rc;
		mcdi->resp_hdr_len = 0;
		mcdi->resp_data_len = 0;
		spin_unlock_bh(&mcdi->iface_lock);
		return 0;
	}

	/* Poll for completion. Poll quickly (once a us) for the 1st jiffy,
	 * because generally mcdi responses are fast. After that, back off
	 * and poll once a jiffy (approximately)
	 */
	spins = TICK_USEC;
	finish = jiffies + MCDI_RPC_TIMEOUT;

	while (1) {
		if (spins != 0) {
			--spins;
			udelay(1);
		} else {
			schedule_timeout_uninterruptible(1);
		}

		time = jiffies;

		if (efx_mcdi_poll_once(efx))
			break;

		if (time_after(time, finish))
			return -ETIMEDOUT;
	}

	/* Return rc=0 like wait_event_timeout() */
	return 0;
}

/* Test and clear MC-rebooted flag for this port/function; reset
 * software state as necessary.
 */
int efx_mcdi_poll_reboot(struct efx_nic *efx)
{
	if (!efx->mcdi)
		return 0;

	return efx->type->mcdi_poll_reboot(efx);
}

static bool efx_mcdi_acquire_async(struct efx_mcdi_iface *mcdi)
{
	return cmpxchg(&mcdi->state,
		       MCDI_STATE_QUIESCENT, MCDI_STATE_RUNNING_ASYNC) ==
		MCDI_STATE_QUIESCENT;
}

static void efx_mcdi_acquire_sync(struct efx_mcdi_iface *mcdi)
{
	/* Wait until the interface becomes QUIESCENT and we win the race
	 * to mark it RUNNING_SYNC.
	 */
	wait_event(mcdi->wq,
		   cmpxchg(&mcdi->state,
			   MCDI_STATE_QUIESCENT, MCDI_STATE_RUNNING_SYNC) ==
		   MCDI_STATE_QUIESCENT);
}

static int efx_mcdi_await_completion(struct efx_nic *efx)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);

	if (wait_event_timeout(mcdi->wq, mcdi->state == MCDI_STATE_COMPLETED,
			       MCDI_RPC_TIMEOUT) == 0)
		return -ETIMEDOUT;

	/* Check if efx_mcdi_set_mode() switched us back to polled completions.
	 * In which case, poll for completions directly. If efx_mcdi_ev_cpl()
	 * completed the request first, then we'll just end up completing the
	 * request again, which is safe.
	 *
	 * We need an smp_rmb() to synchronise with efx_mcdi_mode_poll(), which
	 * wait_event_timeout() implicitly provides.
	 */
	if (mcdi->mode == MCDI_MODE_POLL)
		return efx_mcdi_poll(efx);

	return 0;
}

/* If the interface is RUNNING_SYNC, switch to COMPLETED and wake the
 * requester.  Return whether this was done.  Does not take any locks.
 */
static bool efx_mcdi_complete_sync(struct efx_mcdi_iface *mcdi)
{
	if (cmpxchg(&mcdi->state,
		    MCDI_STATE_RUNNING_SYNC, MCDI_STATE_COMPLETED) ==
	    MCDI_STATE_RUNNING_SYNC) {
		wake_up(&mcdi->wq);
		return true;
	}

	return false;
}

static void efx_mcdi_release(struct efx_mcdi_iface *mcdi)
{
	if (mcdi->mode == MCDI_MODE_EVENTS) {
		struct efx_mcdi_async_param *async;
		struct efx_nic *efx = mcdi->efx;

		/* Process the asynchronous request queue */
		spin_lock_bh(&mcdi->async_lock);
		async = list_first_entry_or_null(
			&mcdi->async_list, struct efx_mcdi_async_param, list);
		if (async) {
			mcdi->state = MCDI_STATE_RUNNING_ASYNC;
			efx_mcdi_send_request(efx, async->cmd,
					      (const efx_dword_t *)(async + 1),
					      async->inlen);
			mod_timer(&mcdi->async_timer,
				  jiffies + MCDI_RPC_TIMEOUT);
		}
		spin_unlock_bh(&mcdi->async_lock);

		if (async)
			return;
	}

	mcdi->state = MCDI_STATE_QUIESCENT;
	wake_up(&mcdi->wq);
}

/* If the interface is RUNNING_ASYNC, switch to COMPLETED, call the
 * asynchronous completion function, and release the interface.
 * Return whether this was done.  Must be called in bh-disabled
 * context.  Will take iface_lock and async_lock.
 */
static bool efx_mcdi_complete_async(struct efx_mcdi_iface *mcdi, bool timeout)
{
	struct efx_nic *efx = mcdi->efx;
	struct efx_mcdi_async_param *async;
	size_t hdr_len, data_len;
	efx_dword_t *outbuf;
	int rc;

	if (cmpxchg(&mcdi->state,
		    MCDI_STATE_RUNNING_ASYNC, MCDI_STATE_COMPLETED) !=
	    MCDI_STATE_RUNNING_ASYNC)
		return false;

	spin_lock(&mcdi->iface_lock);
	if (timeout) {
		/* Ensure that if the completion event arrives later,
		 * the seqno check in efx_mcdi_ev_cpl() will fail
		 */
		++mcdi->seqno;
		++mcdi->credits;
		rc = -ETIMEDOUT;
		hdr_len = 0;
		data_len = 0;
	} else {
		rc = mcdi->resprc;
		hdr_len = mcdi->resp_hdr_len;
		data_len = mcdi->resp_data_len;
	}
	spin_unlock(&mcdi->iface_lock);

	/* Stop the timer.  In case the timer function is running, we
	 * must wait for it to return so that there is no possibility
	 * of it aborting the next request.
	 */
	if (!timeout)
		del_timer_sync(&mcdi->async_timer);

	spin_lock(&mcdi->async_lock);
	async = list_first_entry(&mcdi->async_list,
				 struct efx_mcdi_async_param, list);
	list_del(&async->list);
	spin_unlock(&mcdi->async_lock);

	outbuf = (efx_dword_t *)(async + 1);
	efx->type->mcdi_read_response(efx, outbuf, hdr_len,
				      min(async->outlen, data_len));
	async->complete(efx, async->cookie, rc, outbuf, data_len);
	kfree(async);

	efx_mcdi_release(mcdi);

	return true;
}

static void efx_mcdi_ev_cpl(struct efx_nic *efx, unsigned int seqno,
			    unsigned int datalen, unsigned int mcdi_err)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);
	bool wake = false;

	spin_lock(&mcdi->iface_lock);

	if ((seqno ^ mcdi->seqno) & SEQ_MASK) {
		if (mcdi->credits)
			/* The request has been cancelled */
			--mcdi->credits;
		else
			netif_err(efx, hw, efx->net_dev,
				  "MC response mismatch tx seq 0x%x rx "
				  "seq 0x%x\n", seqno, mcdi->seqno);
	} else {
		if (efx->type->mcdi_max_ver >= 2) {
			/* MCDI v2 responses don't fit in an event */
			efx_mcdi_read_response_header(efx);
		} else {
			mcdi->resprc = efx_mcdi_errno(mcdi_err);
			mcdi->resp_hdr_len = 4;
			mcdi->resp_data_len = datalen;
		}

		wake = true;
	}

	spin_unlock(&mcdi->iface_lock);

	if (wake) {
		if (!efx_mcdi_complete_async(mcdi, false))
			(void) efx_mcdi_complete_sync(mcdi);

		/* If the interface isn't RUNNING_ASYNC or
		 * RUNNING_SYNC then we've received a duplicate
		 * completion after we've already transitioned back to
		 * QUIESCENT. [A subsequent invocation would increment
		 * seqno, so would have failed the seqno check].
		 */
	}
}

static void efx_mcdi_timeout_async(unsigned long context)
{
	struct efx_mcdi_iface *mcdi = (struct efx_mcdi_iface *)context;

	efx_mcdi_complete_async(mcdi, true);
}

static int
efx_mcdi_check_supported(struct efx_nic *efx, unsigned int cmd, size_t inlen)
{
	if (efx->type->mcdi_max_ver < 0 ||
	     (efx->type->mcdi_max_ver < 2 &&
	      cmd > MC_CMD_CMD_SPACE_ESCAPE_7))
		return -EINVAL;

	if (inlen > MCDI_CTL_SDU_LEN_MAX_V2 ||
	    (efx->type->mcdi_max_ver < 2 &&
	     inlen > MCDI_CTL_SDU_LEN_MAX_V1))
		return -EMSGSIZE;

	return 0;
}

int efx_mcdi_rpc(struct efx_nic *efx, unsigned cmd,
		 const efx_dword_t *inbuf, size_t inlen,
		 efx_dword_t *outbuf, size_t outlen,
		 size_t *outlen_actual)
{
	int rc;

	rc = efx_mcdi_rpc_start(efx, cmd, inbuf, inlen);
	if (rc)
		return rc;
	return efx_mcdi_rpc_finish(efx, cmd, inlen,
				   outbuf, outlen, outlen_actual);
}

int efx_mcdi_rpc_start(struct efx_nic *efx, unsigned cmd,
		       const efx_dword_t *inbuf, size_t inlen)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);
	int rc;

	rc = efx_mcdi_check_supported(efx, cmd, inlen);
	if (rc)
		return rc;

	efx_mcdi_acquire_sync(mcdi);
	efx_mcdi_send_request(efx, cmd, inbuf, inlen);
	return 0;
}

/**
 * efx_mcdi_rpc_async - Schedule an MCDI command to run asynchronously
 * @efx: NIC through which to issue the command
 * @cmd: Command type number
 * @inbuf: Command parameters
 * @inlen: Length of command parameters, in bytes
 * @outlen: Length to allocate for response buffer, in bytes
 * @complete: Function to be called on completion or cancellation.
 * @cookie: Arbitrary value to be passed to @complete.
 *
 * This function does not sleep and therefore may be called in atomic
 * context.  It will fail if event queues are disabled or if MCDI
 * event completions have been disabled due to an error.
 *
 * If it succeeds, the @complete function will be called exactly once
 * in atomic context, when one of the following occurs:
 * (a) the completion event is received (in NAPI context)
 * (b) event queues are disabled (in the process that disables them)
 * (c) the request times-out (in timer context)
 */
int
efx_mcdi_rpc_async(struct efx_nic *efx, unsigned int cmd,
		   const efx_dword_t *inbuf, size_t inlen, size_t outlen,
		   efx_mcdi_async_completer *complete, unsigned long cookie)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);
	struct efx_mcdi_async_param *async;
	int rc;

	rc = efx_mcdi_check_supported(efx, cmd, inlen);
	if (rc)
		return rc;

	async = kmalloc(sizeof(*async) + ALIGN(max(inlen, outlen), 4),
			GFP_ATOMIC);
	if (!async)
		return -ENOMEM;

	async->cmd = cmd;
	async->inlen = inlen;
	async->outlen = outlen;
	async->complete = complete;
	async->cookie = cookie;
	memcpy(async + 1, inbuf, inlen);

	spin_lock_bh(&mcdi->async_lock);

	if (mcdi->mode == MCDI_MODE_EVENTS) {
		list_add_tail(&async->list, &mcdi->async_list);

		/* If this is at the front of the queue, try to start it
		 * immediately
		 */
		if (mcdi->async_list.next == &async->list &&
		    efx_mcdi_acquire_async(mcdi)) {
			efx_mcdi_send_request(efx, cmd, inbuf, inlen);
			mod_timer(&mcdi->async_timer,
				  jiffies + MCDI_RPC_TIMEOUT);
		}
	} else {
		kfree(async);
		rc = -ENETDOWN;
	}

	spin_unlock_bh(&mcdi->async_lock);

	return rc;
}

int efx_mcdi_rpc_finish(struct efx_nic *efx, unsigned cmd, size_t inlen,
			efx_dword_t *outbuf, size_t outlen,
			size_t *outlen_actual)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);
	int rc;

	if (mcdi->mode == MCDI_MODE_POLL)
		rc = efx_mcdi_poll(efx);
	else
		rc = efx_mcdi_await_completion(efx);

	if (rc != 0) {
		netif_err(efx, hw, efx->net_dev,
			  "MC command 0x%x inlen %d mode %d timed out\n",
			  cmd, (int)inlen, mcdi->mode);

		if (mcdi->mode == MCDI_MODE_EVENTS && efx_mcdi_poll_once(efx)) {
			netif_err(efx, hw, efx->net_dev,
				  "MCDI request was completed without an event\n");
			rc = 0;
		}

		/* Close the race with efx_mcdi_ev_cpl() executing just too late
		 * and completing a request we've just cancelled, by ensuring
		 * that the seqno check therein fails.
		 */
		spin_lock_bh(&mcdi->iface_lock);
		++mcdi->seqno;
		++mcdi->credits;
		spin_unlock_bh(&mcdi->iface_lock);
	}

	if (rc == 0) {
		size_t hdr_len, data_len;

		/* At the very least we need a memory barrier here to ensure
		 * we pick up changes from efx_mcdi_ev_cpl(). Protect against
		 * a spurious efx_mcdi_ev_cpl() running concurrently by
		 * acquiring the iface_lock. */
		spin_lock_bh(&mcdi->iface_lock);
		rc = mcdi->resprc;
		hdr_len = mcdi->resp_hdr_len;
		data_len = mcdi->resp_data_len;
		spin_unlock_bh(&mcdi->iface_lock);

		BUG_ON(rc > 0);

		if (rc == 0) {
			efx->type->mcdi_read_response(efx, outbuf, hdr_len,
						      min(outlen, data_len));
			if (outlen_actual != NULL)
				*outlen_actual = data_len;
		} else if (cmd == MC_CMD_REBOOT && rc == -EIO)
			; /* Don't reset if MC_CMD_REBOOT returns EIO */
		else if (rc == -EIO || rc == -EINTR) {
			netif_err(efx, hw, efx->net_dev, "MC fatal error %d\n",
				  -rc);
			efx_schedule_reset(efx, RESET_TYPE_MC_FAILURE);
		} else
			netif_dbg(efx, hw, efx->net_dev,
				  "MC command 0x%x inlen %d failed rc=%d\n",
				  cmd, (int)inlen, -rc);

		if (rc == -EIO || rc == -EINTR) {
			msleep(MCDI_STATUS_SLEEP_MS);
			efx_mcdi_poll_reboot(efx);
			mcdi->new_epoch = true;
		}
	}

	efx_mcdi_release(mcdi);
	return rc;
}

/* Switch to polled MCDI completions.  This can be called in various
 * error conditions with various locks held, so it must be lockless.
 * Caller is responsible for flushing asynchronous requests later.
 */
void efx_mcdi_mode_poll(struct efx_nic *efx)
{
	struct efx_mcdi_iface *mcdi;

	if (!efx->mcdi)
		return;

	mcdi = efx_mcdi(efx);
	if (mcdi->mode == MCDI_MODE_POLL)
		return;

	/* We can switch from event completion to polled completion, because
	 * mcdi requests are always completed in shared memory. We do this by
	 * switching the mode to POLL'd then completing the request.
	 * efx_mcdi_await_completion() will then call efx_mcdi_poll().
	 *
	 * We need an smp_wmb() to synchronise with efx_mcdi_await_completion(),
	 * which efx_mcdi_complete_sync() provides for us.
	 */
	mcdi->mode = MCDI_MODE_POLL;

	efx_mcdi_complete_sync(mcdi);
}

/* Flush any running or queued asynchronous requests, after event processing
 * is stopped
 */
void efx_mcdi_flush_async(struct efx_nic *efx)
{
	struct efx_mcdi_async_param *async, *next;
	struct efx_mcdi_iface *mcdi;

	if (!efx->mcdi)
		return;

	mcdi = efx_mcdi(efx);

	/* We must be in polling mode so no more requests can be queued */
	BUG_ON(mcdi->mode != MCDI_MODE_POLL);

	del_timer_sync(&mcdi->async_timer);

	/* If a request is still running, make sure we give the MC
	 * time to complete it so that the response won't overwrite our
	 * next request.
	 */
	if (mcdi->state == MCDI_STATE_RUNNING_ASYNC) {
		efx_mcdi_poll(efx);
		mcdi->state = MCDI_STATE_QUIESCENT;
	}

	/* Nothing else will access the async list now, so it is safe
	 * to walk it without holding async_lock.  If we hold it while
	 * calling a completer then lockdep may warn that we have
	 * acquired locks in the wrong order.
	 */
	list_for_each_entry_safe(async, next, &mcdi->async_list, list) {
		async->complete(efx, async->cookie, -ENETDOWN, NULL, 0);
		list_del(&async->list);
		kfree(async);
	}
}

void efx_mcdi_mode_event(struct efx_nic *efx)
{
	struct efx_mcdi_iface *mcdi;

	if (!efx->mcdi)
		return;

	mcdi = efx_mcdi(efx);

	if (mcdi->mode == MCDI_MODE_EVENTS)
		return;

	/* We can't switch from polled to event completion in the middle of a
	 * request, because the completion method is specified in the request.
	 * So acquire the interface to serialise the requestors. We don't need
	 * to acquire the iface_lock to change the mode here, but we do need a
	 * write memory barrier ensure that efx_mcdi_rpc() sees it, which
	 * efx_mcdi_acquire() provides.
	 */
	efx_mcdi_acquire_sync(mcdi);
	mcdi->mode = MCDI_MODE_EVENTS;
	efx_mcdi_release(mcdi);
}

static void efx_mcdi_ev_death(struct efx_nic *efx, int rc)
{
	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);

	/* If there is an outstanding MCDI request, it has been terminated
	 * either by a BADASSERT or REBOOT event. If the mcdi interface is
	 * in polled mode, then do nothing because the MC reboot handler will
	 * set the header correctly. However, if the mcdi interface is waiting
	 * for a CMDDONE event it won't receive it [and since all MCDI events
	 * are sent to the same queue, we can't be racing with
	 * efx_mcdi_ev_cpl()]
	 *
	 * If there is an outstanding asynchronous request, we can't
	 * complete it now (efx_mcdi_complete() would deadlock).  The
	 * reset process will take care of this.
	 *
	 * There's a race here with efx_mcdi_send_request(), because
	 * we might receive a REBOOT event *before* the request has
	 * been copied out. In polled mode (during startup) this is
	 * irrelevant, because efx_mcdi_complete_sync() is ignored. In
	 * event mode, this condition is just an edge-case of
	 * receiving a REBOOT event after posting the MCDI
	 * request. Did the mc reboot before or after the copyout? The
	 * best we can do always is just return failure.
	 */
	spin_lock(&mcdi->iface_lock);
	if (efx_mcdi_complete_sync(mcdi)) {
		if (mcdi->mode == MCDI_MODE_EVENTS) {
			mcdi->resprc = rc;
			mcdi->resp_hdr_len = 0;
			mcdi->resp_data_len = 0;
			++mcdi->credits;
		}
	} else {
		int count;

		/* Consume the status word since efx_mcdi_rpc_finish() won't */
		for (count = 0; count < MCDI_STATUS_DELAY_COUNT; ++count) {
			if (efx_mcdi_poll_reboot(efx))
				break;
			udelay(MCDI_STATUS_DELAY_US);
		}
		mcdi->new_epoch = true;

		/* Nobody was waiting for an MCDI request, so trigger a reset */
		efx_schedule_reset(efx, RESET_TYPE_MC_FAILURE);
	}

	spin_unlock(&mcdi->iface_lock);
}

/* Called from  falcon_process_eventq for MCDI events */
void efx_mcdi_process_event(struct efx_channel *channel,
			    efx_qword_t *event)
{
	struct efx_nic *efx = channel->efx;
	int code = EFX_QWORD_FIELD(*event, MCDI_EVENT_CODE);
	u32 data = EFX_QWORD_FIELD(*event, MCDI_EVENT_DATA);

	switch (code) {
	case MCDI_EVENT_CODE_BADSSERT:
		netif_err(efx, hw, efx->net_dev,
			  "MC watchdog or assertion failure at 0x%x\n", data);
		efx_mcdi_ev_death(efx, -EINTR);
		break;

	case MCDI_EVENT_CODE_PMNOTICE:
		netif_info(efx, wol, efx->net_dev, "MCDI PM event.\n");
		break;

	case MCDI_EVENT_CODE_CMDDONE:
		efx_mcdi_ev_cpl(efx,
				MCDI_EVENT_FIELD(*event, CMDDONE_SEQ),
				MCDI_EVENT_FIELD(*event, CMDDONE_DATALEN),
				MCDI_EVENT_FIELD(*event, CMDDONE_ERRNO));
		break;

	case MCDI_EVENT_CODE_LINKCHANGE:
		efx_mcdi_process_link_change(efx, event);
		break;
	case MCDI_EVENT_CODE_SENSOREVT:
		efx_mcdi_sensor_event(efx, event);
		break;
	case MCDI_EVENT_CODE_SCHEDERR:
		netif_info(efx, hw, efx->net_dev,
			   "MC Scheduler error address=0x%x\n", data);
		break;
	case MCDI_EVENT_CODE_REBOOT:
	case MCDI_EVENT_CODE_MC_REBOOT:
		netif_info(efx, hw, efx->net_dev, "MC Reboot\n");
		efx_mcdi_ev_death(efx, -EIO);
		break;
	case MCDI_EVENT_CODE_MAC_STATS_DMA:
		/* MAC stats are gather lazily.  We can ignore this. */
		break;
	case MCDI_EVENT_CODE_FLR:
		efx_sriov_flr(efx, MCDI_EVENT_FIELD(*event, FLR_VF));
		break;
	case MCDI_EVENT_CODE_PTP_RX:
	case MCDI_EVENT_CODE_PTP_FAULT:
	case MCDI_EVENT_CODE_PTP_PPS:
		efx_ptp_event(efx, event);
		break;
	case MCDI_EVENT_CODE_TX_FLUSH:
	case MCDI_EVENT_CODE_RX_FLUSH:
		/* Two flush events will be sent: one to the same event
		 * queue as completions, and one to event queue 0.
		 * In the latter case the {RX,TX}_FLUSH_TO_DRIVER
		 * flag will be set, and we should ignore the event
		 * because we want to wait for all completions.
		 */
		BUILD_BUG_ON(MCDI_EVENT_TX_FLUSH_TO_DRIVER_LBN !=
			     MCDI_EVENT_RX_FLUSH_TO_DRIVER_LBN);
		if (!MCDI_EVENT_FIELD(*event, TX_FLUSH_TO_DRIVER))
			efx_ef10_handle_drain_event(efx);
		break;
	case MCDI_EVENT_CODE_TX_ERR:
	case MCDI_EVENT_CODE_RX_ERR:
		netif_err(efx, hw, efx->net_dev,
			  "%s DMA error (event: "EFX_QWORD_FMT")\n",
			  code == MCDI_EVENT_CODE_TX_ERR ? "TX" : "RX",
			  EFX_QWORD_VAL(*event));
		efx_schedule_reset(efx, RESET_TYPE_DMA_ERROR);
		break;
	default:
		netif_err(efx, hw, efx->net_dev, "Unknown MCDI event 0x%x\n",
			  code);
	}
}

/**************************************************************************
 *
 * Specific request functions
 *
 **************************************************************************
 */

void efx_mcdi_print_fwver(struct efx_nic *efx, char *buf, size_t len)
{
	MCDI_DECLARE_BUF(outbuf,
			 max(MC_CMD_GET_VERSION_OUT_LEN,
			     MC_CMD_GET_CAPABILITIES_OUT_LEN));
	size_t outlength;
	const __le16 *ver_words;
	size_t offset;
	int rc;

	BUILD_BUG_ON(MC_CMD_GET_VERSION_IN_LEN != 0);
	rc = efx_mcdi_rpc(efx, MC_CMD_GET_VERSION, NULL, 0,
			  outbuf, sizeof(outbuf), &outlength);
	if (rc)
		goto fail;
	if (outlength < MC_CMD_GET_VERSION_OUT_LEN) {
		rc = -EIO;
		goto fail;
	}

	ver_words = (__le16 *)MCDI_PTR(outbuf, GET_VERSION_OUT_VERSION);
	offset = snprintf(buf, len, "%u.%u.%u.%u",
			  le16_to_cpu(ver_words[0]), le16_to_cpu(ver_words[1]),
			  le16_to_cpu(ver_words[2]), le16_to_cpu(ver_words[3]));

	/* EF10 may have multiple datapath firmware variants within a
	 * single version.  Report which variants are running.
	 */
	if (efx_nic_rev(efx) >= EFX_REV_HUNT_A0) {
		BUILD_BUG_ON(MC_CMD_GET_CAPABILITIES_IN_LEN != 0);
		rc = efx_mcdi_rpc(efx, MC_CMD_GET_CAPABILITIES, NULL, 0,
				  outbuf, sizeof(outbuf), &outlength);
		if (rc || outlength < MC_CMD_GET_CAPABILITIES_OUT_LEN)
			offset += snprintf(
				buf + offset, len - offset, " rx? tx?");
		else
			offset += snprintf(
				buf + offset, len - offset, " rx%x tx%x",
				MCDI_WORD(outbuf,
					  GET_CAPABILITIES_OUT_RX_DPCPU_FW_ID),
				MCDI_WORD(outbuf,
					  GET_CAPABILITIES_OUT_TX_DPCPU_FW_ID));

		/* It's theoretically possible for the string to exceed 31
		 * characters, though in practice the first three version
		 * components are short enough that this doesn't happen.
		 */
		if (WARN_ON(offset >= len))
			buf[0] = 0;
	}

	return;

fail:
	netif_err(efx, probe, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	buf[0] = 0;
}

static int efx_mcdi_drv_attach(struct efx_nic *efx, bool driver_operating,
			       bool *was_attached)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_DRV_ATTACH_IN_LEN);
	MCDI_DECLARE_BUF(outbuf, MC_CMD_DRV_ATTACH_EXT_OUT_LEN);
	size_t outlen;
	int rc;

	MCDI_SET_DWORD(inbuf, DRV_ATTACH_IN_NEW_STATE,
		       driver_operating ? 1 : 0);
	MCDI_SET_DWORD(inbuf, DRV_ATTACH_IN_UPDATE, 1);
	MCDI_SET_DWORD(inbuf, DRV_ATTACH_IN_FIRMWARE_ID, MC_CMD_FW_LOW_LATENCY);

	rc = efx_mcdi_rpc(efx, MC_CMD_DRV_ATTACH, inbuf, sizeof(inbuf),
			  outbuf, sizeof(outbuf), &outlen);
	if (rc)
		goto fail;
	if (outlen < MC_CMD_DRV_ATTACH_OUT_LEN) {
		rc = -EIO;
		goto fail;
	}

	/* We currently assume we have control of the external link
	 * and are completely trusted by firmware.  Abort probing
	 * if that's not true for this function.
	 */
	if (driver_operating &&
	    outlen >= MC_CMD_DRV_ATTACH_EXT_OUT_LEN &&
	    (MCDI_DWORD(outbuf, DRV_ATTACH_EXT_OUT_FUNC_FLAGS) &
	     (1 << MC_CMD_DRV_ATTACH_EXT_OUT_FLAG_LINKCTRL |
	      1 << MC_CMD_DRV_ATTACH_EXT_OUT_FLAG_TRUSTED)) !=
	    (1 << MC_CMD_DRV_ATTACH_EXT_OUT_FLAG_LINKCTRL |
	     1 << MC_CMD_DRV_ATTACH_EXT_OUT_FLAG_TRUSTED)) {
		netif_err(efx, probe, efx->net_dev,
			  "This driver version only supports one function per port\n");
		return -ENODEV;
	}

	if (was_attached != NULL)
		*was_attached = MCDI_DWORD(outbuf, DRV_ATTACH_OUT_OLD_STATE);
	return 0;

fail:
	netif_err(efx, probe, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

int efx_mcdi_get_board_cfg(struct efx_nic *efx, u8 *mac_address,
			   u16 *fw_subtype_list, u32 *capabilities)
{
	MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_BOARD_CFG_OUT_LENMAX);
	size_t outlen, i;
	int port_num = efx_port_num(efx);
	int rc;

	BUILD_BUG_ON(MC_CMD_GET_BOARD_CFG_IN_LEN != 0);

	rc = efx_mcdi_rpc(efx, MC_CMD_GET_BOARD_CFG, NULL, 0,
			  outbuf, sizeof(outbuf), &outlen);
	if (rc)
		goto fail;

	if (outlen < MC_CMD_GET_BOARD_CFG_OUT_LENMIN) {
		rc = -EIO;
		goto fail;
	}

	if (mac_address)
		memcpy(mac_address,
		       port_num ?
		       MCDI_PTR(outbuf, GET_BOARD_CFG_OUT_MAC_ADDR_BASE_PORT1) :
		       MCDI_PTR(outbuf, GET_BOARD_CFG_OUT_MAC_ADDR_BASE_PORT0),
		       ETH_ALEN);
	if (fw_subtype_list) {
		for (i = 0;
		     i < MCDI_VAR_ARRAY_LEN(outlen,
					    GET_BOARD_CFG_OUT_FW_SUBTYPE_LIST);
		     i++)
			fw_subtype_list[i] = MCDI_ARRAY_WORD(
				outbuf, GET_BOARD_CFG_OUT_FW_SUBTYPE_LIST, i);
		for (; i < MC_CMD_GET_BOARD_CFG_OUT_FW_SUBTYPE_LIST_MAXNUM; i++)
			fw_subtype_list[i] = 0;
	}
	if (capabilities) {
		if (port_num)
			*capabilities = MCDI_DWORD(outbuf,
					GET_BOARD_CFG_OUT_CAPABILITIES_PORT1);
		else
			*capabilities = MCDI_DWORD(outbuf,
					GET_BOARD_CFG_OUT_CAPABILITIES_PORT0);
	}

	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d len=%d\n",
		  __func__, rc, (int)outlen);

	return rc;
}

int efx_mcdi_log_ctrl(struct efx_nic *efx, bool evq, bool uart, u32 dest_evq)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_LOG_CTRL_IN_LEN);
	u32 dest = 0;
	int rc;

	if (uart)
		dest |= MC_CMD_LOG_CTRL_IN_LOG_DEST_UART;
	if (evq)
		dest |= MC_CMD_LOG_CTRL_IN_LOG_DEST_EVQ;

	MCDI_SET_DWORD(inbuf, LOG_CTRL_IN_LOG_DEST, dest);
	MCDI_SET_DWORD(inbuf, LOG_CTRL_IN_LOG_DEST_EVQ, dest_evq);

	BUILD_BUG_ON(MC_CMD_LOG_CTRL_OUT_LEN != 0);

	rc = efx_mcdi_rpc(efx, MC_CMD_LOG_CTRL, inbuf, sizeof(inbuf),
			  NULL, 0, NULL);
	if (rc)
		goto fail;

	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

int efx_mcdi_nvram_types(struct efx_nic *efx, u32 *nvram_types_out)
{
	MCDI_DECLARE_BUF(outbuf, MC_CMD_NVRAM_TYPES_OUT_LEN);
	size_t outlen;
	int rc;

	BUILD_BUG_ON(MC_CMD_NVRAM_TYPES_IN_LEN != 0);

	rc = efx_mcdi_rpc(efx, MC_CMD_NVRAM_TYPES, NULL, 0,
			  outbuf, sizeof(outbuf), &outlen);
	if (rc)
		goto fail;
	if (outlen < MC_CMD_NVRAM_TYPES_OUT_LEN) {
		rc = -EIO;
		goto fail;
	}

	*nvram_types_out = MCDI_DWORD(outbuf, NVRAM_TYPES_OUT_TYPES);
	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n",
		  __func__, rc);
	return rc;
}

int efx_mcdi_nvram_info(struct efx_nic *efx, unsigned int type,
			size_t *size_out, size_t *erase_size_out,
			bool *protected_out)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_NVRAM_INFO_IN_LEN);
	MCDI_DECLARE_BUF(outbuf, MC_CMD_NVRAM_INFO_OUT_LEN);
	size_t outlen;
	int rc;

	MCDI_SET_DWORD(inbuf, NVRAM_INFO_IN_TYPE, type);

	rc = efx_mcdi_rpc(efx, MC_CMD_NVRAM_INFO, inbuf, sizeof(inbuf),
			  outbuf, sizeof(outbuf), &outlen);
	if (rc)
		goto fail;
	if (outlen < MC_CMD_NVRAM_INFO_OUT_LEN) {
		rc = -EIO;
		goto fail;
	}

	*size_out = MCDI_DWORD(outbuf, NVRAM_INFO_OUT_SIZE);
	*erase_size_out = MCDI_DWORD(outbuf, NVRAM_INFO_OUT_ERASESIZE);
	*protected_out = !!(MCDI_DWORD(outbuf, NVRAM_INFO_OUT_FLAGS) &
				(1 << MC_CMD_NVRAM_INFO_OUT_PROTECTED_LBN));
	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

static int efx_mcdi_nvram_test(struct efx_nic *efx, unsigned int type)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_NVRAM_TEST_IN_LEN);
	MCDI_DECLARE_BUF(outbuf, MC_CMD_NVRAM_TEST_OUT_LEN);
	int rc;

	MCDI_SET_DWORD(inbuf, NVRAM_TEST_IN_TYPE, type);

	rc = efx_mcdi_rpc(efx, MC_CMD_NVRAM_TEST, inbuf, sizeof(inbuf),
			  outbuf, sizeof(outbuf), NULL);
	if (rc)
		return rc;

	switch (MCDI_DWORD(outbuf, NVRAM_TEST_OUT_RESULT)) {
	case MC_CMD_NVRAM_TEST_PASS:
	case MC_CMD_NVRAM_TEST_NOTSUPP:
		return 0;
	default:
		return -EIO;
	}
}

int efx_mcdi_nvram_test_all(struct efx_nic *efx)
{
	u32 nvram_types;
	unsigned int type;
	int rc;

	rc = efx_mcdi_nvram_types(efx, &nvram_types);
	if (rc)
		goto fail1;

	type = 0;
	while (nvram_types != 0) {
		if (nvram_types & 1) {
			rc = efx_mcdi_nvram_test(efx, type);
			if (rc)
				goto fail2;
		}
		type++;
		nvram_types >>= 1;
	}

	return 0;

fail2:
	netif_err(efx, hw, efx->net_dev, "%s: failed type=%u\n",
		  __func__, type);
fail1:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

static int efx_mcdi_read_assertion(struct efx_nic *efx)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_GET_ASSERTS_IN_LEN);
	MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_ASSERTS_OUT_LEN);
	unsigned int flags, index;
	const char *reason;
	size_t outlen;
	int retry;
	int rc;

	/* Attempt to read any stored assertion state before we reboot
	 * the mcfw out of the assertion handler. Retry twice, once
	 * because a boot-time assertion might cause this command to fail
	 * with EINTR. And once again because GET_ASSERTS can race with
	 * MC_CMD_REBOOT running on the other port. */
	retry = 2;
	do {
		MCDI_SET_DWORD(inbuf, GET_ASSERTS_IN_CLEAR, 1);
		rc = efx_mcdi_rpc(efx, MC_CMD_GET_ASSERTS,
				  inbuf, MC_CMD_GET_ASSERTS_IN_LEN,
				  outbuf, sizeof(outbuf), &outlen);
	} while ((rc == -EINTR || rc == -EIO) && retry-- > 0);

	if (rc)
		return rc;
	if (outlen < MC_CMD_GET_ASSERTS_OUT_LEN)
		return -EIO;

	/* Print out any recorded assertion state */
	flags = MCDI_DWORD(outbuf, GET_ASSERTS_OUT_GLOBAL_FLAGS);
	if (flags == MC_CMD_GET_ASSERTS_FLAGS_NO_FAILS)
		return 0;

	reason = (flags == MC_CMD_GET_ASSERTS_FLAGS_SYS_FAIL)
		? "system-level assertion"
		: (flags == MC_CMD_GET_ASSERTS_FLAGS_THR_FAIL)
		? "thread-level assertion"
		: (flags == MC_CMD_GET_ASSERTS_FLAGS_WDOG_FIRED)
		? "watchdog reset"
		: "unknown assertion";
	netif_err(efx, hw, efx->net_dev,
		  "MCPU %s at PC = 0x%.8x in thread 0x%.8x\n", reason,
		  MCDI_DWORD(outbuf, GET_ASSERTS_OUT_SAVED_PC_OFFS),
		  MCDI_DWORD(outbuf, GET_ASSERTS_OUT_THREAD_OFFS));

	/* Print out the registers */
	for (index = 0;
	     index < MC_CMD_GET_ASSERTS_OUT_GP_REGS_OFFS_NUM;
	     index++)
		netif_err(efx, hw, efx->net_dev, "R%.2d (?): 0x%.8x\n",
			  1 + index,
			  MCDI_ARRAY_DWORD(outbuf, GET_ASSERTS_OUT_GP_REGS_OFFS,
					   index));

	return 0;
}

static void efx_mcdi_exit_assertion(struct efx_nic *efx)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_REBOOT_IN_LEN);

	/* If the MC is running debug firmware, it might now be
	 * waiting for a debugger to attach, but we just want it to
	 * reboot.  We set a flag that makes the command a no-op if it
	 * has already done so.  We don't know what return code to
	 * expect (0 or -EIO), so ignore it.
	 */
	BUILD_BUG_ON(MC_CMD_REBOOT_OUT_LEN != 0);
	MCDI_SET_DWORD(inbuf, REBOOT_IN_FLAGS,
		       MC_CMD_REBOOT_FLAGS_AFTER_ASSERTION);
	(void) efx_mcdi_rpc(efx, MC_CMD_REBOOT, inbuf, MC_CMD_REBOOT_IN_LEN,
			    NULL, 0, NULL);
}

int efx_mcdi_handle_assertion(struct efx_nic *efx)
{
	int rc;

	rc = efx_mcdi_read_assertion(efx);
	if (rc)
		return rc;

	efx_mcdi_exit_assertion(efx);

	return 0;
}

void efx_mcdi_set_id_led(struct efx_nic *efx, enum efx_led_mode mode)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_SET_ID_LED_IN_LEN);
	int rc;

	BUILD_BUG_ON(EFX_LED_OFF != MC_CMD_LED_OFF);
	BUILD_BUG_ON(EFX_LED_ON != MC_CMD_LED_ON);
	BUILD_BUG_ON(EFX_LED_DEFAULT != MC_CMD_LED_DEFAULT);

	BUILD_BUG_ON(MC_CMD_SET_ID_LED_OUT_LEN != 0);

	MCDI_SET_DWORD(inbuf, SET_ID_LED_IN_STATE, mode);

	rc = efx_mcdi_rpc(efx, MC_CMD_SET_ID_LED, inbuf, sizeof(inbuf),
			  NULL, 0, NULL);
	if (rc)
		netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n",
			  __func__, rc);
}

static int efx_mcdi_reset_port(struct efx_nic *efx)
{
	int rc = efx_mcdi_rpc(efx, MC_CMD_ENTITY_RESET, NULL, 0, NULL, 0, NULL);
	if (rc)
		netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n",
			  __func__, rc);
	return rc;
}

static int efx_mcdi_reset_mc(struct efx_nic *efx)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_REBOOT_IN_LEN);
	int rc;

	BUILD_BUG_ON(MC_CMD_REBOOT_OUT_LEN != 0);
	MCDI_SET_DWORD(inbuf, REBOOT_IN_FLAGS, 0);
	rc = efx_mcdi_rpc(efx, MC_CMD_REBOOT, inbuf, sizeof(inbuf),
			  NULL, 0, NULL);
	/* White is black, and up is down */
	if (rc == -EIO)
		return 0;
	if (rc == 0)
		rc = -EIO;
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

enum reset_type efx_mcdi_map_reset_reason(enum reset_type reason)
{
	return RESET_TYPE_RECOVER_OR_ALL;
}

int efx_mcdi_reset(struct efx_nic *efx, enum reset_type method)
{
	int rc;

	/* Recover from a failed assertion pre-reset */
	rc = efx_mcdi_handle_assertion(efx);
	if (rc)
		return rc;

	if (method == RESET_TYPE_WORLD)
		return efx_mcdi_reset_mc(efx);
	else
		return efx_mcdi_reset_port(efx);
}

static int efx_mcdi_wol_filter_set(struct efx_nic *efx, u32 type,
				   const u8 *mac, int *id_out)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_WOL_FILTER_SET_IN_LEN);
	MCDI_DECLARE_BUF(outbuf, MC_CMD_WOL_FILTER_SET_OUT_LEN);
	size_t outlen;
	int rc;

	MCDI_SET_DWORD(inbuf, WOL_FILTER_SET_IN_WOL_TYPE, type);
	MCDI_SET_DWORD(inbuf, WOL_FILTER_SET_IN_FILTER_MODE,
		       MC_CMD_FILTER_MODE_SIMPLE);
	memcpy(MCDI_PTR(inbuf, WOL_FILTER_SET_IN_MAGIC_MAC), mac, ETH_ALEN);

	rc = efx_mcdi_rpc(efx, MC_CMD_WOL_FILTER_SET, inbuf, sizeof(inbuf),
			  outbuf, sizeof(outbuf), &outlen);
	if (rc)
		goto fail;

	if (outlen < MC_CMD_WOL_FILTER_SET_OUT_LEN) {
		rc = -EIO;
		goto fail;
	}

	*id_out = (int)MCDI_DWORD(outbuf, WOL_FILTER_SET_OUT_FILTER_ID);

	return 0;

fail:
	*id_out = -1;
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;

}


int
efx_mcdi_wol_filter_set_magic(struct efx_nic *efx,  const u8 *mac, int *id_out)
{
	return efx_mcdi_wol_filter_set(efx, MC_CMD_WOL_TYPE_MAGIC, mac, id_out);
}


int efx_mcdi_wol_filter_get_magic(struct efx_nic *efx, int *id_out)
{
	MCDI_DECLARE_BUF(outbuf, MC_CMD_WOL_FILTER_GET_OUT_LEN);
	size_t outlen;
	int rc;

	rc = efx_mcdi_rpc(efx, MC_CMD_WOL_FILTER_GET, NULL, 0,
			  outbuf, sizeof(outbuf), &outlen);
	if (rc)
		goto fail;

	if (outlen < MC_CMD_WOL_FILTER_GET_OUT_LEN) {
		rc = -EIO;
		goto fail;
	}

	*id_out = (int)MCDI_DWORD(outbuf, WOL_FILTER_GET_OUT_FILTER_ID);

	return 0;

fail:
	*id_out = -1;
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}


int efx_mcdi_wol_filter_remove(struct efx_nic *efx, int id)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_WOL_FILTER_REMOVE_IN_LEN);
	int rc;

	MCDI_SET_DWORD(inbuf, WOL_FILTER_REMOVE_IN_FILTER_ID, (u32)id);

	rc = efx_mcdi_rpc(efx, MC_CMD_WOL_FILTER_REMOVE, inbuf, sizeof(inbuf),
			  NULL, 0, NULL);
	if (rc)
		goto fail;

	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

int efx_mcdi_flush_rxqs(struct efx_nic *efx)
{
	struct efx_channel *channel;
	struct efx_rx_queue *rx_queue;
	MCDI_DECLARE_BUF(inbuf,
			 MC_CMD_FLUSH_RX_QUEUES_IN_LEN(EFX_MAX_CHANNELS));
	int rc, count;

	BUILD_BUG_ON(EFX_MAX_CHANNELS >
		     MC_CMD_FLUSH_RX_QUEUES_IN_QID_OFST_MAXNUM);

	count = 0;
	efx_for_each_channel(channel, efx) {
		efx_for_each_channel_rx_queue(rx_queue, channel) {
			if (rx_queue->flush_pending) {
				rx_queue->flush_pending = false;
				atomic_dec(&efx->rxq_flush_pending);
				MCDI_SET_ARRAY_DWORD(
					inbuf, FLUSH_RX_QUEUES_IN_QID_OFST,
					count, efx_rx_queue_index(rx_queue));
				count++;
			}
		}
	}

	rc = efx_mcdi_rpc(efx, MC_CMD_FLUSH_RX_QUEUES, inbuf,
			  MC_CMD_FLUSH_RX_QUEUES_IN_LEN(count), NULL, 0, NULL);
	WARN_ON(rc < 0);

	return rc;
}

int efx_mcdi_wol_filter_reset(struct efx_nic *efx)
{
	int rc;

	rc = efx_mcdi_rpc(efx, MC_CMD_WOL_FILTER_RESET, NULL, 0, NULL, 0, NULL);
	if (rc)
		goto fail;

	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

int efx_mcdi_set_workaround(struct efx_nic *efx, u32 type, bool enabled)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_WORKAROUND_IN_LEN);

	BUILD_BUG_ON(MC_CMD_WORKAROUND_OUT_LEN != 0);
	MCDI_SET_DWORD(inbuf, WORKAROUND_IN_TYPE, type);
	MCDI_SET_DWORD(inbuf, WORKAROUND_IN_ENABLED, enabled);
	return efx_mcdi_rpc(efx, MC_CMD_WORKAROUND, inbuf, sizeof(inbuf),
			    NULL, 0, NULL);
}

#ifdef CONFIG_SFC_MTD

#define EFX_MCDI_NVRAM_LEN_MAX 128

static int efx_mcdi_nvram_update_start(struct efx_nic *efx, unsigned int type)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_NVRAM_UPDATE_START_IN_LEN);
	int rc;

	MCDI_SET_DWORD(inbuf, NVRAM_UPDATE_START_IN_TYPE, type);

	BUILD_BUG_ON(MC_CMD_NVRAM_UPDATE_START_OUT_LEN != 0);

	rc = efx_mcdi_rpc(efx, MC_CMD_NVRAM_UPDATE_START, inbuf, sizeof(inbuf),
			  NULL, 0, NULL);
	if (rc)
		goto fail;

	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

static int efx_mcdi_nvram_read(struct efx_nic *efx, unsigned int type,
			       loff_t offset, u8 *buffer, size_t length)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_NVRAM_READ_IN_LEN);
	MCDI_DECLARE_BUF(outbuf,
			 MC_CMD_NVRAM_READ_OUT_LEN(EFX_MCDI_NVRAM_LEN_MAX));
	size_t outlen;
	int rc;

	MCDI_SET_DWORD(inbuf, NVRAM_READ_IN_TYPE, type);
	MCDI_SET_DWORD(inbuf, NVRAM_READ_IN_OFFSET, offset);
	MCDI_SET_DWORD(inbuf, NVRAM_READ_IN_LENGTH, length);

	rc = efx_mcdi_rpc(efx, MC_CMD_NVRAM_READ, inbuf, sizeof(inbuf),
			  outbuf, sizeof(outbuf), &outlen);
	if (rc)
		goto fail;

	memcpy(buffer, MCDI_PTR(outbuf, NVRAM_READ_OUT_READ_BUFFER), length);
	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

static int efx_mcdi_nvram_write(struct efx_nic *efx, unsigned int type,
				loff_t offset, const u8 *buffer, size_t length)
{
	MCDI_DECLARE_BUF(inbuf,
			 MC_CMD_NVRAM_WRITE_IN_LEN(EFX_MCDI_NVRAM_LEN_MAX));
	int rc;

	MCDI_SET_DWORD(inbuf, NVRAM_WRITE_IN_TYPE, type);
	MCDI_SET_DWORD(inbuf, NVRAM_WRITE_IN_OFFSET, offset);
	MCDI_SET_DWORD(inbuf, NVRAM_WRITE_IN_LENGTH, length);
	memcpy(MCDI_PTR(inbuf, NVRAM_WRITE_IN_WRITE_BUFFER), buffer, length);

	BUILD_BUG_ON(MC_CMD_NVRAM_WRITE_OUT_LEN != 0);

	rc = efx_mcdi_rpc(efx, MC_CMD_NVRAM_WRITE, inbuf,
			  ALIGN(MC_CMD_NVRAM_WRITE_IN_LEN(length), 4),
			  NULL, 0, NULL);
	if (rc)
		goto fail;

	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

static int efx_mcdi_nvram_erase(struct efx_nic *efx, unsigned int type,
				loff_t offset, size_t length)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_NVRAM_ERASE_IN_LEN);
	int rc;

	MCDI_SET_DWORD(inbuf, NVRAM_ERASE_IN_TYPE, type);
	MCDI_SET_DWORD(inbuf, NVRAM_ERASE_IN_OFFSET, offset);
	MCDI_SET_DWORD(inbuf, NVRAM_ERASE_IN_LENGTH, length);

	BUILD_BUG_ON(MC_CMD_NVRAM_ERASE_OUT_LEN != 0);

	rc = efx_mcdi_rpc(efx, MC_CMD_NVRAM_ERASE, inbuf, sizeof(inbuf),
			  NULL, 0, NULL);
	if (rc)
		goto fail;

	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

static int efx_mcdi_nvram_update_finish(struct efx_nic *efx, unsigned int type)
{
	MCDI_DECLARE_BUF(inbuf, MC_CMD_NVRAM_UPDATE_FINISH_IN_LEN);
	int rc;

	MCDI_SET_DWORD(inbuf, NVRAM_UPDATE_FINISH_IN_TYPE, type);

	BUILD_BUG_ON(MC_CMD_NVRAM_UPDATE_FINISH_OUT_LEN != 0);

	rc = efx_mcdi_rpc(efx, MC_CMD_NVRAM_UPDATE_FINISH, inbuf, sizeof(inbuf),
			  NULL, 0, NULL);
	if (rc)
		goto fail;

	return 0;

fail:
	netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc);
	return rc;
}

int efx_mcdi_mtd_read(struct mtd_info *mtd, loff_t start,
		      size_t len, size_t *retlen, u8 *buffer)
{
	struct efx_mcdi_mtd_partition *part = to_efx_mcdi_mtd_partition(mtd);
	struct efx_nic *efx = mtd->priv;
	loff_t offset = start;
	loff_t end = min_t(loff_t, start + len, mtd->size);
	size_t chunk;
	int rc = 0;

	while (offset < end) {
		chunk = min_t(size_t, end - offset, EFX_MCDI_NVRAM_LEN_MAX);
		rc = efx_mcdi_nvram_read(efx, part->nvram_type, offset,
					 buffer, chunk);
		if (rc)
			goto out;
		offset += chunk;
		buffer += chunk;
	}
out:
	*retlen = offset - start;
	return rc;
}

int efx_mcdi_mtd_erase(struct mtd_info *mtd, loff_t start, size_t len)
{
	struct efx_mcdi_mtd_partition *part = to_efx_mcdi_mtd_partition(mtd);
	struct efx_nic *efx = mtd->priv;
	loff_t offset = start & ~((loff_t)(mtd->erasesize - 1));
	loff_t end = min_t(loff_t, start + len, mtd->size);
	size_t chunk = part->common.mtd.erasesize;
	int rc = 0;

	if (!part->updating) {
		rc = efx_mcdi_nvram_update_start(efx, part->nvram_type);
		if (rc)
			goto out;
		part->updating = true;
	}

	/* The MCDI interface can in fact do multiple erase blocks at once;
	 * but erasing may be slow, so we make multiple calls here to avoid
	 * tripping the MCDI RPC timeout. */
	while (offset < end) {
		rc = efx_mcdi_nvram_erase(efx, part->nvram_type, offset,
					  chunk);
		if (rc)
			goto out;
		offset += chunk;
	}
out:
	return rc;
}

int efx_mcdi_mtd_write(struct mtd_info *mtd, loff_t start,
		       size_t len, size_t *retlen, const u8 *buffer)
{
	struct efx_mcdi_mtd_partition *part = to_efx_mcdi_mtd_partition(mtd);
	struct efx_nic *efx = mtd->priv;
	loff_t offset = start;
	loff_t end = min_t(loff_t, start + len, mtd->size);
	size_t chunk;
	int rc = 0;

	if (!part->updating) {
		rc = efx_mcdi_nvram_update_start(efx, part->nvram_type);
		if (rc)
			goto out;
		part->updating = true;
	}

	while (offset < end) {
		chunk = min_t(size_t, end - offset, EFX_MCDI_NVRAM_LEN_MAX);
		rc = efx_mcdi_nvram_write(efx, part->nvram_type, offset,
					  buffer, chunk);
		if (rc)
			goto out;
		offset += chunk;
		buffer += chunk;
	}
out:
	*retlen = offset - start;
	return rc;
}

int efx_mcdi_mtd_sync(struct mtd_info *mtd)
{
	struct efx_mcdi_mtd_partition *part = to_efx_mcdi_mtd_partition(mtd);
	struct efx_nic *efx = mtd->priv;
	int rc = 0;

	if (part->updating) {
		part->updating = false;
		rc = efx_mcdi_nvram_update_finish(efx, part->nvram_type);
	}

	return rc;
}

void efx_mcdi_mtd_rename(struct efx_mtd_partition *part)
{
	struct efx_mcdi_mtd_partition *mcdi_part =
		container_of(part, struct efx_mcdi_mtd_partition, common);
	struct efx_nic *efx = part->mtd.priv;

	snprintf(part->name, sizeof(part->name), "%s %s:%02x",
		 efx->name, part->type_name, mcdi_part->fw_subtype);
}

#endif /* CONFIG_SFC_MTD */