summaryrefslogtreecommitdiff
path: root/drivers/ntb/hw/amd/ntb_hw_amd.c
blob: 87847c380051688a27192cbb317de8644461943b (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
/*
 * This file is provided under a dual BSD/GPLv2 license.  When using or
 *   redistributing this file, you may do so under either license.
 *
 *   GPL LICENSE SUMMARY
 *
 *   Copyright (C) 2016 Advanced Micro Devices, Inc. All Rights Reserved.
 *   Copyright (C) 2016 T-Platforms. All Rights Reserved.
 *
 *   This program is free software; you can redistribute it and/or modify
 *   it under the terms of version 2 of the GNU General Public License as
 *   published by the Free Software Foundation.
 *
 *   BSD LICENSE
 *
 *   Copyright (C) 2016 Advanced Micro Devices, Inc. All Rights Reserved.
 *   Copyright (C) 2016 T-Platforms. All Rights Reserved.
 *
 *   Redistribution and use in source and binary forms, with or without
 *   modification, are permitted provided that the following conditions
 *   are met:
 *
 *     * Redistributions of source code must retain the above copyright
 *       notice, this list of conditions and the following disclaimer.
 *     * Redistributions in binary form must reproduce the above copy
 *       notice, this list of conditions and the following disclaimer in
 *       the documentation and/or other materials provided with the
 *       distribution.
 *     * Neither the name of AMD Corporation nor the names of its
 *       contributors may be used to endorse or promote products derived
 *       from this software without specific prior written permission.
 *
 *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
 *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
 *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
 *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
 *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 *
 * AMD PCIe NTB Linux driver
 *
 * Contact Information:
 * Xiangliang Yu <Xiangliang.Yu@amd.com>
 */

#include <linux/debugfs.h>
#include <linux/delay.h>
#include <linux/init.h>
#include <linux/interrupt.h>
#include <linux/module.h>
#include <linux/acpi.h>
#include <linux/pci.h>
#include <linux/random.h>
#include <linux/slab.h>
#include <linux/ntb.h>

#include "ntb_hw_amd.h"

#define NTB_NAME	"ntb_hw_amd"
#define NTB_DESC	"AMD(R) PCI-E Non-Transparent Bridge Driver"
#define NTB_VER		"1.0"

MODULE_DESCRIPTION(NTB_DESC);
MODULE_VERSION(NTB_VER);
MODULE_LICENSE("Dual BSD/GPL");
MODULE_AUTHOR("AMD Inc.");

static const struct file_operations amd_ntb_debugfs_info;
static struct dentry *debugfs_dir;

static int ndev_mw_to_bar(struct amd_ntb_dev *ndev, int idx)
{
	if (idx < 0 || idx > ndev->mw_count)
		return -EINVAL;

	return ndev->dev_data->mw_idx << idx;
}

static int amd_ntb_mw_count(struct ntb_dev *ntb, int pidx)
{
	if (pidx != NTB_DEF_PEER_IDX)
		return -EINVAL;

	return ntb_ndev(ntb)->mw_count;
}

static int amd_ntb_mw_get_align(struct ntb_dev *ntb, int pidx, int idx,
				resource_size_t *addr_align,
				resource_size_t *size_align,
				resource_size_t *size_max)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	int bar;

	if (pidx != NTB_DEF_PEER_IDX)
		return -EINVAL;

	bar = ndev_mw_to_bar(ndev, idx);
	if (bar < 0)
		return bar;

	if (addr_align)
		*addr_align = SZ_4K;

	if (size_align)
		*size_align = 1;

	if (size_max)
		*size_max = pci_resource_len(ndev->ntb.pdev, bar);

	return 0;
}

static int amd_ntb_mw_set_trans(struct ntb_dev *ntb, int pidx, int idx,
				dma_addr_t addr, resource_size_t size)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	unsigned long xlat_reg, limit_reg = 0;
	resource_size_t mw_size;
	void __iomem *mmio, *peer_mmio;
	u64 base_addr, limit, reg_val;
	int bar;

	if (pidx != NTB_DEF_PEER_IDX)
		return -EINVAL;

	bar = ndev_mw_to_bar(ndev, idx);
	if (bar < 0)
		return bar;

	mw_size = pci_resource_len(ntb->pdev, bar);

	/* make sure the range fits in the usable mw size */
	if (size > mw_size)
		return -EINVAL;

	mmio = ndev->self_mmio;
	peer_mmio = ndev->peer_mmio;

	base_addr = pci_resource_start(ntb->pdev, bar);

	if (bar != 1) {
		xlat_reg = AMD_BAR23XLAT_OFFSET + ((bar - 2) << 2);
		limit_reg = AMD_BAR23LMT_OFFSET + ((bar - 2) << 2);

		/* Set the limit if supported */
		limit = size;

		/* set and verify setting the translation address */
		write64(addr, peer_mmio + xlat_reg);
		reg_val = read64(peer_mmio + xlat_reg);
		if (reg_val != addr) {
			write64(0, peer_mmio + xlat_reg);
			return -EIO;
		}

		/* set and verify setting the limit */
		write64(limit, peer_mmio + limit_reg);
		reg_val = read64(peer_mmio + limit_reg);
		if (reg_val != limit) {
			write64(base_addr, mmio + limit_reg);
			write64(0, peer_mmio + xlat_reg);
			return -EIO;
		}
	} else {
		xlat_reg = AMD_BAR1XLAT_OFFSET;
		limit_reg = AMD_BAR1LMT_OFFSET;

		/* Set the limit if supported */
		limit = size;

		/* set and verify setting the translation address */
		write64(addr, peer_mmio + xlat_reg);
		reg_val = read64(peer_mmio + xlat_reg);
		if (reg_val != addr) {
			write64(0, peer_mmio + xlat_reg);
			return -EIO;
		}

		/* set and verify setting the limit */
		writel(limit, peer_mmio + limit_reg);
		reg_val = readl(peer_mmio + limit_reg);
		if (reg_val != limit) {
			writel(base_addr, mmio + limit_reg);
			writel(0, peer_mmio + xlat_reg);
			return -EIO;
		}
	}

	return 0;
}

static int amd_ntb_get_link_status(struct amd_ntb_dev *ndev)
{
	struct pci_dev *pdev = NULL;
	struct pci_dev *pci_swds = NULL;
	struct pci_dev *pci_swus = NULL;
	u32 stat;
	int rc;

	if (ndev->ntb.topo == NTB_TOPO_SEC) {
		/* Locate the pointer to Downstream Switch for this device */
		pci_swds = pci_upstream_bridge(ndev->ntb.pdev);
		if (pci_swds) {
			/*
			 * Locate the pointer to Upstream Switch for
			 * the Downstream Switch.
			 */
			pci_swus = pci_upstream_bridge(pci_swds);
			if (pci_swus) {
				rc = pcie_capability_read_dword(pci_swus,
								PCI_EXP_LNKCTL,
								&stat);
				if (rc)
					return 0;
			} else {
				return 0;
			}
		} else {
			return 0;
		}
	} else if (ndev->ntb.topo == NTB_TOPO_PRI) {
		/*
		 * For NTB primary, we simply read the Link Status and control
		 * register of the NTB device itself.
		 */
		pdev = ndev->ntb.pdev;
		rc = pcie_capability_read_dword(pdev, PCI_EXP_LNKCTL, &stat);
		if (rc)
			return 0;
	} else {
		/* Catch all for everything else */
		return 0;
	}

	ndev->lnk_sta = stat;

	return 1;
}

static int amd_link_is_up(struct amd_ntb_dev *ndev)
{
	int ret;

	/*
	 * We consider the link to be up under two conditions:
	 *
	 *   - When a link-up event is received. This is indicated by
	 *     AMD_LINK_UP_EVENT set in peer_sta.
	 *   - When driver on both sides of the link have been loaded.
	 *     This is indicated by bit 1 being set in the peer
	 *     SIDEINFO register.
	 *
	 * This function should return 1 when the latter of the above
	 * two conditions is true.
	 *
	 * Now consider the sequence of events - Link-Up event occurs,
	 * then the peer side driver loads. In this case, we would have
	 * received LINK_UP event and bit 1 of peer SIDEINFO is also
	 * set. What happens now if the link goes down? Bit 1 of
	 * peer SIDEINFO remains set, but LINK_DOWN bit is set in
	 * peer_sta. So we should return 0 from this function. Not only
	 * that, we clear bit 1 of peer SIDEINFO to 0, since the peer
	 * side driver did not even get a chance to clear it before
	 * the link went down. This can be the case of surprise link
	 * removal.
	 *
	 * LINK_UP event will always occur before the peer side driver
	 * gets loaded the very first time. So there can be a case when
	 * the LINK_UP event has occurred, but the peer side driver hasn't
	 * yet loaded. We return 0 in that case.
	 *
	 * There is also a special case when the primary side driver is
	 * unloaded and then loaded again. Since there is no change in
	 * the status of NTB secondary in this case, there is no Link-Up
	 * or Link-Down notification received. We recognize this condition
	 * with peer_sta being set to 0.
	 *
	 * If bit 1 of peer SIDEINFO register is not set, then we
	 * simply return 0 irrespective of the link up or down status
	 * set in peer_sta.
	 */
	ret = amd_poll_link(ndev);
	if (ret) {
		/*
		 * We need to check the below only for NTB primary. For NTB
		 * secondary, simply checking the result of PSIDE_INFO
		 * register will suffice.
		 */
		if (ndev->ntb.topo == NTB_TOPO_PRI) {
			if ((ndev->peer_sta & AMD_LINK_UP_EVENT) ||
			    (ndev->peer_sta == 0))
				return ret;
			else if (ndev->peer_sta & AMD_LINK_DOWN_EVENT) {
				/* Clear peer sideinfo register */
				amd_clear_side_info_reg(ndev, true);

				return 0;
			}
		} else { /* NTB_TOPO_SEC */
			return ret;
		}
	}

	return 0;
}

static u64 amd_ntb_link_is_up(struct ntb_dev *ntb,
			      enum ntb_speed *speed,
			      enum ntb_width *width)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	int ret = 0;

	if (amd_link_is_up(ndev)) {
		if (speed)
			*speed = NTB_LNK_STA_SPEED(ndev->lnk_sta);
		if (width)
			*width = NTB_LNK_STA_WIDTH(ndev->lnk_sta);

		dev_dbg(&ntb->pdev->dev, "link is up.\n");

		ret = 1;
	} else {
		if (speed)
			*speed = NTB_SPEED_NONE;
		if (width)
			*width = NTB_WIDTH_NONE;

		dev_dbg(&ntb->pdev->dev, "link is down.\n");
	}

	return ret;
}

static int amd_ntb_link_enable(struct ntb_dev *ntb,
			       enum ntb_speed max_speed,
			       enum ntb_width max_width)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;

	/* Enable event interrupt */
	ndev->int_mask &= ~AMD_EVENT_INTMASK;
	writel(ndev->int_mask, mmio + AMD_INTMASK_OFFSET);

	if (ndev->ntb.topo == NTB_TOPO_SEC)
		return -EINVAL;
	dev_dbg(&ntb->pdev->dev, "Enabling Link.\n");

	return 0;
}

static int amd_ntb_link_disable(struct ntb_dev *ntb)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;

	/* Disable event interrupt */
	ndev->int_mask |= AMD_EVENT_INTMASK;
	writel(ndev->int_mask, mmio + AMD_INTMASK_OFFSET);

	if (ndev->ntb.topo == NTB_TOPO_SEC)
		return -EINVAL;
	dev_dbg(&ntb->pdev->dev, "Enabling Link.\n");

	return 0;
}

static int amd_ntb_peer_mw_count(struct ntb_dev *ntb)
{
	/* The same as for inbound MWs */
	return ntb_ndev(ntb)->mw_count;
}

static int amd_ntb_peer_mw_get_addr(struct ntb_dev *ntb, int idx,
				    phys_addr_t *base, resource_size_t *size)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	int bar;

	bar = ndev_mw_to_bar(ndev, idx);
	if (bar < 0)
		return bar;

	if (base)
		*base = pci_resource_start(ndev->ntb.pdev, bar);

	if (size)
		*size = pci_resource_len(ndev->ntb.pdev, bar);

	return 0;
}

static u64 amd_ntb_db_valid_mask(struct ntb_dev *ntb)
{
	return ntb_ndev(ntb)->db_valid_mask;
}

static int amd_ntb_db_vector_count(struct ntb_dev *ntb)
{
	return ntb_ndev(ntb)->db_count;
}

static u64 amd_ntb_db_vector_mask(struct ntb_dev *ntb, int db_vector)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);

	if (db_vector < 0 || db_vector > ndev->db_count)
		return 0;

	return ntb_ndev(ntb)->db_valid_mask & (1ULL << db_vector);
}

static u64 amd_ntb_db_read(struct ntb_dev *ntb)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;

	return (u64)readw(mmio + AMD_DBSTAT_OFFSET);
}

static int amd_ntb_db_clear(struct ntb_dev *ntb, u64 db_bits)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;

	writew((u16)db_bits, mmio + AMD_DBSTAT_OFFSET);

	return 0;
}

static int amd_ntb_db_set_mask(struct ntb_dev *ntb, u64 db_bits)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;
	unsigned long flags;

	if (db_bits & ~ndev->db_valid_mask)
		return -EINVAL;

	spin_lock_irqsave(&ndev->db_mask_lock, flags);
	ndev->db_mask |= db_bits;
	writew((u16)ndev->db_mask, mmio + AMD_DBMASK_OFFSET);
	spin_unlock_irqrestore(&ndev->db_mask_lock, flags);

	return 0;
}

static int amd_ntb_db_clear_mask(struct ntb_dev *ntb, u64 db_bits)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;
	unsigned long flags;

	if (db_bits & ~ndev->db_valid_mask)
		return -EINVAL;

	spin_lock_irqsave(&ndev->db_mask_lock, flags);
	ndev->db_mask &= ~db_bits;
	writew((u16)ndev->db_mask, mmio + AMD_DBMASK_OFFSET);
	spin_unlock_irqrestore(&ndev->db_mask_lock, flags);

	return 0;
}

static int amd_ntb_peer_db_set(struct ntb_dev *ntb, u64 db_bits)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;

	writew((u16)db_bits, mmio + AMD_DBREQ_OFFSET);

	return 0;
}

static int amd_ntb_spad_count(struct ntb_dev *ntb)
{
	return ntb_ndev(ntb)->spad_count;
}

static u32 amd_ntb_spad_read(struct ntb_dev *ntb, int idx)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;
	u32 offset;

	if (idx < 0 || idx >= ndev->spad_count)
		return 0;

	offset = ndev->self_spad + (idx << 2);
	return readl(mmio + AMD_SPAD_OFFSET + offset);
}

static int amd_ntb_spad_write(struct ntb_dev *ntb,
			      int idx, u32 val)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;
	u32 offset;

	if (idx < 0 || idx >= ndev->spad_count)
		return -EINVAL;

	offset = ndev->self_spad + (idx << 2);
	writel(val, mmio + AMD_SPAD_OFFSET + offset);

	return 0;
}

static u32 amd_ntb_peer_spad_read(struct ntb_dev *ntb, int pidx, int sidx)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;
	u32 offset;

	if (sidx < 0 || sidx >= ndev->spad_count)
		return -EINVAL;

	offset = ndev->peer_spad + (sidx << 2);
	return readl(mmio + AMD_SPAD_OFFSET + offset);
}

static int amd_ntb_peer_spad_write(struct ntb_dev *ntb, int pidx,
				   int sidx, u32 val)
{
	struct amd_ntb_dev *ndev = ntb_ndev(ntb);
	void __iomem *mmio = ndev->self_mmio;
	u32 offset;

	if (sidx < 0 || sidx >= ndev->spad_count)
		return -EINVAL;

	offset = ndev->peer_spad + (sidx << 2);
	writel(val, mmio + AMD_SPAD_OFFSET + offset);

	return 0;
}

static const struct ntb_dev_ops amd_ntb_ops = {
	.mw_count		= amd_ntb_mw_count,
	.mw_get_align		= amd_ntb_mw_get_align,
	.mw_set_trans		= amd_ntb_mw_set_trans,
	.peer_mw_count		= amd_ntb_peer_mw_count,
	.peer_mw_get_addr	= amd_ntb_peer_mw_get_addr,
	.link_is_up		= amd_ntb_link_is_up,
	.link_enable		= amd_ntb_link_enable,
	.link_disable		= amd_ntb_link_disable,
	.db_valid_mask		= amd_ntb_db_valid_mask,
	.db_vector_count	= amd_ntb_db_vector_count,
	.db_vector_mask		= amd_ntb_db_vector_mask,
	.db_read		= amd_ntb_db_read,
	.db_clear		= amd_ntb_db_clear,
	.db_set_mask		= amd_ntb_db_set_mask,
	.db_clear_mask		= amd_ntb_db_clear_mask,
	.peer_db_set		= amd_ntb_peer_db_set,
	.spad_count		= amd_ntb_spad_count,
	.spad_read		= amd_ntb_spad_read,
	.spad_write		= amd_ntb_spad_write,
	.peer_spad_read		= amd_ntb_peer_spad_read,
	.peer_spad_write	= amd_ntb_peer_spad_write,
};

static void amd_ack_smu(struct amd_ntb_dev *ndev, u32 bit)
{
	void __iomem *mmio = ndev->self_mmio;
	int reg;

	reg = readl(mmio + AMD_SMUACK_OFFSET);
	reg |= bit;
	writel(reg, mmio + AMD_SMUACK_OFFSET);
}

static void amd_handle_event(struct amd_ntb_dev *ndev, int vec)
{
	void __iomem *mmio = ndev->self_mmio;
	struct device *dev = &ndev->ntb.pdev->dev;
	u32 status;

	status = readl(mmio + AMD_INTSTAT_OFFSET);
	if (!(status & AMD_EVENT_INTMASK))
		return;

	dev_dbg(dev, "status = 0x%x and vec = %d\n", status, vec);

	status &= AMD_EVENT_INTMASK;
	switch (status) {
	case AMD_PEER_FLUSH_EVENT:
		ndev->peer_sta |= AMD_PEER_FLUSH_EVENT;
		dev_info(dev, "Flush is done.\n");
		break;
	case AMD_PEER_RESET_EVENT:
	case AMD_LINK_DOWN_EVENT:
		ndev->peer_sta |= status;
		if (status == AMD_LINK_DOWN_EVENT)
			ndev->peer_sta &= ~AMD_LINK_UP_EVENT;

		amd_ack_smu(ndev, status);

		/* link down first */
		ntb_link_event(&ndev->ntb);
		/* polling peer status */
		schedule_delayed_work(&ndev->hb_timer, AMD_LINK_HB_TIMEOUT);

		break;
	case AMD_PEER_D3_EVENT:
	case AMD_PEER_PMETO_EVENT:
	case AMD_LINK_UP_EVENT:
		ndev->peer_sta |= status;
		if (status == AMD_LINK_UP_EVENT)
			ndev->peer_sta &= ~AMD_LINK_DOWN_EVENT;
		else if (status == AMD_PEER_D3_EVENT)
			ndev->peer_sta &= ~AMD_PEER_D0_EVENT;

		amd_ack_smu(ndev, status);

		/* link down */
		ntb_link_event(&ndev->ntb);

		break;
	case AMD_PEER_D0_EVENT:
		mmio = ndev->peer_mmio;
		status = readl(mmio + AMD_PMESTAT_OFFSET);
		/* check if this is WAKEUP event */
		if (status & 0x1)
			dev_info(dev, "Wakeup is done.\n");

		ndev->peer_sta |= AMD_PEER_D0_EVENT;
		ndev->peer_sta &= ~AMD_PEER_D3_EVENT;
		amd_ack_smu(ndev, AMD_PEER_D0_EVENT);

		/* start a timer to poll link status */
		schedule_delayed_work(&ndev->hb_timer,
				      AMD_LINK_HB_TIMEOUT);
		break;
	default:
		dev_info(dev, "event status = 0x%x.\n", status);
		break;
	}

	/* Clear the interrupt status */
	writel(status, mmio + AMD_INTSTAT_OFFSET);
}

static void amd_handle_db_event(struct amd_ntb_dev *ndev, int vec)
{
	struct device *dev = &ndev->ntb.pdev->dev;
	u64 status;

	status = amd_ntb_db_read(&ndev->ntb);

	dev_dbg(dev, "status = 0x%llx and vec = %d\n", status, vec);

	/*
	 * Since we had reserved highest order bit of DB for signaling peer of
	 * a special event, this is the only status bit we should be concerned
	 * here now.
	 */
	if (status & BIT(ndev->db_last_bit)) {
		ntb_db_clear(&ndev->ntb, BIT(ndev->db_last_bit));
		/* send link down event notification */
		ntb_link_event(&ndev->ntb);

		/*
		 * If we are here, that means the peer has signalled a special
		 * event which notifies that the peer driver has been
		 * un-loaded for some reason. Since there is a chance that the
		 * peer will load its driver again sometime, we schedule link
		 * polling routine.
		 */
		schedule_delayed_work(&ndev->hb_timer, AMD_LINK_HB_TIMEOUT);
	}
}

static irqreturn_t ndev_interrupt(struct amd_ntb_dev *ndev, int vec)
{
	dev_dbg(&ndev->ntb.pdev->dev, "vec %d\n", vec);

	if (vec > (AMD_DB_CNT - 1) || (ndev->msix_vec_count == 1))
		amd_handle_event(ndev, vec);

	if (vec < AMD_DB_CNT) {
		amd_handle_db_event(ndev, vec);
		ntb_db_event(&ndev->ntb, vec);
	}

	return IRQ_HANDLED;
}

static irqreturn_t ndev_vec_isr(int irq, void *dev)
{
	struct amd_ntb_vec *nvec = dev;

	return ndev_interrupt(nvec->ndev, nvec->num);
}

static irqreturn_t ndev_irq_isr(int irq, void *dev)
{
	struct amd_ntb_dev *ndev = dev;

	return ndev_interrupt(ndev, irq - ndev->ntb.pdev->irq);
}

static int ndev_init_isr(struct amd_ntb_dev *ndev,
			 int msix_min, int msix_max)
{
	struct pci_dev *pdev;
	int rc, i, msix_count, node;

	pdev = ndev->ntb.pdev;

	node = dev_to_node(&pdev->dev);

	ndev->db_mask = ndev->db_valid_mask;

	/* Try to set up msix irq */
	ndev->vec = kcalloc_node(msix_max, sizeof(*ndev->vec),
				 GFP_KERNEL, node);
	if (!ndev->vec)
		goto err_msix_vec_alloc;

	ndev->msix = kcalloc_node(msix_max, sizeof(*ndev->msix),
				  GFP_KERNEL, node);
	if (!ndev->msix)
		goto err_msix_alloc;

	for (i = 0; i < msix_max; ++i)
		ndev->msix[i].entry = i;

	msix_count = pci_enable_msix_range(pdev, ndev->msix,
					   msix_min, msix_max);
	if (msix_count < 0)
		goto err_msix_enable;

	/* NOTE: Disable MSIX if msix count is less than 16 because of
	 * hardware limitation.
	 */
	if (msix_count < msix_min) {
		pci_disable_msix(pdev);
		goto err_msix_enable;
	}

	for (i = 0; i < msix_count; ++i) {
		ndev->vec[i].ndev = ndev;
		ndev->vec[i].num = i;
		rc = request_irq(ndev->msix[i].vector, ndev_vec_isr, 0,
				 "ndev_vec_isr", &ndev->vec[i]);
		if (rc)
			goto err_msix_request;
	}

	dev_dbg(&pdev->dev, "Using msix interrupts\n");
	ndev->db_count = msix_min;
	ndev->msix_vec_count = msix_max;
	return 0;

err_msix_request:
	while (i-- > 0)
		free_irq(ndev->msix[i].vector, &ndev->vec[i]);
	pci_disable_msix(pdev);
err_msix_enable:
	kfree(ndev->msix);
err_msix_alloc:
	kfree(ndev->vec);
err_msix_vec_alloc:
	ndev->msix = NULL;
	ndev->vec = NULL;

	/* Try to set up msi irq */
	rc = pci_enable_msi(pdev);
	if (rc)
		goto err_msi_enable;

	rc = request_irq(pdev->irq, ndev_irq_isr, 0,
			 "ndev_irq_isr", ndev);
	if (rc)
		goto err_msi_request;

	dev_dbg(&pdev->dev, "Using msi interrupts\n");
	ndev->db_count = 1;
	ndev->msix_vec_count = 1;
	return 0;

err_msi_request:
	pci_disable_msi(pdev);
err_msi_enable:

	/* Try to set up intx irq */
	pci_intx(pdev, 1);

	rc = request_irq(pdev->irq, ndev_irq_isr, IRQF_SHARED,
			 "ndev_irq_isr", ndev);
	if (rc)
		goto err_intx_request;

	dev_dbg(&pdev->dev, "Using intx interrupts\n");
	ndev->db_count = 1;
	ndev->msix_vec_count = 1;
	return 0;

err_intx_request:
	return rc;
}

static void ndev_deinit_isr(struct amd_ntb_dev *ndev)
{
	struct pci_dev *pdev;
	void __iomem *mmio = ndev->self_mmio;
	int i;

	pdev = ndev->ntb.pdev;

	/* Mask all doorbell interrupts */
	ndev->db_mask = ndev->db_valid_mask;
	writel(ndev->db_mask, mmio + AMD_DBMASK_OFFSET);

	if (ndev->msix) {
		i = ndev->msix_vec_count;
		while (i--)
			free_irq(ndev->msix[i].vector, &ndev->vec[i]);
		pci_disable_msix(pdev);
		kfree(ndev->msix);
		kfree(ndev->vec);
	} else {
		free_irq(pdev->irq, ndev);
		if (pci_dev_msi_enabled(pdev))
			pci_disable_msi(pdev);
		else
			pci_intx(pdev, 0);
	}
}

static ssize_t ndev_debugfs_read(struct file *filp, char __user *ubuf,
				 size_t count, loff_t *offp)
{
	struct amd_ntb_dev *ndev;
	void __iomem *mmio;
	char *buf;
	size_t buf_size;
	ssize_t ret, off;
	union { u64 v64; u32 v32; u16 v16; } u;

	ndev = filp->private_data;
	mmio = ndev->self_mmio;

	buf_size = min(count, 0x800ul);

	buf = kmalloc(buf_size, GFP_KERNEL);
	if (!buf)
		return -ENOMEM;

	off = 0;

	off += scnprintf(buf + off, buf_size - off,
			 "NTB Device Information:\n");

	off += scnprintf(buf + off, buf_size - off,
			 "Connection Topology -\t%s\n",
			 ntb_topo_string(ndev->ntb.topo));

	off += scnprintf(buf + off, buf_size - off,
			 "LNK STA -\t\t%#06x\n", ndev->lnk_sta);

	if (!amd_link_is_up(ndev)) {
		off += scnprintf(buf + off, buf_size - off,
				 "Link Status -\t\tDown\n");
	} else {
		off += scnprintf(buf + off, buf_size - off,
				 "Link Status -\t\tUp\n");
		off += scnprintf(buf + off, buf_size - off,
				 "Link Speed -\t\tPCI-E Gen %u\n",
				 NTB_LNK_STA_SPEED(ndev->lnk_sta));
		off += scnprintf(buf + off, buf_size - off,
				 "Link Width -\t\tx%u\n",
				 NTB_LNK_STA_WIDTH(ndev->lnk_sta));
	}

	off += scnprintf(buf + off, buf_size - off,
			 "Memory Window Count -\t%u\n", ndev->mw_count);
	off += scnprintf(buf + off, buf_size - off,
			 "Scratchpad Count -\t%u\n", ndev->spad_count);
	off += scnprintf(buf + off, buf_size - off,
			 "Doorbell Count -\t%u\n", ndev->db_count);
	off += scnprintf(buf + off, buf_size - off,
			 "MSIX Vector Count -\t%u\n", ndev->msix_vec_count);

	off += scnprintf(buf + off, buf_size - off,
			 "Doorbell Valid Mask -\t%#llx\n", ndev->db_valid_mask);

	u.v32 = readl(ndev->self_mmio + AMD_DBMASK_OFFSET);
	off += scnprintf(buf + off, buf_size - off,
			 "Doorbell Mask -\t\t\t%#06x\n", u.v32);

	u.v32 = readl(mmio + AMD_DBSTAT_OFFSET);
	off += scnprintf(buf + off, buf_size - off,
			 "Doorbell Bell -\t\t\t%#06x\n", u.v32);

	off += scnprintf(buf + off, buf_size - off,
			 "\nNTB Incoming XLAT:\n");

	u.v64 = read64(mmio + AMD_BAR1XLAT_OFFSET);
	off += scnprintf(buf + off, buf_size - off,
			 "XLAT1 -\t\t%#018llx\n", u.v64);

	u.v64 = read64(ndev->self_mmio + AMD_BAR23XLAT_OFFSET);
	off += scnprintf(buf + off, buf_size - off,
			 "XLAT23 -\t\t%#018llx\n", u.v64);

	u.v64 = read64(ndev->self_mmio + AMD_BAR45XLAT_OFFSET);
	off += scnprintf(buf + off, buf_size - off,
			 "XLAT45 -\t\t%#018llx\n", u.v64);

	u.v32 = readl(mmio + AMD_BAR1LMT_OFFSET);
	off += scnprintf(buf + off, buf_size - off,
			 "LMT1 -\t\t\t%#06x\n", u.v32);

	u.v64 = read64(ndev->self_mmio + AMD_BAR23LMT_OFFSET);
	off += scnprintf(buf + off, buf_size - off,
			 "LMT23 -\t\t\t%#018llx\n", u.v64);

	u.v64 = read64(ndev->self_mmio + AMD_BAR45LMT_OFFSET);
	off += scnprintf(buf + off, buf_size - off,
			 "LMT45 -\t\t\t%#018llx\n", u.v64);

	ret = simple_read_from_buffer(ubuf, count, offp, buf, off);
	kfree(buf);
	return ret;
}

static void ndev_init_debugfs(struct amd_ntb_dev *ndev)
{
	if (!debugfs_dir) {
		ndev->debugfs_dir = NULL;
		ndev->debugfs_info = NULL;
	} else {
		ndev->debugfs_dir =
			debugfs_create_dir(pci_name(ndev->ntb.pdev),
					   debugfs_dir);
		if (!ndev->debugfs_dir)
			ndev->debugfs_info = NULL;
		else
			ndev->debugfs_info =
				debugfs_create_file("info", S_IRUSR,
						    ndev->debugfs_dir, ndev,
						    &amd_ntb_debugfs_info);
	}
}

static void ndev_deinit_debugfs(struct amd_ntb_dev *ndev)
{
	debugfs_remove_recursive(ndev->debugfs_dir);
}

static inline void ndev_init_struct(struct amd_ntb_dev *ndev,
				    struct pci_dev *pdev)
{
	ndev->ntb.pdev = pdev;
	ndev->ntb.topo = NTB_TOPO_NONE;
	ndev->ntb.ops = &amd_ntb_ops;
	ndev->int_mask = AMD_EVENT_INTMASK;
	spin_lock_init(&ndev->db_mask_lock);
}

static int amd_poll_link(struct amd_ntb_dev *ndev)
{
	void __iomem *mmio = ndev->peer_mmio;
	u32 reg;

	reg = readl(mmio + AMD_SIDEINFO_OFFSET);
	reg &= AMD_SIDE_READY;

	dev_dbg(&ndev->ntb.pdev->dev, "%s: reg_val = 0x%x.\n", __func__, reg);

	ndev->cntl_sta = reg;

	amd_ntb_get_link_status(ndev);

	return ndev->cntl_sta;
}

static void amd_link_hb(struct work_struct *work)
{
	struct amd_ntb_dev *ndev = hb_ndev(work);

	if (amd_poll_link(ndev))
		ntb_link_event(&ndev->ntb);

	if (!amd_link_is_up(ndev))
		schedule_delayed_work(&ndev->hb_timer, AMD_LINK_HB_TIMEOUT);
}

static int amd_init_isr(struct amd_ntb_dev *ndev)
{
	return ndev_init_isr(ndev, AMD_DB_CNT, AMD_MSIX_VECTOR_CNT);
}

static void amd_set_side_info_reg(struct amd_ntb_dev *ndev, bool peer)
{
	void __iomem *mmio = NULL;
	unsigned int reg;

	if (peer)
		mmio = ndev->peer_mmio;
	else
		mmio = ndev->self_mmio;

	reg = readl(mmio + AMD_SIDEINFO_OFFSET);
	if (!(reg & AMD_SIDE_READY)) {
		reg |= AMD_SIDE_READY;
		writel(reg, mmio + AMD_SIDEINFO_OFFSET);
	}
}

static void amd_clear_side_info_reg(struct amd_ntb_dev *ndev, bool peer)
{
	void __iomem *mmio = NULL;
	unsigned int reg;

	if (peer)
		mmio = ndev->peer_mmio;
	else
		mmio = ndev->self_mmio;

	reg = readl(mmio + AMD_SIDEINFO_OFFSET);
	if (reg & AMD_SIDE_READY) {
		reg &= ~AMD_SIDE_READY;
		writel(reg, mmio + AMD_SIDEINFO_OFFSET);
		readl(mmio + AMD_SIDEINFO_OFFSET);
	}
}

static void amd_init_side_info(struct amd_ntb_dev *ndev)
{
	void __iomem *mmio = ndev->self_mmio;
	u32 ntb_ctl;

	amd_set_side_info_reg(ndev, false);

	ntb_ctl = readl(mmio + AMD_CNTL_OFFSET);
	ntb_ctl |= (PMM_REG_CTL | SMM_REG_CTL);
	writel(ntb_ctl, mmio + AMD_CNTL_OFFSET);
}

static void amd_deinit_side_info(struct amd_ntb_dev *ndev)
{
	void __iomem *mmio = ndev->self_mmio;
	u32 ntb_ctl;

	amd_clear_side_info_reg(ndev, false);

	ntb_ctl = readl(mmio + AMD_CNTL_OFFSET);
	ntb_ctl &= ~(PMM_REG_CTL | SMM_REG_CTL);
	writel(ntb_ctl, mmio + AMD_CNTL_OFFSET);
}

static int amd_init_ntb(struct amd_ntb_dev *ndev)
{
	void __iomem *mmio = ndev->self_mmio;

	ndev->mw_count = ndev->dev_data->mw_count;
	ndev->spad_count = AMD_SPADS_CNT;
	ndev->db_count = AMD_DB_CNT;

	switch (ndev->ntb.topo) {
	case NTB_TOPO_PRI:
	case NTB_TOPO_SEC:
		ndev->spad_count >>= 1;
		if (ndev->ntb.topo == NTB_TOPO_PRI) {
			ndev->self_spad = 0;
			ndev->peer_spad = 0x20;
		} else {
			ndev->self_spad = 0x20;
			ndev->peer_spad = 0;
		}

		INIT_DELAYED_WORK(&ndev->hb_timer, amd_link_hb);
		schedule_delayed_work(&ndev->hb_timer, AMD_LINK_HB_TIMEOUT);

		break;
	default:
		dev_err(&ndev->ntb.pdev->dev,
			"AMD NTB does not support B2B mode.\n");
		return -EINVAL;
	}

	/* Mask event interrupts */
	writel(ndev->int_mask, mmio + AMD_INTMASK_OFFSET);

	return 0;
}

static enum ntb_topo amd_get_topo(struct amd_ntb_dev *ndev)
{
	void __iomem *mmio = ndev->self_mmio;
	u32 info;

	info = readl(mmio + AMD_SIDEINFO_OFFSET);
	if (info & AMD_SIDE_MASK)
		return NTB_TOPO_SEC;
	else
		return NTB_TOPO_PRI;
}

static int amd_init_dev(struct amd_ntb_dev *ndev)
{
	void __iomem *mmio = ndev->self_mmio;
	struct pci_dev *pdev;
	int rc = 0;

	pdev = ndev->ntb.pdev;

	ndev->ntb.topo = amd_get_topo(ndev);
	dev_dbg(&pdev->dev, "AMD NTB topo is %s\n",
		ntb_topo_string(ndev->ntb.topo));

	rc = amd_init_ntb(ndev);
	if (rc)
		return rc;

	rc = amd_init_isr(ndev);
	if (rc) {
		dev_err(&pdev->dev, "fail to init isr.\n");
		return rc;
	}

	ndev->db_valid_mask = BIT_ULL(ndev->db_count) - 1;
	/*
	 * We reserve the highest order bit of the DB register which will
	 * be used to notify peer when the driver on this side is being
	 * un-loaded.
	 */
	ndev->db_last_bit =
			find_last_bit((unsigned long *)&ndev->db_valid_mask,
				      hweight64(ndev->db_valid_mask));
	writew((u16)~BIT(ndev->db_last_bit), mmio + AMD_DBMASK_OFFSET);
	/*
	 * Since now there is one less bit to account for, the DB count
	 * and DB mask should be adjusted accordingly.
	 */
	ndev->db_count -= 1;
	ndev->db_valid_mask = BIT_ULL(ndev->db_count) - 1;

	/* Enable Link-Up and Link-Down event interrupts */
	ndev->int_mask &= ~(AMD_LINK_UP_EVENT | AMD_LINK_DOWN_EVENT);
	writel(ndev->int_mask, mmio + AMD_INTMASK_OFFSET);

	return 0;
}

static void amd_deinit_dev(struct amd_ntb_dev *ndev)
{
	cancel_delayed_work_sync(&ndev->hb_timer);

	ndev_deinit_isr(ndev);
}

static int amd_ntb_init_pci(struct amd_ntb_dev *ndev,
			    struct pci_dev *pdev)
{
	int rc;

	pci_set_drvdata(pdev, ndev);

	rc = pci_enable_device(pdev);
	if (rc)
		goto err_pci_enable;

	rc = pci_request_regions(pdev, NTB_NAME);
	if (rc)
		goto err_pci_regions;

	pci_set_master(pdev);

	rc = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
	if (rc) {
		rc = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
		if (rc)
			goto err_dma_mask;
		dev_warn(&pdev->dev, "Cannot DMA highmem\n");
	}

	ndev->self_mmio = pci_iomap(pdev, 0, 0);
	if (!ndev->self_mmio) {
		rc = -EIO;
		goto err_dma_mask;
	}
	ndev->peer_mmio = ndev->self_mmio + AMD_PEER_OFFSET;

	return 0;

err_dma_mask:
	pci_clear_master(pdev);
	pci_release_regions(pdev);
err_pci_regions:
	pci_disable_device(pdev);
err_pci_enable:
	pci_set_drvdata(pdev, NULL);
	return rc;
}

static void amd_ntb_deinit_pci(struct amd_ntb_dev *ndev)
{
	struct pci_dev *pdev = ndev->ntb.pdev;

	pci_iounmap(pdev, ndev->self_mmio);

	pci_clear_master(pdev);
	pci_release_regions(pdev);
	pci_disable_device(pdev);
	pci_set_drvdata(pdev, NULL);
}

static int amd_ntb_pci_probe(struct pci_dev *pdev,
			     const struct pci_device_id *id)
{
	struct amd_ntb_dev *ndev;
	int rc, node;

	node = dev_to_node(&pdev->dev);

	ndev = kzalloc_node(sizeof(*ndev), GFP_KERNEL, node);
	if (!ndev) {
		rc = -ENOMEM;
		goto err_ndev;
	}

	ndev->dev_data = (struct ntb_dev_data *)id->driver_data;

	ndev_init_struct(ndev, pdev);

	rc = amd_ntb_init_pci(ndev, pdev);
	if (rc)
		goto err_init_pci;

	rc = amd_init_dev(ndev);
	if (rc)
		goto err_init_dev;

	/* write side info */
	amd_init_side_info(ndev);

	amd_poll_link(ndev);

	ndev_init_debugfs(ndev);

	rc = ntb_register_device(&ndev->ntb);
	if (rc)
		goto err_register;

	dev_info(&pdev->dev, "NTB device registered.\n");

	return 0;

err_register:
	ndev_deinit_debugfs(ndev);
	amd_deinit_dev(ndev);
err_init_dev:
	amd_ntb_deinit_pci(ndev);
err_init_pci:
	kfree(ndev);
err_ndev:
	return rc;
}

static void amd_ntb_pci_remove(struct pci_dev *pdev)
{
	struct amd_ntb_dev *ndev = pci_get_drvdata(pdev);

	/*
	 * Clear the READY bit in SIDEINFO register before sending DB event
	 * to the peer. This will make sure that when the peer handles the
	 * DB event, it correctly reads this bit as being 0.
	 */
	amd_deinit_side_info(ndev);
	ntb_peer_db_set(&ndev->ntb, BIT_ULL(ndev->db_last_bit));
	ntb_unregister_device(&ndev->ntb);
	ndev_deinit_debugfs(ndev);
	amd_deinit_dev(ndev);
	amd_ntb_deinit_pci(ndev);
	kfree(ndev);
}

static void amd_ntb_pci_shutdown(struct pci_dev *pdev)
{
	struct amd_ntb_dev *ndev = pci_get_drvdata(pdev);

	/* Send link down notification */
	ntb_link_event(&ndev->ntb);

	amd_deinit_side_info(ndev);
	ntb_peer_db_set(&ndev->ntb, BIT_ULL(ndev->db_last_bit));
	ntb_unregister_device(&ndev->ntb);
	ndev_deinit_debugfs(ndev);
	amd_deinit_dev(ndev);
	amd_ntb_deinit_pci(ndev);
	kfree(ndev);
}

static const struct file_operations amd_ntb_debugfs_info = {
	.owner = THIS_MODULE,
	.open = simple_open,
	.read = ndev_debugfs_read,
};

static const struct ntb_dev_data dev_data[] = {
	{ /* for device 145b */
		.mw_count = 3,
		.mw_idx = 1,
	},
	{ /* for device 148b */
		.mw_count = 2,
		.mw_idx = 2,
	},
};

static const struct pci_device_id amd_ntb_pci_tbl[] = {
	{ PCI_VDEVICE(AMD, 0x145b), (kernel_ulong_t)&dev_data[0] },
	{ PCI_VDEVICE(AMD, 0x148b), (kernel_ulong_t)&dev_data[1] },
	{ PCI_VDEVICE(HYGON, 0x145b), (kernel_ulong_t)&dev_data[0] },
	{ 0, }
};
MODULE_DEVICE_TABLE(pci, amd_ntb_pci_tbl);

static struct pci_driver amd_ntb_pci_driver = {
	.name		= KBUILD_MODNAME,
	.id_table	= amd_ntb_pci_tbl,
	.probe		= amd_ntb_pci_probe,
	.remove		= amd_ntb_pci_remove,
	.shutdown	= amd_ntb_pci_shutdown,
};

static int __init amd_ntb_pci_driver_init(void)
{
	pr_info("%s %s\n", NTB_DESC, NTB_VER);

	if (debugfs_initialized())
		debugfs_dir = debugfs_create_dir(KBUILD_MODNAME, NULL);

	return pci_register_driver(&amd_ntb_pci_driver);
}
module_init(amd_ntb_pci_driver_init);

static void __exit amd_ntb_pci_driver_exit(void)
{
	pci_unregister_driver(&amd_ntb_pci_driver);
	debugfs_remove_recursive(debugfs_dir);
}
module_exit(amd_ntb_pci_driver_exit);