summaryrefslogtreecommitdiffstats
path: root/arch/sparc64/kernel/pci_sun4v.c
blob: 7bee6b19272ebf8e733518efecf71dd60979c125 (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
/* pci_sun4v.c: SUN4V specific PCI controller support.
 *
 * Copyright (C) 2006, 2007 David S. Miller (davem@davemloft.net)
 */

#include <linux/kernel.h>
#include <linux/types.h>
#include <linux/pci.h>
#include <linux/init.h>
#include <linux/slab.h>
#include <linux/interrupt.h>
#include <linux/percpu.h>
#include <linux/irq.h>
#include <linux/msi.h>

#include <asm/pbm.h>
#include <asm/iommu.h>
#include <asm/irq.h>
#include <asm/upa.h>
#include <asm/pstate.h>
#include <asm/oplib.h>
#include <asm/hypervisor.h>
#include <asm/prom.h>

#include "pci_impl.h"
#include "iommu_common.h"

#include "pci_sun4v.h"

#define PGLIST_NENTS	(PAGE_SIZE / sizeof(u64))

struct pci_iommu_batch {
	struct pci_dev	*pdev;		/* Device mapping is for.	*/
	unsigned long	prot;		/* IOMMU page protections	*/
	unsigned long	entry;		/* Index into IOTSB.		*/
	u64		*pglist;	/* List of physical pages	*/
	unsigned long	npages;		/* Number of pages in list.	*/
};

static DEFINE_PER_CPU(struct pci_iommu_batch, pci_iommu_batch);

/* Interrupts must be disabled.  */
static inline void pci_iommu_batch_start(struct pci_dev *pdev, unsigned long prot, unsigned long entry)
{
	struct pci_iommu_batch *p = &__get_cpu_var(pci_iommu_batch);

	p->pdev		= pdev;
	p->prot		= prot;
	p->entry	= entry;
	p->npages	= 0;
}

/* Interrupts must be disabled.  */
static long pci_iommu_batch_flush(struct pci_iommu_batch *p)
{
	struct pci_pbm_info *pbm = p->pdev->dev.archdata.host_controller;
	unsigned long devhandle = pbm->devhandle;
	unsigned long prot = p->prot;
	unsigned long entry = p->entry;
	u64 *pglist = p->pglist;
	unsigned long npages = p->npages;

	while (npages != 0) {
		long num;

		num = pci_sun4v_iommu_map(devhandle, HV_PCI_TSBID(0, entry),
					  npages, prot, __pa(pglist));
		if (unlikely(num < 0)) {
			if (printk_ratelimit())
				printk("pci_iommu_batch_flush: IOMMU map of "
				       "[%08lx:%08lx:%lx:%lx:%lx] failed with "
				       "status %ld\n",
				       devhandle, HV_PCI_TSBID(0, entry),
				       npages, prot, __pa(pglist), num);
			return -1;
		}

		entry += num;
		npages -= num;
		pglist += num;
	}

	p->entry = entry;
	p->npages = 0;

	return 0;
}

/* Interrupts must be disabled.  */
static inline long pci_iommu_batch_add(u64 phys_page)
{
	struct pci_iommu_batch *p = &__get_cpu_var(pci_iommu_batch);

	BUG_ON(p->npages >= PGLIST_NENTS);

	p->pglist[p->npages++] = phys_page;
	if (p->npages == PGLIST_NENTS)
		return pci_iommu_batch_flush(p);

	return 0;
}

/* Interrupts must be disabled.  */
static inline long pci_iommu_batch_end(void)
{
	struct pci_iommu_batch *p = &__get_cpu_var(pci_iommu_batch);

	BUG_ON(p->npages >= PGLIST_NENTS);

	return pci_iommu_batch_flush(p);
}

static long pci_arena_alloc(struct pci_iommu_arena *arena, unsigned long npages)
{
	unsigned long n, i, start, end, limit;
	int pass;

	limit = arena->limit;
	start = arena->hint;
	pass = 0;

again:
	n = find_next_zero_bit(arena->map, limit, start);
	end = n + npages;
	if (unlikely(end >= limit)) {
		if (likely(pass < 1)) {
			limit = start;
			start = 0;
			pass++;
			goto again;
		} else {
			/* Scanned the whole thing, give up. */
			return -1;
		}
	}

	for (i = n; i < end; i++) {
		if (test_bit(i, arena->map)) {
			start = i + 1;
			goto again;
		}
	}

	for (i = n; i < end; i++)
		__set_bit(i, arena->map);

	arena->hint = end;

	return n;
}

static void pci_arena_free(struct pci_iommu_arena *arena, unsigned long base, unsigned long npages)
{
	unsigned long i;

	for (i = base; i < (base + npages); i++)
		__clear_bit(i, arena->map);
}

static void *pci_4v_alloc_consistent(struct pci_dev *pdev, size_t size, dma_addr_t *dma_addrp, gfp_t gfp)
{
	struct pci_iommu *iommu;
	unsigned long flags, order, first_page, npages, n;
	void *ret;
	long entry;

	size = IO_PAGE_ALIGN(size);
	order = get_order(size);
	if (unlikely(order >= MAX_ORDER))
		return NULL;

	npages = size >> IO_PAGE_SHIFT;

	first_page = __get_free_pages(gfp, order);
	if (unlikely(first_page == 0UL))
		return NULL;

	memset((char *)first_page, 0, PAGE_SIZE << order);

	iommu = pdev->dev.archdata.iommu;

	spin_lock_irqsave(&iommu->lock, flags);
	entry = pci_arena_alloc(&iommu->arena, npages);
	spin_unlock_irqrestore(&iommu->lock, flags);

	if (unlikely(entry < 0L))
		goto arena_alloc_fail;

	*dma_addrp = (iommu->page_table_map_base +
		      (entry << IO_PAGE_SHIFT));
	ret = (void *) first_page;
	first_page = __pa(first_page);

	local_irq_save(flags);

	pci_iommu_batch_start(pdev,
			      (HV_PCI_MAP_ATTR_READ |
			       HV_PCI_MAP_ATTR_WRITE),
			      entry);

	for (n = 0; n < npages; n++) {
		long err = pci_iommu_batch_add(first_page + (n * PAGE_SIZE));
		if (unlikely(err < 0L))
			goto iommu_map_fail;
	}

	if (unlikely(pci_iommu_batch_end() < 0L))
		goto iommu_map_fail;

	local_irq_restore(flags);

	return ret;

iommu_map_fail:
	/* Interrupts are disabled.  */
	spin_lock(&iommu->lock);
	pci_arena_free(&iommu->arena, entry, npages);
	spin_unlock_irqrestore(&iommu->lock, flags);

arena_alloc_fail:
	free_pages(first_page, order);
	return NULL;
}

static void pci_4v_free_consistent(struct pci_dev *pdev, size_t size, void *cpu, dma_addr_t dvma)
{
	struct pci_pbm_info *pbm;
	struct pci_iommu *iommu;
	unsigned long flags, order, npages, entry;
	u32 devhandle;

	npages = IO_PAGE_ALIGN(size) >> IO_PAGE_SHIFT;
	iommu = pdev->dev.archdata.iommu;
	pbm = pdev->dev.archdata.host_controller;
	devhandle = pbm->devhandle;
	entry = ((dvma - iommu->page_table_map_base) >> IO_PAGE_SHIFT);

	spin_lock_irqsave(&iommu->lock, flags);

	pci_arena_free(&iommu->arena, entry, npages);

	do {
		unsigned long num;

		num = pci_sun4v_iommu_demap(devhandle, HV_PCI_TSBID(0, entry),
					    npages);
		entry += num;
		npages -= num;
	} while (npages != 0);

	spin_unlock_irqrestore(&iommu->lock, flags);

	order = get_order(size);
	if (order < 10)
		free_pages((unsigned long)cpu, order);
}

static dma_addr_t pci_4v_map_single(struct pci_dev *pdev, void *ptr, size_t sz, int direction)
{
	struct pci_iommu *iommu;
	unsigned long flags, npages, oaddr;
	unsigned long i, base_paddr;
	u32 bus_addr, ret;
	unsigned long prot;
	long entry;

	iommu = pdev->dev.archdata.iommu;

	if (unlikely(direction == PCI_DMA_NONE))
		goto bad;

	oaddr = (unsigned long)ptr;
	npages = IO_PAGE_ALIGN(oaddr + sz) - (oaddr & IO_PAGE_MASK);
	npages >>= IO_PAGE_SHIFT;

	spin_lock_irqsave(&iommu->lock, flags);
	entry = pci_arena_alloc(&iommu->arena, npages);
	spin_unlock_irqrestore(&iommu->lock, flags);

	if (unlikely(entry < 0L))
		goto bad;

	bus_addr = (iommu->page_table_map_base +
		    (entry << IO_PAGE_SHIFT));
	ret = bus_addr | (oaddr & ~IO_PAGE_MASK);
	base_paddr = __pa(oaddr & IO_PAGE_MASK);
	prot = HV_PCI_MAP_ATTR_READ;
	if (direction != PCI_DMA_TODEVICE)
		prot |= HV_PCI_MAP_ATTR_WRITE;

	local_irq_save(flags);

	pci_iommu_batch_start(pdev, prot, entry);

	for (i = 0; i < npages; i++, base_paddr += IO_PAGE_SIZE) {
		long err = pci_iommu_batch_add(base_paddr);
		if (unlikely(err < 0L))
			goto iommu_map_fail;
	}
	if (unlikely(pci_iommu_batch_end() < 0L))
		goto iommu_map_fail;

	local_irq_restore(flags);

	return ret;

bad:
	if (printk_ratelimit())
		WARN_ON(1);
	return PCI_DMA_ERROR_CODE;

iommu_map_fail:
	/* Interrupts are disabled.  */
	spin_lock(&iommu->lock);
	pci_arena_free(&iommu->arena, entry, npages);
	spin_unlock_irqrestore(&iommu->lock, flags);

	return PCI_DMA_ERROR_CODE;
}

static void pci_4v_unmap_single(struct pci_dev *pdev, dma_addr_t bus_addr, size_t sz, int direction)
{
	struct pci_pbm_info *pbm;
	struct pci_iommu *iommu;
	unsigned long flags, npages;
	long entry;
	u32 devhandle;

	if (unlikely(direction == PCI_DMA_NONE)) {
		if (printk_ratelimit())
			WARN_ON(1);
		return;
	}

	iommu = pdev->dev.archdata.iommu;
	pbm = pdev->dev.archdata.host_controller;
	devhandle = pbm->devhandle;

	npages = IO_PAGE_ALIGN(bus_addr + sz) - (bus_addr & IO_PAGE_MASK);
	npages >>= IO_PAGE_SHIFT;
	bus_addr &= IO_PAGE_MASK;

	spin_lock_irqsave(&iommu->lock, flags);

	entry = (bus_addr - iommu->page_table_map_base) >> IO_PAGE_SHIFT;
	pci_arena_free(&iommu->arena, entry, npages);

	do {
		unsigned long num;

		num = pci_sun4v_iommu_demap(devhandle, HV_PCI_TSBID(0, entry),
					    npages);
		entry += num;
		npages -= num;
	} while (npages != 0);

	spin_unlock_irqrestore(&iommu->lock, flags);
}

#define SG_ENT_PHYS_ADDRESS(SG)	\
	(__pa(page_address((SG)->page)) + (SG)->offset)

static inline long fill_sg(long entry, struct pci_dev *pdev,
			   struct scatterlist *sg,
			   int nused, int nelems, unsigned long prot)
{
	struct scatterlist *dma_sg = sg;
	struct scatterlist *sg_end = sg + nelems;
	unsigned long flags;
	int i;

	local_irq_save(flags);

	pci_iommu_batch_start(pdev, prot, entry);

	for (i = 0; i < nused; i++) {
		unsigned long pteval = ~0UL;
		u32 dma_npages;

		dma_npages = ((dma_sg->dma_address & (IO_PAGE_SIZE - 1UL)) +
			      dma_sg->dma_length +
			      ((IO_PAGE_SIZE - 1UL))) >> IO_PAGE_SHIFT;
		do {
			unsigned long offset;
			signed int len;

			/* If we are here, we know we have at least one
			 * more page to map.  So walk forward until we
			 * hit a page crossing, and begin creating new
			 * mappings from that spot.
			 */
			for (;;) {
				unsigned long tmp;

				tmp = SG_ENT_PHYS_ADDRESS(sg);
				len = sg->length;
				if (((tmp ^ pteval) >> IO_PAGE_SHIFT) != 0UL) {
					pteval = tmp & IO_PAGE_MASK;
					offset = tmp & (IO_PAGE_SIZE - 1UL);
					break;
				}
				if (((tmp ^ (tmp + len - 1UL)) >> IO_PAGE_SHIFT) != 0UL) {
					pteval = (tmp + IO_PAGE_SIZE) & IO_PAGE_MASK;
					offset = 0UL;
					len -= (IO_PAGE_SIZE - (tmp & (IO_PAGE_SIZE - 1UL)));
					break;
				}
				sg++;
			}

			pteval = (pteval & IOPTE_PAGE);
			while (len > 0) {
				long err;

				err = pci_iommu_batch_add(pteval);
				if (unlikely(err < 0L))
					goto iommu_map_failed;

				pteval += IO_PAGE_SIZE;
				len -= (IO_PAGE_SIZE - offset);
				offset = 0;
				dma_npages--;
			}

			pteval = (pteval & IOPTE_PAGE) + len;
			sg++;

			/* Skip over any tail mappings we've fully mapped,
			 * adjusting pteval along the way.  Stop when we
			 * detect a page crossing event.
			 */
			while (sg < sg_end &&
			       (pteval << (64 - IO_PAGE_SHIFT)) != 0UL &&
			       (pteval == SG_ENT_PHYS_ADDRESS(sg)) &&
			       ((pteval ^
				 (SG_ENT_PHYS_ADDRESS(sg) + sg->length - 1UL)) >> IO_PAGE_SHIFT) == 0UL) {
				pteval += sg->length;
				sg++;
			}
			if ((pteval << (64 - IO_PAGE_SHIFT)) == 0UL)
				pteval = ~0UL;
		} while (dma_npages != 0);
		dma_sg++;
	}

	if (unlikely(pci_iommu_batch_end() < 0L))
		goto iommu_map_failed;

	local_irq_restore(flags);
	return 0;

iommu_map_failed:
	local_irq_restore(flags);
	return -1L;
}

static int pci_4v_map_sg(struct pci_dev *pdev, struct scatterlist *sglist, int nelems, int direction)
{
	struct pci_iommu *iommu;
	unsigned long flags, npages, prot;
	u32 dma_base;
	struct scatterlist *sgtmp;
	long entry, err;
	int used;

	/* Fast path single entry scatterlists. */
	if (nelems == 1) {
		sglist->dma_address =
			pci_4v_map_single(pdev,
					  (page_address(sglist->page) + sglist->offset),
					  sglist->length, direction);
		if (unlikely(sglist->dma_address == PCI_DMA_ERROR_CODE))
			return 0;
		sglist->dma_length = sglist->length;
		return 1;
	}

	iommu = pdev->dev.archdata.iommu;
	
	if (unlikely(direction == PCI_DMA_NONE))
		goto bad;

	/* Step 1: Prepare scatter list. */
	npages = prepare_sg(sglist, nelems);

	/* Step 2: Allocate a cluster and context, if necessary. */
	spin_lock_irqsave(&iommu->lock, flags);
	entry = pci_arena_alloc(&iommu->arena, npages);
	spin_unlock_irqrestore(&iommu->lock, flags);

	if (unlikely(entry < 0L))
		goto bad;

	dma_base = iommu->page_table_map_base +
		(entry << IO_PAGE_SHIFT);

	/* Step 3: Normalize DMA addresses. */
	used = nelems;

	sgtmp = sglist;
	while (used && sgtmp->dma_length) {
		sgtmp->dma_address += dma_base;
		sgtmp++;
		used--;
	}
	used = nelems - used;

	/* Step 4: Create the mappings. */
	prot = HV_PCI_MAP_ATTR_READ;
	if (direction != PCI_DMA_TODEVICE)
		prot |= HV_PCI_MAP_ATTR_WRITE;

	err = fill_sg(entry, pdev, sglist, used, nelems, prot);
	if (unlikely(err < 0L))
		goto iommu_map_failed;

	return used;

bad:
	if (printk_ratelimit())
		WARN_ON(1);
	return 0;

iommu_map_failed:
	spin_lock_irqsave(&iommu->lock, flags);
	pci_arena_free(&iommu->arena, entry, npages);
	spin_unlock_irqrestore(&iommu->lock, flags);

	return 0;
}

static void pci_4v_unmap_sg(struct pci_dev *pdev, struct scatterlist *sglist, int nelems, int direction)
{
	struct pci_pbm_info *pbm;
	struct pci_iommu *iommu;
	unsigned long flags, i, npages;
	long entry;
	u32 devhandle, bus_addr;

	if (unlikely(direction == PCI_DMA_NONE)) {
		if (printk_ratelimit())
			WARN_ON(1);
	}

	iommu = pdev->dev.archdata.iommu;
	pbm = pdev->dev.archdata.host_controller;
	devhandle = pbm->devhandle;
	
	bus_addr = sglist->dma_address & IO_PAGE_MASK;

	for (i = 1; i < nelems; i++)
		if (sglist[i].dma_length == 0)
			break;
	i--;
	npages = (IO_PAGE_ALIGN(sglist[i].dma_address + sglist[i].dma_length) -
		  bus_addr) >> IO_PAGE_SHIFT;

	entry = ((bus_addr - iommu->page_table_map_base) >> IO_PAGE_SHIFT);

	spin_lock_irqsave(&iommu->lock, flags);

	pci_arena_free(&iommu->arena, entry, npages);

	do {
		unsigned long num;

		num = pci_sun4v_iommu_demap(devhandle, HV_PCI_TSBID(0, entry),
					    npages);
		entry += num;
		npages -= num;
	} while (npages != 0);

	spin_unlock_irqrestore(&iommu->lock, flags);
}

static void pci_4v_dma_sync_single_for_cpu(struct pci_dev *pdev, dma_addr_t bus_addr, size_t sz, int direction)
{
	/* Nothing to do... */
}

static void pci_4v_dma_sync_sg_for_cpu(struct pci_dev *pdev, struct scatterlist *sglist, int nelems, int direction)
{
	/* Nothing to do... */
}

struct pci_iommu_ops pci_sun4v_iommu_ops = {
	.alloc_consistent		= pci_4v_alloc_consistent,
	.free_consistent		= pci_4v_free_consistent,
	.map_single			= pci_4v_map_single,
	.unmap_single			= pci_4v_unmap_single,
	.map_sg				= pci_4v_map_sg,
	.unmap_sg			= pci_4v_unmap_sg,
	.dma_sync_single_for_cpu	= pci_4v_dma_sync_single_for_cpu,
	.dma_sync_sg_for_cpu		= pci_4v_dma_sync_sg_for_cpu,
};

static inline int pci_sun4v_out_of_range(struct pci_pbm_info *pbm, unsigned int bus, unsigned int device, unsigned int func)
{
	if (bus < pbm->pci_first_busno ||
	    bus > pbm->pci_last_busno)
		return 1;
	return 0;
}

static int pci_sun4v_read_pci_cfg(struct pci_bus *bus_dev, unsigned int devfn,
				  int where, int size, u32 *value)
{
	struct pci_pbm_info *pbm = bus_dev->sysdata;
	u32 devhandle = pbm->devhandle;
	unsigned int bus = bus_dev->number;
	unsigned int device = PCI_SLOT(devfn);
	unsigned int func = PCI_FUNC(devfn);
	unsigned long ret;

	if (pci_sun4v_out_of_range(pbm, bus, device, func)) {
		ret = ~0UL;
	} else {
		ret = pci_sun4v_config_get(devhandle,
				HV_PCI_DEVICE_BUILD(bus, device, func),
				where, size);
#if 0
		printk("rcfg: [%x:%x:%x:%d]=[%lx]\n",
		       devhandle, HV_PCI_DEVICE_BUILD(bus, device, func),
		       where, size, ret);
#endif
	}
	switch (size) {
	case 1:
		*value = ret & 0xff;
		break;
	case 2:
		*value = ret & 0xffff;
		break;
	case 4:
		*value = ret & 0xffffffff;
		break;
	};


	return PCIBIOS_SUCCESSFUL;
}

static int pci_sun4v_write_pci_cfg(struct pci_bus *bus_dev, unsigned int devfn,
				   int where, int size, u32 value)
{
	struct pci_pbm_info *pbm = bus_dev->sysdata;
	u32 devhandle = pbm->devhandle;
	unsigned int bus = bus_dev->number;
	unsigned int device = PCI_SLOT(devfn);
	unsigned int func = PCI_FUNC(devfn);
	unsigned long ret;

	if (pci_sun4v_out_of_range(pbm, bus, device, func)) {
		/* Do nothing. */
	} else {
		ret = pci_sun4v_config_put(devhandle,
				HV_PCI_DEVICE_BUILD(bus, device, func),
				where, size, value);
#if 0
		printk("wcfg: [%x:%x:%x:%d] v[%x] == [%lx]\n",
		       devhandle, HV_PCI_DEVICE_BUILD(bus, device, func),
		       where, size, value, ret);
#endif
	}
	return PCIBIOS_SUCCESSFUL;
}

static struct pci_ops pci_sun4v_ops = {
	.read =		pci_sun4v_read_pci_cfg,
	.write =	pci_sun4v_write_pci_cfg,
};


static void pbm_scan_bus(struct pci_controller_info *p,
			 struct pci_pbm_info *pbm)
{
	pbm->pci_bus = pci_scan_one_pbm(pbm);
}

static void pci_sun4v_scan_bus(struct pci_controller_info *p)
{
	struct property *prop;
	struct device_node *dp;

	if ((dp = p->pbm_A.prom_node) != NULL) {
		prop = of_find_property(dp, "66mhz-capable", NULL);
		p->pbm_A.is_66mhz_capable = (prop != NULL);

		pbm_scan_bus(p, &p->pbm_A);
	}
	if ((dp = p->pbm_B.prom_node) != NULL) {
		prop = of_find_property(dp, "66mhz-capable", NULL);
		p->pbm_B.is_66mhz_capable = (prop != NULL);

		pbm_scan_bus(p, &p->pbm_B);
	}

	/* XXX register error interrupt handlers XXX */
}

static void pci_sun4v_base_address_update(struct pci_dev *pdev, int resource)
{
	struct pci_pbm_info *pbm = pdev->dev.archdata.host_controller;
	struct resource *res, *root;
	u32 reg;
	int where, size, is_64bit;

	res = &pdev->resource[resource];
	if (resource < 6) {
		where = PCI_BASE_ADDRESS_0 + (resource * 4);
	} else if (resource == PCI_ROM_RESOURCE) {
		where = pdev->rom_base_reg;
	} else {
		/* Somebody might have asked allocation of a non-standard resource */
		return;
	}

	/* XXX 64-bit MEM handling is not %100 correct... XXX */
	is_64bit = 0;
	if (res->flags & IORESOURCE_IO)
		root = &pbm->io_space;
	else {
		root = &pbm->mem_space;
		if ((res->flags & PCI_BASE_ADDRESS_MEM_TYPE_MASK)
		    == PCI_BASE_ADDRESS_MEM_TYPE_64)
			is_64bit = 1;
	}

	size = res->end - res->start;
	pci_read_config_dword(pdev, where, &reg);
	reg = ((reg & size) |
	       (((u32)(res->start - root->start)) & ~size));
	if (resource == PCI_ROM_RESOURCE) {
		reg |= PCI_ROM_ADDRESS_ENABLE;
		res->flags |= IORESOURCE_ROM_ENABLE;
	}
	pci_write_config_dword(pdev, where, reg);

	/* This knows that the upper 32-bits of the address
	 * must be zero.  Our PCI common layer enforces this.
	 */
	if (is_64bit)
		pci_write_config_dword(pdev, where + 4, 0);
}

static void pci_sun4v_resource_adjust(struct pci_dev *pdev,
				      struct resource *res,
				      struct resource *root)
{
	res->start += root->start;
	res->end += root->start;
}

static unsigned long probe_existing_entries(struct pci_pbm_info *pbm,
					    struct pci_iommu *iommu)
{
	struct pci_iommu_arena *arena = &iommu->arena;
	unsigned long i, cnt = 0;
	u32 devhandle;

	devhandle = pbm->devhandle;
	for (i = 0; i < arena->limit; i++) {
		unsigned long ret, io_attrs, ra;

		ret = pci_sun4v_iommu_getmap(devhandle,
					     HV_PCI_TSBID(0, i),
					     &io_attrs, &ra);
		if (ret == HV_EOK) {
			if (page_in_phys_avail(ra)) {
				pci_sun4v_iommu_demap(devhandle,
						      HV_PCI_TSBID(0, i), 1);
			} else {
				cnt++;
				__set_bit(i, arena->map);
			}
		}
	}

	return cnt;
}

static void pci_sun4v_iommu_init(struct pci_pbm_info *pbm)
{
	struct pci_iommu *iommu = pbm->iommu;
	struct property *prop;
	unsigned long num_tsb_entries, sz;
	u32 vdma[2], dma_mask, dma_offset;
	int tsbsize;

	prop = of_find_property(pbm->prom_node, "virtual-dma", NULL);
	if (prop) {
		u32 *val = prop->value;

		vdma[0] = val[0];
		vdma[1] = val[1];
	} else {
		/* No property, use default values. */
		vdma[0] = 0x80000000;
		vdma[1] = 0x80000000;
	}

	dma_mask = vdma[0];
	switch (vdma[1]) {
		case 0x20000000:
			dma_mask |= 0x1fffffff;
			tsbsize = 64;
			break;

		case 0x40000000:
			dma_mask |= 0x3fffffff;
			tsbsize = 128;
			break;

		case 0x80000000:
			dma_mask |= 0x7fffffff;
			tsbsize = 256;
			break;

		default:
			prom_printf("PCI-SUN4V: strange virtual-dma size.\n");
			prom_halt();
	};

	tsbsize *= (8 * 1024);

	num_tsb_entries = tsbsize / sizeof(iopte_t);

	dma_offset = vdma[0];

	/* Setup initial software IOMMU state. */
	spin_lock_init(&iommu->lock);
	iommu->ctx_lowest_free = 1;
	iommu->page_table_map_base = dma_offset;
	iommu->dma_addr_mask = dma_mask;

	/* Allocate and initialize the free area map.  */
	sz = num_tsb_entries / 8;
	sz = (sz + 7UL) & ~7UL;
	iommu->arena.map = kzalloc(sz, GFP_KERNEL);
	if (!iommu->arena.map) {
		prom_printf("PCI_IOMMU: Error, kmalloc(arena.map) failed.\n");
		prom_halt();
	}
	iommu->arena.limit = num_tsb_entries;

	sz = probe_existing_entries(pbm, iommu);
	if (sz)
		printk("%s: Imported %lu TSB entries from OBP\n",
		       pbm->name, sz);
}

static void pci_sun4v_get_bus_range(struct pci_pbm_info *pbm)
{
	struct property *prop;
	unsigned int *busrange;

	prop = of_find_property(pbm->prom_node, "bus-range", NULL);

	busrange = prop->value;

	pbm->pci_first_busno = busrange[0];
	pbm->pci_last_busno = busrange[1];

}

#ifdef CONFIG_PCI_MSI
struct pci_sun4v_msiq_entry {
	u64		version_type;
#define MSIQ_VERSION_MASK		0xffffffff00000000UL
#define MSIQ_VERSION_SHIFT		32
#define MSIQ_TYPE_MASK			0x00000000000000ffUL
#define MSIQ_TYPE_SHIFT			0
#define MSIQ_TYPE_NONE			0x00
#define MSIQ_TYPE_MSG			0x01
#define MSIQ_TYPE_MSI32			0x02
#define MSIQ_TYPE_MSI64			0x03
#define MSIQ_TYPE_INTX			0x08
#define MSIQ_TYPE_NONE2			0xff

	u64		intx_sysino;
	u64		reserved1;
	u64		stick;
	u64		req_id;  /* bus/device/func */
#define MSIQ_REQID_BUS_MASK		0xff00UL
#define MSIQ_REQID_BUS_SHIFT		8
#define MSIQ_REQID_DEVICE_MASK		0x00f8UL
#define MSIQ_REQID_DEVICE_SHIFT		3
#define MSIQ_REQID_FUNC_MASK		0x0007UL
#define MSIQ_REQID_FUNC_SHIFT		0

	u64		msi_address;

	/* The format of this value is message type dependant.
	 * For MSI bits 15:0 are the data from the MSI packet.
	 * For MSI-X bits 31:0 are the data from the MSI packet.
	 * For MSG, the message code and message routing code where:
	 * 	bits 39:32 is the bus/device/fn of the msg target-id
	 *	bits 18:16 is the message routing code
	 *	bits 7:0 is the message code
	 * For INTx the low order 2-bits are:
	 *	00 - INTA
	 *	01 - INTB
	 *	10 - INTC
	 *	11 - INTD
	 */
	u64		msi_data;

	u64		reserved2;
};

/* For now this just runs as a pre-handler for the real interrupt handler.
 * So we just walk through the queue and ACK all the entries, update the
 * head pointer, and return.
 *
 * In the longer term it would be nice to do something more integrated
 * wherein we can pass in some of this MSI info to the drivers.  This
 * would be most useful for PCIe fabric error messages, although we could
 * invoke those directly from the loop here in order to pass the info around.
 */
static void pci_sun4v_msi_prehandler(unsigned int ino, void *data1, void *data2)
{
	struct pci_pbm_info *pbm = data1;
	struct pci_sun4v_msiq_entry *base, *ep;
	unsigned long msiqid, orig_head, head, type, err;

	msiqid = (unsigned long) data2;

	head = 0xdeadbeef;
	err = pci_sun4v_msiq_gethead(pbm->devhandle, msiqid, &head);
	if (unlikely(err))
		goto hv_error_get;

	if (unlikely(head >= (pbm->msiq_ent_count * sizeof(struct pci_sun4v_msiq_entry))))
		goto bad_offset;

	head /= sizeof(struct pci_sun4v_msiq_entry);
	orig_head = head;
	base = (pbm->msi_queues + ((msiqid - pbm->msiq_first) *
				   (pbm->msiq_ent_count *
				    sizeof(struct pci_sun4v_msiq_entry))));
	ep = &base[head];
	while ((ep->version_type & MSIQ_TYPE_MASK) != 0) {
		type = (ep->version_type & MSIQ_TYPE_MASK) >> MSIQ_TYPE_SHIFT;
		if (unlikely(type != MSIQ_TYPE_MSI32 &&
			     type != MSIQ_TYPE_MSI64))
			goto bad_type;

		pci_sun4v_msi_setstate(pbm->devhandle,
				       ep->msi_data /* msi_num */,
				       HV_MSISTATE_IDLE);

		/* Clear the entry.  */
		ep->version_type &= ~MSIQ_TYPE_MASK;

		/* Go to next entry in ring.  */
		head++;
		if (head >= pbm->msiq_ent_count)
			head = 0;
		ep = &base[head];
	}

	if (likely(head != orig_head)) {
		/* ACK entries by updating head pointer.  */
		head *= sizeof(struct pci_sun4v_msiq_entry);
		err = pci_sun4v_msiq_sethead(pbm->devhandle, msiqid, head);
		if (unlikely(err))
			goto hv_error_set;
	}
	return;

hv_error_set:
	printk(KERN_EMERG "MSI: Hypervisor set head gives error %lu\n", err);
	goto hv_error_cont;

hv_error_get:
	printk(KERN_EMERG "MSI: Hypervisor get head gives error %lu\n", err);

hv_error_cont:
	printk(KERN_EMERG "MSI: devhandle[%x] msiqid[%lx] head[%lu]\n",
	       pbm->devhandle, msiqid, head);
	return;

bad_offset:
	printk(KERN_EMERG "MSI: Hypervisor gives bad offset %lx max(%lx)\n",
	       head, pbm->msiq_ent_count * sizeof(struct pci_sun4v_msiq_entry));
	return;

bad_type:
	printk(KERN_EMERG "MSI: Entry has bad type %lx\n", type);
	return;
}

static int msi_bitmap_alloc(struct pci_pbm_info *pbm)
{
	unsigned long size, bits_per_ulong;

	bits_per_ulong = sizeof(unsigned long) * 8;
	size = (pbm->msi_num + (bits_per_ulong - 1)) & ~(bits_per_ulong - 1);
	size /= 8;
	BUG_ON(size % sizeof(unsigned long));

	pbm->msi_bitmap = kzalloc(size, GFP_KERNEL);
	if (!pbm->msi_bitmap)
		return -ENOMEM;

	return 0;
}

static void msi_bitmap_free(struct pci_pbm_info *pbm)
{
	kfree(pbm->msi_bitmap);
	pbm->msi_bitmap = NULL;
}

static int msi_queue_alloc(struct pci_pbm_info *pbm)
{
	unsigned long q_size, alloc_size, pages, order;
	int i;

	q_size = pbm->msiq_ent_count * sizeof(struct pci_sun4v_msiq_entry);
	alloc_size = (pbm->msiq_num * q_size);
	order = get_order(alloc_size);
	pages = __get_free_pages(GFP_KERNEL | __GFP_COMP, order);
	if (pages == 0UL) {
		printk(KERN_ERR "MSI: Cannot allocate MSI queues (o=%lu).\n",
		       order);
		return -ENOMEM;
	}
	memset((char *)pages, 0, PAGE_SIZE << order);
	pbm->msi_queues = (void *) pages;

	for (i = 0; i < pbm->msiq_num; i++) {
		unsigned long err, base = __pa(pages + (i * q_size));
		unsigned long ret1, ret2;

		err = pci_sun4v_msiq_conf(pbm->devhandle,
					  pbm->msiq_first + i,
					  base, pbm->msiq_ent_count);
		if (err) {
			printk(KERN_ERR "MSI: msiq register fails (err=%lu)\n",
			       err);
			goto h_error;
		}

		err = pci_sun4v_msiq_info(pbm->devhandle,
					  pbm->msiq_first + i,
					  &ret1, &ret2);
		if (err) {
			printk(KERN_ERR "MSI: Cannot read msiq (err=%lu)\n",
			       err);
			goto h_error;
		}
		if (ret1 != base || ret2 != pbm->msiq_ent_count) {
			printk(KERN_ERR "MSI: Bogus qconf "
			       "expected[%lx:%x] got[%lx:%lx]\n",
			       base, pbm->msiq_ent_count,
			       ret1, ret2);
			goto h_error;
		}
	}

	return 0;

h_error:
	free_pages(pages, order);
	return -EINVAL;
}

static void pci_sun4v_msi_init(struct pci_pbm_info *pbm)
{
	const u32 *val;
	int len;

	val = of_get_property(pbm->prom_node, "#msi-eqs", &len);
	if (!val || len != 4)
		goto no_msi;
	pbm->msiq_num = *val;
	if (pbm->msiq_num) {
		const struct msiq_prop {
			u32 first_msiq;
			u32 num_msiq;
			u32 first_devino;
		} *mqp;
		const struct msi_range_prop {
			u32 first_msi;
			u32 num_msi;
		} *mrng;
		const struct addr_range_prop {
			u32 msi32_high;
			u32 msi32_low;
			u32 msi32_len;
			u32 msi64_high;
			u32 msi64_low;
			u32 msi64_len;
		} *arng;

		val = of_get_property(pbm->prom_node, "msi-eq-size", &len);
		if (!val || len != 4)
			goto no_msi;

		pbm->msiq_ent_count = *val;

		mqp = of_get_property(pbm->prom_node,
				      "msi-eq-to-devino", &len);
		if (!mqp || len != sizeof(struct msiq_prop))
			goto no_msi;

		pbm->msiq_first = mqp->first_msiq;
		pbm->msiq_first_devino = mqp->first_devino;

		val = of_get_property(pbm->prom_node, "#msi", &len);
		if (!val || len != 4)
			goto no_msi;
		pbm->msi_num = *val;

		mrng = of_get_property(pbm->prom_node, "msi-ranges", &len);
		if (!mrng || len != sizeof(struct msi_range_prop))
			goto no_msi;
		pbm->msi_first = mrng->first_msi;

		val = of_get_property(pbm->prom_node, "msi-data-mask", &len);
		if (!val || len != 4)
			goto no_msi;
		pbm->msi_data_mask = *val;

		val = of_get_property(pbm->prom_node, "msix-data-width", &len);
		if (!val || len != 4)
			goto no_msi;
		pbm->msix_data_width = *val;

		arng = of_get_property(pbm->prom_node, "msi-address-ranges",
				       &len);
		if (!arng || len != sizeof(struct addr_range_prop))
			goto no_msi;
		pbm->msi32_start = ((u64)arng->msi32_high << 32) |
			(u64) arng->msi32_low;
		pbm->msi64_start = ((u64)arng->msi64_high << 32) |
			(u64) arng->msi64_low;
		pbm->msi32_len = arng->msi32_len;
		pbm->msi64_len = arng->msi64_len;

		if (msi_bitmap_alloc(pbm))
			goto no_msi;

		if (msi_queue_alloc(pbm)) {
			msi_bitmap_free(pbm);
			goto no_msi;
		}

		printk(KERN_INFO "%s: MSI Queue first[%u] num[%u] count[%u] "
		       "devino[0x%x]\n",
		       pbm->name,
		       pbm->msiq_first, pbm->msiq_num,
		       pbm->msiq_ent_count,
		       pbm->msiq_first_devino);
		printk(KERN_INFO "%s: MSI first[%u] num[%u] mask[0x%x] "
		       "width[%u]\n",
		       pbm->name,
		       pbm->msi_first, pbm->msi_num, pbm->msi_data_mask,
		       pbm->msix_data_width);
		printk(KERN_INFO "%s: MSI addr32[0x%lx:0x%x] "
		       "addr64[0x%lx:0x%x]\n",
		       pbm->name,
		       pbm->msi32_start, pbm->msi32_len,
		       pbm->msi64_start, pbm->msi64_len);
		printk(KERN_INFO "%s: MSI queues at RA [%p]\n",
		       pbm->name,
		       pbm->msi_queues);
	}

	return;

no_msi:
	pbm->msiq_num = 0;
	printk(KERN_INFO "%s: No MSI support.\n", pbm->name);
}

static int alloc_msi(struct pci_pbm_info *pbm)
{
	int i;

	for (i = 0; i < pbm->msi_num; i++) {
		if (!test_and_set_bit(i, pbm->msi_bitmap))
			return i + pbm->msi_first;
	}

	return -ENOENT;
}

static void free_msi(struct pci_pbm_info *pbm, int msi_num)
{
	msi_num -= pbm->msi_first;
	clear_bit(msi_num, pbm->msi_bitmap);
}

static int pci_sun4v_setup_msi_irq(unsigned int *virt_irq_p,
				   struct pci_dev *pdev,
				   struct msi_desc *entry)
{
	struct pci_pbm_info *pbm = pdev->dev.archdata.host_controller;
	unsigned long devino, msiqid;
	struct msi_msg msg;
	int msi_num, err;

	*virt_irq_p = 0;

	msi_num = alloc_msi(pbm);
	if (msi_num < 0)
		return msi_num;

	devino = sun4v_build_msi(pbm->devhandle, virt_irq_p,
				 pbm->msiq_first_devino,
				 (pbm->msiq_first_devino +
				  pbm->msiq_num));
	err = -ENOMEM;
	if (!devino)
		goto out_err;

	set_irq_msi(*virt_irq_p, entry);

	msiqid = ((devino - pbm->msiq_first_devino) +
		  pbm->msiq_first);

	err = -EINVAL;
	if (pci_sun4v_msiq_setstate(pbm->devhandle, msiqid, HV_MSIQSTATE_IDLE))
	if (err)
		goto out_err;

	if (pci_sun4v_msiq_setvalid(pbm->devhandle, msiqid, HV_MSIQ_VALID))
		goto out_err;

	if (pci_sun4v_msi_setmsiq(pbm->devhandle,
				  msi_num, msiqid,
				  (entry->msi_attrib.is_64 ?
				   HV_MSITYPE_MSI64 : HV_MSITYPE_MSI32)))
		goto out_err;

	if (pci_sun4v_msi_setstate(pbm->devhandle, msi_num, HV_MSISTATE_IDLE))
		goto out_err;

	if (pci_sun4v_msi_setvalid(pbm->devhandle, msi_num, HV_MSIVALID_VALID))
		goto out_err;

	pdev->dev.archdata.msi_num = msi_num;

	if (entry->msi_attrib.is_64) {
		msg.address_hi = pbm->msi64_start >> 32;
		msg.address_lo = pbm->msi64_start & 0xffffffff;
	} else {
		msg.address_hi = 0;
		msg.address_lo = pbm->msi32_start;
	}
	msg.data = msi_num;
	write_msi_msg(*virt_irq_p, &msg);

	irq_install_pre_handler(*virt_irq_p,
				pci_sun4v_msi_prehandler,
				pbm, (void *) msiqid);

	return 0;

out_err:
	free_msi(pbm, msi_num);
	sun4v_destroy_msi(*virt_irq_p);
	*virt_irq_p = 0;
	return err;

}

static void pci_sun4v_teardown_msi_irq(unsigned int virt_irq,
				       struct pci_dev *pdev)
{
	struct pci_pbm_info *pbm = pdev->dev.archdata.host_controller;
	unsigned long msiqid, err;
	unsigned int msi_num;

	msi_num = pdev->dev.archdata.msi_num;
	err = pci_sun4v_msi_getmsiq(pbm->devhandle, msi_num, &msiqid);
	if (err) {
		printk(KERN_ERR "%s: getmsiq gives error %lu\n",
		       pbm->name, err);
		return;
	}

	pci_sun4v_msi_setvalid(pbm->devhandle, msi_num, HV_MSIVALID_INVALID);
	pci_sun4v_msiq_setvalid(pbm->devhandle, msiqid, HV_MSIQ_INVALID);

	free_msi(pbm, msi_num);

	/* The sun4v_destroy_msi() will liberate the devino and thus the MSIQ
	 * allocation.
	 */
	sun4v_destroy_msi(virt_irq);
}
#else /* CONFIG_PCI_MSI */
static void pci_sun4v_msi_init(struct pci_pbm_info *pbm)
{
}
#endif /* !(CONFIG_PCI_MSI) */

static void pci_sun4v_pbm_init(struct pci_controller_info *p, struct device_node *dp, u32 devhandle)
{
	struct pci_pbm_info *pbm;
	struct property *prop;
	int len, i;

	if (devhandle & 0x40)
		pbm = &p->pbm_B;
	else
		pbm = &p->pbm_A;

	pbm->parent = p;
	pbm->prom_node = dp;
	pbm->pci_first_slot = 1;

	pbm->devhandle = devhandle;

	pbm->name = dp->full_name;

	printk("%s: SUN4V PCI Bus Module\n", pbm->name);

	prop = of_find_property(dp, "ranges", &len);
	pbm->pbm_ranges = prop->value;
	pbm->num_pbm_ranges =
		(len / sizeof(struct linux_prom_pci_ranges));

	/* Mask out the top 8 bits of the ranges, leaving the real
	 * physical address.
	 */
	for (i = 0; i < pbm->num_pbm_ranges; i++)
		pbm->pbm_ranges[i].parent_phys_hi &= 0x0fffffff;

	pci_determine_mem_io_space(pbm);

	pci_sun4v_get_bus_range(pbm);
	pci_sun4v_iommu_init(pbm);
	pci_sun4v_msi_init(pbm);
}

void sun4v_pci_init(struct device_node *dp, char *model_name)
{
	struct pci_controller_info *p;
	struct pci_iommu *iommu;
	struct property *prop;
	struct linux_prom64_registers *regs;
	u32 devhandle;
	int i;

	prop = of_find_property(dp, "reg", NULL);
	regs = prop->value;

	devhandle = (regs->phys_addr >> 32UL) & 0x0fffffff;

	for (p = pci_controller_root; p; p = p->next) {
		struct pci_pbm_info *pbm;

		if (p->pbm_A.prom_node && p->pbm_B.prom_node)
			continue;

		pbm = (p->pbm_A.prom_node ?
		       &p->pbm_A :
		       &p->pbm_B);

		if (pbm->devhandle == (devhandle ^ 0x40)) {
			pci_sun4v_pbm_init(p, dp, devhandle);
			return;
		}
	}

	for_each_possible_cpu(i) {
		unsigned long page = get_zeroed_page(GFP_ATOMIC);

		if (!page)
			goto fatal_memory_error;

		per_cpu(pci_iommu_batch, i).pglist = (u64 *) page;
	}

	p = kzalloc(sizeof(struct pci_controller_info), GFP_ATOMIC);
	if (!p)
		goto fatal_memory_error;

	iommu = kzalloc(sizeof(struct pci_iommu), GFP_ATOMIC);
	if (!iommu)
		goto fatal_memory_error;

	p->pbm_A.iommu = iommu;

	iommu = kzalloc(sizeof(struct pci_iommu), GFP_ATOMIC);
	if (!iommu)
		goto fatal_memory_error;

	p->pbm_B.iommu = iommu;

	p->next = pci_controller_root;
	pci_controller_root = p;

	p->index = pci_num_controllers++;
	p->pbms_same_domain = 0;

	p->scan_bus = pci_sun4v_scan_bus;
	p->base_address_update = pci_sun4v_base_address_update;
	p->resource_adjust = pci_sun4v_resource_adjust;
#ifdef CONFIG_PCI_MSI
	p->setup_msi_irq = pci_sun4v_setup_msi_irq;
	p->teardown_msi_irq = pci_sun4v_teardown_msi_irq;
#endif
	p->pci_ops = &pci_sun4v_ops;

	/* Like PSYCHO and SCHIZO we have a 2GB aligned area
	 * for memory space.
	 */
	pci_memspace_mask = 0x7fffffffUL;

	pci_sun4v_pbm_init(p, dp, devhandle);
	return;

fatal_memory_error:
	prom_printf("SUN4V_PCI: Fatal memory allocation error.\n");
	prom_halt();
}