summaryrefslogtreecommitdiff
path: root/arch/x86/kernel/amd_iommu.c
blob: 331b318304eb180d5e4f022253360d2c4efc49e6 (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
/*
 * Copyright (C) 2007-2008 Advanced Micro Devices, Inc.
 * Author: Joerg Roedel <joerg.roedel@amd.com>
 *         Leo Duran <leo.duran@amd.com>
 *
 * 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.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
 */

#include <linux/pci.h>
#include <linux/gfp.h>
#include <linux/bitops.h>
#include <linux/scatterlist.h>
#include <linux/iommu-helper.h>
#include <asm/proto.h>
#include <asm/iommu.h>
#include <asm/amd_iommu_types.h>
#include <asm/amd_iommu.h>

#define CMD_SET_TYPE(cmd, t) ((cmd)->data[1] |= ((t) << 28))

#define EXIT_LOOP_COUNT 10000000

static DEFINE_RWLOCK(amd_iommu_devtable_lock);

/* A list of preallocated protection domains */
static LIST_HEAD(iommu_pd_list);
static DEFINE_SPINLOCK(iommu_pd_list_lock);

/*
 * general struct to manage commands send to an IOMMU
 */
struct iommu_cmd {
	u32 data[4];
};

static int dma_ops_unity_map(struct dma_ops_domain *dma_dom,
			     struct unity_map_entry *e);

/* returns !0 if the IOMMU is caching non-present entries in its TLB */
static int iommu_has_npcache(struct amd_iommu *iommu)
{
	return iommu->cap & (1UL << IOMMU_CAP_NPCACHE);
}

/****************************************************************************
 *
 * Interrupt handling functions
 *
 ****************************************************************************/

static void iommu_print_event(void *__evt)
{
	u32 *event = __evt;
	int type  = (event[1] >> EVENT_TYPE_SHIFT)  & EVENT_TYPE_MASK;
	int devid = (event[0] >> EVENT_DEVID_SHIFT) & EVENT_DEVID_MASK;
	int domid = (event[1] >> EVENT_DOMID_SHIFT) & EVENT_DOMID_MASK;
	int flags = (event[1] >> EVENT_FLAGS_SHIFT) & EVENT_FLAGS_MASK;
	u64 address = (u64)(((u64)event[3]) << 32) | event[2];

	printk(KERN_ERR "AMD IOMMU: Event logged [");

	switch (type) {
	case EVENT_TYPE_ILL_DEV:
		printk("ILLEGAL_DEV_TABLE_ENTRY device=%02x:%02x.%x "
		       "address=0x%016llx flags=0x%04x]\n",
		       PCI_BUS(devid), PCI_SLOT(devid), PCI_FUNC(devid),
		       address, flags);
		break;
	case EVENT_TYPE_IO_FAULT:
		printk("IO_PAGE_FAULT device=%02x:%02x.%x "
		       "domain=0x%04x address=0x%016llx flags=0x%04x]\n",
		       PCI_BUS(devid), PCI_SLOT(devid), PCI_FUNC(devid),
		       domid, address, flags);
		break;
	case EVENT_TYPE_DEV_TAB_ERR:
		printk("DEV_TAB_HARDWARE_ERROR device=%02x:%02x.%x "
		       "address=0x%016llx flags=0x%04x]\n",
		       PCI_BUS(devid), PCI_SLOT(devid), PCI_FUNC(devid),
		       address, flags);
		break;
	case EVENT_TYPE_PAGE_TAB_ERR:
		printk("PAGE_TAB_HARDWARE_ERROR device=%02x:%02x.%x "
		       "domain=0x%04x address=0x%016llx flags=0x%04x]\n",
		       PCI_BUS(devid), PCI_SLOT(devid), PCI_FUNC(devid),
		       domid, address, flags);
		break;
	case EVENT_TYPE_ILL_CMD:
		printk("ILLEGAL_COMMAND_ERROR address=0x%016llx]\n", address);
		break;
	case EVENT_TYPE_CMD_HARD_ERR:
		printk("COMMAND_HARDWARE_ERROR address=0x%016llx "
		       "flags=0x%04x]\n", address, flags);
		break;
	case EVENT_TYPE_IOTLB_INV_TO:
		printk("IOTLB_INV_TIMEOUT device=%02x:%02x.%x "
		       "address=0x%016llx]\n",
		       PCI_BUS(devid), PCI_SLOT(devid), PCI_FUNC(devid),
		       address);
		break;
	case EVENT_TYPE_INV_DEV_REQ:
		printk("INVALID_DEVICE_REQUEST device=%02x:%02x.%x "
		       "address=0x%016llx flags=0x%04x]\n",
		       PCI_BUS(devid), PCI_SLOT(devid), PCI_FUNC(devid),
		       address, flags);
		break;
	default:
		printk(KERN_ERR "UNKNOWN type=0x%02x]\n", type);
	}
}

static void iommu_poll_events(struct amd_iommu *iommu)
{
	u32 head, tail;
	unsigned long flags;

	spin_lock_irqsave(&iommu->lock, flags);

	head = readl(iommu->mmio_base + MMIO_EVT_HEAD_OFFSET);
	tail = readl(iommu->mmio_base + MMIO_EVT_TAIL_OFFSET);

	while (head != tail) {
		iommu_print_event(iommu->evt_buf + head);
		head = (head + EVENT_ENTRY_SIZE) % iommu->evt_buf_size;
	}

	writel(head, iommu->mmio_base + MMIO_EVT_HEAD_OFFSET);

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

irqreturn_t amd_iommu_int_handler(int irq, void *data)
{
	struct amd_iommu *iommu;

	list_for_each_entry(iommu, &amd_iommu_list, list)
		iommu_poll_events(iommu);

	return IRQ_HANDLED;
}

/****************************************************************************
 *
 * IOMMU command queuing functions
 *
 ****************************************************************************/

/*
 * Writes the command to the IOMMUs command buffer and informs the
 * hardware about the new command. Must be called with iommu->lock held.
 */
static int __iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
{
	u32 tail, head;
	u8 *target;

	tail = readl(iommu->mmio_base + MMIO_CMD_TAIL_OFFSET);
	target = iommu->cmd_buf + tail;
	memcpy_toio(target, cmd, sizeof(*cmd));
	tail = (tail + sizeof(*cmd)) % iommu->cmd_buf_size;
	head = readl(iommu->mmio_base + MMIO_CMD_HEAD_OFFSET);
	if (tail == head)
		return -ENOMEM;
	writel(tail, iommu->mmio_base + MMIO_CMD_TAIL_OFFSET);

	return 0;
}

/*
 * General queuing function for commands. Takes iommu->lock and calls
 * __iommu_queue_command().
 */
static int iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd)
{
	unsigned long flags;
	int ret;

	spin_lock_irqsave(&iommu->lock, flags);
	ret = __iommu_queue_command(iommu, cmd);
	spin_unlock_irqrestore(&iommu->lock, flags);

	return ret;
}

/*
 * This function is called whenever we need to ensure that the IOMMU has
 * completed execution of all commands we sent. It sends a
 * COMPLETION_WAIT command and waits for it to finish. The IOMMU informs
 * us about that by writing a value to a physical address we pass with
 * the command.
 */
static int iommu_completion_wait(struct amd_iommu *iommu)
{
	int ret = 0, ready = 0;
	unsigned status = 0;
	struct iommu_cmd cmd;
	unsigned long flags, i = 0;

	memset(&cmd, 0, sizeof(cmd));
	cmd.data[0] = CMD_COMPL_WAIT_INT_MASK;
	CMD_SET_TYPE(&cmd, CMD_COMPL_WAIT);

	iommu->need_sync = 0;

	spin_lock_irqsave(&iommu->lock, flags);

	ret = __iommu_queue_command(iommu, &cmd);

	if (ret)
		goto out;

	while (!ready && (i < EXIT_LOOP_COUNT)) {
		++i;
		/* wait for the bit to become one */
		status = readl(iommu->mmio_base + MMIO_STATUS_OFFSET);
		ready = status & MMIO_STATUS_COM_WAIT_INT_MASK;
	}

	/* set bit back to zero */
	status &= ~MMIO_STATUS_COM_WAIT_INT_MASK;
	writel(status, iommu->mmio_base + MMIO_STATUS_OFFSET);

	if (unlikely((i == EXIT_LOOP_COUNT) && printk_ratelimit()))
		printk(KERN_WARNING "AMD IOMMU: Completion wait loop failed\n");
out:
	spin_unlock_irqrestore(&iommu->lock, flags);

	return 0;
}

/*
 * Command send function for invalidating a device table entry
 */
static int iommu_queue_inv_dev_entry(struct amd_iommu *iommu, u16 devid)
{
	struct iommu_cmd cmd;
	int ret;

	BUG_ON(iommu == NULL);

	memset(&cmd, 0, sizeof(cmd));
	CMD_SET_TYPE(&cmd, CMD_INV_DEV_ENTRY);
	cmd.data[0] = devid;

	ret = iommu_queue_command(iommu, &cmd);

	iommu->need_sync = 1;

	return ret;
}

/*
 * Generic command send function for invalidaing TLB entries
 */
static int iommu_queue_inv_iommu_pages(struct amd_iommu *iommu,
		u64 address, u16 domid, int pde, int s)
{
	struct iommu_cmd cmd;
	int ret;

	memset(&cmd, 0, sizeof(cmd));
	address &= PAGE_MASK;
	CMD_SET_TYPE(&cmd, CMD_INV_IOMMU_PAGES);
	cmd.data[1] |= domid;
	cmd.data[2] = lower_32_bits(address);
	cmd.data[3] = upper_32_bits(address);
	if (s) /* size bit - we flush more than one 4kb page */
		cmd.data[2] |= CMD_INV_IOMMU_PAGES_SIZE_MASK;
	if (pde) /* PDE bit - we wan't flush everything not only the PTEs */
		cmd.data[2] |= CMD_INV_IOMMU_PAGES_PDE_MASK;

	ret = iommu_queue_command(iommu, &cmd);

	iommu->need_sync = 1;

	return ret;
}

/*
 * TLB invalidation function which is called from the mapping functions.
 * It invalidates a single PTE if the range to flush is within a single
 * page. Otherwise it flushes the whole TLB of the IOMMU.
 */
static int iommu_flush_pages(struct amd_iommu *iommu, u16 domid,
		u64 address, size_t size)
{
	int s = 0;
	unsigned pages = iommu_num_pages(address, size, PAGE_SIZE);

	address &= PAGE_MASK;

	if (pages > 1) {
		/*
		 * If we have to flush more than one page, flush all
		 * TLB entries for this domain
		 */
		address = CMD_INV_IOMMU_ALL_PAGES_ADDRESS;
		s = 1;
	}

	iommu_queue_inv_iommu_pages(iommu, address, domid, 0, s);

	return 0;
}

/* Flush the whole IO/TLB for a given protection domain */
static void iommu_flush_tlb(struct amd_iommu *iommu, u16 domid)
{
	u64 address = CMD_INV_IOMMU_ALL_PAGES_ADDRESS;

	iommu_queue_inv_iommu_pages(iommu, address, domid, 0, 1);
}

/****************************************************************************
 *
 * The functions below are used the create the page table mappings for
 * unity mapped regions.
 *
 ****************************************************************************/

/*
 * Generic mapping functions. It maps a physical address into a DMA
 * address space. It allocates the page table pages if necessary.
 * In the future it can be extended to a generic mapping function
 * supporting all features of AMD IOMMU page tables like level skipping
 * and full 64 bit address spaces.
 */
static int iommu_map(struct protection_domain *dom,
		     unsigned long bus_addr,
		     unsigned long phys_addr,
		     int prot)
{
	u64 __pte, *pte, *page;

	bus_addr  = PAGE_ALIGN(bus_addr);
	phys_addr = PAGE_ALIGN(bus_addr);

	/* only support 512GB address spaces for now */
	if (bus_addr > IOMMU_MAP_SIZE_L3 || !(prot & IOMMU_PROT_MASK))
		return -EINVAL;

	pte = &dom->pt_root[IOMMU_PTE_L2_INDEX(bus_addr)];

	if (!IOMMU_PTE_PRESENT(*pte)) {
		page = (u64 *)get_zeroed_page(GFP_KERNEL);
		if (!page)
			return -ENOMEM;
		*pte = IOMMU_L2_PDE(virt_to_phys(page));
	}

	pte = IOMMU_PTE_PAGE(*pte);
	pte = &pte[IOMMU_PTE_L1_INDEX(bus_addr)];

	if (!IOMMU_PTE_PRESENT(*pte)) {
		page = (u64 *)get_zeroed_page(GFP_KERNEL);
		if (!page)
			return -ENOMEM;
		*pte = IOMMU_L1_PDE(virt_to_phys(page));
	}

	pte = IOMMU_PTE_PAGE(*pte);
	pte = &pte[IOMMU_PTE_L0_INDEX(bus_addr)];

	if (IOMMU_PTE_PRESENT(*pte))
		return -EBUSY;

	__pte = phys_addr | IOMMU_PTE_P;
	if (prot & IOMMU_PROT_IR)
		__pte |= IOMMU_PTE_IR;
	if (prot & IOMMU_PROT_IW)
		__pte |= IOMMU_PTE_IW;

	*pte = __pte;

	return 0;
}

/*
 * This function checks if a specific unity mapping entry is needed for
 * this specific IOMMU.
 */
static int iommu_for_unity_map(struct amd_iommu *iommu,
			       struct unity_map_entry *entry)
{
	u16 bdf, i;

	for (i = entry->devid_start; i <= entry->devid_end; ++i) {
		bdf = amd_iommu_alias_table[i];
		if (amd_iommu_rlookup_table[bdf] == iommu)
			return 1;
	}

	return 0;
}

/*
 * Init the unity mappings for a specific IOMMU in the system
 *
 * Basically iterates over all unity mapping entries and applies them to
 * the default domain DMA of that IOMMU if necessary.
 */
static int iommu_init_unity_mappings(struct amd_iommu *iommu)
{
	struct unity_map_entry *entry;
	int ret;

	list_for_each_entry(entry, &amd_iommu_unity_map, list) {
		if (!iommu_for_unity_map(iommu, entry))
			continue;
		ret = dma_ops_unity_map(iommu->default_dom, entry);
		if (ret)
			return ret;
	}

	return 0;
}

/*
 * This function actually applies the mapping to the page table of the
 * dma_ops domain.
 */
static int dma_ops_unity_map(struct dma_ops_domain *dma_dom,
			     struct unity_map_entry *e)
{
	u64 addr;
	int ret;

	for (addr = e->address_start; addr < e->address_end;
	     addr += PAGE_SIZE) {
		ret = iommu_map(&dma_dom->domain, addr, addr, e->prot);
		if (ret)
			return ret;
		/*
		 * if unity mapping is in aperture range mark the page
		 * as allocated in the aperture
		 */
		if (addr < dma_dom->aperture_size)
			__set_bit(addr >> PAGE_SHIFT, dma_dom->bitmap);
	}

	return 0;
}

/*
 * Inits the unity mappings required for a specific device
 */
static int init_unity_mappings_for_device(struct dma_ops_domain *dma_dom,
					  u16 devid)
{
	struct unity_map_entry *e;
	int ret;

	list_for_each_entry(e, &amd_iommu_unity_map, list) {
		if (!(devid >= e->devid_start && devid <= e->devid_end))
			continue;
		ret = dma_ops_unity_map(dma_dom, e);
		if (ret)
			return ret;
	}

	return 0;
}

/****************************************************************************
 *
 * The next functions belong to the address allocator for the dma_ops
 * interface functions. They work like the allocators in the other IOMMU
 * drivers. Its basically a bitmap which marks the allocated pages in
 * the aperture. Maybe it could be enhanced in the future to a more
 * efficient allocator.
 *
 ****************************************************************************/

/*
 * The address allocator core function.
 *
 * called with domain->lock held
 */
static unsigned long dma_ops_alloc_addresses(struct device *dev,
					     struct dma_ops_domain *dom,
					     unsigned int pages,
					     unsigned long align_mask,
					     u64 dma_mask)
{
	unsigned long limit;
	unsigned long address;
	unsigned long boundary_size;

	boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1,
			PAGE_SIZE) >> PAGE_SHIFT;
	limit = iommu_device_max_index(dom->aperture_size >> PAGE_SHIFT, 0,
				       dma_mask >> PAGE_SHIFT);

	if (dom->next_bit >= limit) {
		dom->next_bit = 0;
		dom->need_flush = true;
	}

	address = iommu_area_alloc(dom->bitmap, limit, dom->next_bit, pages,
				   0 , boundary_size, align_mask);
	if (address == -1) {
		address = iommu_area_alloc(dom->bitmap, limit, 0, pages,
				0, boundary_size, align_mask);
		dom->need_flush = true;
	}

	if (likely(address != -1)) {
		dom->next_bit = address + pages;
		address <<= PAGE_SHIFT;
	} else
		address = bad_dma_address;

	WARN_ON((address + (PAGE_SIZE*pages)) > dom->aperture_size);

	return address;
}

/*
 * The address free function.
 *
 * called with domain->lock held
 */
static void dma_ops_free_addresses(struct dma_ops_domain *dom,
				   unsigned long address,
				   unsigned int pages)
{
	address >>= PAGE_SHIFT;
	iommu_area_free(dom->bitmap, address, pages);

	if (address + pages >= dom->next_bit)
		dom->need_flush = true;
}

/****************************************************************************
 *
 * The next functions belong to the domain allocation. A domain is
 * allocated for every IOMMU as the default domain. If device isolation
 * is enabled, every device get its own domain. The most important thing
 * about domains is the page table mapping the DMA address space they
 * contain.
 *
 ****************************************************************************/

static u16 domain_id_alloc(void)
{
	unsigned long flags;
	int id;

	write_lock_irqsave(&amd_iommu_devtable_lock, flags);
	id = find_first_zero_bit(amd_iommu_pd_alloc_bitmap, MAX_DOMAIN_ID);
	BUG_ON(id == 0);
	if (id > 0 && id < MAX_DOMAIN_ID)
		__set_bit(id, amd_iommu_pd_alloc_bitmap);
	else
		id = 0;
	write_unlock_irqrestore(&amd_iommu_devtable_lock, flags);

	return id;
}

/*
 * Used to reserve address ranges in the aperture (e.g. for exclusion
 * ranges.
 */
static void dma_ops_reserve_addresses(struct dma_ops_domain *dom,
				      unsigned long start_page,
				      unsigned int pages)
{
	unsigned int last_page = dom->aperture_size >> PAGE_SHIFT;

	if (start_page + pages > last_page)
		pages = last_page - start_page;

	iommu_area_reserve(dom->bitmap, start_page, pages);
}

static void dma_ops_free_pagetable(struct dma_ops_domain *dma_dom)
{
	int i, j;
	u64 *p1, *p2, *p3;

	p1 = dma_dom->domain.pt_root;

	if (!p1)
		return;

	for (i = 0; i < 512; ++i) {
		if (!IOMMU_PTE_PRESENT(p1[i]))
			continue;

		p2 = IOMMU_PTE_PAGE(p1[i]);
		for (j = 0; j < 512; ++i) {
			if (!IOMMU_PTE_PRESENT(p2[j]))
				continue;
			p3 = IOMMU_PTE_PAGE(p2[j]);
			free_page((unsigned long)p3);
		}

		free_page((unsigned long)p2);
	}

	free_page((unsigned long)p1);
}

/*
 * Free a domain, only used if something went wrong in the
 * allocation path and we need to free an already allocated page table
 */
static void dma_ops_domain_free(struct dma_ops_domain *dom)
{
	if (!dom)
		return;

	dma_ops_free_pagetable(dom);

	kfree(dom->pte_pages);

	kfree(dom->bitmap);

	kfree(dom);
}

/*
 * Allocates a new protection domain usable for the dma_ops functions.
 * It also intializes the page table and the address allocator data
 * structures required for the dma_ops interface
 */
static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu,
						   unsigned order)
{
	struct dma_ops_domain *dma_dom;
	unsigned i, num_pte_pages;
	u64 *l2_pde;
	u64 address;

	/*
	 * Currently the DMA aperture must be between 32 MB and 1GB in size
	 */
	if ((order < 25) || (order > 30))
		return NULL;

	dma_dom = kzalloc(sizeof(struct dma_ops_domain), GFP_KERNEL);
	if (!dma_dom)
		return NULL;

	spin_lock_init(&dma_dom->domain.lock);

	dma_dom->domain.id = domain_id_alloc();
	if (dma_dom->domain.id == 0)
		goto free_dma_dom;
	dma_dom->domain.mode = PAGE_MODE_3_LEVEL;
	dma_dom->domain.pt_root = (void *)get_zeroed_page(GFP_KERNEL);
	dma_dom->domain.priv = dma_dom;
	if (!dma_dom->domain.pt_root)
		goto free_dma_dom;
	dma_dom->aperture_size = (1ULL << order);
	dma_dom->bitmap = kzalloc(dma_dom->aperture_size / (PAGE_SIZE * 8),
				  GFP_KERNEL);
	if (!dma_dom->bitmap)
		goto free_dma_dom;
	/*
	 * mark the first page as allocated so we never return 0 as
	 * a valid dma-address. So we can use 0 as error value
	 */
	dma_dom->bitmap[0] = 1;
	dma_dom->next_bit = 0;

	dma_dom->need_flush = false;
	dma_dom->target_dev = 0xffff;

	/* Intialize the exclusion range if necessary */
	if (iommu->exclusion_start &&
	    iommu->exclusion_start < dma_dom->aperture_size) {
		unsigned long startpage = iommu->exclusion_start >> PAGE_SHIFT;
		int pages = iommu_num_pages(iommu->exclusion_start,
					    iommu->exclusion_length,
					    PAGE_SIZE);
		dma_ops_reserve_addresses(dma_dom, startpage, pages);
	}

	/*
	 * At the last step, build the page tables so we don't need to
	 * allocate page table pages in the dma_ops mapping/unmapping
	 * path.
	 */
	num_pte_pages = dma_dom->aperture_size / (PAGE_SIZE * 512);
	dma_dom->pte_pages = kzalloc(num_pte_pages * sizeof(void *),
			GFP_KERNEL);
	if (!dma_dom->pte_pages)
		goto free_dma_dom;

	l2_pde = (u64 *)get_zeroed_page(GFP_KERNEL);
	if (l2_pde == NULL)
		goto free_dma_dom;

	dma_dom->domain.pt_root[0] = IOMMU_L2_PDE(virt_to_phys(l2_pde));

	for (i = 0; i < num_pte_pages; ++i) {
		dma_dom->pte_pages[i] = (u64 *)get_zeroed_page(GFP_KERNEL);
		if (!dma_dom->pte_pages[i])
			goto free_dma_dom;
		address = virt_to_phys(dma_dom->pte_pages[i]);
		l2_pde[i] = IOMMU_L1_PDE(address);
	}

	return dma_dom;

free_dma_dom:
	dma_ops_domain_free(dma_dom);

	return NULL;
}

/*
 * Find out the protection domain structure for a given PCI device. This
 * will give us the pointer to the page table root for example.
 */
static struct protection_domain *domain_for_device(u16 devid)
{
	struct protection_domain *dom;
	unsigned long flags;

	read_lock_irqsave(&amd_iommu_devtable_lock, flags);
	dom = amd_iommu_pd_table[devid];
	read_unlock_irqrestore(&amd_iommu_devtable_lock, flags);

	return dom;
}

/*
 * If a device is not yet associated with a domain, this function does
 * assigns it visible for the hardware
 */
static void set_device_domain(struct amd_iommu *iommu,
			      struct protection_domain *domain,
			      u16 devid)
{
	unsigned long flags;

	u64 pte_root = virt_to_phys(domain->pt_root);

	pte_root |= (domain->mode & DEV_ENTRY_MODE_MASK)
		    << DEV_ENTRY_MODE_SHIFT;
	pte_root |= IOMMU_PTE_IR | IOMMU_PTE_IW | IOMMU_PTE_P | IOMMU_PTE_TV;

	write_lock_irqsave(&amd_iommu_devtable_lock, flags);
	amd_iommu_dev_table[devid].data[0] = lower_32_bits(pte_root);
	amd_iommu_dev_table[devid].data[1] = upper_32_bits(pte_root);
	amd_iommu_dev_table[devid].data[2] = domain->id;

	amd_iommu_pd_table[devid] = domain;
	write_unlock_irqrestore(&amd_iommu_devtable_lock, flags);

	iommu_queue_inv_dev_entry(iommu, devid);

	iommu->need_sync = 1;
}

/*****************************************************************************
 *
 * The next functions belong to the dma_ops mapping/unmapping code.
 *
 *****************************************************************************/

/*
 * This function checks if the driver got a valid device from the caller to
 * avoid dereferencing invalid pointers.
 */
static bool check_device(struct device *dev)
{
	if (!dev || !dev->dma_mask)
		return false;

	return true;
}

/*
 * In this function the list of preallocated protection domains is traversed to
 * find the domain for a specific device
 */
static struct dma_ops_domain *find_protection_domain(u16 devid)
{
	struct dma_ops_domain *entry, *ret = NULL;
	unsigned long flags;

	if (list_empty(&iommu_pd_list))
		return NULL;

	spin_lock_irqsave(&iommu_pd_list_lock, flags);

	list_for_each_entry(entry, &iommu_pd_list, list) {
		if (entry->target_dev == devid) {
			ret = entry;
			list_del(&ret->list);
			break;
		}
	}

	spin_unlock_irqrestore(&iommu_pd_list_lock, flags);

	return ret;
}

/*
 * In the dma_ops path we only have the struct device. This function
 * finds the corresponding IOMMU, the protection domain and the
 * requestor id for a given device.
 * If the device is not yet associated with a domain this is also done
 * in this function.
 */
static int get_device_resources(struct device *dev,
				struct amd_iommu **iommu,
				struct protection_domain **domain,
				u16 *bdf)
{
	struct dma_ops_domain *dma_dom;
	struct pci_dev *pcidev;
	u16 _bdf;

	*iommu = NULL;
	*domain = NULL;
	*bdf = 0xffff;

	if (dev->bus != &pci_bus_type)
		return 0;

	pcidev = to_pci_dev(dev);
	_bdf = calc_devid(pcidev->bus->number, pcidev->devfn);

	/* device not translated by any IOMMU in the system? */
	if (_bdf > amd_iommu_last_bdf)
		return 0;

	*bdf = amd_iommu_alias_table[_bdf];

	*iommu = amd_iommu_rlookup_table[*bdf];
	if (*iommu == NULL)
		return 0;
	*domain = domain_for_device(*bdf);
	if (*domain == NULL) {
		dma_dom = find_protection_domain(*bdf);
		if (!dma_dom)
			dma_dom = (*iommu)->default_dom;
		*domain = &dma_dom->domain;
		set_device_domain(*iommu, *domain, *bdf);
		printk(KERN_INFO "AMD IOMMU: Using protection domain %d for "
				"device ", (*domain)->id);
		print_devid(_bdf, 1);
	}

	return 1;
}

/*
 * This is the generic map function. It maps one 4kb page at paddr to
 * the given address in the DMA address space for the domain.
 */
static dma_addr_t dma_ops_domain_map(struct amd_iommu *iommu,
				     struct dma_ops_domain *dom,
				     unsigned long address,
				     phys_addr_t paddr,
				     int direction)
{
	u64 *pte, __pte;

	WARN_ON(address > dom->aperture_size);

	paddr &= PAGE_MASK;

	pte  = dom->pte_pages[IOMMU_PTE_L1_INDEX(address)];
	pte += IOMMU_PTE_L0_INDEX(address);

	__pte = paddr | IOMMU_PTE_P | IOMMU_PTE_FC;

	if (direction == DMA_TO_DEVICE)
		__pte |= IOMMU_PTE_IR;
	else if (direction == DMA_FROM_DEVICE)
		__pte |= IOMMU_PTE_IW;
	else if (direction == DMA_BIDIRECTIONAL)
		__pte |= IOMMU_PTE_IR | IOMMU_PTE_IW;

	WARN_ON(*pte);

	*pte = __pte;

	return (dma_addr_t)address;
}

/*
 * The generic unmapping function for on page in the DMA address space.
 */
static void dma_ops_domain_unmap(struct amd_iommu *iommu,
				 struct dma_ops_domain *dom,
				 unsigned long address)
{
	u64 *pte;

	if (address >= dom->aperture_size)
		return;

	WARN_ON(address & 0xfffULL || address > dom->aperture_size);

	pte  = dom->pte_pages[IOMMU_PTE_L1_INDEX(address)];
	pte += IOMMU_PTE_L0_INDEX(address);

	WARN_ON(!*pte);

	*pte = 0ULL;
}

/*
 * This function contains common code for mapping of a physically
 * contiguous memory region into DMA address space. It is uses by all
 * mapping functions provided by this IOMMU driver.
 * Must be called with the domain lock held.
 */
static dma_addr_t __map_single(struct device *dev,
			       struct amd_iommu *iommu,
			       struct dma_ops_domain *dma_dom,
			       phys_addr_t paddr,
			       size_t size,
			       int dir,
			       bool align,
			       u64 dma_mask)
{
	dma_addr_t offset = paddr & ~PAGE_MASK;
	dma_addr_t address, start;
	unsigned int pages;
	unsigned long align_mask = 0;
	int i;

	pages = iommu_num_pages(paddr, size, PAGE_SIZE);
	paddr &= PAGE_MASK;

	if (align)
		align_mask = (1UL << get_order(size)) - 1;

	address = dma_ops_alloc_addresses(dev, dma_dom, pages, align_mask,
					  dma_mask);
	if (unlikely(address == bad_dma_address))
		goto out;

	start = address;
	for (i = 0; i < pages; ++i) {
		dma_ops_domain_map(iommu, dma_dom, start, paddr, dir);
		paddr += PAGE_SIZE;
		start += PAGE_SIZE;
	}
	address += offset;

	if (unlikely(dma_dom->need_flush && !amd_iommu_unmap_flush)) {
		iommu_flush_tlb(iommu, dma_dom->domain.id);
		dma_dom->need_flush = false;
	} else if (unlikely(iommu_has_npcache(iommu)))
		iommu_flush_pages(iommu, dma_dom->domain.id, address, size);

out:
	return address;
}

/*
 * Does the reverse of the __map_single function. Must be called with
 * the domain lock held too
 */
static void __unmap_single(struct amd_iommu *iommu,
			   struct dma_ops_domain *dma_dom,
			   dma_addr_t dma_addr,
			   size_t size,
			   int dir)
{
	dma_addr_t i, start;
	unsigned int pages;

	if ((dma_addr == 0) || (dma_addr + size > dma_dom->aperture_size))
		return;

	pages = iommu_num_pages(dma_addr, size, PAGE_SIZE);
	dma_addr &= PAGE_MASK;
	start = dma_addr;

	for (i = 0; i < pages; ++i) {
		dma_ops_domain_unmap(iommu, dma_dom, start);
		start += PAGE_SIZE;
	}

	dma_ops_free_addresses(dma_dom, dma_addr, pages);

	if (amd_iommu_unmap_flush || dma_dom->need_flush) {
		iommu_flush_pages(iommu, dma_dom->domain.id, dma_addr, size);
		dma_dom->need_flush = false;
	}
}

/*
 * The exported map_single function for dma_ops.
 */
static dma_addr_t map_single(struct device *dev, phys_addr_t paddr,
			     size_t size, int dir)
{
	unsigned long flags;
	struct amd_iommu *iommu;
	struct protection_domain *domain;
	u16 devid;
	dma_addr_t addr;
	u64 dma_mask;

	if (!check_device(dev))
		return bad_dma_address;

	dma_mask = *dev->dma_mask;

	get_device_resources(dev, &iommu, &domain, &devid);

	if (iommu == NULL || domain == NULL)
		/* device not handled by any AMD IOMMU */
		return (dma_addr_t)paddr;

	spin_lock_irqsave(&domain->lock, flags);
	addr = __map_single(dev, iommu, domain->priv, paddr, size, dir, false,
			    dma_mask);
	if (addr == bad_dma_address)
		goto out;

	if (unlikely(iommu->need_sync))
		iommu_completion_wait(iommu);

out:
	spin_unlock_irqrestore(&domain->lock, flags);

	return addr;
}

/*
 * The exported unmap_single function for dma_ops.
 */
static void unmap_single(struct device *dev, dma_addr_t dma_addr,
			 size_t size, int dir)
{
	unsigned long flags;
	struct amd_iommu *iommu;
	struct protection_domain *domain;
	u16 devid;

	if (!check_device(dev) ||
	    !get_device_resources(dev, &iommu, &domain, &devid))
		/* device not handled by any AMD IOMMU */
		return;

	spin_lock_irqsave(&domain->lock, flags);

	__unmap_single(iommu, domain->priv, dma_addr, size, dir);

	if (unlikely(iommu->need_sync))
		iommu_completion_wait(iommu);

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

/*
 * This is a special map_sg function which is used if we should map a
 * device which is not handled by an AMD IOMMU in the system.
 */
static int map_sg_no_iommu(struct device *dev, struct scatterlist *sglist,
			   int nelems, int dir)
{
	struct scatterlist *s;
	int i;

	for_each_sg(sglist, s, nelems, i) {
		s->dma_address = (dma_addr_t)sg_phys(s);
		s->dma_length  = s->length;
	}

	return nelems;
}

/*
 * The exported map_sg function for dma_ops (handles scatter-gather
 * lists).
 */
static int map_sg(struct device *dev, struct scatterlist *sglist,
		  int nelems, int dir)
{
	unsigned long flags;
	struct amd_iommu *iommu;
	struct protection_domain *domain;
	u16 devid;
	int i;
	struct scatterlist *s;
	phys_addr_t paddr;
	int mapped_elems = 0;
	u64 dma_mask;

	if (!check_device(dev))
		return 0;

	dma_mask = *dev->dma_mask;

	get_device_resources(dev, &iommu, &domain, &devid);

	if (!iommu || !domain)
		return map_sg_no_iommu(dev, sglist, nelems, dir);

	spin_lock_irqsave(&domain->lock, flags);

	for_each_sg(sglist, s, nelems, i) {
		paddr = sg_phys(s);

		s->dma_address = __map_single(dev, iommu, domain->priv,
					      paddr, s->length, dir, false,
					      dma_mask);

		if (s->dma_address) {
			s->dma_length = s->length;
			mapped_elems++;
		} else
			goto unmap;
	}

	if (unlikely(iommu->need_sync))
		iommu_completion_wait(iommu);

out:
	spin_unlock_irqrestore(&domain->lock, flags);

	return mapped_elems;
unmap:
	for_each_sg(sglist, s, mapped_elems, i) {
		if (s->dma_address)
			__unmap_single(iommu, domain->priv, s->dma_address,
				       s->dma_length, dir);
		s->dma_address = s->dma_length = 0;
	}

	mapped_elems = 0;

	goto out;
}

/*
 * The exported map_sg function for dma_ops (handles scatter-gather
 * lists).
 */
static void unmap_sg(struct device *dev, struct scatterlist *sglist,
		     int nelems, int dir)
{
	unsigned long flags;
	struct amd_iommu *iommu;
	struct protection_domain *domain;
	struct scatterlist *s;
	u16 devid;
	int i;

	if (!check_device(dev) ||
	    !get_device_resources(dev, &iommu, &domain, &devid))
		return;

	spin_lock_irqsave(&domain->lock, flags);

	for_each_sg(sglist, s, nelems, i) {
		__unmap_single(iommu, domain->priv, s->dma_address,
			       s->dma_length, dir);
		s->dma_address = s->dma_length = 0;
	}

	if (unlikely(iommu->need_sync))
		iommu_completion_wait(iommu);

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

/*
 * The exported alloc_coherent function for dma_ops.
 */
static void *alloc_coherent(struct device *dev, size_t size,
			    dma_addr_t *dma_addr, gfp_t flag)
{
	unsigned long flags;
	void *virt_addr;
	struct amd_iommu *iommu;
	struct protection_domain *domain;
	u16 devid;
	phys_addr_t paddr;
	u64 dma_mask = dev->coherent_dma_mask;

	if (!check_device(dev))
		return NULL;

	if (!get_device_resources(dev, &iommu, &domain, &devid))
		flag &= ~(__GFP_DMA | __GFP_HIGHMEM | __GFP_DMA32);

	flag |= __GFP_ZERO;
	virt_addr = (void *)__get_free_pages(flag, get_order(size));
	if (!virt_addr)
		return 0;

	paddr = virt_to_phys(virt_addr);

	if (!iommu || !domain) {
		*dma_addr = (dma_addr_t)paddr;
		return virt_addr;
	}

	if (!dma_mask)
		dma_mask = *dev->dma_mask;

	spin_lock_irqsave(&domain->lock, flags);

	*dma_addr = __map_single(dev, iommu, domain->priv, paddr,
				 size, DMA_BIDIRECTIONAL, true, dma_mask);

	if (*dma_addr == bad_dma_address) {
		free_pages((unsigned long)virt_addr, get_order(size));
		virt_addr = NULL;
		goto out;
	}

	if (unlikely(iommu->need_sync))
		iommu_completion_wait(iommu);

out:
	spin_unlock_irqrestore(&domain->lock, flags);

	return virt_addr;
}

/*
 * The exported free_coherent function for dma_ops.
 */
static void free_coherent(struct device *dev, size_t size,
			  void *virt_addr, dma_addr_t dma_addr)
{
	unsigned long flags;
	struct amd_iommu *iommu;
	struct protection_domain *domain;
	u16 devid;

	if (!check_device(dev))
		return;

	get_device_resources(dev, &iommu, &domain, &devid);

	if (!iommu || !domain)
		goto free_mem;

	spin_lock_irqsave(&domain->lock, flags);

	__unmap_single(iommu, domain->priv, dma_addr, size, DMA_BIDIRECTIONAL);

	if (unlikely(iommu->need_sync))
		iommu_completion_wait(iommu);

	spin_unlock_irqrestore(&domain->lock, flags);

free_mem:
	free_pages((unsigned long)virt_addr, get_order(size));
}

/*
 * This function is called by the DMA layer to find out if we can handle a
 * particular device. It is part of the dma_ops.
 */
static int amd_iommu_dma_supported(struct device *dev, u64 mask)
{
	u16 bdf;
	struct pci_dev *pcidev;

	/* No device or no PCI device */
	if (!dev || dev->bus != &pci_bus_type)
		return 0;

	pcidev = to_pci_dev(dev);

	bdf = calc_devid(pcidev->bus->number, pcidev->devfn);

	/* Out of our scope? */
	if (bdf > amd_iommu_last_bdf)
		return 0;

	return 1;
}

/*
 * The function for pre-allocating protection domains.
 *
 * If the driver core informs the DMA layer if a driver grabs a device
 * we don't need to preallocate the protection domains anymore.
 * For now we have to.
 */
void prealloc_protection_domains(void)
{
	struct pci_dev *dev = NULL;
	struct dma_ops_domain *dma_dom;
	struct amd_iommu *iommu;
	int order = amd_iommu_aperture_order;
	u16 devid;

	while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
		devid = (dev->bus->number << 8) | dev->devfn;
		if (devid > amd_iommu_last_bdf)
			continue;
		devid = amd_iommu_alias_table[devid];
		if (domain_for_device(devid))
			continue;
		iommu = amd_iommu_rlookup_table[devid];
		if (!iommu)
			continue;
		dma_dom = dma_ops_domain_alloc(iommu, order);
		if (!dma_dom)
			continue;
		init_unity_mappings_for_device(dma_dom, devid);
		dma_dom->target_dev = devid;

		list_add_tail(&dma_dom->list, &iommu_pd_list);
	}
}

static struct dma_mapping_ops amd_iommu_dma_ops = {
	.alloc_coherent = alloc_coherent,
	.free_coherent = free_coherent,
	.map_single = map_single,
	.unmap_single = unmap_single,
	.map_sg = map_sg,
	.unmap_sg = unmap_sg,
	.dma_supported = amd_iommu_dma_supported,
};

/*
 * The function which clues the AMD IOMMU driver into dma_ops.
 */
int __init amd_iommu_init_dma_ops(void)
{
	struct amd_iommu *iommu;
	int order = amd_iommu_aperture_order;
	int ret;

	/*
	 * first allocate a default protection domain for every IOMMU we
	 * found in the system. Devices not assigned to any other
	 * protection domain will be assigned to the default one.
	 */
	list_for_each_entry(iommu, &amd_iommu_list, list) {
		iommu->default_dom = dma_ops_domain_alloc(iommu, order);
		if (iommu->default_dom == NULL)
			return -ENOMEM;
		ret = iommu_init_unity_mappings(iommu);
		if (ret)
			goto free_domains;
	}

	/*
	 * If device isolation is enabled, pre-allocate the protection
	 * domains for each device.
	 */
	if (amd_iommu_isolate)
		prealloc_protection_domains();

	iommu_detected = 1;
	force_iommu = 1;
	bad_dma_address = 0;
#ifdef CONFIG_GART_IOMMU
	gart_iommu_aperture_disabled = 1;
	gart_iommu_aperture = 0;
#endif

	/* Make the driver finally visible to the drivers */
	dma_ops = &amd_iommu_dma_ops;

	return 0;

free_domains:

	list_for_each_entry(iommu, &amd_iommu_list, list) {
		if (iommu->default_dom)
			dma_ops_domain_free(iommu->default_dom);
	}

	return ret;
}