aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
blob: b11f4e8868d7652503713b508da2c8820a99c36b (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
/*
 * Copyright 2011 Advanced Micro Devices, Inc.
 * All Rights Reserved.
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the
 * "Software"), to deal in the Software without restriction, including
 * without limitation the rights to use, copy, modify, merge, publish,
 * distribute, sub license, and/or sell copies of the Software, and to
 * permit persons to whom the Software is furnished to do so, subject to
 * the following conditions:
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
 * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
 * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
 * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
 * USE OR OTHER DEALINGS IN THE SOFTWARE.
 *
 * The above copyright notice and this permission notice (including the
 * next paragraph) shall be included in all copies or substantial portions
 * of the Software.
 *
 */
/*
 * Authors:
 *    Christian König <deathsimple@vodafone.de>
 */

#include <linux/firmware.h>
#include <linux/module.h>
#include <drm/drmP.h>
#include <drm/drm.h>

#include "amdgpu.h"
#include "amdgpu_pm.h"
#include "amdgpu_uvd.h"
#include "cikd.h"
#include "uvd/uvd_4_2_d.h"

/* 1 second timeout */
#define UVD_IDLE_TIMEOUT	msecs_to_jiffies(1000)

/* Firmware versions for VI */
#define FW_1_65_10	((1 << 24) | (65 << 16) | (10 << 8))
#define FW_1_87_11	((1 << 24) | (87 << 16) | (11 << 8))
#define FW_1_87_12	((1 << 24) | (87 << 16) | (12 << 8))
#define FW_1_37_15	((1 << 24) | (37 << 16) | (15 << 8))

/* Polaris10/11 firmware version */
#define FW_1_66_16	((1 << 24) | (66 << 16) | (16 << 8))

/* Firmware Names */
#ifdef CONFIG_DRM_AMDGPU_CIK
#define FIRMWARE_BONAIRE	"radeon/bonaire_uvd.bin"
#define FIRMWARE_KABINI	"radeon/kabini_uvd.bin"
#define FIRMWARE_KAVERI	"radeon/kaveri_uvd.bin"
#define FIRMWARE_HAWAII	"radeon/hawaii_uvd.bin"
#define FIRMWARE_MULLINS	"radeon/mullins_uvd.bin"
#endif
#define FIRMWARE_TONGA		"amdgpu/tonga_uvd.bin"
#define FIRMWARE_CARRIZO	"amdgpu/carrizo_uvd.bin"
#define FIRMWARE_FIJI		"amdgpu/fiji_uvd.bin"
#define FIRMWARE_STONEY		"amdgpu/stoney_uvd.bin"
#define FIRMWARE_POLARIS10	"amdgpu/polaris10_uvd.bin"
#define FIRMWARE_POLARIS11	"amdgpu/polaris11_uvd.bin"

/**
 * amdgpu_uvd_cs_ctx - Command submission parser context
 *
 * Used for emulating virtual memory support on UVD 4.2.
 */
struct amdgpu_uvd_cs_ctx {
	struct amdgpu_cs_parser *parser;
	unsigned reg, count;
	unsigned data0, data1;
	unsigned idx;
	unsigned ib_idx;

	/* does the IB has a msg command */
	bool has_msg_cmd;

	/* minimum buffer sizes */
	unsigned *buf_sizes;
};

#ifdef CONFIG_DRM_AMDGPU_CIK
MODULE_FIRMWARE(FIRMWARE_BONAIRE);
MODULE_FIRMWARE(FIRMWARE_KABINI);
MODULE_FIRMWARE(FIRMWARE_KAVERI);
MODULE_FIRMWARE(FIRMWARE_HAWAII);
MODULE_FIRMWARE(FIRMWARE_MULLINS);
#endif
MODULE_FIRMWARE(FIRMWARE_TONGA);
MODULE_FIRMWARE(FIRMWARE_CARRIZO);
MODULE_FIRMWARE(FIRMWARE_FIJI);
MODULE_FIRMWARE(FIRMWARE_STONEY);
MODULE_FIRMWARE(FIRMWARE_POLARIS10);
MODULE_FIRMWARE(FIRMWARE_POLARIS11);

static void amdgpu_uvd_idle_work_handler(struct work_struct *work);

int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
{
	struct amdgpu_ring *ring;
	struct amd_sched_rq *rq;
	unsigned long bo_size;
	const char *fw_name;
	const struct common_firmware_header *hdr;
	unsigned version_major, version_minor, family_id;
	int i, r;

	INIT_DELAYED_WORK(&adev->uvd.idle_work, amdgpu_uvd_idle_work_handler);

	switch (adev->asic_type) {
#ifdef CONFIG_DRM_AMDGPU_CIK
	case CHIP_BONAIRE:
		fw_name = FIRMWARE_BONAIRE;
		break;
	case CHIP_KABINI:
		fw_name = FIRMWARE_KABINI;
		break;
	case CHIP_KAVERI:
		fw_name = FIRMWARE_KAVERI;
		break;
	case CHIP_HAWAII:
		fw_name = FIRMWARE_HAWAII;
		break;
	case CHIP_MULLINS:
		fw_name = FIRMWARE_MULLINS;
		break;
#endif
	case CHIP_TONGA:
		fw_name = FIRMWARE_TONGA;
		break;
	case CHIP_FIJI:
		fw_name = FIRMWARE_FIJI;
		break;
	case CHIP_CARRIZO:
		fw_name = FIRMWARE_CARRIZO;
		break;
	case CHIP_STONEY:
		fw_name = FIRMWARE_STONEY;
		break;
	case CHIP_POLARIS10:
		fw_name = FIRMWARE_POLARIS10;
		break;
	case CHIP_POLARIS11:
		fw_name = FIRMWARE_POLARIS11;
		break;
	default:
		return -EINVAL;
	}

	r = request_firmware(&adev->uvd.fw, fw_name, adev->dev);
	if (r) {
		dev_err(adev->dev, "amdgpu_uvd: Can't load firmware \"%s\"\n",
			fw_name);
		return r;
	}

	r = amdgpu_ucode_validate(adev->uvd.fw);
	if (r) {
		dev_err(adev->dev, "amdgpu_uvd: Can't validate firmware \"%s\"\n",
			fw_name);
		release_firmware(adev->uvd.fw);
		adev->uvd.fw = NULL;
		return r;
	}

	/* Set the default UVD handles that the firmware can handle */
	adev->uvd.max_handles = AMDGPU_DEFAULT_UVD_HANDLES;

	hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
	family_id = le32_to_cpu(hdr->ucode_version) & 0xff;
	version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff;
	version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
	DRM_INFO("Found UVD firmware Version: %hu.%hu Family ID: %hu\n",
		version_major, version_minor, family_id);

	/*
	 * Limit the number of UVD handles depending on microcode major
	 * and minor versions. The firmware version which has 40 UVD
	 * instances support is 1.80. So all subsequent versions should
	 * also have the same support.
	 */
	if ((version_major > 0x01) ||
	    ((version_major == 0x01) && (version_minor >= 0x50)))
		adev->uvd.max_handles = AMDGPU_MAX_UVD_HANDLES;

	adev->uvd.fw_version = ((version_major << 24) | (version_minor << 16) |
				(family_id << 8));

	if ((adev->asic_type == CHIP_POLARIS10 ||
	     adev->asic_type == CHIP_POLARIS11) &&
	    (adev->uvd.fw_version < FW_1_66_16))
		DRM_ERROR("POLARIS10/11 UVD firmware version %hu.%hu is too old.\n",
			  version_major, version_minor);

	bo_size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8)
		  +  AMDGPU_UVD_STACK_SIZE + AMDGPU_UVD_HEAP_SIZE
		  +  AMDGPU_UVD_SESSION_SIZE * adev->uvd.max_handles;
	r = amdgpu_bo_create(adev, bo_size, PAGE_SIZE, true,
			     AMDGPU_GEM_DOMAIN_VRAM,
			     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
			     NULL, NULL, &adev->uvd.vcpu_bo);
	if (r) {
		dev_err(adev->dev, "(%d) failed to allocate UVD bo\n", r);
		return r;
	}

	r = amdgpu_bo_reserve(adev->uvd.vcpu_bo, false);
	if (r) {
		amdgpu_bo_unref(&adev->uvd.vcpu_bo);
		dev_err(adev->dev, "(%d) failed to reserve UVD bo\n", r);
		return r;
	}

	r = amdgpu_bo_pin(adev->uvd.vcpu_bo, AMDGPU_GEM_DOMAIN_VRAM,
			  &adev->uvd.gpu_addr);
	if (r) {
		amdgpu_bo_unreserve(adev->uvd.vcpu_bo);
		amdgpu_bo_unref(&adev->uvd.vcpu_bo);
		dev_err(adev->dev, "(%d) UVD bo pin failed\n", r);
		return r;
	}

	r = amdgpu_bo_kmap(adev->uvd.vcpu_bo, &adev->uvd.cpu_addr);
	if (r) {
		dev_err(adev->dev, "(%d) UVD map failed\n", r);
		return r;
	}

	amdgpu_bo_unreserve(adev->uvd.vcpu_bo);

	ring = &adev->uvd.ring;
	rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_NORMAL];
	r = amd_sched_entity_init(&ring->sched, &adev->uvd.entity,
				  rq, amdgpu_sched_jobs);
	if (r != 0) {
		DRM_ERROR("Failed setting up UVD run queue.\n");
		return r;
	}

	for (i = 0; i < adev->uvd.max_handles; ++i) {
		atomic_set(&adev->uvd.handles[i], 0);
		adev->uvd.filp[i] = NULL;
	}

	/* from uvd v5.0 HW addressing capacity increased to 64 bits */
	if (!amdgpu_ip_block_version_cmp(adev, AMD_IP_BLOCK_TYPE_UVD, 5, 0))
		adev->uvd.address_64_bit = true;

	switch (adev->asic_type) {
	case CHIP_TONGA:
		adev->uvd.use_ctx_buf = adev->uvd.fw_version >= FW_1_65_10;
		break;
	case CHIP_CARRIZO:
		adev->uvd.use_ctx_buf = adev->uvd.fw_version >= FW_1_87_11;
		break;
	case CHIP_FIJI:
		adev->uvd.use_ctx_buf = adev->uvd.fw_version >= FW_1_87_12;
		break;
	case CHIP_STONEY:
		adev->uvd.use_ctx_buf = adev->uvd.fw_version >= FW_1_37_15;
		break;
	default:
		adev->uvd.use_ctx_buf = adev->asic_type >= CHIP_POLARIS10;
	}

	return 0;
}

int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
{
	int r;

	kfree(adev->uvd.saved_bo);

	amd_sched_entity_fini(&adev->uvd.ring.sched, &adev->uvd.entity);

	if (adev->uvd.vcpu_bo) {
		r = amdgpu_bo_reserve(adev->uvd.vcpu_bo, false);
		if (!r) {
			amdgpu_bo_kunmap(adev->uvd.vcpu_bo);
			amdgpu_bo_unpin(adev->uvd.vcpu_bo);
			amdgpu_bo_unreserve(adev->uvd.vcpu_bo);
		}

		amdgpu_bo_unref(&adev->uvd.vcpu_bo);
	}

	amdgpu_ring_fini(&adev->uvd.ring);

	release_firmware(adev->uvd.fw);

	return 0;
}

int amdgpu_uvd_suspend(struct amdgpu_device *adev)
{
	unsigned size;
	void *ptr;
	int i;

	if (adev->uvd.vcpu_bo == NULL)
		return 0;

	for (i = 0; i < adev->uvd.max_handles; ++i)
		if (atomic_read(&adev->uvd.handles[i]))
			break;

	if (i == AMDGPU_MAX_UVD_HANDLES)
		return 0;

	cancel_delayed_work_sync(&adev->uvd.idle_work);

	size = amdgpu_bo_size(adev->uvd.vcpu_bo);
	ptr = adev->uvd.cpu_addr;

	adev->uvd.saved_bo = kmalloc(size, GFP_KERNEL);
	if (!adev->uvd.saved_bo)
		return -ENOMEM;

	memcpy(adev->uvd.saved_bo, ptr, size);

	return 0;
}

int amdgpu_uvd_resume(struct amdgpu_device *adev)
{
	unsigned size;
	void *ptr;

	if (adev->uvd.vcpu_bo == NULL)
		return -EINVAL;

	size = amdgpu_bo_size(adev->uvd.vcpu_bo);
	ptr = adev->uvd.cpu_addr;

	if (adev->uvd.saved_bo != NULL) {
		memcpy(ptr, adev->uvd.saved_bo, size);
		kfree(adev->uvd.saved_bo);
		adev->uvd.saved_bo = NULL;
	} else {
		const struct common_firmware_header *hdr;
		unsigned offset;

		hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
		offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
		memcpy(adev->uvd.cpu_addr, (adev->uvd.fw->data) + offset,
			(adev->uvd.fw->size) - offset);
		size -= le32_to_cpu(hdr->ucode_size_bytes);
		ptr += le32_to_cpu(hdr->ucode_size_bytes);
		memset(ptr, 0, size);
	}

	return 0;
}

void amdgpu_uvd_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
{
	struct amdgpu_ring *ring = &adev->uvd.ring;
	int i, r;

	for (i = 0; i < adev->uvd.max_handles; ++i) {
		uint32_t handle = atomic_read(&adev->uvd.handles[i]);
		if (handle != 0 && adev->uvd.filp[i] == filp) {
			struct fence *fence;

			r = amdgpu_uvd_get_destroy_msg(ring, handle,
						       false, &fence);
			if (r) {
				DRM_ERROR("Error destroying UVD (%d)!\n", r);
				continue;
			}

			fence_wait(fence, false);
			fence_put(fence);

			adev->uvd.filp[i] = NULL;
			atomic_set(&adev->uvd.handles[i], 0);
		}
	}
}

static void amdgpu_uvd_force_into_uvd_segment(struct amdgpu_bo *rbo)
{
	int i;
	for (i = 0; i < rbo->placement.num_placement; ++i) {
		rbo->placements[i].fpfn = 0 >> PAGE_SHIFT;
		rbo->placements[i].lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
	}
}

/**
 * amdgpu_uvd_cs_pass1 - first parsing round
 *
 * @ctx: UVD parser context
 *
 * Make sure UVD message and feedback buffers are in VRAM and
 * nobody is violating an 256MB boundary.
 */
static int amdgpu_uvd_cs_pass1(struct amdgpu_uvd_cs_ctx *ctx)
{
	struct amdgpu_bo_va_mapping *mapping;
	struct amdgpu_bo *bo;
	uint32_t cmd, lo, hi;
	uint64_t addr;
	int r = 0;

	lo = amdgpu_get_ib_value(ctx->parser, ctx->ib_idx, ctx->data0);
	hi = amdgpu_get_ib_value(ctx->parser, ctx->ib_idx, ctx->data1);
	addr = ((uint64_t)lo) | (((uint64_t)hi) << 32);

	mapping = amdgpu_cs_find_mapping(ctx->parser, addr, &bo);
	if (mapping == NULL) {
		DRM_ERROR("Can't find BO for addr 0x%08Lx\n", addr);
		return -EINVAL;
	}

	if (!ctx->parser->adev->uvd.address_64_bit) {
		/* check if it's a message or feedback command */
		cmd = amdgpu_get_ib_value(ctx->parser, ctx->ib_idx, ctx->idx) >> 1;
		if (cmd == 0x0 || cmd == 0x3) {
			/* yes, force it into VRAM */
			uint32_t domain = AMDGPU_GEM_DOMAIN_VRAM;
			amdgpu_ttm_placement_from_domain(bo, domain);
		}
		amdgpu_uvd_force_into_uvd_segment(bo);

		r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
	}

	return r;
}

/**
 * amdgpu_uvd_cs_msg_decode - handle UVD decode message
 *
 * @msg: pointer to message structure
 * @buf_sizes: returned buffer sizes
 *
 * Peek into the decode message and calculate the necessary buffer sizes.
 */
static int amdgpu_uvd_cs_msg_decode(struct amdgpu_device *adev, uint32_t *msg,
	unsigned buf_sizes[])
{
	unsigned stream_type = msg[4];
	unsigned width = msg[6];
	unsigned height = msg[7];
	unsigned dpb_size = msg[9];
	unsigned pitch = msg[28];
	unsigned level = msg[57];

	unsigned width_in_mb = width / 16;
	unsigned height_in_mb = ALIGN(height / 16, 2);
	unsigned fs_in_mb = width_in_mb * height_in_mb;

	unsigned image_size, tmp, min_dpb_size, num_dpb_buffer;
	unsigned min_ctx_size = ~0;

	image_size = width * height;
	image_size += image_size / 2;
	image_size = ALIGN(image_size, 1024);

	switch (stream_type) {
	case 0: /* H264 */
		switch(level) {
		case 30:
			num_dpb_buffer = 8100 / fs_in_mb;
			break;
		case 31:
			num_dpb_buffer = 18000 / fs_in_mb;
			break;
		case 32:
			num_dpb_buffer = 20480 / fs_in_mb;
			break;
		case 41:
			num_dpb_buffer = 32768 / fs_in_mb;
			break;
		case 42:
			num_dpb_buffer = 34816 / fs_in_mb;
			break;
		case 50:
			num_dpb_buffer = 110400 / fs_in_mb;
			break;
		case 51:
			num_dpb_buffer = 184320 / fs_in_mb;
			break;
		default:
			num_dpb_buffer = 184320 / fs_in_mb;
			break;
		}
		num_dpb_buffer++;
		if (num_dpb_buffer > 17)
			num_dpb_buffer = 17;

		/* reference picture buffer */
		min_dpb_size = image_size * num_dpb_buffer;

		/* macroblock context buffer */
		min_dpb_size += width_in_mb * height_in_mb * num_dpb_buffer * 192;

		/* IT surface buffer */
		min_dpb_size += width_in_mb * height_in_mb * 32;
		break;

	case 1: /* VC1 */

		/* reference picture buffer */
		min_dpb_size = image_size * 3;

		/* CONTEXT_BUFFER */
		min_dpb_size += width_in_mb * height_in_mb * 128;

		/* IT surface buffer */
		min_dpb_size += width_in_mb * 64;

		/* DB surface buffer */
		min_dpb_size += width_in_mb * 128;

		/* BP */
		tmp = max(width_in_mb, height_in_mb);
		min_dpb_size += ALIGN(tmp * 7 * 16, 64);
		break;

	case 3: /* MPEG2 */

		/* reference picture buffer */
		min_dpb_size = image_size * 3;
		break;

	case 4: /* MPEG4 */

		/* reference picture buffer */
		min_dpb_size = image_size * 3;

		/* CM */
		min_dpb_size += width_in_mb * height_in_mb * 64;

		/* IT surface buffer */
		min_dpb_size += ALIGN(width_in_mb * height_in_mb * 32, 64);
		break;

	case 7: /* H264 Perf */
		switch(level) {
		case 30:
			num_dpb_buffer = 8100 / fs_in_mb;
			break;
		case 31:
			num_dpb_buffer = 18000 / fs_in_mb;
			break;
		case 32:
			num_dpb_buffer = 20480 / fs_in_mb;
			break;
		case 41:
			num_dpb_buffer = 32768 / fs_in_mb;
			break;
		case 42:
			num_dpb_buffer = 34816 / fs_in_mb;
			break;
		case 50:
			num_dpb_buffer = 110400 / fs_in_mb;
			break;
		case 51:
			num_dpb_buffer = 184320 / fs_in_mb;
			break;
		default:
			num_dpb_buffer = 184320 / fs_in_mb;
			break;
		}
		num_dpb_buffer++;
		if (num_dpb_buffer > 17)
			num_dpb_buffer = 17;

		/* reference picture buffer */
		min_dpb_size = image_size * num_dpb_buffer;

		if (!adev->uvd.use_ctx_buf){
			/* macroblock context buffer */
			min_dpb_size +=
				width_in_mb * height_in_mb * num_dpb_buffer * 192;

			/* IT surface buffer */
			min_dpb_size += width_in_mb * height_in_mb * 32;
		} else {
			/* macroblock context buffer */
			min_ctx_size =
				width_in_mb * height_in_mb * num_dpb_buffer * 192;
		}
		break;

	case 16: /* H265 */
		image_size = (ALIGN(width, 16) * ALIGN(height, 16) * 3) / 2;
		image_size = ALIGN(image_size, 256);

		num_dpb_buffer = (le32_to_cpu(msg[59]) & 0xff) + 2;
		min_dpb_size = image_size * num_dpb_buffer;
		min_ctx_size = ((width + 255) / 16) * ((height + 255) / 16)
					   * 16 * num_dpb_buffer + 52 * 1024;
		break;

	default:
		DRM_ERROR("UVD codec not handled %d!\n", stream_type);
		return -EINVAL;
	}

	if (width > pitch) {
		DRM_ERROR("Invalid UVD decoding target pitch!\n");
		return -EINVAL;
	}

	if (dpb_size < min_dpb_size) {
		DRM_ERROR("Invalid dpb_size in UVD message (%d / %d)!\n",
			  dpb_size, min_dpb_size);
		return -EINVAL;
	}

	buf_sizes[0x1] = dpb_size;
	buf_sizes[0x2] = image_size;
	buf_sizes[0x4] = min_ctx_size;
	return 0;
}

/**
 * amdgpu_uvd_cs_msg - handle UVD message
 *
 * @ctx: UVD parser context
 * @bo: buffer object containing the message
 * @offset: offset into the buffer object
 *
 * Peek into the UVD message and extract the session id.
 * Make sure that we don't open up to many sessions.
 */
static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
			     struct amdgpu_bo *bo, unsigned offset)
{
	struct amdgpu_device *adev = ctx->parser->adev;
	int32_t *msg, msg_type, handle;
	void *ptr;
	long r;
	int i;

	if (offset & 0x3F) {
		DRM_ERROR("UVD messages must be 64 byte aligned!\n");
		return -EINVAL;
	}

	r = amdgpu_bo_kmap(bo, &ptr);
	if (r) {
		DRM_ERROR("Failed mapping the UVD message (%ld)!\n", r);
		return r;
	}

	msg = ptr + offset;

	msg_type = msg[1];
	handle = msg[2];

	if (handle == 0) {
		DRM_ERROR("Invalid UVD handle!\n");
		return -EINVAL;
	}

	switch (msg_type) {
	case 0:
		/* it's a create msg, calc image size (width * height) */
		amdgpu_bo_kunmap(bo);

		/* try to alloc a new handle */
		for (i = 0; i < adev->uvd.max_handles; ++i) {
			if (atomic_read(&adev->uvd.handles[i]) == handle) {
				DRM_ERROR("Handle 0x%x already in use!\n", handle);
				return -EINVAL;
			}

			if (!atomic_cmpxchg(&adev->uvd.handles[i], 0, handle)) {
				adev->uvd.filp[i] = ctx->parser->filp;
				return 0;
			}
		}

		DRM_ERROR("No more free UVD handles!\n");
		return -ENOSPC;

	case 1:
		/* it's a decode msg, calc buffer sizes */
		r = amdgpu_uvd_cs_msg_decode(adev, msg, ctx->buf_sizes);
		amdgpu_bo_kunmap(bo);
		if (r)
			return r;

		/* validate the handle */
		for (i = 0; i < adev->uvd.max_handles; ++i) {
			if (atomic_read(&adev->uvd.handles[i]) == handle) {
				if (adev->uvd.filp[i] != ctx->parser->filp) {
					DRM_ERROR("UVD handle collision detected!\n");
					return -EINVAL;
				}
				return 0;
			}
		}

		DRM_ERROR("Invalid UVD handle 0x%x!\n", handle);
		return -ENOENT;

	case 2:
		/* it's a destroy msg, free the handle */
		for (i = 0; i < adev->uvd.max_handles; ++i)
			atomic_cmpxchg(&adev->uvd.handles[i], handle, 0);
		amdgpu_bo_kunmap(bo);
		return 0;

	default:
		DRM_ERROR("Illegal UVD message type (%d)!\n", msg_type);
		return -EINVAL;
	}
	BUG();
	return -EINVAL;
}

/**
 * amdgpu_uvd_cs_pass2 - second parsing round
 *
 * @ctx: UVD parser context
 *
 * Patch buffer addresses, make sure buffer sizes are correct.
 */
static int amdgpu_uvd_cs_pass2(struct amdgpu_uvd_cs_ctx *ctx)
{
	struct amdgpu_bo_va_mapping *mapping;
	struct amdgpu_bo *bo;
	uint32_t cmd, lo, hi;
	uint64_t start, end;
	uint64_t addr;
	int r;

	lo = amdgpu_get_ib_value(ctx->parser, ctx->ib_idx, ctx->data0);
	hi = amdgpu_get_ib_value(ctx->parser, ctx->ib_idx, ctx->data1);
	addr = ((uint64_t)lo) | (((uint64_t)hi) << 32);

	mapping = amdgpu_cs_find_mapping(ctx->parser, addr, &bo);
	if (mapping == NULL)
		return -EINVAL;

	start = amdgpu_bo_gpu_offset(bo);

	end = (mapping->it.last + 1 - mapping->it.start);
	end = end * AMDGPU_GPU_PAGE_SIZE + start;

	addr -= ((uint64_t)mapping->it.start) * AMDGPU_GPU_PAGE_SIZE;
	start += addr;

	amdgpu_set_ib_value(ctx->parser, ctx->ib_idx, ctx->data0,
			    lower_32_bits(start));
	amdgpu_set_ib_value(ctx->parser, ctx->ib_idx, ctx->data1,
			    upper_32_bits(start));

	cmd = amdgpu_get_ib_value(ctx->parser, ctx->ib_idx, ctx->idx) >> 1;
	if (cmd < 0x4) {
		if ((end - start) < ctx->buf_sizes[cmd]) {
			DRM_ERROR("buffer (%d) to small (%d / %d)!\n", cmd,
				  (unsigned)(end - start),
				  ctx->buf_sizes[cmd]);
			return -EINVAL;
		}

	} else if (cmd == 0x206) {
		if ((end - start) < ctx->buf_sizes[4]) {
			DRM_ERROR("buffer (%d) to small (%d / %d)!\n", cmd,
					  (unsigned)(end - start),
					  ctx->buf_sizes[4]);
			return -EINVAL;
		}
	} else if ((cmd != 0x100) && (cmd != 0x204)) {
		DRM_ERROR("invalid UVD command %X!\n", cmd);
		return -EINVAL;
	}

	if (!ctx->parser->adev->uvd.address_64_bit) {
		if ((start >> 28) != ((end - 1) >> 28)) {
			DRM_ERROR("reloc %LX-%LX crossing 256MB boundary!\n",
				  start, end);
			return -EINVAL;
		}

		if ((cmd == 0 || cmd == 0x3) &&
		    (start >> 28) != (ctx->parser->adev->uvd.gpu_addr >> 28)) {
			DRM_ERROR("msg/fb buffer %LX-%LX out of 256MB segment!\n",
				  start, end);
			return -EINVAL;
		}
	}

	if (cmd == 0) {
		ctx->has_msg_cmd = true;
		r = amdgpu_uvd_cs_msg(ctx, bo, addr);
		if (r)
			return r;
	} else if (!ctx->has_msg_cmd) {
		DRM_ERROR("Message needed before other commands are send!\n");
		return -EINVAL;
	}

	return 0;
}

/**
 * amdgpu_uvd_cs_reg - parse register writes
 *
 * @ctx: UVD parser context
 * @cb: callback function
 *
 * Parse the register writes, call cb on each complete command.
 */
static int amdgpu_uvd_cs_reg(struct amdgpu_uvd_cs_ctx *ctx,
			     int (*cb)(struct amdgpu_uvd_cs_ctx *ctx))
{
	struct amdgpu_ib *ib = &ctx->parser->job->ibs[ctx->ib_idx];
	int i, r;

	ctx->idx++;
	for (i = 0; i <= ctx->count; ++i) {
		unsigned reg = ctx->reg + i;

		if (ctx->idx >= ib->length_dw) {
			DRM_ERROR("Register command after end of CS!\n");
			return -EINVAL;
		}

		switch (reg) {
		case mmUVD_GPCOM_VCPU_DATA0:
			ctx->data0 = ctx->idx;
			break;
		case mmUVD_GPCOM_VCPU_DATA1:
			ctx->data1 = ctx->idx;
			break;
		case mmUVD_GPCOM_VCPU_CMD:
			r = cb(ctx);
			if (r)
				return r;
			break;
		case mmUVD_ENGINE_CNTL:
			break;
		default:
			DRM_ERROR("Invalid reg 0x%X!\n", reg);
			return -EINVAL;
		}
		ctx->idx++;
	}
	return 0;
}

/**
 * amdgpu_uvd_cs_packets - parse UVD packets
 *
 * @ctx: UVD parser context
 * @cb: callback function
 *
 * Parse the command stream packets.
 */
static int amdgpu_uvd_cs_packets(struct amdgpu_uvd_cs_ctx *ctx,
				 int (*cb)(struct amdgpu_uvd_cs_ctx *ctx))
{
	struct amdgpu_ib *ib = &ctx->parser->job->ibs[ctx->ib_idx];
	int r;

	for (ctx->idx = 0 ; ctx->idx < ib->length_dw; ) {
		uint32_t cmd = amdgpu_get_ib_value(ctx->parser, ctx->ib_idx, ctx->idx);
		unsigned type = CP_PACKET_GET_TYPE(cmd);
		switch (type) {
		case PACKET_TYPE0:
			ctx->reg = CP_PACKET0_GET_REG(cmd);
			ctx->count = CP_PACKET_GET_COUNT(cmd);
			r = amdgpu_uvd_cs_reg(ctx, cb);
			if (r)
				return r;
			break;
		case PACKET_TYPE2:
			++ctx->idx;
			break;
		default:
			DRM_ERROR("Unknown packet type %d !\n", type);
			return -EINVAL;
		}
	}
	return 0;
}

/**
 * amdgpu_uvd_ring_parse_cs - UVD command submission parser
 *
 * @parser: Command submission parser context
 *
 * Parse the command stream, patch in addresses as necessary.
 */
int amdgpu_uvd_ring_parse_cs(struct amdgpu_cs_parser *parser, uint32_t ib_idx)
{
	struct amdgpu_uvd_cs_ctx ctx = {};
	unsigned buf_sizes[] = {
		[0x00000000]	=	2048,
		[0x00000001]	=	0xFFFFFFFF,
		[0x00000002]	=	0xFFFFFFFF,
		[0x00000003]	=	2048,
		[0x00000004]	=	0xFFFFFFFF,
	};
	struct amdgpu_ib *ib = &parser->job->ibs[ib_idx];
	int r;

	if (ib->length_dw % 16) {
		DRM_ERROR("UVD IB length (%d) not 16 dwords aligned!\n",
			  ib->length_dw);
		return -EINVAL;
	}

	ctx.parser = parser;
	ctx.buf_sizes = buf_sizes;
	ctx.ib_idx = ib_idx;

	/* first round, make sure the buffers are actually in the UVD segment */
	r = amdgpu_uvd_cs_packets(&ctx, amdgpu_uvd_cs_pass1);
	if (r)
		return r;

	/* second round, patch buffer addresses into the command stream */
	r = amdgpu_uvd_cs_packets(&ctx, amdgpu_uvd_cs_pass2);
	if (r)
		return r;

	if (!ctx.has_msg_cmd) {
		DRM_ERROR("UVD-IBs need a msg command!\n");
		return -EINVAL;
	}

	return 0;
}

static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
			       bool direct, struct fence **fence)
{
	struct ttm_validate_buffer tv;
	struct ww_acquire_ctx ticket;
	struct list_head head;
	struct amdgpu_job *job;
	struct amdgpu_ib *ib;
	struct fence *f = NULL;
	struct amdgpu_device *adev = ring->adev;
	uint64_t addr;
	int i, r;

	memset(&tv, 0, sizeof(tv));
	tv.bo = &bo->tbo;

	INIT_LIST_HEAD(&head);
	list_add(&tv.head, &head);

	r = ttm_eu_reserve_buffers(&ticket, &head, true, NULL);
	if (r)
		return r;

	if (!bo->adev->uvd.address_64_bit) {
		amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM);
		amdgpu_uvd_force_into_uvd_segment(bo);
	}

	r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
	if (r)
		goto err;

	r = amdgpu_job_alloc_with_ib(adev, 64, &job);
	if (r)
		goto err;

	ib = &job->ibs[0];
	addr = amdgpu_bo_gpu_offset(bo);
	ib->ptr[0] = PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0);
	ib->ptr[1] = addr;
	ib->ptr[2] = PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0);
	ib->ptr[3] = addr >> 32;
	ib->ptr[4] = PACKET0(mmUVD_GPCOM_VCPU_CMD, 0);
	ib->ptr[5] = 0;
	for (i = 6; i < 16; ++i)
		ib->ptr[i] = PACKET2(0);
	ib->length_dw = 16;

	if (direct) {
		r = amdgpu_ib_schedule(ring, 1, ib, NULL, NULL, &f);
		job->fence = fence_get(f);
		if (r)
			goto err_free;

		amdgpu_job_free(job);
	} else {
		r = amdgpu_job_submit(job, ring, &adev->uvd.entity,
				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
		if (r)
			goto err_free;
	}

	ttm_eu_fence_buffer_objects(&ticket, &head, f);

	if (fence)
		*fence = fence_get(f);
	amdgpu_bo_unref(&bo);
	fence_put(f);

	return 0;

err_free:
	amdgpu_job_free(job);

err:
	ttm_eu_backoff_reservation(&ticket, &head);
	return r;
}

/* multiple fence commands without any stream commands in between can
   crash the vcpu so just try to emmit a dummy create/destroy msg to
   avoid this */
int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
			      struct fence **fence)
{
	struct amdgpu_device *adev = ring->adev;
	struct amdgpu_bo *bo;
	uint32_t *msg;
	int r, i;

	r = amdgpu_bo_create(adev, 1024, PAGE_SIZE, true,
			     AMDGPU_GEM_DOMAIN_VRAM,
			     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
			     NULL, NULL, &bo);
	if (r)
		return r;

	r = amdgpu_bo_reserve(bo, false);
	if (r) {
		amdgpu_bo_unref(&bo);
		return r;
	}

	r = amdgpu_bo_kmap(bo, (void **)&msg);
	if (r) {
		amdgpu_bo_unreserve(bo);
		amdgpu_bo_unref(&bo);
		return r;
	}

	/* stitch together an UVD create msg */
	msg[0] = cpu_to_le32(0x00000de4);
	msg[1] = cpu_to_le32(0x00000000);
	msg[2] = cpu_to_le32(handle);
	msg[3] = cpu_to_le32(0x00000000);
	msg[4] = cpu_to_le32(0x00000000);
	msg[5] = cpu_to_le32(0x00000000);
	msg[6] = cpu_to_le32(0x00000000);
	msg[7] = cpu_to_le32(0x00000780);
	msg[8] = cpu_to_le32(0x00000440);
	msg[9] = cpu_to_le32(0x00000000);
	msg[10] = cpu_to_le32(0x01b37000);
	for (i = 11; i < 1024; ++i)
		msg[i] = cpu_to_le32(0x0);

	amdgpu_bo_kunmap(bo);
	amdgpu_bo_unreserve(bo);

	return amdgpu_uvd_send_msg(ring, bo, true, fence);
}

int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
			       bool direct, struct fence **fence)
{
	struct amdgpu_device *adev = ring->adev;
	struct amdgpu_bo *bo;
	uint32_t *msg;
	int r, i;

	r = amdgpu_bo_create(adev, 1024, PAGE_SIZE, true,
			     AMDGPU_GEM_DOMAIN_VRAM,
			     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
			     NULL, NULL, &bo);
	if (r)
		return r;

	r = amdgpu_bo_reserve(bo, false);
	if (r) {
		amdgpu_bo_unref(&bo);
		return r;
	}

	r = amdgpu_bo_kmap(bo, (void **)&msg);
	if (r) {
		amdgpu_bo_unreserve(bo);
		amdgpu_bo_unref(&bo);
		return r;
	}

	/* stitch together an UVD destroy msg */
	msg[0] = cpu_to_le32(0x00000de4);
	msg[1] = cpu_to_le32(0x00000002);
	msg[2] = cpu_to_le32(handle);
	msg[3] = cpu_to_le32(0x00000000);
	for (i = 4; i < 1024; ++i)
		msg[i] = cpu_to_le32(0x0);

	amdgpu_bo_kunmap(bo);
	amdgpu_bo_unreserve(bo);

	return amdgpu_uvd_send_msg(ring, bo, direct, fence);
}

static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
{
	struct amdgpu_device *adev =
		container_of(work, struct amdgpu_device, uvd.idle_work.work);
	unsigned i, fences, handles = 0;

	fences = amdgpu_fence_count_emitted(&adev->uvd.ring);

	for (i = 0; i < adev->uvd.max_handles; ++i)
		if (atomic_read(&adev->uvd.handles[i]))
			++handles;

	if (fences == 0 && handles == 0) {
		if (adev->pm.dpm_enabled) {
			amdgpu_dpm_enable_uvd(adev, false);
		} else {
			amdgpu_asic_set_uvd_clocks(adev, 0, 0);
		}
	} else {
		schedule_delayed_work(&adev->uvd.idle_work, UVD_IDLE_TIMEOUT);
	}
}

void amdgpu_uvd_ring_begin_use(struct amdgpu_ring *ring)
{
	struct amdgpu_device *adev = ring->adev;
	bool set_clocks = !cancel_delayed_work_sync(&adev->uvd.idle_work);

	if (set_clocks) {
		if (adev->pm.dpm_enabled) {
			amdgpu_dpm_enable_uvd(adev, true);
		} else {
			amdgpu_asic_set_uvd_clocks(adev, 53300, 40000);
		}
	}
}

void amdgpu_uvd_ring_end_use(struct amdgpu_ring *ring)
{
	schedule_delayed_work(&ring->adev->uvd.idle_work, UVD_IDLE_TIMEOUT);
}

/**
 * amdgpu_uvd_ring_test_ib - test ib execution
 *
 * @ring: amdgpu_ring pointer
 *
 * Test if we can successfully execute an IB
 */
int amdgpu_uvd_ring_test_ib(struct amdgpu_ring *ring, long timeout)
{
	struct fence *fence;
	long r;

	r = amdgpu_uvd_get_create_msg(ring, 1, NULL);
	if (r) {
		DRM_ERROR("amdgpu: failed to get create msg (%ld).\n", r);
		goto error;
	}

	r = amdgpu_uvd_get_destroy_msg(ring, 1, true, &fence);
	if (r) {
		DRM_ERROR("amdgpu: failed to get destroy ib (%ld).\n", r);
		goto error;
	}

	r = fence_wait_timeout(fence, false, timeout);
	if (r == 0) {
		DRM_ERROR("amdgpu: IB test timed out.\n");
		r = -ETIMEDOUT;
	} else if (r < 0) {
		DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
	} else {
		DRM_INFO("ib test on ring %d succeeded\n",  ring->idx);
		r = 0;
	}

error:
	fence_put(fence);
	return r;
}