summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
blob: c853400805d1c66c98496b548ae2c30c3f640987 (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
/*
 * Copyright 2013 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 <christian.koenig@amd.com>
 */

#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_vce.h"
#include "cikd.h"

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

/* Firmware Names */
#ifdef CONFIG_DRM_AMDGPU_CIK
#define FIRMWARE_BONAIRE	"radeon/bonaire_vce.bin"
#define FIRMWARE_KABINI	"radeon/kabini_vce.bin"
#define FIRMWARE_KAVERI	"radeon/kaveri_vce.bin"
#define FIRMWARE_HAWAII	"radeon/hawaii_vce.bin"
#define FIRMWARE_MULLINS	"radeon/mullins_vce.bin"
#endif
#define FIRMWARE_TONGA		"amdgpu/tonga_vce.bin"
#define FIRMWARE_CARRIZO	"amdgpu/carrizo_vce.bin"
#define FIRMWARE_FIJI		"amdgpu/fiji_vce.bin"
#define FIRMWARE_STONEY		"amdgpu/stoney_vce.bin"
#define FIRMWARE_POLARIS10	"amdgpu/polaris10_vce.bin"
#define FIRMWARE_POLARIS11         "amdgpu/polaris11_vce.bin"
#define FIRMWARE_POLARIS12         "amdgpu/polaris12_vce.bin"

#define FIRMWARE_VEGA10		"amdgpu/vega10_vce.bin"

#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);
MODULE_FIRMWARE(FIRMWARE_POLARIS12);

MODULE_FIRMWARE(FIRMWARE_VEGA10);

static void amdgpu_vce_idle_work_handler(struct work_struct *work);

/**
 * amdgpu_vce_init - allocate memory, load vce firmware
 *
 * @adev: amdgpu_device pointer
 *
 * First step to get VCE online, allocate memory and load the firmware
 */
int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
{
	struct amdgpu_ring *ring;
	struct amd_sched_rq *rq;
	const char *fw_name;
	const struct common_firmware_header *hdr;
	unsigned ucode_version, version_major, version_minor, binary_id;
	int i, r;

	switch (adev->asic_type) {
#ifdef CONFIG_DRM_AMDGPU_CIK
	case CHIP_BONAIRE:
		fw_name = FIRMWARE_BONAIRE;
		break;
	case CHIP_KAVERI:
		fw_name = FIRMWARE_KAVERI;
		break;
	case CHIP_KABINI:
		fw_name = FIRMWARE_KABINI;
		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_CARRIZO:
		fw_name = FIRMWARE_CARRIZO;
		break;
	case CHIP_FIJI:
		fw_name = FIRMWARE_FIJI;
		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;
	case CHIP_VEGA10:
		fw_name = FIRMWARE_VEGA10;
		break;
	case CHIP_POLARIS12:
		fw_name = FIRMWARE_POLARIS12;
		break;

	default:
		return -EINVAL;
	}

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

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

	hdr = (const struct common_firmware_header *)adev->vce.fw->data;

	ucode_version = le32_to_cpu(hdr->ucode_version);
	version_major = (ucode_version >> 20) & 0xfff;
	version_minor = (ucode_version >> 8) & 0xfff;
	binary_id = ucode_version & 0xff;
	DRM_INFO("Found VCE firmware Version: %hhd.%hhd Binary ID: %hhd\n",
		version_major, version_minor, binary_id);
	adev->vce.fw_version = ((version_major << 24) | (version_minor << 16) |
				(binary_id << 8));

	/* allocate firmware, stack and heap BO */

	r = amdgpu_bo_create(adev, size, PAGE_SIZE, true,
			     AMDGPU_GEM_DOMAIN_VRAM,
			     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
			     AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
			     NULL, NULL, &adev->vce.vcpu_bo);
	if (r) {
		dev_err(adev->dev, "(%d) failed to allocate VCE bo\n", r);
		return r;
	}

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

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


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

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

	INIT_DELAYED_WORK(&adev->vce.idle_work, amdgpu_vce_idle_work_handler);
	mutex_init(&adev->vce.idle_mutex);

	return 0;
}

/**
 * amdgpu_vce_fini - free memory
 *
 * @adev: amdgpu_device pointer
 *
 * Last step on VCE teardown, free firmware memory
 */
int amdgpu_vce_sw_fini(struct amdgpu_device *adev)
{
	unsigned i;

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

	amd_sched_entity_fini(&adev->vce.ring[0].sched, &adev->vce.entity);

	amdgpu_bo_unref(&adev->vce.vcpu_bo);

	for (i = 0; i < adev->vce.num_rings; i++)
		amdgpu_ring_fini(&adev->vce.ring[i]);

	release_firmware(adev->vce.fw);
	mutex_destroy(&adev->vce.idle_mutex);

	return 0;
}

/**
 * amdgpu_vce_suspend - unpin VCE fw memory
 *
 * @adev: amdgpu_device pointer
 *
 */
int amdgpu_vce_suspend(struct amdgpu_device *adev)
{
	int i;

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

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

	if (i == AMDGPU_MAX_VCE_HANDLES)
		return 0;

	cancel_delayed_work_sync(&adev->vce.idle_work);
	/* TODO: suspending running encoding sessions isn't supported */
	return -EINVAL;
}

/**
 * amdgpu_vce_resume - pin VCE fw memory
 *
 * @adev: amdgpu_device pointer
 *
 */
int amdgpu_vce_resume(struct amdgpu_device *adev)
{
	void *cpu_addr;
	const struct common_firmware_header *hdr;
	unsigned offset;
	int r;

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

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

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

	hdr = (const struct common_firmware_header *)adev->vce.fw->data;
	offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
	memcpy_toio(cpu_addr, adev->vce.fw->data + offset,
		    adev->vce.fw->size - offset);

	amdgpu_bo_kunmap(adev->vce.vcpu_bo);

	amdgpu_bo_unreserve(adev->vce.vcpu_bo);

	return 0;
}

/**
 * amdgpu_vce_idle_work_handler - power off VCE
 *
 * @work: pointer to work structure
 *
 * power of VCE when it's not used any more
 */
static void amdgpu_vce_idle_work_handler(struct work_struct *work)
{
	struct amdgpu_device *adev =
		container_of(work, struct amdgpu_device, vce.idle_work.work);
	unsigned i, count = 0;

	if (amdgpu_sriov_vf(adev))
		return;

	for (i = 0; i < adev->vce.num_rings; i++)
		count += amdgpu_fence_count_emitted(&adev->vce.ring[i]);

	if (count == 0) {
		if (adev->pm.dpm_enabled) {
			amdgpu_dpm_enable_vce(adev, false);
		} else {
			amdgpu_asic_set_vce_clocks(adev, 0, 0);
			amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
							    AMD_PG_STATE_GATE);
			amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
							    AMD_CG_STATE_GATE);
		}
	} else {
		schedule_delayed_work(&adev->vce.idle_work, VCE_IDLE_TIMEOUT);
	}
}

/**
 * amdgpu_vce_ring_begin_use - power up VCE
 *
 * @ring: amdgpu ring
 *
 * Make sure VCE is powerd up when we want to use it
 */
void amdgpu_vce_ring_begin_use(struct amdgpu_ring *ring)
{
	struct amdgpu_device *adev = ring->adev;
	bool set_clocks;

	if (amdgpu_sriov_vf(adev))
		return;

	mutex_lock(&adev->vce.idle_mutex);
	set_clocks = !cancel_delayed_work_sync(&adev->vce.idle_work);
	if (set_clocks) {
		if (adev->pm.dpm_enabled) {
			amdgpu_dpm_enable_vce(adev, true);
		} else {
			amdgpu_asic_set_vce_clocks(adev, 53300, 40000);
			amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
							    AMD_CG_STATE_UNGATE);
			amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
							    AMD_PG_STATE_UNGATE);

		}
	}
	mutex_unlock(&adev->vce.idle_mutex);
}

/**
 * amdgpu_vce_ring_end_use - power VCE down
 *
 * @ring: amdgpu ring
 *
 * Schedule work to power VCE down again
 */
void amdgpu_vce_ring_end_use(struct amdgpu_ring *ring)
{
	schedule_delayed_work(&ring->adev->vce.idle_work, VCE_IDLE_TIMEOUT);
}

/**
 * amdgpu_vce_free_handles - free still open VCE handles
 *
 * @adev: amdgpu_device pointer
 * @filp: drm file pointer
 *
 * Close all VCE handles still open by this file pointer
 */
void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
{
	struct amdgpu_ring *ring = &adev->vce.ring[0];
	int i, r;
	for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i) {
		uint32_t handle = atomic_read(&adev->vce.handles[i]);

		if (!handle || adev->vce.filp[i] != filp)
			continue;

		r = amdgpu_vce_get_destroy_msg(ring, handle, false, NULL);
		if (r)
			DRM_ERROR("Error destroying VCE handle (%d)!\n", r);

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

/**
 * amdgpu_vce_get_create_msg - generate a VCE create msg
 *
 * @adev: amdgpu_device pointer
 * @ring: ring we should submit the msg to
 * @handle: VCE session handle to use
 * @fence: optional fence to return
 *
 * Open up a stream for HW test
 */
int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
			      struct dma_fence **fence)
{
	const unsigned ib_size_dw = 1024;
	struct amdgpu_job *job;
	struct amdgpu_ib *ib;
	struct dma_fence *f = NULL;
	uint64_t dummy;
	int i, r;

	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
	if (r)
		return r;

	ib = &job->ibs[0];

	dummy = ib->gpu_addr + 1024;

	/* stitch together an VCE create msg */
	ib->length_dw = 0;
	ib->ptr[ib->length_dw++] = 0x0000000c; /* len */
	ib->ptr[ib->length_dw++] = 0x00000001; /* session cmd */
	ib->ptr[ib->length_dw++] = handle;

	if ((ring->adev->vce.fw_version >> 24) >= 52)
		ib->ptr[ib->length_dw++] = 0x00000040; /* len */
	else
		ib->ptr[ib->length_dw++] = 0x00000030; /* len */
	ib->ptr[ib->length_dw++] = 0x01000001; /* create cmd */
	ib->ptr[ib->length_dw++] = 0x00000000;
	ib->ptr[ib->length_dw++] = 0x00000042;
	ib->ptr[ib->length_dw++] = 0x0000000a;
	ib->ptr[ib->length_dw++] = 0x00000001;
	ib->ptr[ib->length_dw++] = 0x00000080;
	ib->ptr[ib->length_dw++] = 0x00000060;
	ib->ptr[ib->length_dw++] = 0x00000100;
	ib->ptr[ib->length_dw++] = 0x00000100;
	ib->ptr[ib->length_dw++] = 0x0000000c;
	ib->ptr[ib->length_dw++] = 0x00000000;
	if ((ring->adev->vce.fw_version >> 24) >= 52) {
		ib->ptr[ib->length_dw++] = 0x00000000;
		ib->ptr[ib->length_dw++] = 0x00000000;
		ib->ptr[ib->length_dw++] = 0x00000000;
		ib->ptr[ib->length_dw++] = 0x00000000;
	}

	ib->ptr[ib->length_dw++] = 0x00000014; /* len */
	ib->ptr[ib->length_dw++] = 0x05000005; /* feedback buffer */
	ib->ptr[ib->length_dw++] = upper_32_bits(dummy);
	ib->ptr[ib->length_dw++] = dummy;
	ib->ptr[ib->length_dw++] = 0x00000001;

	for (i = ib->length_dw; i < ib_size_dw; ++i)
		ib->ptr[i] = 0x0;

	r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
	job->fence = dma_fence_get(f);
	if (r)
		goto err;

	amdgpu_job_free(job);
	if (fence)
		*fence = dma_fence_get(f);
	dma_fence_put(f);
	return 0;

err:
	amdgpu_job_free(job);
	return r;
}

/**
 * amdgpu_vce_get_destroy_msg - generate a VCE destroy msg
 *
 * @adev: amdgpu_device pointer
 * @ring: ring we should submit the msg to
 * @handle: VCE session handle to use
 * @fence: optional fence to return
 *
 * Close up a stream for HW test or if userspace failed to do so
 */
int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
			       bool direct, struct dma_fence **fence)
{
	const unsigned ib_size_dw = 1024;
	struct amdgpu_job *job;
	struct amdgpu_ib *ib;
	struct dma_fence *f = NULL;
	int i, r;

	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
	if (r)
		return r;

	ib = &job->ibs[0];

	/* stitch together an VCE destroy msg */
	ib->length_dw = 0;
	ib->ptr[ib->length_dw++] = 0x0000000c; /* len */
	ib->ptr[ib->length_dw++] = 0x00000001; /* session cmd */
	ib->ptr[ib->length_dw++] = handle;

	ib->ptr[ib->length_dw++] = 0x00000020; /* len */
	ib->ptr[ib->length_dw++] = 0x00000002; /* task info */
	ib->ptr[ib->length_dw++] = 0xffffffff; /* next task info, set to 0xffffffff if no */
	ib->ptr[ib->length_dw++] = 0x00000001; /* destroy session */
	ib->ptr[ib->length_dw++] = 0x00000000;
	ib->ptr[ib->length_dw++] = 0x00000000;
	ib->ptr[ib->length_dw++] = 0xffffffff; /* feedback is not needed, set to 0xffffffff and firmware will not output feedback */
	ib->ptr[ib->length_dw++] = 0x00000000;

	ib->ptr[ib->length_dw++] = 0x00000008; /* len */
	ib->ptr[ib->length_dw++] = 0x02000001; /* destroy cmd */

	for (i = ib->length_dw; i < ib_size_dw; ++i)
		ib->ptr[i] = 0x0;

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

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

	if (fence)
		*fence = dma_fence_get(f);
	dma_fence_put(f);
	return 0;

err:
	amdgpu_job_free(job);
	return r;
}

/**
 * amdgpu_vce_cs_reloc - command submission relocation
 *
 * @p: parser context
 * @lo: address of lower dword
 * @hi: address of higher dword
 * @size: minimum size
 *
 * Patch relocation inside command stream with real buffer address
 */
static int amdgpu_vce_cs_reloc(struct amdgpu_cs_parser *p, uint32_t ib_idx,
			       int lo, int hi, unsigned size, uint32_t index)
{
	struct amdgpu_bo_va_mapping *mapping;
	struct amdgpu_bo *bo;
	uint64_t addr;

	if (index == 0xffffffff)
		index = 0;

	addr = ((uint64_t)amdgpu_get_ib_value(p, ib_idx, lo)) |
	       ((uint64_t)amdgpu_get_ib_value(p, ib_idx, hi)) << 32;
	addr += ((uint64_t)size) * ((uint64_t)index);

	mapping = amdgpu_cs_find_mapping(p, addr, &bo);
	if (mapping == NULL) {
		DRM_ERROR("Can't find BO for addr 0x%010Lx %d %d %d %d\n",
			  addr, lo, hi, size, index);
		return -EINVAL;
	}

	if ((addr + (uint64_t)size) >
	    (mapping->last + 1) * AMDGPU_GPU_PAGE_SIZE) {
		DRM_ERROR("BO to small for addr 0x%010Lx %d %d\n",
			  addr, lo, hi);
		return -EINVAL;
	}

	addr -= mapping->start * AMDGPU_GPU_PAGE_SIZE;
	addr += amdgpu_bo_gpu_offset(bo);
	addr -= ((uint64_t)size) * ((uint64_t)index);

	amdgpu_set_ib_value(p, ib_idx, lo, lower_32_bits(addr));
	amdgpu_set_ib_value(p, ib_idx, hi, upper_32_bits(addr));

	return 0;
}

/**
 * amdgpu_vce_validate_handle - validate stream handle
 *
 * @p: parser context
 * @handle: handle to validate
 * @allocated: allocated a new handle?
 *
 * Validates the handle and return the found session index or -EINVAL
 * we we don't have another free session index.
 */
static int amdgpu_vce_validate_handle(struct amdgpu_cs_parser *p,
				      uint32_t handle, uint32_t *allocated)
{
	unsigned i;

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

	/* handle not found try to alloc a new one */
	for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i) {
		if (!atomic_cmpxchg(&p->adev->vce.handles[i], 0, handle)) {
			p->adev->vce.filp[i] = p->filp;
			p->adev->vce.img_size[i] = 0;
			*allocated |= 1 << i;
			return i;
		}
	}

	DRM_ERROR("No more free VCE handles!\n");
	return -EINVAL;
}

/**
 * amdgpu_vce_cs_parse - parse and validate the command stream
 *
 * @p: parser context
 *
 */
int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
{
	struct amdgpu_ib *ib = &p->job->ibs[ib_idx];
	unsigned fb_idx = 0, bs_idx = 0;
	int session_idx = -1;
	uint32_t destroyed = 0;
	uint32_t created = 0;
	uint32_t allocated = 0;
	uint32_t tmp, handle = 0;
	uint32_t *size = &tmp;
	int i, r, idx = 0;

	p->job->vm = NULL;
	ib->gpu_addr = amdgpu_sa_bo_gpu_addr(ib->sa_bo);

	r = amdgpu_cs_sysvm_access_required(p);
	if (r)
		return r;

	while (idx < ib->length_dw) {
		uint32_t len = amdgpu_get_ib_value(p, ib_idx, idx);
		uint32_t cmd = amdgpu_get_ib_value(p, ib_idx, idx + 1);

		if ((len < 8) || (len & 3)) {
			DRM_ERROR("invalid VCE command length (%d)!\n", len);
			r = -EINVAL;
			goto out;
		}

		switch (cmd) {
		case 0x00000001: /* session */
			handle = amdgpu_get_ib_value(p, ib_idx, idx + 2);
			session_idx = amdgpu_vce_validate_handle(p, handle,
								 &allocated);
			if (session_idx < 0) {
				r = session_idx;
				goto out;
			}
			size = &p->adev->vce.img_size[session_idx];
			break;

		case 0x00000002: /* task info */
			fb_idx = amdgpu_get_ib_value(p, ib_idx, idx + 6);
			bs_idx = amdgpu_get_ib_value(p, ib_idx, idx + 7);
			break;

		case 0x01000001: /* create */
			created |= 1 << session_idx;
			if (destroyed & (1 << session_idx)) {
				destroyed &= ~(1 << session_idx);
				allocated |= 1 << session_idx;

			} else if (!(allocated & (1 << session_idx))) {
				DRM_ERROR("Handle already in use!\n");
				r = -EINVAL;
				goto out;
			}

			*size = amdgpu_get_ib_value(p, ib_idx, idx + 8) *
				amdgpu_get_ib_value(p, ib_idx, idx + 10) *
				8 * 3 / 2;
			break;

		case 0x04000001: /* config extension */
		case 0x04000002: /* pic control */
		case 0x04000005: /* rate control */
		case 0x04000007: /* motion estimation */
		case 0x04000008: /* rdo */
		case 0x04000009: /* vui */
		case 0x05000002: /* auxiliary buffer */
		case 0x05000009: /* clock table */
			break;

		case 0x0500000c: /* hw config */
			switch (p->adev->asic_type) {
#ifdef CONFIG_DRM_AMDGPU_CIK
			case CHIP_KAVERI:
			case CHIP_MULLINS:
#endif
			case CHIP_CARRIZO:
				break;
			default:
				r = -EINVAL;
				goto out;
			}
			break;

		case 0x03000001: /* encode */
			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 10, idx + 9,
						*size, 0);
			if (r)
				goto out;

			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 12, idx + 11,
						*size / 3, 0);
			if (r)
				goto out;
			break;

		case 0x02000001: /* destroy */
			destroyed |= 1 << session_idx;
			break;

		case 0x05000001: /* context buffer */
			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 3, idx + 2,
						*size * 2, 0);
			if (r)
				goto out;
			break;

		case 0x05000004: /* video bitstream buffer */
			tmp = amdgpu_get_ib_value(p, ib_idx, idx + 4);
			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 3, idx + 2,
						tmp, bs_idx);
			if (r)
				goto out;
			break;

		case 0x05000005: /* feedback buffer */
			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 3, idx + 2,
						4096, fb_idx);
			if (r)
				goto out;
			break;

		default:
			DRM_ERROR("invalid VCE command (0x%x)!\n", cmd);
			r = -EINVAL;
			goto out;
		}

		if (session_idx == -1) {
			DRM_ERROR("no session command at start of IB\n");
			r = -EINVAL;
			goto out;
		}

		idx += len / 4;
	}

	if (allocated & ~created) {
		DRM_ERROR("New session without create command!\n");
		r = -ENOENT;
	}

out:
	if (!r) {
		/* No error, free all destroyed handle slots */
		tmp = destroyed;
	} else {
		/* Error during parsing, free all allocated handle slots */
		tmp = allocated;
	}

	for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i)
		if (tmp & (1 << i))
			atomic_set(&p->adev->vce.handles[i], 0);

	return r;
}

/**
 * amdgpu_vce_cs_parse_vm - parse the command stream in VM mode
 *
 * @p: parser context
 *
 */
int amdgpu_vce_ring_parse_cs_vm(struct amdgpu_cs_parser *p, uint32_t ib_idx)
{
	struct amdgpu_ib *ib = &p->job->ibs[ib_idx];
	int session_idx = -1;
	uint32_t destroyed = 0;
	uint32_t created = 0;
	uint32_t allocated = 0;
	uint32_t tmp, handle = 0;
	int i, r = 0, idx = 0;

	while (idx < ib->length_dw) {
		uint32_t len = amdgpu_get_ib_value(p, ib_idx, idx);
		uint32_t cmd = amdgpu_get_ib_value(p, ib_idx, idx + 1);

		if ((len < 8) || (len & 3)) {
			DRM_ERROR("invalid VCE command length (%d)!\n", len);
			r = -EINVAL;
			goto out;
		}

		switch (cmd) {
		case 0x00000001: /* session */
			handle = amdgpu_get_ib_value(p, ib_idx, idx + 2);
			session_idx = amdgpu_vce_validate_handle(p, handle,
								 &allocated);
			if (session_idx < 0) {
				r = session_idx;
				goto out;
			}
			break;

		case 0x01000001: /* create */
			created |= 1 << session_idx;
			if (destroyed & (1 << session_idx)) {
				destroyed &= ~(1 << session_idx);
				allocated |= 1 << session_idx;

			} else if (!(allocated & (1 << session_idx))) {
				DRM_ERROR("Handle already in use!\n");
				r = -EINVAL;
				goto out;
			}

			break;

		case 0x02000001: /* destroy */
			destroyed |= 1 << session_idx;
			break;

		default:
			break;
		}

		if (session_idx == -1) {
			DRM_ERROR("no session command at start of IB\n");
			r = -EINVAL;
			goto out;
		}

		idx += len / 4;
	}

	if (allocated & ~created) {
		DRM_ERROR("New session without create command!\n");
		r = -ENOENT;
	}

out:
	if (!r) {
		/* No error, free all destroyed handle slots */
		tmp = destroyed;
		amdgpu_ib_free(p->adev, ib, NULL);
	} else {
		/* Error during parsing, free all allocated handle slots */
		tmp = allocated;
	}

	for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i)
		if (tmp & (1 << i))
			atomic_set(&p->adev->vce.handles[i], 0);

	return r;
}

/**
 * amdgpu_vce_ring_emit_ib - execute indirect buffer
 *
 * @ring: engine to use
 * @ib: the IB to execute
 *
 */
void amdgpu_vce_ring_emit_ib(struct amdgpu_ring *ring, struct amdgpu_ib *ib,
			     unsigned vm_id, bool ctx_switch)
{
	amdgpu_ring_write(ring, VCE_CMD_IB);
	amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
	amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr));
	amdgpu_ring_write(ring, ib->length_dw);
}

/**
 * amdgpu_vce_ring_emit_fence - add a fence command to the ring
 *
 * @ring: engine to use
 * @fence: the fence
 *
 */
void amdgpu_vce_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
				unsigned flags)
{
	WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);

	amdgpu_ring_write(ring, VCE_CMD_FENCE);
	amdgpu_ring_write(ring, addr);
	amdgpu_ring_write(ring, upper_32_bits(addr));
	amdgpu_ring_write(ring, seq);
	amdgpu_ring_write(ring, VCE_CMD_TRAP);
	amdgpu_ring_write(ring, VCE_CMD_END);
}

/**
 * amdgpu_vce_ring_test_ring - test if VCE ring is working
 *
 * @ring: the engine to test on
 *
 */
int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring)
{
	struct amdgpu_device *adev = ring->adev;
	uint32_t rptr = amdgpu_ring_get_rptr(ring);
	unsigned i;
	int r;

	/* TODO: remove it if VCE can work for sriov */
	if (amdgpu_sriov_vf(adev))
		return 0;

	r = amdgpu_ring_alloc(ring, 16);
	if (r) {
		DRM_ERROR("amdgpu: vce failed to lock ring %d (%d).\n",
			  ring->idx, r);
		return r;
	}
	amdgpu_ring_write(ring, VCE_CMD_END);
	amdgpu_ring_commit(ring);

	for (i = 0; i < adev->usec_timeout; i++) {
		if (amdgpu_ring_get_rptr(ring) != rptr)
			break;
		DRM_UDELAY(1);
	}

	if (i < adev->usec_timeout) {
		DRM_INFO("ring test on %d succeeded in %d usecs\n",
			 ring->idx, i);
	} else {
		DRM_ERROR("amdgpu: ring %d test failed\n",
			  ring->idx);
		r = -ETIMEDOUT;
	}

	return r;
}

/**
 * amdgpu_vce_ring_test_ib - test if VCE IBs are working
 *
 * @ring: the engine to test on
 *
 */
int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout)
{
	struct dma_fence *fence = NULL;
	long r;

	/* TODO: remove it if VCE can work for sriov */
	if (amdgpu_sriov_vf(ring->adev))
		return 0;

	/* skip vce ring1/2 ib test for now, since it's not reliable */
	if (ring != &ring->adev->vce.ring[0])
		return 0;

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

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

	r = dma_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:
	dma_fence_put(fence);
	return r;
}