summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
blob: 6b4b30a8dce5ead531d534840d81328c990a8c62 (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
/*
 * Copyright 2018 Advanced Micro Devices, Inc.
 *
 * 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, sublicense,
 * 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 above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * 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 NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) 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.
 *
 */
#include "amdgpu.h"
#include "df_v3_6.h"

#include "df/df_3_6_default.h"
#include "df/df_3_6_offset.h"
#include "df/df_3_6_sh_mask.h"

#define DF_3_6_SMN_REG_INST_DIST        0x8
#define DF_3_6_INST_CNT                 8

/* Defined in global_features.h as FTI_PERFMON_VISIBLE */
#define DF_V3_6_MAX_COUNTERS		4

/* get flags from df perfmon config */
#define DF_V3_6_GET_EVENT(x)		(x & 0xFFUL)
#define DF_V3_6_GET_INSTANCE(x)		((x >> 8) & 0xFFUL)
#define DF_V3_6_GET_UNITMASK(x)		((x >> 16) & 0xFFUL)
#define DF_V3_6_PERFMON_OVERFLOW	0xFFFFFFFFFFFFULL

static u32 df_v3_6_channel_number[] = {1, 2, 0, 4, 0, 8, 0,
				       16, 32, 0, 0, 0, 2, 4, 8};

static uint64_t df_v3_6_get_fica(struct amdgpu_device *adev,
				 uint32_t ficaa_val)
{
	unsigned long flags, address, data;
	uint32_t ficadl_val, ficadh_val;

	address = adev->nbio.funcs->get_pcie_index_offset(adev);
	data = adev->nbio.funcs->get_pcie_data_offset(adev);

	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
	WREG32(address, smnDF_PIE_AON_FabricIndirectConfigAccessAddress3);
	WREG32(data, ficaa_val);

	WREG32(address, smnDF_PIE_AON_FabricIndirectConfigAccessDataLo3);
	ficadl_val = RREG32(data);

	WREG32(address, smnDF_PIE_AON_FabricIndirectConfigAccessDataHi3);
	ficadh_val = RREG32(data);

	spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);

	return (((ficadh_val & 0xFFFFFFFFFFFFFFFF) << 32) | ficadl_val);
}

static void df_v3_6_set_fica(struct amdgpu_device *adev, uint32_t ficaa_val,
			     uint32_t ficadl_val, uint32_t ficadh_val)
{
	unsigned long flags, address, data;

	address = adev->nbio.funcs->get_pcie_index_offset(adev);
	data = adev->nbio.funcs->get_pcie_data_offset(adev);

	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
	WREG32(address, smnDF_PIE_AON_FabricIndirectConfigAccessAddress3);
	WREG32(data, ficaa_val);

	WREG32(address, smnDF_PIE_AON_FabricIndirectConfigAccessDataLo3);
	WREG32(data, ficadl_val);

	WREG32(address, smnDF_PIE_AON_FabricIndirectConfigAccessDataHi3);
	WREG32(data, ficadh_val);

	spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
}

/*
 * df_v3_6_perfmon_rreg - read perfmon lo and hi
 *
 * required to be atomic.  no mmio method provided so subsequent reads for lo
 * and hi require to preserve df finite state machine
 */
static void df_v3_6_perfmon_rreg(struct amdgpu_device *adev,
			    uint32_t lo_addr, uint32_t *lo_val,
			    uint32_t hi_addr, uint32_t *hi_val)
{
	unsigned long flags, address, data;

	address = adev->nbio.funcs->get_pcie_index_offset(adev);
	data = adev->nbio.funcs->get_pcie_data_offset(adev);

	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
	WREG32(address, lo_addr);
	*lo_val = RREG32(data);
	WREG32(address, hi_addr);
	*hi_val = RREG32(data);
	spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
}

/*
 * df_v3_6_perfmon_wreg - write to perfmon lo and hi
 *
 * required to be atomic.  no mmio method provided so subsequent reads after
 * data writes cannot occur to preserve data fabrics finite state machine.
 */
static void df_v3_6_perfmon_wreg(struct amdgpu_device *adev, uint32_t lo_addr,
			    uint32_t lo_val, uint32_t hi_addr, uint32_t hi_val)
{
	unsigned long flags, address, data;

	address = adev->nbio.funcs->get_pcie_index_offset(adev);
	data = adev->nbio.funcs->get_pcie_data_offset(adev);

	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
	WREG32(address, lo_addr);
	WREG32(data, lo_val);
	WREG32(address, hi_addr);
	WREG32(data, hi_val);
	spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
}

/* same as perfmon_wreg but return status on write value check */
static int df_v3_6_perfmon_arm_with_status(struct amdgpu_device *adev,
					  uint32_t lo_addr, uint32_t lo_val,
					  uint32_t hi_addr, uint32_t  hi_val)
{
	unsigned long flags, address, data;
	uint32_t lo_val_rb, hi_val_rb;

	address = adev->nbio.funcs->get_pcie_index_offset(adev);
	data = adev->nbio.funcs->get_pcie_data_offset(adev);

	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
	WREG32(address, lo_addr);
	WREG32(data, lo_val);
	WREG32(address, hi_addr);
	WREG32(data, hi_val);

	WREG32(address, lo_addr);
	lo_val_rb = RREG32(data);
	WREG32(address, hi_addr);
	hi_val_rb = RREG32(data);
	spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);

	if (!(lo_val == lo_val_rb && hi_val == hi_val_rb))
		return -EBUSY;

	return 0;
}


/*
 * retry arming counters every 100 usecs within 1 millisecond interval.
 * if retry fails after time out, return error.
 */
#define ARM_RETRY_USEC_TIMEOUT	1000
#define ARM_RETRY_USEC_INTERVAL	100
static int df_v3_6_perfmon_arm_with_retry(struct amdgpu_device *adev,
					  uint32_t lo_addr, uint32_t lo_val,
					  uint32_t hi_addr, uint32_t  hi_val)
{
	int countdown = ARM_RETRY_USEC_TIMEOUT;

	while (countdown) {

		if (!df_v3_6_perfmon_arm_with_status(adev, lo_addr, lo_val,
						     hi_addr, hi_val))
			break;

		countdown -= ARM_RETRY_USEC_INTERVAL;
		udelay(ARM_RETRY_USEC_INTERVAL);
	}

	return countdown > 0 ? 0 : -ETIME;
}

/* get the number of df counters available */
static ssize_t df_v3_6_get_df_cntr_avail(struct device *dev,
		struct device_attribute *attr,
		char *buf)
{
	struct amdgpu_device *adev;
	struct drm_device *ddev;
	int i, count;

	ddev = dev_get_drvdata(dev);
	adev = drm_to_adev(ddev);
	count = 0;

	for (i = 0; i < DF_V3_6_MAX_COUNTERS; i++) {
		if (adev->df_perfmon_config_assign_mask[i] == 0)
			count++;
	}

	return snprintf(buf, PAGE_SIZE,	"%i\n", count);
}

/* device attr for available perfmon counters */
static DEVICE_ATTR(df_cntr_avail, S_IRUGO, df_v3_6_get_df_cntr_avail, NULL);

static void df_v3_6_query_hashes(struct amdgpu_device *adev)
{
	u32 tmp;

	adev->df.hash_status.hash_64k = false;
	adev->df.hash_status.hash_2m = false;
	adev->df.hash_status.hash_1g = false;

	if (adev->asic_type != CHIP_ARCTURUS)
		return;

	/* encoding for hash-enabled on Arcturus */
	if (adev->df.funcs->get_fb_channel_number(adev) == 0xe) {
		tmp = RREG32_SOC15(DF, 0, mmDF_CS_UMC_AON0_DfGlobalCtrl);
		adev->df.hash_status.hash_64k = REG_GET_FIELD(tmp,
						DF_CS_UMC_AON0_DfGlobalCtrl,
						GlbHashIntlvCtl64K);
		adev->df.hash_status.hash_2m = REG_GET_FIELD(tmp,
						DF_CS_UMC_AON0_DfGlobalCtrl,
						GlbHashIntlvCtl2M);
		adev->df.hash_status.hash_1g = REG_GET_FIELD(tmp,
						DF_CS_UMC_AON0_DfGlobalCtrl,
						GlbHashIntlvCtl1G);
	}
}

/* init perfmons */
static void df_v3_6_sw_init(struct amdgpu_device *adev)
{
	int i, ret;

	ret = device_create_file(adev->dev, &dev_attr_df_cntr_avail);
	if (ret)
		DRM_ERROR("failed to create file for available df counters\n");

	for (i = 0; i < AMDGPU_MAX_DF_PERFMONS; i++)
		adev->df_perfmon_config_assign_mask[i] = 0;

	df_v3_6_query_hashes(adev);
}

static void df_v3_6_sw_fini(struct amdgpu_device *adev)
{

	device_remove_file(adev->dev, &dev_attr_df_cntr_avail);

}

static void df_v3_6_enable_broadcast_mode(struct amdgpu_device *adev,
					  bool enable)
{
	u32 tmp;

	if (enable) {
		tmp = RREG32_SOC15(DF, 0, mmFabricConfigAccessControl);
		tmp &= ~FabricConfigAccessControl__CfgRegInstAccEn_MASK;
		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl, tmp);
	} else
		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl,
			     mmFabricConfigAccessControl_DEFAULT);
}

static u32 df_v3_6_get_fb_channel_number(struct amdgpu_device *adev)
{
	u32 tmp;

	tmp = RREG32_SOC15(DF, 0, mmDF_CS_UMC_AON0_DramBaseAddress0);
	tmp &= DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan_MASK;
	tmp >>= DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;

	return tmp;
}

static u32 df_v3_6_get_hbm_channel_number(struct amdgpu_device *adev)
{
	int fb_channel_number;

	fb_channel_number = adev->df.funcs->get_fb_channel_number(adev);
	if (fb_channel_number >= ARRAY_SIZE(df_v3_6_channel_number))
		fb_channel_number = 0;

	return df_v3_6_channel_number[fb_channel_number];
}

static void df_v3_6_update_medium_grain_clock_gating(struct amdgpu_device *adev,
						     bool enable)
{
	u32 tmp;

	if (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG) {
		/* Put DF on broadcast mode */
		adev->df.funcs->enable_broadcast_mode(adev, true);

		if (enable) {
			tmp = RREG32_SOC15(DF, 0,
					mmDF_PIE_AON0_DfGlobalClkGater);
			tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
			tmp |= DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY;
			WREG32_SOC15(DF, 0,
					mmDF_PIE_AON0_DfGlobalClkGater, tmp);
		} else {
			tmp = RREG32_SOC15(DF, 0,
					mmDF_PIE_AON0_DfGlobalClkGater);
			tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
			tmp |= DF_V3_6_MGCG_DISABLE;
			WREG32_SOC15(DF, 0,
					mmDF_PIE_AON0_DfGlobalClkGater, tmp);
		}

		/* Exit broadcast mode */
		adev->df.funcs->enable_broadcast_mode(adev, false);
	}
}

static void df_v3_6_get_clockgating_state(struct amdgpu_device *adev,
					  u32 *flags)
{
	u32 tmp;

	/* AMD_CG_SUPPORT_DF_MGCG */
	tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
	if (tmp & DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY)
		*flags |= AMD_CG_SUPPORT_DF_MGCG;
}

/* get assigned df perfmon ctr as int */
static bool df_v3_6_pmc_has_counter(struct amdgpu_device *adev,
				      uint64_t config,
				      int counter_idx)
{

	return ((config & 0x0FFFFFFUL) ==
			adev->df_perfmon_config_assign_mask[counter_idx]);

}

/* get address based on counter assignment */
static void df_v3_6_pmc_get_addr(struct amdgpu_device *adev,
				 uint64_t config,
				 int counter_idx,
				 int is_ctrl,
				 uint32_t *lo_base_addr,
				 uint32_t *hi_base_addr)
{
	if (!df_v3_6_pmc_has_counter(adev, config, counter_idx))
		return;

	switch (counter_idx) {

	case 0:
		*lo_base_addr = is_ctrl ? smnPerfMonCtlLo4 : smnPerfMonCtrLo4;
		*hi_base_addr = is_ctrl ? smnPerfMonCtlHi4 : smnPerfMonCtrHi4;
		break;
	case 1:
		*lo_base_addr = is_ctrl ? smnPerfMonCtlLo5 : smnPerfMonCtrLo5;
		*hi_base_addr = is_ctrl ? smnPerfMonCtlHi5 : smnPerfMonCtrHi5;
		break;
	case 2:
		*lo_base_addr = is_ctrl ? smnPerfMonCtlLo6 : smnPerfMonCtrLo6;
		*hi_base_addr = is_ctrl ? smnPerfMonCtlHi6 : smnPerfMonCtrHi6;
		break;
	case 3:
		*lo_base_addr = is_ctrl ? smnPerfMonCtlLo7 : smnPerfMonCtrLo7;
		*hi_base_addr = is_ctrl ? smnPerfMonCtlHi7 : smnPerfMonCtrHi7;
		break;

	}

}

/* get read counter address */
static void df_v3_6_pmc_get_read_settings(struct amdgpu_device *adev,
					  uint64_t config,
					  int counter_idx,
					  uint32_t *lo_base_addr,
					  uint32_t *hi_base_addr)
{
	df_v3_6_pmc_get_addr(adev, config, counter_idx, 0, lo_base_addr,
								hi_base_addr);
}

/* get control counter settings i.e. address and values to set */
static int df_v3_6_pmc_get_ctrl_settings(struct amdgpu_device *adev,
					  uint64_t config,
					  int counter_idx,
					  uint32_t *lo_base_addr,
					  uint32_t *hi_base_addr,
					  uint32_t *lo_val,
					  uint32_t *hi_val,
					  bool is_enable)
{

	uint32_t eventsel, instance, unitmask;
	uint32_t instance_10, instance_5432, instance_76;

	df_v3_6_pmc_get_addr(adev, config, counter_idx, 1, lo_base_addr,
				hi_base_addr);

	if ((*lo_base_addr == 0) || (*hi_base_addr == 0)) {
		DRM_ERROR("[DF PMC] addressing not retrieved! Lo: %x, Hi: %x",
				*lo_base_addr, *hi_base_addr);
		return -ENXIO;
	}

	eventsel = DF_V3_6_GET_EVENT(config) & 0x3f;
	unitmask = DF_V3_6_GET_UNITMASK(config) & 0xf;
	instance = DF_V3_6_GET_INSTANCE(config);

	instance_10 = instance & 0x3;
	instance_5432 = (instance >> 2) & 0xf;
	instance_76 = (instance >> 6) & 0x3;

	*lo_val = (unitmask << 8) | (instance_10 << 6) | eventsel;
	*lo_val = is_enable ? *lo_val | (1 << 22) : *lo_val & ~(1 << 22);
	*hi_val = (instance_76 << 29) | instance_5432;

	DRM_DEBUG_DRIVER("config=%llx addr=%08x:%08x val=%08x:%08x",
		config, *lo_base_addr, *hi_base_addr, *lo_val, *hi_val);

	return 0;
}

/* add df performance counters for read */
static int df_v3_6_pmc_add_cntr(struct amdgpu_device *adev,
				   uint64_t config)
{
	int i;

	for (i = 0; i < DF_V3_6_MAX_COUNTERS; i++) {
		if (adev->df_perfmon_config_assign_mask[i] == 0U) {
			adev->df_perfmon_config_assign_mask[i] =
							config & 0x0FFFFFFUL;
			return i;
		}
	}

	return -ENOSPC;
}

#define DEFERRED_ARM_MASK	(1 << 31)
static int df_v3_6_pmc_set_deferred(struct amdgpu_device *adev,
				    int counter_idx, uint64_t config,
				    bool is_deferred)
{

	if (!df_v3_6_pmc_has_counter(adev, config, counter_idx))
		return -EINVAL;

	if (is_deferred)
		adev->df_perfmon_config_assign_mask[counter_idx] |=
							DEFERRED_ARM_MASK;
	else
		adev->df_perfmon_config_assign_mask[counter_idx] &=
							~DEFERRED_ARM_MASK;

	return 0;
}

static bool df_v3_6_pmc_is_deferred(struct amdgpu_device *adev,
				    int counter_idx,
				    uint64_t config)
{
	return	(df_v3_6_pmc_has_counter(adev, config, counter_idx) &&
			(adev->df_perfmon_config_assign_mask[counter_idx]
				& DEFERRED_ARM_MASK));

}

/* release performance counter */
static void df_v3_6_pmc_release_cntr(struct amdgpu_device *adev,
				     uint64_t config,
				     int counter_idx)
{
	if (df_v3_6_pmc_has_counter(adev, config, counter_idx))
		adev->df_perfmon_config_assign_mask[counter_idx] = 0ULL;
}


static void df_v3_6_reset_perfmon_cntr(struct amdgpu_device *adev,
					 uint64_t config,
					 int counter_idx)
{
	uint32_t lo_base_addr = 0, hi_base_addr = 0;

	df_v3_6_pmc_get_read_settings(adev, config, counter_idx, &lo_base_addr,
				      &hi_base_addr);

	if ((lo_base_addr == 0) || (hi_base_addr == 0))
		return;

	df_v3_6_perfmon_wreg(adev, lo_base_addr, 0, hi_base_addr, 0);
}

/* return available counter if is_add == 1 otherwise return error status. */
static int df_v3_6_pmc_start(struct amdgpu_device *adev, uint64_t config,
			     int counter_idx, int is_add)
{
	uint32_t lo_base_addr, hi_base_addr, lo_val, hi_val;
	int err = 0, ret = 0;

	switch (adev->asic_type) {
	case CHIP_VEGA20:
	case CHIP_ARCTURUS:
		if (is_add)
			return df_v3_6_pmc_add_cntr(adev, config);

		ret = df_v3_6_pmc_get_ctrl_settings(adev,
					config,
					counter_idx,
					&lo_base_addr,
					&hi_base_addr,
					&lo_val,
					&hi_val,
					true);

		if (ret)
			return ret;

		err = df_v3_6_perfmon_arm_with_retry(adev,
						     lo_base_addr,
						     lo_val,
						     hi_base_addr,
						     hi_val);

		if (err)
			ret = df_v3_6_pmc_set_deferred(adev, config,
							counter_idx, true);

		break;
	default:
		break;
	}

	return ret;
}

static int df_v3_6_pmc_stop(struct amdgpu_device *adev, uint64_t config,
			    int counter_idx, int is_remove)
{
	uint32_t lo_base_addr, hi_base_addr, lo_val, hi_val;
	int ret = 0;

	switch (adev->asic_type) {
	case CHIP_VEGA20:
	case CHIP_ARCTURUS:
		ret = df_v3_6_pmc_get_ctrl_settings(adev,
			config,
			counter_idx,
			&lo_base_addr,
			&hi_base_addr,
			&lo_val,
			&hi_val,
			false);

		if (ret)
			return ret;


		if (is_remove) {
			df_v3_6_reset_perfmon_cntr(adev, config, counter_idx);
			df_v3_6_pmc_release_cntr(adev, config, counter_idx);
		}

		break;
	default:
		break;
	}

	return ret;
}

static void df_v3_6_pmc_get_count(struct amdgpu_device *adev,
				  uint64_t config,
				  int counter_idx,
				  uint64_t *count)
{
	uint32_t lo_base_addr = 0, hi_base_addr = 0, lo_val = 0, hi_val = 0;
	*count = 0;

	switch (adev->asic_type) {
	case CHIP_VEGA20:
	case CHIP_ARCTURUS:
		df_v3_6_pmc_get_read_settings(adev, config, counter_idx,
						&lo_base_addr, &hi_base_addr);

		if ((lo_base_addr == 0) || (hi_base_addr == 0))
			return;

		/* rearm the counter or throw away count value on failure */
		if (df_v3_6_pmc_is_deferred(adev, config, counter_idx)) {
			int rearm_err = df_v3_6_perfmon_arm_with_status(adev,
							lo_base_addr, lo_val,
							hi_base_addr, hi_val);

			if (rearm_err)
				return;

			df_v3_6_pmc_set_deferred(adev, config, counter_idx,
									false);
		}

		df_v3_6_perfmon_rreg(adev, lo_base_addr, &lo_val,
				hi_base_addr, &hi_val);

		*count  = ((hi_val | 0ULL) << 32) | (lo_val | 0ULL);

		if (*count >= DF_V3_6_PERFMON_OVERFLOW)
			*count = 0;

		DRM_DEBUG_DRIVER("config=%llx addr=%08x:%08x val=%08x:%08x",
			 config, lo_base_addr, hi_base_addr, lo_val, hi_val);

		break;
	default:
		break;
	}
}

const struct amdgpu_df_funcs df_v3_6_funcs = {
	.sw_init = df_v3_6_sw_init,
	.sw_fini = df_v3_6_sw_fini,
	.enable_broadcast_mode = df_v3_6_enable_broadcast_mode,
	.get_fb_channel_number = df_v3_6_get_fb_channel_number,
	.get_hbm_channel_number = df_v3_6_get_hbm_channel_number,
	.update_medium_grain_clock_gating =
			df_v3_6_update_medium_grain_clock_gating,
	.get_clockgating_state = df_v3_6_get_clockgating_state,
	.pmc_start = df_v3_6_pmc_start,
	.pmc_stop = df_v3_6_pmc_stop,
	.pmc_get_count = df_v3_6_pmc_get_count,
	.get_fica = df_v3_6_get_fica,
	.set_fica = df_v3_6_set_fica,
};