uvd_v6_0.c 32.4 KB
Newer Older
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
/*
 * Copyright 2014 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.
 *
 * Authors: Christian König <christian.koenig@amd.com>
 */

#include <linux/firmware.h>
#include <drm/drmP.h>
#include "amdgpu.h"
#include "amdgpu_uvd.h"
#include "vid.h"
#include "uvd/uvd_6_0_d.h"
#include "uvd/uvd_6_0_sh_mask.h"
#include "oss/oss_2_0_d.h"
#include "oss/oss_2_0_sh_mask.h"
34 35
#include "smu/smu_7_1_3_d.h"
#include "smu/smu_7_1_3_sh_mask.h"
36
#include "bif/bif_5_1_d.h"
37
#include "gmc/gmc_8_1_d.h"
38
#include "vi.h"
39 40 41 42 43

static void uvd_v6_0_set_ring_funcs(struct amdgpu_device *adev);
static void uvd_v6_0_set_irq_funcs(struct amdgpu_device *adev);
static int uvd_v6_0_start(struct amdgpu_device *adev);
static void uvd_v6_0_stop(struct amdgpu_device *adev);
44
static void uvd_v6_0_set_sw_clock_gating(struct amdgpu_device *adev);
45 46 47 48
static int uvd_v6_0_set_clockgating_state(void *handle,
					  enum amd_clockgating_state state);
static void uvd_v6_0_enable_mgcg(struct amdgpu_device *adev,
				 bool enable);
49 50 51 52 53 54 55 56

/**
 * uvd_v6_0_ring_get_rptr - get read pointer
 *
 * @ring: amdgpu_ring pointer
 *
 * Returns the current hardware read pointer
 */
57
static uint64_t uvd_v6_0_ring_get_rptr(struct amdgpu_ring *ring)
58 59 60 61 62 63 64 65 66 67 68 69 70
{
	struct amdgpu_device *adev = ring->adev;

	return RREG32(mmUVD_RBC_RB_RPTR);
}

/**
 * uvd_v6_0_ring_get_wptr - get write pointer
 *
 * @ring: amdgpu_ring pointer
 *
 * Returns the current hardware write pointer
 */
71
static uint64_t uvd_v6_0_ring_get_wptr(struct amdgpu_ring *ring)
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
{
	struct amdgpu_device *adev = ring->adev;

	return RREG32(mmUVD_RBC_RB_WPTR);
}

/**
 * uvd_v6_0_ring_set_wptr - set write pointer
 *
 * @ring: amdgpu_ring pointer
 *
 * Commits the write pointer to the hardware
 */
static void uvd_v6_0_ring_set_wptr(struct amdgpu_ring *ring)
{
	struct amdgpu_device *adev = ring->adev;

89
	WREG32(mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr));
90 91
}

92
static int uvd_v6_0_early_init(void *handle)
93
{
94 95
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

96 97 98 99 100 101
	uvd_v6_0_set_ring_funcs(adev);
	uvd_v6_0_set_irq_funcs(adev);

	return 0;
}

102
static int uvd_v6_0_sw_init(void *handle)
103 104 105
{
	struct amdgpu_ring *ring;
	int r;
106
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
107 108

	/* UVD TRAP */
109
	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.irq);
110 111 112 113 114 115 116 117 118 119 120 121 122
	if (r)
		return r;

	r = amdgpu_uvd_sw_init(adev);
	if (r)
		return r;

	r = amdgpu_uvd_resume(adev);
	if (r)
		return r;

	ring = &adev->uvd.ring;
	sprintf(ring->name, "uvd");
123
	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
124 125 126 127

	return r;
}

128
static int uvd_v6_0_sw_fini(void *handle)
129 130
{
	int r;
131
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150

	r = amdgpu_uvd_suspend(adev);
	if (r)
		return r;

	r = amdgpu_uvd_sw_fini(adev);
	if (r)
		return r;

	return r;
}

/**
 * uvd_v6_0_hw_init - start and test UVD block
 *
 * @adev: amdgpu_device pointer
 *
 * Initialize the hardware, boot up the VCPU and do some testing
 */
151
static int uvd_v6_0_hw_init(void *handle)
152
{
153
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
154 155 156 157
	struct amdgpu_ring *ring = &adev->uvd.ring;
	uint32_t tmp;
	int r;

R
Rex Zhu 已提交
158 159 160
	amdgpu_asic_set_uvd_clocks(adev, 10000, 10000);
	uvd_v6_0_set_clockgating_state(adev, AMD_CG_STATE_UNGATE);
	uvd_v6_0_enable_mgcg(adev, true);
161 162 163 164 165 166 167 168

	ring->ready = true;
	r = amdgpu_ring_test_ring(ring);
	if (r) {
		ring->ready = false;
		goto done;
	}

169
	r = amdgpu_ring_alloc(ring, 10);
170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193
	if (r) {
		DRM_ERROR("amdgpu: ring failed to lock UVD ring (%d).\n", r);
		goto done;
	}

	tmp = PACKET0(mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL, 0);
	amdgpu_ring_write(ring, tmp);
	amdgpu_ring_write(ring, 0xFFFFF);

	tmp = PACKET0(mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL, 0);
	amdgpu_ring_write(ring, tmp);
	amdgpu_ring_write(ring, 0xFFFFF);

	tmp = PACKET0(mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL, 0);
	amdgpu_ring_write(ring, tmp);
	amdgpu_ring_write(ring, 0xFFFFF);

	/* Clear timeout status bits */
	amdgpu_ring_write(ring, PACKET0(mmUVD_SEMA_TIMEOUT_STATUS, 0));
	amdgpu_ring_write(ring, 0x8);

	amdgpu_ring_write(ring, PACKET0(mmUVD_SEMA_CNTL, 0));
	amdgpu_ring_write(ring, 3);

194
	amdgpu_ring_commit(ring);
195 196 197 198 199 200 201 202 203 204 205 206 207 208 209

done:
	if (!r)
		DRM_INFO("UVD initialized successfully.\n");

	return r;
}

/**
 * uvd_v6_0_hw_fini - stop the hardware block
 *
 * @adev: amdgpu_device pointer
 *
 * Stop the UVD block, mark ring as not ready any more
 */
210
static int uvd_v6_0_hw_fini(void *handle)
211
{
212
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
213 214
	struct amdgpu_ring *ring = &adev->uvd.ring;

R
Rex Zhu 已提交
215 216 217
	if (RREG32(mmUVD_STATUS) != 0)
		uvd_v6_0_stop(adev);

218 219 220 221 222
	ring->ready = false;

	return 0;
}

223
static int uvd_v6_0_suspend(void *handle)
224 225
{
	int r;
226
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
227

228 229 230 231
	r = uvd_v6_0_hw_fini(adev);
	if (r)
		return r;

232 233 234 235 236 237
	/* Skip this for APU for now */
	if (!(adev->flags & AMD_IS_APU)) {
		r = amdgpu_uvd_suspend(adev);
		if (r)
			return r;
	}
238 239 240 241

	return r;
}

242
static int uvd_v6_0_resume(void *handle)
243 244
{
	int r;
245
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
246

247 248 249 250 251 252
	/* Skip this for APU for now */
	if (!(adev->flags & AMD_IS_APU)) {
		r = amdgpu_uvd_resume(adev);
		if (r)
			return r;
	}
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
	r = uvd_v6_0_hw_init(adev);
	if (r)
		return r;

	return r;
}

/**
 * uvd_v6_0_mc_resume - memory controller programming
 *
 * @adev: amdgpu_device pointer
 *
 * Let the UVD memory controller know it's offsets
 */
static void uvd_v6_0_mc_resume(struct amdgpu_device *adev)
{
	uint64_t offset;
	uint32_t size;

	/* programm memory controller bits 0-27 */
	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
			lower_32_bits(adev->uvd.gpu_addr));
	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
			upper_32_bits(adev->uvd.gpu_addr));

	offset = AMDGPU_UVD_FIRMWARE_OFFSET;
	size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4);
	WREG32(mmUVD_VCPU_CACHE_OFFSET0, offset >> 3);
	WREG32(mmUVD_VCPU_CACHE_SIZE0, size);

	offset += size;
284
	size = AMDGPU_UVD_HEAP_SIZE;
285 286 287 288
	WREG32(mmUVD_VCPU_CACHE_OFFSET1, offset >> 3);
	WREG32(mmUVD_VCPU_CACHE_SIZE1, size);

	offset += size;
289 290
	size = AMDGPU_UVD_STACK_SIZE +
	       (AMDGPU_UVD_SESSION_SIZE * adev->uvd.max_handles);
291 292
	WREG32(mmUVD_VCPU_CACHE_OFFSET2, offset >> 3);
	WREG32(mmUVD_VCPU_CACHE_SIZE2, size);
293 294 295 296

	WREG32(mmUVD_UDEC_ADDR_CONFIG, adev->gfx.config.gb_addr_config);
	WREG32(mmUVD_UDEC_DB_ADDR_CONFIG, adev->gfx.config.gb_addr_config);
	WREG32(mmUVD_UDEC_DBW_ADDR_CONFIG, adev->gfx.config.gb_addr_config);
297 298

	WREG32(mmUVD_GP_SCRATCH4, adev->uvd.max_handles);
299 300
}

301
#if 0
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
static void cz_set_uvd_clock_gating_branches(struct amdgpu_device *adev,
		bool enable)
{
	u32 data, data1;

	data = RREG32(mmUVD_CGC_GATE);
	data1 = RREG32(mmUVD_SUVD_CGC_GATE);
	if (enable) {
		data |= UVD_CGC_GATE__SYS_MASK |
				UVD_CGC_GATE__UDEC_MASK |
				UVD_CGC_GATE__MPEG2_MASK |
				UVD_CGC_GATE__RBC_MASK |
				UVD_CGC_GATE__LMI_MC_MASK |
				UVD_CGC_GATE__IDCT_MASK |
				UVD_CGC_GATE__MPRD_MASK |
				UVD_CGC_GATE__MPC_MASK |
				UVD_CGC_GATE__LBSI_MASK |
				UVD_CGC_GATE__LRBBM_MASK |
				UVD_CGC_GATE__UDEC_RE_MASK |
				UVD_CGC_GATE__UDEC_CM_MASK |
				UVD_CGC_GATE__UDEC_IT_MASK |
				UVD_CGC_GATE__UDEC_DB_MASK |
				UVD_CGC_GATE__UDEC_MP_MASK |
				UVD_CGC_GATE__WCB_MASK |
				UVD_CGC_GATE__VCPU_MASK |
				UVD_CGC_GATE__SCPU_MASK;
		data1 |= UVD_SUVD_CGC_GATE__SRE_MASK |
				UVD_SUVD_CGC_GATE__SIT_MASK |
				UVD_SUVD_CGC_GATE__SMP_MASK |
				UVD_SUVD_CGC_GATE__SCM_MASK |
				UVD_SUVD_CGC_GATE__SDB_MASK |
				UVD_SUVD_CGC_GATE__SRE_H264_MASK |
				UVD_SUVD_CGC_GATE__SRE_HEVC_MASK |
				UVD_SUVD_CGC_GATE__SIT_H264_MASK |
				UVD_SUVD_CGC_GATE__SIT_HEVC_MASK |
				UVD_SUVD_CGC_GATE__SCM_H264_MASK |
				UVD_SUVD_CGC_GATE__SCM_HEVC_MASK |
				UVD_SUVD_CGC_GATE__SDB_H264_MASK |
				UVD_SUVD_CGC_GATE__SDB_HEVC_MASK;
	} else {
		data &= ~(UVD_CGC_GATE__SYS_MASK |
				UVD_CGC_GATE__UDEC_MASK |
				UVD_CGC_GATE__MPEG2_MASK |
				UVD_CGC_GATE__RBC_MASK |
				UVD_CGC_GATE__LMI_MC_MASK |
				UVD_CGC_GATE__LMI_UMC_MASK |
				UVD_CGC_GATE__IDCT_MASK |
				UVD_CGC_GATE__MPRD_MASK |
				UVD_CGC_GATE__MPC_MASK |
				UVD_CGC_GATE__LBSI_MASK |
				UVD_CGC_GATE__LRBBM_MASK |
				UVD_CGC_GATE__UDEC_RE_MASK |
				UVD_CGC_GATE__UDEC_CM_MASK |
				UVD_CGC_GATE__UDEC_IT_MASK |
				UVD_CGC_GATE__UDEC_DB_MASK |
				UVD_CGC_GATE__UDEC_MP_MASK |
				UVD_CGC_GATE__WCB_MASK |
				UVD_CGC_GATE__VCPU_MASK |
				UVD_CGC_GATE__SCPU_MASK);
		data1 &= ~(UVD_SUVD_CGC_GATE__SRE_MASK |
				UVD_SUVD_CGC_GATE__SIT_MASK |
				UVD_SUVD_CGC_GATE__SMP_MASK |
				UVD_SUVD_CGC_GATE__SCM_MASK |
				UVD_SUVD_CGC_GATE__SDB_MASK |
				UVD_SUVD_CGC_GATE__SRE_H264_MASK |
				UVD_SUVD_CGC_GATE__SRE_HEVC_MASK |
				UVD_SUVD_CGC_GATE__SIT_H264_MASK |
				UVD_SUVD_CGC_GATE__SIT_HEVC_MASK |
				UVD_SUVD_CGC_GATE__SCM_H264_MASK |
				UVD_SUVD_CGC_GATE__SCM_HEVC_MASK |
				UVD_SUVD_CGC_GATE__SDB_H264_MASK |
				UVD_SUVD_CGC_GATE__SDB_HEVC_MASK);
	}
	WREG32(mmUVD_CGC_GATE, data);
	WREG32(mmUVD_SUVD_CGC_GATE, data1);
}
378
#endif
379

380 381 382 383 384 385 386 387 388 389 390 391 392 393 394
/**
 * uvd_v6_0_start - start UVD block
 *
 * @adev: amdgpu_device pointer
 *
 * Setup and start the UVD block
 */
static int uvd_v6_0_start(struct amdgpu_device *adev)
{
	struct amdgpu_ring *ring = &adev->uvd.ring;
	uint32_t rb_bufsz, tmp;
	uint32_t lmi_swap_cntl;
	uint32_t mp_swap_cntl;
	int i, j, r;

395 396
	/* disable DPG */
	WREG32_P(mmUVD_POWER_STATUS, 0, ~UVD_POWER_STATUS__UVD_PG_MODE_MASK);
397 398 399 400 401 402 403 404

	/* disable byte swapping */
	lmi_swap_cntl = 0;
	mp_swap_cntl = 0;

	uvd_v6_0_mc_resume(adev);

	/* disable interupt */
405
	WREG32_FIELD(UVD_MASTINT_EN, VCPU_EN, 0);
406 407

	/* stall UMC and register bus before resetting VCPU */
408
	WREG32_FIELD(UVD_LMI_CTRL2, STALL_ARB_UMC, 1);
409 410 411
	mdelay(1);

	/* put LMI, VCPU, RBC etc... into reset */
412 413 414 415 416 417 418 419
	WREG32(mmUVD_SOFT_RESET,
		UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
		UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
		UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
		UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
		UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
		UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
		UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
420 421 422 423
		UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK);
	mdelay(5);

	/* take UVD block out of reset */
424
	WREG32_FIELD(SRBM_SOFT_RESET, SOFT_RESET_UVD, 0);
425 426 427
	mdelay(5);

	/* initialize UVD memory controller */
428 429 430 431 432 433 434
	WREG32(mmUVD_LMI_CTRL,
		(0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
		UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
		UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
		UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
		UVD_LMI_CTRL__REQ_MODE_MASK |
		UVD_LMI_CTRL__DISABLE_ON_FWV_FAIL_MASK);
435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455

#ifdef __BIG_ENDIAN
	/* swap (8 in 32) RB and IB */
	lmi_swap_cntl = 0xa;
	mp_swap_cntl = 0;
#endif
	WREG32(mmUVD_LMI_SWAP_CNTL, lmi_swap_cntl);
	WREG32(mmUVD_MP_SWAP_CNTL, mp_swap_cntl);

	WREG32(mmUVD_MPC_SET_MUXA0, 0x40c2040);
	WREG32(mmUVD_MPC_SET_MUXA1, 0x0);
	WREG32(mmUVD_MPC_SET_MUXB0, 0x40c2040);
	WREG32(mmUVD_MPC_SET_MUXB1, 0x0);
	WREG32(mmUVD_MPC_SET_ALU, 0);
	WREG32(mmUVD_MPC_SET_MUX, 0x88);

	/* take all subblocks out of reset, except VCPU */
	WREG32(mmUVD_SOFT_RESET, UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
	mdelay(5);

	/* enable VCPU clock */
456
	WREG32(mmUVD_VCPU_CNTL, UVD_VCPU_CNTL__CLK_EN_MASK);
457 458

	/* enable UMC */
459
	WREG32_FIELD(UVD_LMI_CTRL2, STALL_ARB_UMC, 0);
460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478

	/* boot up the VCPU */
	WREG32(mmUVD_SOFT_RESET, 0);
	mdelay(10);

	for (i = 0; i < 10; ++i) {
		uint32_t status;

		for (j = 0; j < 100; ++j) {
			status = RREG32(mmUVD_STATUS);
			if (status & 2)
				break;
			mdelay(10);
		}
		r = 0;
		if (status & 2)
			break;

		DRM_ERROR("UVD not responding, trying to reset the VCPU!!!\n");
479
		WREG32_FIELD(UVD_SOFT_RESET, VCPU_SOFT_RESET, 1);
480
		mdelay(10);
481
		WREG32_FIELD(UVD_SOFT_RESET, VCPU_SOFT_RESET, 0);
482 483 484 485 486 487 488 489 490
		mdelay(10);
		r = -1;
	}

	if (r) {
		DRM_ERROR("UVD not responding, giving up!!!\n");
		return r;
	}
	/* enable master interrupt */
491 492 493
	WREG32_P(mmUVD_MASTINT_EN,
		(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
		~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
494 495

	/* clear the bit 4 of UVD_STATUS */
496
	WREG32_P(mmUVD_STATUS, 0, ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT));
497

498
	/* force RBC into idle state */
499
	rb_bufsz = order_base_2(ring->ring_size);
500
	tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, rb_bufsz);
501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523
	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BLKSZ, 1);
	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_WPTR_POLL_EN, 0);
	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1);
	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1);
	WREG32(mmUVD_RBC_RB_CNTL, tmp);

	/* set the write pointer delay */
	WREG32(mmUVD_RBC_RB_WPTR_CNTL, 0);

	/* set the wb address */
	WREG32(mmUVD_RBC_RB_RPTR_ADDR, (upper_32_bits(ring->gpu_addr) >> 2));

	/* programm the RB_BASE for ring buffer */
	WREG32(mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
			lower_32_bits(ring->gpu_addr));
	WREG32(mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
			upper_32_bits(ring->gpu_addr));

	/* Initialize the ring buffer's read and write pointers */
	WREG32(mmUVD_RBC_RB_RPTR, 0);

	ring->wptr = RREG32(mmUVD_RBC_RB_RPTR);
524
	WREG32(mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr));
525

526
	WREG32_FIELD(UVD_RBC_RB_CNTL, RB_NO_FETCH, 0);
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

	return 0;
}

/**
 * uvd_v6_0_stop - stop UVD block
 *
 * @adev: amdgpu_device pointer
 *
 * stop the UVD block
 */
static void uvd_v6_0_stop(struct amdgpu_device *adev)
{
	/* force RBC into idle state */
	WREG32(mmUVD_RBC_RB_CNTL, 0x11010101);

	/* Stall UMC and register bus before resetting VCPU */
	WREG32_P(mmUVD_LMI_CTRL2, 1 << 8, ~(1 << 8));
	mdelay(1);

	/* put VCPU into reset */
	WREG32(mmUVD_SOFT_RESET, UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
	mdelay(5);

	/* disable VCPU clock */
	WREG32(mmUVD_VCPU_CNTL, 0x0);

	/* Unstall UMC and register bus */
	WREG32_P(mmUVD_LMI_CTRL2, 0, ~(1 << 8));
R
Rex Zhu 已提交
556 557

	WREG32(mmUVD_STATUS, 0);
558 559 560 561 562 563 564 565 566 567 568
}

/**
 * uvd_v6_0_ring_emit_fence - emit an fence & trap command
 *
 * @ring: amdgpu_ring pointer
 * @fence: fence to emit
 *
 * Write a fence and a trap command to the ring.
 */
static void uvd_v6_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
569
				     unsigned flags)
570
{
571
	WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589

	amdgpu_ring_write(ring, PACKET0(mmUVD_CONTEXT_ID, 0));
	amdgpu_ring_write(ring, seq);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0));
	amdgpu_ring_write(ring, addr & 0xffffffff);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0));
	amdgpu_ring_write(ring, upper_32_bits(addr) & 0xff);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD, 0));
	amdgpu_ring_write(ring, 0);

	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0));
	amdgpu_ring_write(ring, 0);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0));
	amdgpu_ring_write(ring, 0);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD, 0));
	amdgpu_ring_write(ring, 2);
}

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
/**
 * uvd_v6_0_ring_emit_hdp_flush - emit an hdp flush
 *
 * @ring: amdgpu_ring pointer
 *
 * Emits an hdp flush.
 */
static void uvd_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
{
	amdgpu_ring_write(ring, PACKET0(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0));
	amdgpu_ring_write(ring, 0);
}

/**
 * uvd_v6_0_ring_hdp_invalidate - emit an hdp invalidate
 *
 * @ring: amdgpu_ring pointer
 *
 * Emits an hdp invalidate.
 */
static void uvd_v6_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring)
{
	amdgpu_ring_write(ring, PACKET0(mmHDP_DEBUG0, 0));
	amdgpu_ring_write(ring, 1);
}

616 617 618 619 620 621 622 623 624 625 626 627 628 629 630
/**
 * uvd_v6_0_ring_test_ring - register write test
 *
 * @ring: amdgpu_ring pointer
 *
 * Test if we can successfully write to the context register
 */
static int uvd_v6_0_ring_test_ring(struct amdgpu_ring *ring)
{
	struct amdgpu_device *adev = ring->adev;
	uint32_t tmp = 0;
	unsigned i;
	int r;

	WREG32(mmUVD_CONTEXT_ID, 0xCAFEDEAD);
631
	r = amdgpu_ring_alloc(ring, 3);
632 633 634 635 636 637 638
	if (r) {
		DRM_ERROR("amdgpu: cp failed to lock ring %d (%d).\n",
			  ring->idx, r);
		return r;
	}
	amdgpu_ring_write(ring, PACKET0(mmUVD_CONTEXT_ID, 0));
	amdgpu_ring_write(ring, 0xDEADBEEF);
639
	amdgpu_ring_commit(ring);
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
	for (i = 0; i < adev->usec_timeout; i++) {
		tmp = RREG32(mmUVD_CONTEXT_ID);
		if (tmp == 0xDEADBEEF)
			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 (0x%08X)\n",
			  ring->idx, tmp);
		r = -EINVAL;
	}
	return r;
}

/**
 * uvd_v6_0_ring_emit_ib - execute indirect buffer
 *
 * @ring: amdgpu_ring pointer
 * @ib: indirect buffer to execute
 *
 * Write ring commands to execute the indirect buffer
 */
static void uvd_v6_0_ring_emit_ib(struct amdgpu_ring *ring,
667 668
				  struct amdgpu_ib *ib,
				  unsigned vm_id, bool ctx_switch)
669
{
670 671 672
	amdgpu_ring_write(ring, PACKET0(mmUVD_LMI_RBC_IB_VMID, 0));
	amdgpu_ring_write(ring, vm_id);

673 674 675 676 677 678 679 680
	amdgpu_ring_write(ring, PACKET0(mmUVD_LMI_RBC_IB_64BIT_BAR_LOW, 0));
	amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
	amdgpu_ring_write(ring, PACKET0(mmUVD_LMI_RBC_IB_64BIT_BAR_HIGH, 0));
	amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr));
	amdgpu_ring_write(ring, PACKET0(mmUVD_RBC_IB_SIZE, 0));
	amdgpu_ring_write(ring, ib->length_dw);
}

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
static void uvd_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
					 unsigned vm_id, uint64_t pd_addr)
{
	uint32_t reg;

	if (vm_id < 8)
		reg = mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vm_id;
	else
		reg = mmVM_CONTEXT8_PAGE_TABLE_BASE_ADDR + vm_id - 8;

	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0));
	amdgpu_ring_write(ring, reg << 2);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0));
	amdgpu_ring_write(ring, pd_addr >> 12);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD, 0));
	amdgpu_ring_write(ring, 0x8);

	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0));
	amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST << 2);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0));
	amdgpu_ring_write(ring, 1 << vm_id);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD, 0));
	amdgpu_ring_write(ring, 0x8);

	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0));
	amdgpu_ring_write(ring, mmVM_INVALIDATE_REQUEST << 2);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0));
	amdgpu_ring_write(ring, 0);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GP_SCRATCH8, 0));
	amdgpu_ring_write(ring, 1 << vm_id); /* mask */
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD, 0));
	amdgpu_ring_write(ring, 0xC);
}

static void uvd_v6_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
{
	uint32_t seq = ring->fence_drv.sync_seq;
	uint64_t addr = ring->fence_drv.gpu_addr;

	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0));
	amdgpu_ring_write(ring, lower_32_bits(addr));
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0));
	amdgpu_ring_write(ring, upper_32_bits(addr));
	amdgpu_ring_write(ring, PACKET0(mmUVD_GP_SCRATCH8, 0));
	amdgpu_ring_write(ring, 0xffffffff); /* mask */
	amdgpu_ring_write(ring, PACKET0(mmUVD_GP_SCRATCH9, 0));
	amdgpu_ring_write(ring, seq);
	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD, 0));
	amdgpu_ring_write(ring, 0xE);
}

732
static bool uvd_v6_0_is_idle(void *handle)
733
{
734 735
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

736 737 738
	return !(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK);
}

739
static int uvd_v6_0_wait_for_idle(void *handle)
740 741
{
	unsigned i;
742
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
743 744

	for (i = 0; i < adev->usec_timeout; i++) {
745
		if (uvd_v6_0_is_idle(handle))
746 747 748 749 750
			return 0;
	}
	return -ETIMEDOUT;
}

751
#define AMDGPU_UVD_STATUS_BUSY_MASK    0xfd
752
static bool uvd_v6_0_check_soft_reset(void *handle)
753 754 755 756 757 758 759 760 761 762 763 764
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
	u32 srbm_soft_reset = 0;
	u32 tmp = RREG32(mmSRBM_STATUS);

	if (REG_GET_FIELD(tmp, SRBM_STATUS, UVD_RQ_PENDING) ||
	    REG_GET_FIELD(tmp, SRBM_STATUS, UVD_BUSY) ||
	    (RREG32(mmUVD_STATUS) & AMDGPU_UVD_STATUS_BUSY_MASK))
		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_UVD, 1);

	if (srbm_soft_reset) {
		adev->uvd.srbm_soft_reset = srbm_soft_reset;
765
		return true;
766 767
	} else {
		adev->uvd.srbm_soft_reset = 0;
768
		return false;
769 770
	}
}
771

772
static int uvd_v6_0_pre_soft_reset(void *handle)
773
{
774 775
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

776
	if (!adev->uvd.srbm_soft_reset)
777 778
		return 0;

779
	uvd_v6_0_stop(adev);
780 781 782 783 784 785 786 787
	return 0;
}

static int uvd_v6_0_soft_reset(void *handle)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
	u32 srbm_soft_reset;

788
	if (!adev->uvd.srbm_soft_reset)
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
		return 0;
	srbm_soft_reset = adev->uvd.srbm_soft_reset;

	if (srbm_soft_reset) {
		u32 tmp;

		tmp = RREG32(mmSRBM_SOFT_RESET);
		tmp |= srbm_soft_reset;
		dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp);
		WREG32(mmSRBM_SOFT_RESET, tmp);
		tmp = RREG32(mmSRBM_SOFT_RESET);

		udelay(50);

		tmp &= ~srbm_soft_reset;
		WREG32(mmSRBM_SOFT_RESET, tmp);
		tmp = RREG32(mmSRBM_SOFT_RESET);

		/* Wait a little for things to settle down */
		udelay(50);
	}

	return 0;
}

static int uvd_v6_0_post_soft_reset(void *handle)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

818
	if (!adev->uvd.srbm_soft_reset)
819
		return 0;
820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843

	mdelay(5);

	return uvd_v6_0_start(adev);
}

static int uvd_v6_0_set_interrupt_state(struct amdgpu_device *adev,
					struct amdgpu_irq_src *source,
					unsigned type,
					enum amdgpu_interrupt_state state)
{
	// TODO
	return 0;
}

static int uvd_v6_0_process_interrupt(struct amdgpu_device *adev,
				      struct amdgpu_irq_src *source,
				      struct amdgpu_iv_entry *entry)
{
	DRM_DEBUG("IH: UVD TRAP\n");
	amdgpu_fence_process(&adev->uvd.ring);
	return 0;
}

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
static void uvd_v6_0_enable_clock_gating(struct amdgpu_device *adev, bool enable)
{
	uint32_t data1, data3;

	data1 = RREG32(mmUVD_SUVD_CGC_GATE);
	data3 = RREG32(mmUVD_CGC_GATE);

	data1 |= UVD_SUVD_CGC_GATE__SRE_MASK |
		     UVD_SUVD_CGC_GATE__SIT_MASK |
		     UVD_SUVD_CGC_GATE__SMP_MASK |
		     UVD_SUVD_CGC_GATE__SCM_MASK |
		     UVD_SUVD_CGC_GATE__SDB_MASK |
		     UVD_SUVD_CGC_GATE__SRE_H264_MASK |
		     UVD_SUVD_CGC_GATE__SRE_HEVC_MASK |
		     UVD_SUVD_CGC_GATE__SIT_H264_MASK |
		     UVD_SUVD_CGC_GATE__SIT_HEVC_MASK |
		     UVD_SUVD_CGC_GATE__SCM_H264_MASK |
		     UVD_SUVD_CGC_GATE__SCM_HEVC_MASK |
		     UVD_SUVD_CGC_GATE__SDB_H264_MASK |
		     UVD_SUVD_CGC_GATE__SDB_HEVC_MASK;

	if (enable) {
		data3 |= (UVD_CGC_GATE__SYS_MASK       |
			UVD_CGC_GATE__UDEC_MASK      |
			UVD_CGC_GATE__MPEG2_MASK     |
			UVD_CGC_GATE__RBC_MASK       |
			UVD_CGC_GATE__LMI_MC_MASK    |
			UVD_CGC_GATE__LMI_UMC_MASK   |
			UVD_CGC_GATE__IDCT_MASK      |
			UVD_CGC_GATE__MPRD_MASK      |
			UVD_CGC_GATE__MPC_MASK       |
			UVD_CGC_GATE__LBSI_MASK      |
			UVD_CGC_GATE__LRBBM_MASK     |
			UVD_CGC_GATE__UDEC_RE_MASK   |
			UVD_CGC_GATE__UDEC_CM_MASK   |
			UVD_CGC_GATE__UDEC_IT_MASK   |
			UVD_CGC_GATE__UDEC_DB_MASK   |
			UVD_CGC_GATE__UDEC_MP_MASK   |
			UVD_CGC_GATE__WCB_MASK       |
			UVD_CGC_GATE__JPEG_MASK      |
			UVD_CGC_GATE__SCPU_MASK      |
			UVD_CGC_GATE__JPEG2_MASK);
886 887 888 889
		/* only in pg enabled, we can gate clock to vcpu*/
		if (adev->pg_flags & AMD_PG_SUPPORT_UVD)
			data3 |= UVD_CGC_GATE__VCPU_MASK;

890 891 892 893 894 895 896 897 898
		data3 &= ~UVD_CGC_GATE__REGS_MASK;
	} else {
		data3 = 0;
	}

	WREG32(mmUVD_SUVD_CGC_GATE, data1);
	WREG32(mmUVD_CGC_GATE, data3);
}

899 900
static void uvd_v6_0_set_sw_clock_gating(struct amdgpu_device *adev)
{
901
	uint32_t data, data2;
902 903 904 905

	data = RREG32(mmUVD_CGC_CTRL);
	data2 = RREG32(mmUVD_SUVD_CGC_CTRL);

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
	data &= ~(UVD_CGC_CTRL__CLK_OFF_DELAY_MASK |
		  UVD_CGC_CTRL__CLK_GATE_DLY_TIMER_MASK);


	data |= UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK |
		(1 << REG_FIELD_SHIFT(UVD_CGC_CTRL, CLK_GATE_DLY_TIMER)) |
		(4 << REG_FIELD_SHIFT(UVD_CGC_CTRL, CLK_OFF_DELAY));

	data &= ~(UVD_CGC_CTRL__UDEC_RE_MODE_MASK |
			UVD_CGC_CTRL__UDEC_CM_MODE_MASK |
			UVD_CGC_CTRL__UDEC_IT_MODE_MASK |
			UVD_CGC_CTRL__UDEC_DB_MODE_MASK |
			UVD_CGC_CTRL__UDEC_MP_MODE_MASK |
			UVD_CGC_CTRL__SYS_MODE_MASK |
			UVD_CGC_CTRL__UDEC_MODE_MASK |
			UVD_CGC_CTRL__MPEG2_MODE_MASK |
			UVD_CGC_CTRL__REGS_MODE_MASK |
			UVD_CGC_CTRL__RBC_MODE_MASK |
			UVD_CGC_CTRL__LMI_MC_MODE_MASK |
			UVD_CGC_CTRL__LMI_UMC_MODE_MASK |
			UVD_CGC_CTRL__IDCT_MODE_MASK |
			UVD_CGC_CTRL__MPRD_MODE_MASK |
			UVD_CGC_CTRL__MPC_MODE_MASK |
			UVD_CGC_CTRL__LBSI_MODE_MASK |
			UVD_CGC_CTRL__LRBBM_MODE_MASK |
			UVD_CGC_CTRL__WCB_MODE_MASK |
			UVD_CGC_CTRL__VCPU_MODE_MASK |
			UVD_CGC_CTRL__JPEG_MODE_MASK |
			UVD_CGC_CTRL__SCPU_MODE_MASK |
			UVD_CGC_CTRL__JPEG2_MODE_MASK);
	data2 &= ~(UVD_SUVD_CGC_CTRL__SRE_MODE_MASK |
			UVD_SUVD_CGC_CTRL__SIT_MODE_MASK |
			UVD_SUVD_CGC_CTRL__SMP_MODE_MASK |
			UVD_SUVD_CGC_CTRL__SCM_MODE_MASK |
			UVD_SUVD_CGC_CTRL__SDB_MODE_MASK);

	WREG32(mmUVD_CGC_CTRL, data);
	WREG32(mmUVD_SUVD_CGC_CTRL, data2);
}

#if 0
static void uvd_v6_0_set_hw_clock_gating(struct amdgpu_device *adev)
{
	uint32_t data, data1, cgc_flags, suvd_flags;

	data = RREG32(mmUVD_CGC_GATE);
	data1 = RREG32(mmUVD_SUVD_CGC_GATE);

	cgc_flags = UVD_CGC_GATE__SYS_MASK |
		UVD_CGC_GATE__UDEC_MASK |
		UVD_CGC_GATE__MPEG2_MASK |
		UVD_CGC_GATE__RBC_MASK |
		UVD_CGC_GATE__LMI_MC_MASK |
		UVD_CGC_GATE__IDCT_MASK |
		UVD_CGC_GATE__MPRD_MASK |
		UVD_CGC_GATE__MPC_MASK |
		UVD_CGC_GATE__LBSI_MASK |
		UVD_CGC_GATE__LRBBM_MASK |
		UVD_CGC_GATE__UDEC_RE_MASK |
		UVD_CGC_GATE__UDEC_CM_MASK |
		UVD_CGC_GATE__UDEC_IT_MASK |
		UVD_CGC_GATE__UDEC_DB_MASK |
		UVD_CGC_GATE__UDEC_MP_MASK |
		UVD_CGC_GATE__WCB_MASK |
		UVD_CGC_GATE__VCPU_MASK |
		UVD_CGC_GATE__SCPU_MASK |
		UVD_CGC_GATE__JPEG_MASK |
		UVD_CGC_GATE__JPEG2_MASK;

	suvd_flags = UVD_SUVD_CGC_GATE__SRE_MASK |
				UVD_SUVD_CGC_GATE__SIT_MASK |
				UVD_SUVD_CGC_GATE__SMP_MASK |
				UVD_SUVD_CGC_GATE__SCM_MASK |
				UVD_SUVD_CGC_GATE__SDB_MASK;

	data |= cgc_flags;
	data1 |= suvd_flags;

	WREG32(mmUVD_CGC_GATE, data);
	WREG32(mmUVD_SUVD_CGC_GATE, data1);
}
#endif

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
static void uvd_v6_0_enable_mgcg(struct amdgpu_device *adev,
				 bool enable)
{
	u32 orig, data;

	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_UVD_MGCG)) {
		data = RREG32_UVD_CTX(ixUVD_CGC_MEM_CTRL);
		data |= 0xfff;
		WREG32_UVD_CTX(ixUVD_CGC_MEM_CTRL, data);

		orig = data = RREG32(mmUVD_CGC_CTRL);
		data |= UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK;
		if (orig != data)
			WREG32(mmUVD_CGC_CTRL, data);
	} else {
		data = RREG32_UVD_CTX(ixUVD_CGC_MEM_CTRL);
		data &= ~0xfff;
		WREG32_UVD_CTX(ixUVD_CGC_MEM_CTRL, data);

		orig = data = RREG32(mmUVD_CGC_CTRL);
		data &= ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK;
		if (orig != data)
			WREG32(mmUVD_CGC_CTRL, data);
	}
}

1016 1017
static int uvd_v6_0_set_clockgating_state(void *handle,
					  enum amd_clockgating_state state)
1018
{
1019
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
1020
	bool enable = (state == AMD_CG_STATE_GATE) ? true : false;
1021

1022
	if (enable) {
1023 1024 1025
		/* wait for STATUS to clear */
		if (uvd_v6_0_wait_for_idle(handle))
			return -EBUSY;
1026
		uvd_v6_0_enable_clock_gating(adev, true);
1027 1028
		/* enable HW gates because UVD is idle */
/*		uvd_v6_0_set_hw_clock_gating(adev); */
1029 1030 1031
	} else {
		/* disable HW gating and enable Sw gating */
		uvd_v6_0_enable_clock_gating(adev, false);
1032
	}
1033
	uvd_v6_0_set_sw_clock_gating(adev);
1034 1035 1036
	return 0;
}

1037 1038
static int uvd_v6_0_set_powergating_state(void *handle,
					  enum amd_powergating_state state)
1039 1040 1041 1042 1043 1044 1045 1046
{
	/* This doesn't actually powergate the UVD block.
	 * That's done in the dpm code via the SMC.  This
	 * just re-inits the block as necessary.  The actual
	 * gating still happens in the dpm code.  We should
	 * revisit this when there is a cleaner line between
	 * the smc and the hw blocks
	 */
1047
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
1048
	int ret = 0;
1049

1050 1051
	WREG32(mmUVD_POWER_STATUS, UVD_POWER_STATUS__UVD_PG_EN_MASK);

1052
	if (state == AMD_PG_STATE_GATE) {
1053 1054
		uvd_v6_0_stop(adev);
	} else {
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070
		ret = uvd_v6_0_start(adev);
		if (ret)
			goto out;
	}

out:
	return ret;
}

static void uvd_v6_0_get_clockgating_state(void *handle, u32 *flags)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
	int data;

	mutex_lock(&adev->pm.mutex);

1071 1072 1073 1074 1075 1076
	if (adev->flags & AMD_IS_APU)
		data = RREG32_SMC(ixCURRENT_PG_STATUS_APU);
	else
		data = RREG32_SMC(ixCURRENT_PG_STATUS);

	if (data & CURRENT_PG_STATUS__UVD_PG_STATUS_MASK) {
1077 1078
		DRM_INFO("Cannot get clockgating state when UVD is powergated.\n");
		goto out;
1079
	}
1080 1081 1082 1083 1084 1085 1086 1087

	/* AMD_CG_SUPPORT_UVD_MGCG */
	data = RREG32(mmUVD_CGC_CTRL);
	if (data & UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK)
		*flags |= AMD_CG_SUPPORT_UVD_MGCG;

out:
	mutex_unlock(&adev->pm.mutex);
1088 1089
}

1090
static const struct amd_ip_funcs uvd_v6_0_ip_funcs = {
1091
	.name = "uvd_v6_0",
1092 1093 1094 1095 1096 1097 1098 1099 1100 1101
	.early_init = uvd_v6_0_early_init,
	.late_init = NULL,
	.sw_init = uvd_v6_0_sw_init,
	.sw_fini = uvd_v6_0_sw_fini,
	.hw_init = uvd_v6_0_hw_init,
	.hw_fini = uvd_v6_0_hw_fini,
	.suspend = uvd_v6_0_suspend,
	.resume = uvd_v6_0_resume,
	.is_idle = uvd_v6_0_is_idle,
	.wait_for_idle = uvd_v6_0_wait_for_idle,
1102 1103
	.check_soft_reset = uvd_v6_0_check_soft_reset,
	.pre_soft_reset = uvd_v6_0_pre_soft_reset,
1104
	.soft_reset = uvd_v6_0_soft_reset,
1105
	.post_soft_reset = uvd_v6_0_post_soft_reset,
1106 1107
	.set_clockgating_state = uvd_v6_0_set_clockgating_state,
	.set_powergating_state = uvd_v6_0_set_powergating_state,
1108
	.get_clockgating_state = uvd_v6_0_get_clockgating_state,
1109 1110
};

1111
static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = {
1112
	.type = AMDGPU_RING_TYPE_UVD,
1113 1114
	.align_mask = 0xf,
	.nop = PACKET0(mmUVD_NO_OP, 0),
1115
	.support_64bit_ptrs = false,
1116 1117 1118 1119
	.get_rptr = uvd_v6_0_ring_get_rptr,
	.get_wptr = uvd_v6_0_ring_get_wptr,
	.set_wptr = uvd_v6_0_ring_set_wptr,
	.parse_cs = amdgpu_uvd_ring_parse_cs,
1120 1121 1122 1123 1124 1125
	.emit_frame_size =
		2 + /* uvd_v6_0_ring_emit_hdp_flush */
		2 + /* uvd_v6_0_ring_emit_hdp_invalidate */
		10 + /* uvd_v6_0_ring_emit_pipeline_sync */
		14, /* uvd_v6_0_ring_emit_fence x1 no user fence */
	.emit_ib_size = 8, /* uvd_v6_0_ring_emit_ib */
1126 1127
	.emit_ib = uvd_v6_0_ring_emit_ib,
	.emit_fence = uvd_v6_0_ring_emit_fence,
1128 1129
	.emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush,
	.emit_hdp_invalidate = uvd_v6_0_ring_emit_hdp_invalidate,
1130
	.test_ring = uvd_v6_0_ring_test_ring,
1131
	.test_ib = amdgpu_uvd_ring_test_ib,
1132
	.insert_nop = amdgpu_ring_insert_nop,
1133
	.pad_ib = amdgpu_ring_generic_pad_ib,
1134 1135
	.begin_use = amdgpu_uvd_ring_begin_use,
	.end_use = amdgpu_uvd_ring_end_use,
1136 1137
};

1138
static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = {
1139
	.type = AMDGPU_RING_TYPE_UVD,
1140 1141
	.align_mask = 0xf,
	.nop = PACKET0(mmUVD_NO_OP, 0),
1142
	.support_64bit_ptrs = false,
1143 1144 1145
	.get_rptr = uvd_v6_0_ring_get_rptr,
	.get_wptr = uvd_v6_0_ring_get_wptr,
	.set_wptr = uvd_v6_0_ring_set_wptr,
1146 1147 1148 1149 1150 1151 1152
	.emit_frame_size =
		2 + /* uvd_v6_0_ring_emit_hdp_flush */
		2 + /* uvd_v6_0_ring_emit_hdp_invalidate */
		10 + /* uvd_v6_0_ring_emit_pipeline_sync */
		20 + /* uvd_v6_0_ring_emit_vm_flush */
		14 + 14, /* uvd_v6_0_ring_emit_fence x2 vm fence */
	.emit_ib_size = 8, /* uvd_v6_0_ring_emit_ib */
1153 1154 1155 1156 1157 1158 1159
	.emit_ib = uvd_v6_0_ring_emit_ib,
	.emit_fence = uvd_v6_0_ring_emit_fence,
	.emit_vm_flush = uvd_v6_0_ring_emit_vm_flush,
	.emit_pipeline_sync = uvd_v6_0_ring_emit_pipeline_sync,
	.emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush,
	.emit_hdp_invalidate = uvd_v6_0_ring_emit_hdp_invalidate,
	.test_ring = uvd_v6_0_ring_test_ring,
1160
	.test_ib = amdgpu_uvd_ring_test_ib,
1161 1162
	.insert_nop = amdgpu_ring_insert_nop,
	.pad_ib = amdgpu_ring_generic_pad_ib,
1163 1164
	.begin_use = amdgpu_uvd_ring_begin_use,
	.end_use = amdgpu_uvd_ring_end_use,
1165 1166
};

1167 1168
static void uvd_v6_0_set_ring_funcs(struct amdgpu_device *adev)
{
1169
	if (adev->asic_type >= CHIP_POLARIS10) {
1170 1171 1172 1173 1174 1175
		adev->uvd.ring.funcs = &uvd_v6_0_ring_vm_funcs;
		DRM_INFO("UVD is enabled in VM mode\n");
	} else {
		adev->uvd.ring.funcs = &uvd_v6_0_ring_phys_funcs;
		DRM_INFO("UVD is enabled in physical mode\n");
	}
1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
}

static const struct amdgpu_irq_src_funcs uvd_v6_0_irq_funcs = {
	.set = uvd_v6_0_set_interrupt_state,
	.process = uvd_v6_0_process_interrupt,
};

static void uvd_v6_0_set_irq_funcs(struct amdgpu_device *adev)
{
	adev->uvd.irq.num_types = 1;
	adev->uvd.irq.funcs = &uvd_v6_0_irq_funcs;
}
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

const struct amdgpu_ip_block_version uvd_v6_0_ip_block =
{
		.type = AMD_IP_BLOCK_TYPE_UVD,
		.major = 6,
		.minor = 0,
		.rev = 0,
		.funcs = &uvd_v6_0_ip_funcs,
};

const struct amdgpu_ip_block_version uvd_v6_2_ip_block =
{
		.type = AMD_IP_BLOCK_TYPE_UVD,
		.major = 6,
		.minor = 2,
		.rev = 0,
		.funcs = &uvd_v6_0_ip_funcs,
};

const struct amdgpu_ip_block_version uvd_v6_3_ip_block =
{
		.type = AMD_IP_BLOCK_TYPE_UVD,
		.major = 6,
		.minor = 3,
		.rev = 0,
		.funcs = &uvd_v6_0_ip_funcs,
};