navi10_ih.c 21.8 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
/*
 * Copyright 2019 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.
 *
 */

24 25
#include <linux/pci.h>

26 27 28 29 30 31 32 33 34
#include "amdgpu.h"
#include "amdgpu_ih.h"

#include "oss/osssys_5_0_0_offset.h"
#include "oss/osssys_5_0_0_sh_mask.h"

#include "soc15_common.h"
#include "navi10_ih.h"

35
#define MAX_REARM_RETRY 10
36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

static void navi10_ih_set_interrupt_funcs(struct amdgpu_device *adev);

/**
 * navi10_ih_enable_interrupts - Enable the interrupt ring buffer
 *
 * @adev: amdgpu_device pointer
 *
 * Enable the interrupt ring buffer (NAVI10).
 */
static void navi10_ih_enable_interrupts(struct amdgpu_device *adev)
{
	u32 ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL);

	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, RB_ENABLE, 1);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, ENABLE_INTR, 1);
52
	if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
53 54 55 56 57 58 59 60
		if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL, ih_rb_cntl)) {
			DRM_ERROR("PSP program IH_RB_CNTL failed!\n");
			return;
		}
	} else {
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL, ih_rb_cntl);
	}

61
	adev->irq.ih.enabled = true;
62 63 64 65 66

	if (adev->irq.ih1.ring_size) {
		ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1);
		ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL_RING1,
					   RB_ENABLE, 1);
67
		if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
68 69 70 71 72 73 74 75
			if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL_RING1,
						ih_rb_cntl)) {
				DRM_ERROR("program IH_RB_CNTL_RING1 failed!\n");
				return;
			}
		} else {
			WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1, ih_rb_cntl);
		}
76 77 78 79 80 81 82
		adev->irq.ih1.enabled = true;
	}

	if (adev->irq.ih2.ring_size) {
		ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2);
		ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL_RING2,
					   RB_ENABLE, 1);
83
		if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
84 85 86 87 88 89 90 91
			if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL_RING2,
						ih_rb_cntl)) {
				DRM_ERROR("program IH_RB_CNTL_RING2 failed!\n");
				return;
			}
		} else {
			WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2, ih_rb_cntl);
		}
92 93
		adev->irq.ih2.enabled = true;
	}
94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
}

/**
 * navi10_ih_disable_interrupts - Disable the interrupt ring buffer
 *
 * @adev: amdgpu_device pointer
 *
 * Disable the interrupt ring buffer (NAVI10).
 */
static void navi10_ih_disable_interrupts(struct amdgpu_device *adev)
{
	u32 ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL);

	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, RB_ENABLE, 0);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, ENABLE_INTR, 0);
109
	if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
110 111 112 113 114 115 116 117
		if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL, ih_rb_cntl)) {
			DRM_ERROR("PSP program IH_RB_CNTL failed!\n");
			return;
		}
	} else {
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL, ih_rb_cntl);
	}

118 119 120 121 122
	/* set rptr, wptr to 0 */
	WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR, 0);
	WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR, 0);
	adev->irq.ih.enabled = false;
	adev->irq.ih.rptr = 0;
123 124 125 126 127

	if (adev->irq.ih1.ring_size) {
		ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1);
		ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL_RING1,
					   RB_ENABLE, 0);
128
		if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
129 130 131 132 133 134 135 136
			if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL_RING1,
						ih_rb_cntl)) {
				DRM_ERROR("program IH_RB_CNTL_RING1 failed!\n");
				return;
			}
		} else {
			WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1, ih_rb_cntl);
		}
137 138 139 140 141 142 143 144 145 146 147
		/* set rptr, wptr to 0 */
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING1, 0);
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR_RING1, 0);
		adev->irq.ih1.enabled = false;
		adev->irq.ih1.rptr = 0;
	}

	if (adev->irq.ih2.ring_size) {
		ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2);
		ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL_RING2,
					   RB_ENABLE, 0);
148
		if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
149 150 151 152 153 154 155 156
			if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL_RING2,
						ih_rb_cntl)) {
				DRM_ERROR("program IH_RB_CNTL_RING2 failed!\n");
				return;
			}
		} else {
			WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2, ih_rb_cntl);
		}
157 158 159 160 161 162 163
		/* set rptr, wptr to 0 */
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING2, 0);
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR_RING2, 0);
		adev->irq.ih2.enabled = false;
		adev->irq.ih2.rptr = 0;
	}

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
}

static uint32_t navi10_ih_rb_cntl(struct amdgpu_ih_ring *ih, uint32_t ih_rb_cntl)
{
	int rb_bufsz = order_base_2(ih->ring_size / 4);

	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL,
				   MC_SPACE, ih->use_bus_addr ? 1 : 4);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL,
				   WPTR_OVERFLOW_CLEAR, 1);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL,
				   WPTR_OVERFLOW_ENABLE, 1);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, RB_SIZE, rb_bufsz);
	/* Ring Buffer write pointer writeback. If enabled, IH_RB_WPTR register
	 * value is written to memory
	 */
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL,
				   WPTR_WRITEBACK_ENABLE, 1);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, MC_SNOOP, 1);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, MC_RO, 0);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, MC_VMID, 0);

	return ih_rb_cntl;
}

189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
static uint32_t navi10_ih_doorbell_rptr(struct amdgpu_ih_ring *ih)
{
	u32 ih_doorbell_rtpr = 0;

	if (ih->use_doorbell) {
		ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr,
						 IH_DOORBELL_RPTR, OFFSET,
						 ih->doorbell_index);
		ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr,
						 IH_DOORBELL_RPTR,
						 ENABLE, 1);
	} else {
		ih_doorbell_rtpr = REG_SET_FIELD(ih_doorbell_rtpr,
						 IH_DOORBELL_RPTR,
						 ENABLE, 0);
	}
	return ih_doorbell_rtpr;
}

208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
static void navi10_ih_reroute_ih(struct amdgpu_device *adev)
{
	uint32_t tmp;

	/* Reroute to IH ring 1 for VMC */
	WREG32_SOC15(OSSSYS, 0, mmIH_CLIENT_CFG_INDEX, 0x12);
	tmp = RREG32_SOC15(OSSSYS, 0, mmIH_CLIENT_CFG_DATA);
	tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, CLIENT_TYPE, 1);
	tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, RING_ID, 1);
	WREG32_SOC15(OSSSYS, 0, mmIH_CLIENT_CFG_DATA, tmp);

	/* Reroute IH ring 1 for UMC */
	WREG32_SOC15(OSSSYS, 0, mmIH_CLIENT_CFG_INDEX, 0x1B);
	tmp = RREG32_SOC15(OSSSYS, 0, mmIH_CLIENT_CFG_DATA);
	tmp = REG_SET_FIELD(tmp, IH_CLIENT_CFG_DATA, RING_ID, 1);
	WREG32_SOC15(OSSSYS, 0, mmIH_CLIENT_CFG_DATA, tmp);
}

226 227 228 229 230 231 232 233 234 235 236 237 238 239
/**
 * navi10_ih_irq_init - init and enable the interrupt ring
 *
 * @adev: amdgpu_device pointer
 *
 * Allocate a ring buffer for the interrupt controller,
 * enable the RLC, disable interrupts, enable the IH
 * ring buffer and enable it (NAVI).
 * Called at device load and reume.
 * Returns 0 for success, errors for failure.
 */
static int navi10_ih_irq_init(struct amdgpu_device *adev)
{
	struct amdgpu_ih_ring *ih = &adev->irq.ih;
240
	u32 ih_rb_cntl, ih_chicken;
241 242 243 244 245
	u32 tmp;

	/* disable irqs */
	navi10_ih_disable_interrupts(adev);

246
	adev->nbio.funcs->ih_control(adev);
247 248 249 250 251 252 253 254 255

	/* Ring Buffer base. [39:8] of 40-bit address of the beginning of the ring buffer*/
	WREG32_SOC15(OSSSYS, 0, mmIH_RB_BASE, ih->gpu_addr >> 8);
	WREG32_SOC15(OSSSYS, 0, mmIH_RB_BASE_HI, (ih->gpu_addr >> 40) & 0xff);

	ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL);
	ih_rb_cntl = navi10_ih_rb_cntl(ih, ih_rb_cntl);
	ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, RPTR_REARM,
				   !!adev->irq.msi_enabled);
256
	if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
257 258 259 260 261 262 263
		if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL, ih_rb_cntl)) {
			DRM_ERROR("PSP program IH_RB_CNTL failed!\n");
			return -ETIMEDOUT;
		}
	} else {
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL, ih_rb_cntl);
	}
264
	navi10_ih_reroute_ih(adev);
265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284

	if (unlikely(adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT)) {
		if (ih->use_bus_addr) {
			ih_chicken = RREG32_SOC15(OSSSYS, 0, mmIH_CHICKEN);
			ih_chicken = REG_SET_FIELD(ih_chicken,
					IH_CHICKEN, MC_SPACE_GPA_ENABLE, 1);
			WREG32_SOC15(OSSSYS, 0, mmIH_CHICKEN, ih_chicken);
		}
	}

	/* set the writeback address whether it's enabled or not */
	WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR_ADDR_LO,
		     lower_32_bits(ih->wptr_addr));
	WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR_ADDR_HI,
		     upper_32_bits(ih->wptr_addr) & 0xFFFF);

	/* set rptr, wptr to 0 */
	WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR, 0);
	WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR, 0);

285 286
	WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR,
			navi10_ih_doorbell_rptr(ih));
287

288
	adev->nbio.funcs->ih_doorbell_range(adev, ih->use_doorbell,
289 290
					    ih->doorbell_index);

291 292 293 294 295 296 297 298 299 300 301 302
	ih = &adev->irq.ih1;
	if (ih->ring_size) {
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_BASE_RING1, ih->gpu_addr >> 8);
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_BASE_HI_RING1,
			     (ih->gpu_addr >> 40) & 0xff);

		ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1);
		ih_rb_cntl = navi10_ih_rb_cntl(ih, ih_rb_cntl);
		ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL,
					   WPTR_OVERFLOW_ENABLE, 0);
		ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL,
					   RB_FULL_DRAIN_ENABLE, 1);
303
		if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
304 305 306 307 308 309 310 311
			if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL_RING1,
						ih_rb_cntl)) {
				DRM_ERROR("program IH_RB_CNTL_RING1 failed!\n");
				return -ETIMEDOUT;
			}
		} else {
			WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1, ih_rb_cntl);
		}
312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328
		/* set rptr, wptr to 0 */
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR_RING1, 0);
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING1, 0);

		WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR_RING1,
				navi10_ih_doorbell_rptr(ih));
	}

	ih = &adev->irq.ih2;
	if (ih->ring_size) {
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_BASE_RING2, ih->gpu_addr >> 8);
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_BASE_HI_RING2,
			     (ih->gpu_addr >> 40) & 0xff);

		ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2);
		ih_rb_cntl = navi10_ih_rb_cntl(ih, ih_rb_cntl);

329
		if (amdgpu_sriov_vf(adev) && adev->asic_type < CHIP_NAVI10) {
330 331 332 333 334 335 336 337
			if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL_RING2,
						ih_rb_cntl)) {
				DRM_ERROR("program IH_RB_CNTL_RING2 failed!\n");
				return -ETIMEDOUT;
			}
		} else {
			WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2, ih_rb_cntl);
		}
338 339 340 341 342 343 344 345 346
		/* set rptr, wptr to 0 */
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_WPTR_RING2, 0);
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING2, 0);

		WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR_RING2,
			     navi10_ih_doorbell_rptr(ih));
	}


347 348 349 350 351 352 353 354 355 356 357 358 359 360
	tmp = RREG32_SOC15(OSSSYS, 0, mmIH_STORM_CLIENT_LIST_CNTL);
	tmp = REG_SET_FIELD(tmp, IH_STORM_CLIENT_LIST_CNTL,
			    CLIENT18_IS_STORM_CLIENT, 1);
	WREG32_SOC15(OSSSYS, 0, mmIH_STORM_CLIENT_LIST_CNTL, tmp);

	tmp = RREG32_SOC15(OSSSYS, 0, mmIH_INT_FLOOD_CNTL);
	tmp = REG_SET_FIELD(tmp, IH_INT_FLOOD_CNTL, FLOOD_CNTL_ENABLE, 1);
	WREG32_SOC15(OSSSYS, 0, mmIH_INT_FLOOD_CNTL, tmp);

	pci_set_master(adev->pdev);

	/* enable interrupts */
	navi10_ih_enable_interrupts(adev);

361
	return 0;
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
}

/**
 * navi10_ih_irq_disable - disable interrupts
 *
 * @adev: amdgpu_device pointer
 *
 * Disable interrupts on the hw (NAVI10).
 */
static void navi10_ih_irq_disable(struct amdgpu_device *adev)
{
	navi10_ih_disable_interrupts(adev);

	/* Wait and acknowledge irq */
	mdelay(1);
}

/**
 * navi10_ih_get_wptr - get the IH ring buffer wptr
 *
 * @adev: amdgpu_device pointer
 *
 * Get the IH ring buffer wptr from either the register
 * or the writeback memory buffer (NAVI10).  Also check for
 * ring buffer overflow and deal with it.
 * Returns the value of the wptr.
 */
static u32 navi10_ih_get_wptr(struct amdgpu_device *adev,
			      struct amdgpu_ih_ring *ih)
{
	u32 wptr, reg, tmp;

	wptr = le32_to_cpu(*ih->wptr_cpu);

	if (!REG_GET_FIELD(wptr, IH_RB_WPTR, RB_OVERFLOW))
		goto out;

399 400 401 402 403 404 405 406 407
	if (ih == &adev->irq.ih)
		reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_WPTR);
	else if (ih == &adev->irq.ih1)
		reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_WPTR_RING1);
	else if (ih == &adev->irq.ih2)
		reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_WPTR_RING2);
	else
		BUG();

408 409 410 411 412 413 414 415 416 417 418 419 420 421 422
	wptr = RREG32_NO_KIQ(reg);
	if (!REG_GET_FIELD(wptr, IH_RB_WPTR, RB_OVERFLOW))
		goto out;
	wptr = REG_SET_FIELD(wptr, IH_RB_WPTR, RB_OVERFLOW, 0);

	/* When a ring buffer overflow happen start parsing interrupt
	 * from the last not overwritten vector (wptr + 32). Hopefully
	 * this should allow us to catch up.
	 */
	tmp = (wptr + 32) & ih->ptr_mask;
	dev_warn(adev->dev, "IH ring buffer overflow "
		 "(0x%08X, 0x%08X, 0x%08X)\n",
		 wptr, ih->rptr, tmp);
	ih->rptr = tmp;

423 424 425 426 427 428 429 430 431
	if (ih == &adev->irq.ih)
		reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_CNTL);
	else if (ih == &adev->irq.ih1)
		reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_CNTL_RING1);
	else if (ih == &adev->irq.ih2)
		reg = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_CNTL_RING2);
	else
		BUG();

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
	tmp = RREG32_NO_KIQ(reg);
	tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, WPTR_OVERFLOW_CLEAR, 1);
	WREG32_NO_KIQ(reg, tmp);
out:
	return (wptr & ih->ptr_mask);
}

/**
 * navi10_ih_decode_iv - decode an interrupt vector
 *
 * @adev: amdgpu_device pointer
 *
 * Decodes the interrupt vector at the current rptr
 * position and also advance the position.
 */
static void navi10_ih_decode_iv(struct amdgpu_device *adev,
				struct amdgpu_ih_ring *ih,
				struct amdgpu_iv_entry *entry)
{
	/* wptr/rptr are in bytes! */
	u32 ring_index = ih->rptr >> 2;
	uint32_t dw[8];

	dw[0] = le32_to_cpu(ih->ring[ring_index + 0]);
	dw[1] = le32_to_cpu(ih->ring[ring_index + 1]);
	dw[2] = le32_to_cpu(ih->ring[ring_index + 2]);
	dw[3] = le32_to_cpu(ih->ring[ring_index + 3]);
	dw[4] = le32_to_cpu(ih->ring[ring_index + 4]);
	dw[5] = le32_to_cpu(ih->ring[ring_index + 5]);
	dw[6] = le32_to_cpu(ih->ring[ring_index + 6]);
	dw[7] = le32_to_cpu(ih->ring[ring_index + 7]);

	entry->client_id = dw[0] & 0xff;
	entry->src_id = (dw[0] >> 8) & 0xff;
	entry->ring_id = (dw[0] >> 16) & 0xff;
	entry->vmid = (dw[0] >> 24) & 0xf;
	entry->vmid_src = (dw[0] >> 31);
	entry->timestamp = dw[1] | ((u64)(dw[2] & 0xffff) << 32);
	entry->timestamp_src = dw[2] >> 31;
	entry->pasid = dw[3] & 0xffff;
	entry->pasid_src = dw[3] >> 31;
	entry->src_data[0] = dw[4];
	entry->src_data[1] = dw[5];
	entry->src_data[2] = dw[6];
	entry->src_data[3] = dw[7];

	/* wptr/rptr are in bytes! */
	ih->rptr += 32;
}

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
/**
 * navi10_ih_irq_rearm - rearm IRQ if lost
 *
 * @adev: amdgpu_device pointer
 *
 */
static void navi10_ih_irq_rearm(struct amdgpu_device *adev,
			       struct amdgpu_ih_ring *ih)
{
	uint32_t reg_rptr = 0;
	uint32_t v = 0;
	uint32_t i = 0;

	if (ih == &adev->irq.ih)
		reg_rptr = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_RPTR);
	else if (ih == &adev->irq.ih1)
		reg_rptr = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_RPTR_RING1);
	else if (ih == &adev->irq.ih2)
		reg_rptr = SOC15_REG_OFFSET(OSSSYS, 0, mmIH_RB_RPTR_RING2);
	else
		return;

	/* Rearm IRQ / re-write doorbell if doorbell write is lost */
	for (i = 0; i < MAX_REARM_RETRY; i++) {
		v = RREG32_NO_KIQ(reg_rptr);
		if ((v < ih->ring_size) && (v != ih->rptr))
			WDOORBELL32(ih->doorbell_index, ih->rptr);
		else
			break;
	}
}

514 515 516 517 518 519 520 521 522 523 524 525 526 527
/**
 * navi10_ih_set_rptr - set the IH ring buffer rptr
 *
 * @adev: amdgpu_device pointer
 *
 * Set the IH ring buffer rptr.
 */
static void navi10_ih_set_rptr(struct amdgpu_device *adev,
			       struct amdgpu_ih_ring *ih)
{
	if (ih->use_doorbell) {
		/* XXX check if swapping is necessary on BE */
		*ih->rptr_cpu = ih->rptr;
		WDOORBELL32(ih->doorbell_index, ih->rptr);
528 529 530

		if (amdgpu_sriov_vf(adev))
			navi10_ih_irq_rearm(adev, ih);
531
	} else if (ih == &adev->irq.ih) {
532
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR, ih->rptr);
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
	} else if (ih == &adev->irq.ih1) {
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING1, ih->rptr);
	} else if (ih == &adev->irq.ih2) {
		WREG32_SOC15(OSSSYS, 0, mmIH_RB_RPTR_RING2, ih->rptr);
	}
}

/**
 * navi10_ih_self_irq - dispatch work for ring 1 and 2
 *
 * @adev: amdgpu_device pointer
 * @source: irq source
 * @entry: IV with WPTR update
 *
 * Update the WPTR from the IV and schedule work to handle the entries.
 */
static int navi10_ih_self_irq(struct amdgpu_device *adev,
			      struct amdgpu_irq_src *source,
			      struct amdgpu_iv_entry *entry)
{
	uint32_t wptr = cpu_to_le32(entry->src_data[0]);

	switch (entry->ring_id) {
	case 1:
		*adev->irq.ih1.wptr_cpu = wptr;
		schedule_work(&adev->irq.ih1_work);
		break;
	case 2:
		*adev->irq.ih2.wptr_cpu = wptr;
		schedule_work(&adev->irq.ih2_work);
		break;
	default: break;
	}
	return 0;
}

static const struct amdgpu_irq_src_funcs navi10_ih_self_irq_funcs = {
	.process = navi10_ih_self_irq,
};

static void navi10_ih_set_self_irq_funcs(struct amdgpu_device *adev)
{
	adev->irq.self_irq.num_types = 0;
	adev->irq.self_irq.funcs = &navi10_ih_self_irq_funcs;
577 578 579 580 581 582 583
}

static int navi10_ih_early_init(void *handle)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

	navi10_ih_set_interrupt_funcs(adev);
584
	navi10_ih_set_self_irq_funcs(adev);
585 586 587 588 589 590 591 592 593
	return 0;
}

static int navi10_ih_sw_init(void *handle)
{
	int r;
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
	bool use_bus_addr;

594 595 596 597 598 599
	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_IH, 0,
				&adev->irq.self_irq);

	if (r)
		return r;

600 601 602 603 604 605 606 607 608 609 610 611
	/* use gpu virtual address for ih ring
	 * until ih_checken is programmed to allow
	 * use bus address for ih ring by psp bl */
	use_bus_addr =
		(adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) ? false : true;
	r = amdgpu_ih_ring_init(adev, &adev->irq.ih, 256 * 1024, use_bus_addr);
	if (r)
		return r;

	adev->irq.ih.use_doorbell = true;
	adev->irq.ih.doorbell_index = adev->doorbell_index.ih << 1;

612 613 614 615 616 617 618 619 620 621 622 623 624 625
	r = amdgpu_ih_ring_init(adev, &adev->irq.ih1, PAGE_SIZE, true);
	if (r)
		return r;

	adev->irq.ih1.use_doorbell = true;
	adev->irq.ih1.doorbell_index = (adev->doorbell_index.ih + 1) << 1;

	r = amdgpu_ih_ring_init(adev, &adev->irq.ih2, PAGE_SIZE, true);
	if (r)
		return r;

	adev->irq.ih2.use_doorbell = true;
	adev->irq.ih2.doorbell_index = (adev->doorbell_index.ih + 2) << 1;

626 627 628 629 630 631 632 633 634 635
	r = amdgpu_irq_init(adev);

	return r;
}

static int navi10_ih_sw_fini(void *handle)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

	amdgpu_irq_fini(adev);
636 637
	amdgpu_ih_ring_fini(adev, &adev->irq.ih2);
	amdgpu_ih_ring_fini(adev, &adev->irq.ih1);
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
	amdgpu_ih_ring_fini(adev, &adev->irq.ih);

	return 0;
}

static int navi10_ih_hw_init(void *handle)
{
	int r;
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

	r = navi10_ih_irq_init(adev);
	if (r)
		return r;

	return 0;
}

static int navi10_ih_hw_fini(void *handle)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

	navi10_ih_irq_disable(adev);

	return 0;
}

static int navi10_ih_suspend(void *handle)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

	return navi10_ih_hw_fini(adev);
}

static int navi10_ih_resume(void *handle)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

	return navi10_ih_hw_init(adev);
}

static bool navi10_ih_is_idle(void *handle)
{
	/* todo */
	return true;
}

static int navi10_ih_wait_for_idle(void *handle)
{
	/* todo */
	return -ETIMEDOUT;
}

static int navi10_ih_soft_reset(void *handle)
{
	/* todo */
	return 0;
}

static void navi10_ih_update_clockgating_state(struct amdgpu_device *adev,
					       bool enable)
{
	uint32_t data, def, field_val;

	if (adev->cg_flags & AMD_CG_SUPPORT_IH_CG) {
		def = data = RREG32_SOC15(OSSSYS, 0, mmIH_CLK_CTRL);
		field_val = enable ? 0 : 1;
		data = REG_SET_FIELD(data, IH_CLK_CTRL,
				     DBUS_MUX_CLK_SOFT_OVERRIDE, field_val);
		data = REG_SET_FIELD(data, IH_CLK_CTRL,
				     OSSSYS_SHARE_CLK_SOFT_OVERRIDE, field_val);
		data = REG_SET_FIELD(data, IH_CLK_CTRL,
				     LIMIT_SMN_CLK_SOFT_OVERRIDE, field_val);
		data = REG_SET_FIELD(data, IH_CLK_CTRL,
				     DYN_CLK_SOFT_OVERRIDE, field_val);
		data = REG_SET_FIELD(data, IH_CLK_CTRL,
				     REG_CLK_SOFT_OVERRIDE, field_val);
		if (def != data)
			WREG32_SOC15(OSSSYS, 0, mmIH_CLK_CTRL, data);
	}

	return;
}

static int navi10_ih_set_clockgating_state(void *handle,
					   enum amd_clockgating_state state)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

	navi10_ih_update_clockgating_state(adev,
727
				state == AMD_CG_STATE_GATE);
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
	return 0;
}

static int navi10_ih_set_powergating_state(void *handle,
					   enum amd_powergating_state state)
{
	return 0;
}

static void navi10_ih_get_clockgating_state(void *handle, u32 *flags)
{
	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

	if (!RREG32_SOC15(OSSSYS, 0, mmIH_CLK_CTRL))
		*flags |= AMD_CG_SUPPORT_IH_CG;

	return;
}

static const struct amd_ip_funcs navi10_ih_ip_funcs = {
	.name = "navi10_ih",
	.early_init = navi10_ih_early_init,
	.late_init = NULL,
	.sw_init = navi10_ih_sw_init,
	.sw_fini = navi10_ih_sw_fini,
	.hw_init = navi10_ih_hw_init,
	.hw_fini = navi10_ih_hw_fini,
	.suspend = navi10_ih_suspend,
	.resume = navi10_ih_resume,
	.is_idle = navi10_ih_is_idle,
	.wait_for_idle = navi10_ih_wait_for_idle,
	.soft_reset = navi10_ih_soft_reset,
	.set_clockgating_state = navi10_ih_set_clockgating_state,
	.set_powergating_state = navi10_ih_set_powergating_state,
	.get_clockgating_state = navi10_ih_get_clockgating_state,
};

static const struct amdgpu_ih_funcs navi10_ih_funcs = {
	.get_wptr = navi10_ih_get_wptr,
	.decode_iv = navi10_ih_decode_iv,
	.set_rptr = navi10_ih_set_rptr
};

static void navi10_ih_set_interrupt_funcs(struct amdgpu_device *adev)
{
	if (adev->irq.ih_funcs == NULL)
		adev->irq.ih_funcs = &navi10_ih_funcs;
}

const struct amdgpu_ip_block_version navi10_ih_ip_block =
{
	.type = AMD_IP_BLOCK_TYPE_IH,
	.major = 5,
	.minor = 0,
	.rev = 0,
	.funcs = &navi10_ih_ip_funcs,
};