* [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5
@ 2020-04-30 18:37 Alex Deucher
2020-04-30 18:37 ` [PATCH 2/8] drm/amdgpu: check SMU NULL ptr on gfx hw init Alex Deucher
` (6 more replies)
0 siblings, 7 replies; 8+ messages in thread
From: Alex Deucher @ 2020-04-30 18:37 UTC (permalink / raw)
To: amd-gfx; +Cc: Alex Deucher, Boyuan Zhang, James Zhu
From: Boyuan Zhang <boyuan.zhang@amd.com>
Rename hardware IP name from UVD to VCN to reduce confusion.
Hardware IP name UVD and VCN are equivalent for VCN2.5 asics.
Use name VCN for future VCN based asics.
V2: update description
V3: rebase
Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: James Zhu <james.zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c | 392 +++++++++++++-------------
1 file changed, 196 insertions(+), 196 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
index 5c1b9ed76534..34ed906645c9 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c
@@ -86,7 +86,7 @@ static int vcn_v2_5_early_init(void *handle)
adev->vcn.num_vcn_inst = VCN25_MAX_HW_INSTANCES_ARCTURUS;
for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
- harvest = RREG32_SOC15(UVD, i, mmCC_UVD_HARVESTING);
+ harvest = RREG32_SOC15(VCN, i, mmCC_UVD_HARVESTING);
if (harvest & CC_UVD_HARVESTING__UVD_DISABLE_MASK)
adev->vcn.harvest_config |= 1 << i;
}
@@ -177,15 +177,15 @@ static int vcn_v2_5_sw_init(void *handle)
adev->vcn.internal.gp_scratch8 = mmUVD_GP_SCRATCH8_INTERNAL_OFFSET;
adev->vcn.internal.scratch9 = mmUVD_SCRATCH9_INTERNAL_OFFSET;
- adev->vcn.inst[j].external.scratch9 = SOC15_REG_OFFSET(UVD, j, mmUVD_SCRATCH9);
+ adev->vcn.inst[j].external.scratch9 = SOC15_REG_OFFSET(VCN, j, mmUVD_SCRATCH9);
adev->vcn.internal.data0 = mmUVD_GPCOM_VCPU_DATA0_INTERNAL_OFFSET;
- adev->vcn.inst[j].external.data0 = SOC15_REG_OFFSET(UVD, j, mmUVD_GPCOM_VCPU_DATA0);
+ adev->vcn.inst[j].external.data0 = SOC15_REG_OFFSET(VCN, j, mmUVD_GPCOM_VCPU_DATA0);
adev->vcn.internal.data1 = mmUVD_GPCOM_VCPU_DATA1_INTERNAL_OFFSET;
- adev->vcn.inst[j].external.data1 = SOC15_REG_OFFSET(UVD, j, mmUVD_GPCOM_VCPU_DATA1);
+ adev->vcn.inst[j].external.data1 = SOC15_REG_OFFSET(VCN, j, mmUVD_GPCOM_VCPU_DATA1);
adev->vcn.internal.cmd = mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET;
- adev->vcn.inst[j].external.cmd = SOC15_REG_OFFSET(UVD, j, mmUVD_GPCOM_VCPU_CMD);
+ adev->vcn.inst[j].external.cmd = SOC15_REG_OFFSET(VCN, j, mmUVD_GPCOM_VCPU_CMD);
adev->vcn.internal.nop = mmUVD_NO_OP_INTERNAL_OFFSET;
- adev->vcn.inst[j].external.nop = SOC15_REG_OFFSET(UVD, j, mmUVD_NO_OP);
+ adev->vcn.inst[j].external.nop = SOC15_REG_OFFSET(VCN, j, mmUVD_NO_OP);
ring = &adev->vcn.inst[j].ring_dec;
ring->use_doorbell = true;
@@ -399,46 +399,46 @@ static void vcn_v2_5_mc_resume(struct amdgpu_device *adev)
continue;
/* cache window 0: fw */
if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
(adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + i].tmr_mc_addr_lo));
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
(adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + i].tmr_mc_addr_hi));
- WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET0, 0);
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_OFFSET0, 0);
offset = 0;
} else {
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
lower_32_bits(adev->vcn.inst[i].gpu_addr));
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
upper_32_bits(adev->vcn.inst[i].gpu_addr));
offset = size;
- WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET0,
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_OFFSET0,
AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
}
- WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE0, size);
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_SIZE0, size);
/* cache window 1: stack */
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW,
lower_32_bits(adev->vcn.inst[i].gpu_addr + offset));
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH,
upper_32_bits(adev->vcn.inst[i].gpu_addr + offset));
- WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET1, 0);
- WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_VCN_STACK_SIZE);
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_OFFSET1, 0);
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_VCN_STACK_SIZE);
/* cache window 2: context */
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW,
lower_32_bits(adev->vcn.inst[i].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE));
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH,
upper_32_bits(adev->vcn.inst[i].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE));
- WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET2, 0);
- WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE2, AMDGPU_VCN_CONTEXT_SIZE);
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_OFFSET2, 0);
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_SIZE2, AMDGPU_VCN_CONTEXT_SIZE);
/* non-cache window */
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_NC0_64BIT_BAR_LOW,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_NC0_64BIT_BAR_LOW,
lower_32_bits(adev->vcn.inst[i].fw_shared_gpu_addr));
- WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_NC0_64BIT_BAR_HIGH,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_NC0_64BIT_BAR_HIGH,
upper_32_bits(adev->vcn.inst[i].fw_shared_gpu_addr));
- WREG32_SOC15(UVD, i, mmUVD_VCPU_NONCACHE_OFFSET0, 0);
- WREG32_SOC15(UVD, i, mmUVD_VCPU_NONCACHE_SIZE0,
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_NONCACHE_OFFSET0, 0);
+ WREG32_SOC15(VCN, i, mmUVD_VCPU_NONCACHE_SIZE0,
AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_fw_shared)));
}
}
@@ -452,91 +452,91 @@ static void vcn_v2_5_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_idx
if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
if (!indirect) {
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
+ VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
(adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + inst_idx].tmr_mc_addr_lo), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
+ VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
(adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + inst_idx].tmr_mc_addr_hi), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_OFFSET0), 0, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_OFFSET0), 0, 0, indirect);
} else {
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), 0, 0, indirect);
+ VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), 0, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), 0, 0, indirect);
+ VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), 0, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_OFFSET0), 0, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_OFFSET0), 0, 0, indirect);
}
offset = 0;
} else {
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
+ VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
lower_32_bits(adev->vcn.inst[inst_idx].gpu_addr), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
+ VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
upper_32_bits(adev->vcn.inst[inst_idx].gpu_addr), 0, indirect);
offset = size;
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_OFFSET0),
+ VCN, 0, mmUVD_VCPU_CACHE_OFFSET0),
AMDGPU_UVD_FIRMWARE_OFFSET >> 3, 0, indirect);
}
if (!indirect)
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_SIZE0), size, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_SIZE0), size, 0, indirect);
else
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_SIZE0), 0, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_SIZE0), 0, 0, indirect);
/* cache window 1: stack */
if (!indirect) {
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW),
+ VCN, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW),
lower_32_bits(adev->vcn.inst[inst_idx].gpu_addr + offset), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH),
+ VCN, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH),
upper_32_bits(adev->vcn.inst[inst_idx].gpu_addr + offset), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), 0, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_OFFSET1), 0, 0, indirect);
} else {
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), 0, 0, indirect);
+ VCN, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), 0, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), 0, 0, indirect);
+ VCN, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), 0, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), 0, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_OFFSET1), 0, 0, indirect);
}
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_VCN_STACK_SIZE, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_VCN_STACK_SIZE, 0, indirect);
/* cache window 2: context */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW),
+ VCN, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW),
lower_32_bits(adev->vcn.inst[inst_idx].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH),
+ VCN, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH),
upper_32_bits(adev->vcn.inst[inst_idx].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_OFFSET2), 0, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_OFFSET2), 0, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CACHE_SIZE2), AMDGPU_VCN_CONTEXT_SIZE, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CACHE_SIZE2), AMDGPU_VCN_CONTEXT_SIZE, 0, indirect);
/* non-cache window */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_NC0_64BIT_BAR_LOW),
+ VCN, 0, mmUVD_LMI_VCPU_NC0_64BIT_BAR_LOW),
lower_32_bits(adev->vcn.inst[inst_idx].fw_shared_gpu_addr), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_VCPU_NC0_64BIT_BAR_HIGH),
+ VCN, 0, mmUVD_LMI_VCPU_NC0_64BIT_BAR_HIGH),
upper_32_bits(adev->vcn.inst[inst_idx].fw_shared_gpu_addr), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_NONCACHE_OFFSET0), 0, 0, indirect);
+ VCN, 0, mmUVD_VCPU_NONCACHE_OFFSET0), 0, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_NONCACHE_SIZE0),
+ VCN, 0, mmUVD_VCPU_NONCACHE_SIZE0),
AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_fw_shared)), 0, indirect);
/* VCN global tiling registers */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_GFX8_ADDR_CONFIG), adev->gfx.config.gb_addr_config, 0, indirect);
+ VCN, 0, mmUVD_GFX8_ADDR_CONFIG), adev->gfx.config.gb_addr_config, 0, indirect);
}
/**
@@ -690,19 +690,19 @@ static void vcn_v2_5_clock_gating_dpg_mode(struct amdgpu_device *adev,
UVD_CGC_CTRL__VCPU_MODE_MASK |
UVD_CGC_CTRL__MMSCH_MODE_MASK);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_CGC_CTRL), reg_data, sram_sel, indirect);
+ VCN, 0, mmUVD_CGC_CTRL), reg_data, sram_sel, indirect);
/* turn off clock gating */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_CGC_GATE), 0, sram_sel, indirect);
+ VCN, 0, mmUVD_CGC_GATE), 0, sram_sel, indirect);
/* turn on SUVD clock gating */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_SUVD_CGC_GATE), 1, sram_sel, indirect);
+ VCN, 0, mmUVD_SUVD_CGC_GATE), 1, sram_sel, indirect);
/* turn on sw mode in UVD_SUVD_CGC_CTRL */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_SUVD_CGC_CTRL), 0, sram_sel, indirect);
+ VCN, 0, mmUVD_SUVD_CGC_CTRL), 0, sram_sel, indirect);
}
/**
@@ -774,13 +774,13 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo
uint32_t rb_bufsz, tmp;
/* disable register anti-hang mechanism */
- WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS), 1,
+ WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS), 1,
~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK);
/* enable dynamic power gating mode */
- tmp = RREG32_SOC15(UVD, inst_idx, mmUVD_POWER_STATUS);
+ tmp = RREG32_SOC15(VCN, inst_idx, mmUVD_POWER_STATUS);
tmp |= UVD_POWER_STATUS__UVD_PG_MODE_MASK;
tmp |= UVD_POWER_STATUS__UVD_PG_EN_MASK;
- WREG32_SOC15(UVD, inst_idx, mmUVD_POWER_STATUS, tmp);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_POWER_STATUS, tmp);
if (indirect)
adev->vcn.inst[inst_idx].dpg_sram_curr_addr = (uint32_t*)adev->vcn.inst[inst_idx].dpg_sram_cpu_addr;
@@ -793,11 +793,11 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo
tmp |= UVD_VCPU_CNTL__CLK_EN_MASK;
tmp |= UVD_VCPU_CNTL__BLK_RST_MASK;
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CNTL), tmp, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CNTL), tmp, 0, indirect);
/* disable master interupt */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_MASTINT_EN), 0, 0, indirect);
+ VCN, 0, mmUVD_MASTINT_EN), 0, 0, indirect);
/* setup mmUVD_LMI_CTRL */
tmp = (0x8 | UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
@@ -809,28 +809,28 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo
(8 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
0x00100000L);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_CTRL), tmp, 0, indirect);
+ VCN, 0, mmUVD_LMI_CTRL), tmp, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_MPC_CNTL),
+ VCN, 0, mmUVD_MPC_CNTL),
0x2 << UVD_MPC_CNTL__REPLACEMENT_MODE__SHIFT, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_MPC_SET_MUXA0),
+ VCN, 0, mmUVD_MPC_SET_MUXA0),
((0x1 << UVD_MPC_SET_MUXA0__VARA_1__SHIFT) |
(0x2 << UVD_MPC_SET_MUXA0__VARA_2__SHIFT) |
(0x3 << UVD_MPC_SET_MUXA0__VARA_3__SHIFT) |
(0x4 << UVD_MPC_SET_MUXA0__VARA_4__SHIFT)), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_MPC_SET_MUXB0),
+ VCN, 0, mmUVD_MPC_SET_MUXB0),
((0x1 << UVD_MPC_SET_MUXB0__VARB_1__SHIFT) |
(0x2 << UVD_MPC_SET_MUXB0__VARB_2__SHIFT) |
(0x3 << UVD_MPC_SET_MUXB0__VARB_3__SHIFT) |
(0x4 << UVD_MPC_SET_MUXB0__VARB_4__SHIFT)), 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_MPC_SET_MUX),
+ VCN, 0, mmUVD_MPC_SET_MUX),
((0x0 << UVD_MPC_SET_MUX__SET_0__SHIFT) |
(0x1 << UVD_MPC_SET_MUX__SET_1__SHIFT) |
(0x2 << UVD_MPC_SET_MUX__SET_2__SHIFT)), 0, indirect);
@@ -838,26 +838,26 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo
vcn_v2_5_mc_resume_dpg_mode(adev, inst_idx, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_REG_XX_MASK), 0x10, 0, indirect);
+ VCN, 0, mmUVD_REG_XX_MASK), 0x10, 0, indirect);
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_RBC_XX_IB_REG_CHECK), 0x3, 0, indirect);
+ VCN, 0, mmUVD_RBC_XX_IB_REG_CHECK), 0x3, 0, indirect);
/* enable LMI MC and UMC channels */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_LMI_CTRL2), 0, 0, indirect);
+ VCN, 0, mmUVD_LMI_CTRL2), 0, 0, indirect);
/* unblock VCPU register access */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_RB_ARB_CTRL), 0, 0, indirect);
+ VCN, 0, mmUVD_RB_ARB_CTRL), 0, 0, indirect);
tmp = (0xFF << UVD_VCPU_CNTL__PRB_TIMEOUT_VAL__SHIFT);
tmp |= UVD_VCPU_CNTL__CLK_EN_MASK;
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_VCPU_CNTL), tmp, 0, indirect);
+ VCN, 0, mmUVD_VCPU_CNTL), tmp, 0, indirect);
/* enable master interrupt */
WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0(
- UVD, 0, mmUVD_MASTINT_EN),
+ VCN, 0, mmUVD_MASTINT_EN),
UVD_MASTINT_EN__VCPU_EN_MASK, 0, indirect);
if (indirect)
@@ -873,39 +873,39 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
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_SOC15(UVD, inst_idx, mmUVD_RBC_RB_CNTL, tmp);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_CNTL, tmp);
/* Stall DPG before WPTR/RPTR reset */
- WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS),
+ WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS),
UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK,
~UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK);
fw_shared->multi_queue.decode_queue_mode |= FW_QUEUE_RING_RESET;
/* set the write pointer delay */
- WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_WPTR_CNTL, 0);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_WPTR_CNTL, 0);
/* set the wb address */
- WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_RPTR_ADDR,
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_RPTR_ADDR,
(upper_32_bits(ring->gpu_addr) >> 2));
/* programm the RB_BASE for ring buffer */
- WREG32_SOC15(UVD, inst_idx, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
+ WREG32_SOC15(VCN, inst_idx, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
lower_32_bits(ring->gpu_addr));
- WREG32_SOC15(UVD, inst_idx, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
+ WREG32_SOC15(VCN, inst_idx, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
upper_32_bits(ring->gpu_addr));
/* Initialize the ring buffer's read and write pointers */
- WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_RPTR, 0);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_RPTR, 0);
- WREG32_SOC15(UVD, inst_idx, mmUVD_SCRATCH2, 0);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_SCRATCH2, 0);
- ring->wptr = RREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_RPTR);
- WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_WPTR,
+ ring->wptr = RREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_RPTR);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_WPTR,
lower_32_bits(ring->wptr));
fw_shared->multi_queue.decode_queue_mode &= ~FW_QUEUE_RING_RESET;
/* Unstall DPG */
- WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS),
+ WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS),
0, ~UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK);
return 0;
@@ -929,12 +929,12 @@ static int vcn_v2_5_start(struct amdgpu_device *adev)
}
/* disable register anti-hang mechanism */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_POWER_STATUS), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_POWER_STATUS), 0,
~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK);
/* set uvd status busy */
- tmp = RREG32_SOC15(UVD, i, mmUVD_STATUS) | UVD_STATUS__UVD_BUSY;
- WREG32_SOC15(UVD, i, mmUVD_STATUS, tmp);
+ tmp = RREG32_SOC15(VCN, i, mmUVD_STATUS) | UVD_STATUS__UVD_BUSY;
+ WREG32_SOC15(VCN, i, mmUVD_STATUS, tmp);
}
if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG)
@@ -947,44 +947,44 @@ static int vcn_v2_5_start(struct amdgpu_device *adev)
if (adev->vcn.harvest_config & (1 << i))
continue;
/* enable VCPU clock */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL),
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL),
UVD_VCPU_CNTL__CLK_EN_MASK, ~UVD_VCPU_CNTL__CLK_EN_MASK);
/* disable master interrupt */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_MASTINT_EN), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_MASTINT_EN), 0,
~UVD_MASTINT_EN__VCPU_EN_MASK);
/* setup mmUVD_LMI_CTRL */
- tmp = RREG32_SOC15(UVD, i, mmUVD_LMI_CTRL);
+ tmp = RREG32_SOC15(VCN, i, mmUVD_LMI_CTRL);
tmp &= ~0xff;
- WREG32_SOC15(UVD, i, mmUVD_LMI_CTRL, tmp | 0x8|
+ WREG32_SOC15(VCN, i, mmUVD_LMI_CTRL, tmp | 0x8|
UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
UVD_LMI_CTRL__MASK_MC_URGENT_MASK |
UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK);
/* setup mmUVD_MPC_CNTL */
- tmp = RREG32_SOC15(UVD, i, mmUVD_MPC_CNTL);
+ tmp = RREG32_SOC15(VCN, i, mmUVD_MPC_CNTL);
tmp &= ~UVD_MPC_CNTL__REPLACEMENT_MODE_MASK;
tmp |= 0x2 << UVD_MPC_CNTL__REPLACEMENT_MODE__SHIFT;
WREG32_SOC15(VCN, i, mmUVD_MPC_CNTL, tmp);
/* setup UVD_MPC_SET_MUXA0 */
- WREG32_SOC15(UVD, i, mmUVD_MPC_SET_MUXA0,
+ WREG32_SOC15(VCN, i, mmUVD_MPC_SET_MUXA0,
((0x1 << UVD_MPC_SET_MUXA0__VARA_1__SHIFT) |
(0x2 << UVD_MPC_SET_MUXA0__VARA_2__SHIFT) |
(0x3 << UVD_MPC_SET_MUXA0__VARA_3__SHIFT) |
(0x4 << UVD_MPC_SET_MUXA0__VARA_4__SHIFT)));
/* setup UVD_MPC_SET_MUXB0 */
- WREG32_SOC15(UVD, i, mmUVD_MPC_SET_MUXB0,
+ WREG32_SOC15(VCN, i, mmUVD_MPC_SET_MUXB0,
((0x1 << UVD_MPC_SET_MUXB0__VARB_1__SHIFT) |
(0x2 << UVD_MPC_SET_MUXB0__VARB_2__SHIFT) |
(0x3 << UVD_MPC_SET_MUXB0__VARB_3__SHIFT) |
(0x4 << UVD_MPC_SET_MUXB0__VARB_4__SHIFT)));
/* setup mmUVD_MPC_SET_MUX */
- WREG32_SOC15(UVD, i, mmUVD_MPC_SET_MUX,
+ WREG32_SOC15(VCN, i, mmUVD_MPC_SET_MUX,
((0x0 << UVD_MPC_SET_MUX__SET_0__SHIFT) |
(0x1 << UVD_MPC_SET_MUX__SET_1__SHIFT) |
(0x2 << UVD_MPC_SET_MUX__SET_2__SHIFT)));
@@ -997,27 +997,27 @@ static int vcn_v2_5_start(struct amdgpu_device *adev)
if (adev->vcn.harvest_config & (1 << i))
continue;
/* VCN global tiling registers */
- WREG32_SOC15(UVD, i, mmUVD_GFX8_ADDR_CONFIG,
+ WREG32_SOC15(VCN, i, mmUVD_GFX8_ADDR_CONFIG,
adev->gfx.config.gb_addr_config);
- WREG32_SOC15(UVD, i, mmUVD_GFX8_ADDR_CONFIG,
+ WREG32_SOC15(VCN, i, mmUVD_GFX8_ADDR_CONFIG,
adev->gfx.config.gb_addr_config);
/* enable LMI MC and UMC channels */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_CTRL2), 0,
~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
/* unblock VCPU register access */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_ARB_CTRL), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_RB_ARB_CTRL), 0,
~UVD_RB_ARB_CTRL__VCPU_DIS_MASK);
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), 0,
~UVD_VCPU_CNTL__BLK_RST_MASK);
for (k = 0; k < 10; ++k) {
uint32_t status;
for (j = 0; j < 100; ++j) {
- status = RREG32_SOC15(UVD, i, mmUVD_STATUS);
+ status = RREG32_SOC15(VCN, i, mmUVD_STATUS);
if (status & 2)
break;
if (amdgpu_emu_mode == 1)
@@ -1030,11 +1030,11 @@ static int vcn_v2_5_start(struct amdgpu_device *adev)
break;
DRM_ERROR("VCN decode not responding, trying to reset the VCPU!!!\n");
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL),
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL),
UVD_VCPU_CNTL__BLK_RST_MASK,
~UVD_VCPU_CNTL__BLK_RST_MASK);
mdelay(10);
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), 0,
~UVD_VCPU_CNTL__BLK_RST_MASK);
mdelay(10);
@@ -1047,15 +1047,15 @@ static int vcn_v2_5_start(struct amdgpu_device *adev)
}
/* enable master interrupt */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_MASTINT_EN),
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_MASTINT_EN),
UVD_MASTINT_EN__VCPU_EN_MASK,
~UVD_MASTINT_EN__VCPU_EN_MASK);
/* clear the busy bit of VCN_STATUS */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_STATUS), 0,
~(2 << UVD_STATUS__VCPU_REPORT__SHIFT));
- WREG32_SOC15(UVD, i, mmUVD_LMI_RBC_RB_VMID, 0);
+ WREG32_SOC15(VCN, i, mmUVD_LMI_RBC_RB_VMID, 0);
ring = &adev->vcn.inst[i].ring_dec;
/* force RBC into idle state */
@@ -1065,39 +1065,39 @@ static int vcn_v2_5_start(struct amdgpu_device *adev)
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
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_SOC15(UVD, i, mmUVD_RBC_RB_CNTL, tmp);
+ WREG32_SOC15(VCN, i, mmUVD_RBC_RB_CNTL, tmp);
fw_shared->multi_queue.decode_queue_mode |= FW_QUEUE_RING_RESET;
/* programm the RB_BASE for ring buffer */
- WREG32_SOC15(UVD, i, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
lower_32_bits(ring->gpu_addr));
- WREG32_SOC15(UVD, i, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
+ WREG32_SOC15(VCN, i, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
upper_32_bits(ring->gpu_addr));
/* Initialize the ring buffer's read and write pointers */
- WREG32_SOC15(UVD, i, mmUVD_RBC_RB_RPTR, 0);
+ WREG32_SOC15(VCN, i, mmUVD_RBC_RB_RPTR, 0);
- ring->wptr = RREG32_SOC15(UVD, i, mmUVD_RBC_RB_RPTR);
- WREG32_SOC15(UVD, i, mmUVD_RBC_RB_WPTR,
+ ring->wptr = RREG32_SOC15(VCN, i, mmUVD_RBC_RB_RPTR);
+ WREG32_SOC15(VCN, i, mmUVD_RBC_RB_WPTR,
lower_32_bits(ring->wptr));
fw_shared->multi_queue.decode_queue_mode &= ~FW_QUEUE_RING_RESET;
fw_shared->multi_queue.encode_generalpurpose_queue_mode |= FW_QUEUE_RING_RESET;
ring = &adev->vcn.inst[i].ring_enc[0];
- WREG32_SOC15(UVD, i, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
- WREG32_SOC15(UVD, i, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
- WREG32_SOC15(UVD, i, mmUVD_RB_BASE_LO, ring->gpu_addr);
- WREG32_SOC15(UVD, i, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
- WREG32_SOC15(UVD, i, mmUVD_RB_SIZE, ring->ring_size / 4);
+ WREG32_SOC15(VCN, i, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, i, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, i, mmUVD_RB_BASE_LO, ring->gpu_addr);
+ WREG32_SOC15(VCN, i, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
+ WREG32_SOC15(VCN, i, mmUVD_RB_SIZE, ring->ring_size / 4);
fw_shared->multi_queue.encode_generalpurpose_queue_mode &= ~FW_QUEUE_RING_RESET;
fw_shared->multi_queue.encode_lowlatency_queue_mode |= FW_QUEUE_RING_RESET;
ring = &adev->vcn.inst[i].ring_enc[1];
- WREG32_SOC15(UVD, i, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
- WREG32_SOC15(UVD, i, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
- WREG32_SOC15(UVD, i, mmUVD_RB_BASE_LO2, ring->gpu_addr);
- WREG32_SOC15(UVD, i, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
- WREG32_SOC15(UVD, i, mmUVD_RB_SIZE2, ring->ring_size / 4);
+ WREG32_SOC15(VCN, i, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, i, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, i, mmUVD_RB_BASE_LO2, ring->gpu_addr);
+ WREG32_SOC15(VCN, i, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
+ WREG32_SOC15(VCN, i, mmUVD_RB_SIZE2, ring->ring_size / 4);
fw_shared->multi_queue.encode_lowlatency_queue_mode &= ~FW_QUEUE_RING_RESET;
}
@@ -1118,33 +1118,33 @@ static int vcn_v2_5_mmsch_start(struct amdgpu_device *adev,
* 1, write to vce_mmsch_vf_ctx_addr_lo/hi register with GPU mc addr of
* memory descriptor location
*/
- WREG32_SOC15(UVD, 0, mmMMSCH_VF_CTX_ADDR_LO, lower_32_bits(addr));
- WREG32_SOC15(UVD, 0, mmMMSCH_VF_CTX_ADDR_HI, upper_32_bits(addr));
+ WREG32_SOC15(VCN, 0, mmMMSCH_VF_CTX_ADDR_LO, lower_32_bits(addr));
+ WREG32_SOC15(VCN, 0, mmMMSCH_VF_CTX_ADDR_HI, upper_32_bits(addr));
/* 2, update vmid of descriptor */
- data = RREG32_SOC15(UVD, 0, mmMMSCH_VF_VMID);
+ data = RREG32_SOC15(VCN, 0, mmMMSCH_VF_VMID);
data &= ~MMSCH_VF_VMID__VF_CTX_VMID_MASK;
/* use domain0 for MM scheduler */
data |= (0 << MMSCH_VF_VMID__VF_CTX_VMID__SHIFT);
- WREG32_SOC15(UVD, 0, mmMMSCH_VF_VMID, data);
+ WREG32_SOC15(VCN, 0, mmMMSCH_VF_VMID, data);
/* 3, notify mmsch about the size of this descriptor */
- WREG32_SOC15(UVD, 0, mmMMSCH_VF_CTX_SIZE, size);
+ WREG32_SOC15(VCN, 0, mmMMSCH_VF_CTX_SIZE, size);
/* 4, set resp to zero */
- WREG32_SOC15(UVD, 0, mmMMSCH_VF_MAILBOX_RESP, 0);
+ WREG32_SOC15(VCN, 0, mmMMSCH_VF_MAILBOX_RESP, 0);
/*
* 5, kick off the initialization and wait until
* VCE_MMSCH_VF_MAILBOX_RESP becomes non-zero
*/
- WREG32_SOC15(UVD, 0, mmMMSCH_VF_MAILBOX_HOST, 0x10000001);
+ WREG32_SOC15(VCN, 0, mmMMSCH_VF_MAILBOX_HOST, 0x10000001);
- data = RREG32_SOC15(UVD, 0, mmMMSCH_VF_MAILBOX_RESP);
+ data = RREG32_SOC15(VCN, 0, mmMMSCH_VF_MAILBOX_RESP);
loop = 10;
while ((data & 0x10000002) != 0x10000002) {
udelay(100);
- data = RREG32_SOC15(UVD, 0, mmMMSCH_VF_MAILBOX_RESP);
+ data = RREG32_SOC15(VCN, 0, mmMMSCH_VF_MAILBOX_RESP);
loop--;
if (!loop)
break;
@@ -1187,93 +1187,93 @@ static int vcn_v2_5_sriov_start(struct amdgpu_device *adev)
table_size = 0;
MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_STATUS),
~UVD_STATUS__UVD_BUSY, UVD_STATUS__UVD_BUSY);
size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4);
/* mc resume*/
if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + i].tmr_mc_addr_lo);
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + i].tmr_mc_addr_hi);
offset = 0;
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET0), 0);
+ SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_OFFSET0), 0);
} else {
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
lower_32_bits(adev->vcn.inst[i].gpu_addr));
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
upper_32_bits(adev->vcn.inst[i].gpu_addr));
offset = size;
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET0),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_OFFSET0),
AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
}
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE0),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_SIZE0),
size);
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW),
lower_32_bits(adev->vcn.inst[i].gpu_addr + offset));
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH),
upper_32_bits(adev->vcn.inst[i].gpu_addr + offset));
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET1),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_OFFSET1),
0);
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE1),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_SIZE1),
AMDGPU_VCN_STACK_SIZE);
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW),
lower_32_bits(adev->vcn.inst[i].gpu_addr + offset +
AMDGPU_VCN_STACK_SIZE));
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH),
upper_32_bits(adev->vcn.inst[i].gpu_addr + offset +
AMDGPU_VCN_STACK_SIZE));
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET2),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_OFFSET2),
0);
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE2),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_SIZE2),
AMDGPU_VCN_CONTEXT_SIZE);
ring = &adev->vcn.inst[i].ring_enc[0];
ring->wptr = 0;
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_RB_BASE_LO),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_RB_BASE_LO),
lower_32_bits(ring->gpu_addr));
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_RB_BASE_HI),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_RB_BASE_HI),
upper_32_bits(ring->gpu_addr));
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_RB_SIZE),
+ SOC15_REG_OFFSET(VCN, i, mmUVD_RB_SIZE),
ring->ring_size / 4);
ring = &adev->vcn.inst[i].ring_dec;
ring->wptr = 0;
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_RBC_RB_64BIT_BAR_LOW),
lower_32_bits(ring->gpu_addr));
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i,
+ SOC15_REG_OFFSET(VCN, i,
mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH),
upper_32_bits(ring->gpu_addr));
@@ -1285,7 +1285,7 @@ static int vcn_v2_5_sriov_start(struct amdgpu_device *adev)
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);
MMSCH_V1_0_INSERT_DIRECT_WT(
- SOC15_REG_OFFSET(UVD, i, mmUVD_RBC_RB_CNTL), tmp);
+ SOC15_REG_OFFSET(VCN, i, mmUVD_RBC_RB_CNTL), tmp);
/* add end packet */
memcpy((void *)init_table, &end, sizeof(struct mmsch_v1_0_cmd_end));
@@ -1306,24 +1306,24 @@ static int vcn_v2_5_stop_dpg_mode(struct amdgpu_device *adev, int inst_idx)
uint32_t tmp;
/* Wait for power status to be 1 */
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, 1,
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 1,
UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code);
/* wait for read ptr to be equal to write ptr */
- tmp = RREG32_SOC15(UVD, inst_idx, mmUVD_RB_WPTR);
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_RB_RPTR, tmp, 0xFFFFFFFF, ret_code);
+ tmp = RREG32_SOC15(VCN, inst_idx, mmUVD_RB_WPTR);
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_RB_RPTR, tmp, 0xFFFFFFFF, ret_code);
- tmp = RREG32_SOC15(UVD, inst_idx, mmUVD_RB_WPTR2);
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_RB_RPTR2, tmp, 0xFFFFFFFF, ret_code);
+ tmp = RREG32_SOC15(VCN, inst_idx, mmUVD_RB_WPTR2);
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_RB_RPTR2, tmp, 0xFFFFFFFF, ret_code);
- tmp = RREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_WPTR) & 0x7FFFFFFF;
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_RBC_RB_RPTR, tmp, 0xFFFFFFFF, ret_code);
+ tmp = RREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_WPTR) & 0x7FFFFFFF;
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_RBC_RB_RPTR, tmp, 0xFFFFFFFF, ret_code);
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, 1,
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 1,
UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code);
/* disable dynamic power gating mode */
- WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS), 0,
~UVD_POWER_STATUS__UVD_PG_MODE_MASK);
return 0;
@@ -1367,17 +1367,17 @@ static int vcn_v2_5_stop(struct amdgpu_device *adev)
return r;
/* block VCPU register access */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_ARB_CTRL),
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_RB_ARB_CTRL),
UVD_RB_ARB_CTRL__VCPU_DIS_MASK,
~UVD_RB_ARB_CTRL__VCPU_DIS_MASK);
/* reset VCPU */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL),
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL),
UVD_VCPU_CNTL__BLK_RST_MASK,
~UVD_VCPU_CNTL__BLK_RST_MASK);
/* disable VCPU clock */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), 0,
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), 0,
~(UVD_VCPU_CNTL__CLK_EN_MASK));
/* clear status */
@@ -1386,7 +1386,7 @@ static int vcn_v2_5_stop(struct amdgpu_device *adev)
vcn_v2_5_enable_clock_gating(adev);
/* enable register anti-hang mechanism */
- WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_POWER_STATUS),
+ WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_POWER_STATUS),
UVD_POWER_STATUS__UVD_POWER_STATUS_MASK,
~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK);
}
@@ -1408,11 +1408,11 @@ static int vcn_v2_5_pause_dpg_mode(struct amdgpu_device *adev,
if (adev->vcn.inst[inst_idx].pause_state.fw_based != new_state->fw_based) {
DRM_DEBUG("dpg pause state changed %d -> %d",
adev->vcn.inst[inst_idx].pause_state.fw_based, new_state->fw_based);
- reg_data = RREG32_SOC15(UVD, inst_idx, mmUVD_DPG_PAUSE) &
+ reg_data = RREG32_SOC15(VCN, inst_idx, mmUVD_DPG_PAUSE) &
(~UVD_DPG_PAUSE__NJ_PAUSE_DPG_ACK_MASK);
if (new_state->fw_based == VCN_DPG_STATE__PAUSE) {
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, 0x1,
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 0x1,
UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code);
if (!ret_code) {
@@ -1420,15 +1420,15 @@ static int vcn_v2_5_pause_dpg_mode(struct amdgpu_device *adev,
/* pause DPG */
reg_data |= UVD_DPG_PAUSE__NJ_PAUSE_DPG_REQ_MASK;
- WREG32_SOC15(UVD, inst_idx, mmUVD_DPG_PAUSE, reg_data);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_DPG_PAUSE, reg_data);
/* wait for ACK */
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_DPG_PAUSE,
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_DPG_PAUSE,
UVD_DPG_PAUSE__NJ_PAUSE_DPG_ACK_MASK,
UVD_DPG_PAUSE__NJ_PAUSE_DPG_ACK_MASK, ret_code);
/* Stall DPG before WPTR/RPTR reset */
- WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS),
+ WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS),
UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK,
~UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK);
@@ -1436,39 +1436,39 @@ static int vcn_v2_5_pause_dpg_mode(struct amdgpu_device *adev,
fw_shared->multi_queue.encode_generalpurpose_queue_mode |= FW_QUEUE_RING_RESET;
ring = &adev->vcn.inst[inst_idx].ring_enc[0];
ring->wptr = 0;
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_BASE_LO, ring->gpu_addr);
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_SIZE, ring->ring_size / 4);
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_BASE_LO, ring->gpu_addr);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_SIZE, ring->ring_size / 4);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
fw_shared->multi_queue.encode_generalpurpose_queue_mode &= ~FW_QUEUE_RING_RESET;
fw_shared->multi_queue.encode_lowlatency_queue_mode |= FW_QUEUE_RING_RESET;
ring = &adev->vcn.inst[inst_idx].ring_enc[1];
ring->wptr = 0;
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_BASE_LO2, ring->gpu_addr);
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_SIZE2, ring->ring_size / 4);
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
- WREG32_SOC15(UVD, inst_idx, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_BASE_LO2, ring->gpu_addr);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_SIZE2, ring->ring_size / 4);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
fw_shared->multi_queue.encode_lowlatency_queue_mode &= ~FW_QUEUE_RING_RESET;
fw_shared->multi_queue.decode_queue_mode |= FW_QUEUE_RING_RESET;
- WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_WPTR,
- RREG32_SOC15(UVD, inst_idx, mmUVD_SCRATCH2) & 0x7FFFFFFF);
+ WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_WPTR,
+ RREG32_SOC15(VCN, inst_idx, mmUVD_SCRATCH2) & 0x7FFFFFFF);
fw_shared->multi_queue.decode_queue_mode &= ~FW_QUEUE_RING_RESET;
/* Unstall DPG */
- WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS),
+ WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS),
0, ~UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK);
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS,
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS,
UVD_PGFSM_CONFIG__UVDM_UVDU_PWR_ON, UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code);
}
} else {
reg_data &= ~UVD_DPG_PAUSE__NJ_PAUSE_DPG_REQ_MASK;
- WREG32_SOC15(UVD, inst_idx, mmUVD_DPG_PAUSE, reg_data);
- SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, 0x1,
+ WREG32_SOC15(VCN, inst_idx, mmUVD_DPG_PAUSE, reg_data);
+ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 0x1,
UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code);
}
adev->vcn.inst[inst_idx].pause_state.fw_based = new_state->fw_based;
@@ -1488,7 +1488,7 @@ static uint64_t vcn_v2_5_dec_ring_get_rptr(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
- return RREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_RPTR);
+ return RREG32_SOC15(VCN, ring->me, mmUVD_RBC_RB_RPTR);
}
/**
@@ -1505,7 +1505,7 @@ static uint64_t vcn_v2_5_dec_ring_get_wptr(struct amdgpu_ring *ring)
if (ring->use_doorbell)
return adev->wb.wb[ring->wptr_offs];
else
- return RREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_WPTR);
+ return RREG32_SOC15(VCN, ring->me, mmUVD_RBC_RB_WPTR);
}
/**
@@ -1520,14 +1520,14 @@ static void vcn_v2_5_dec_ring_set_wptr(struct amdgpu_ring *ring)
struct amdgpu_device *adev = ring->adev;
if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG)
- WREG32_SOC15(UVD, ring->me, mmUVD_SCRATCH2,
+ WREG32_SOC15(VCN, ring->me, mmUVD_SCRATCH2,
lower_32_bits(ring->wptr) | 0x80000000);
if (ring->use_doorbell) {
adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr);
WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr));
} else {
- WREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, ring->me, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr));
}
}
@@ -1573,9 +1573,9 @@ static uint64_t vcn_v2_5_enc_ring_get_rptr(struct amdgpu_ring *ring)
struct amdgpu_device *adev = ring->adev;
if (ring == &adev->vcn.inst[ring->me].ring_enc[0])
- return RREG32_SOC15(UVD, ring->me, mmUVD_RB_RPTR);
+ return RREG32_SOC15(VCN, ring->me, mmUVD_RB_RPTR);
else
- return RREG32_SOC15(UVD, ring->me, mmUVD_RB_RPTR2);
+ return RREG32_SOC15(VCN, ring->me, mmUVD_RB_RPTR2);
}
/**
@@ -1593,12 +1593,12 @@ static uint64_t vcn_v2_5_enc_ring_get_wptr(struct amdgpu_ring *ring)
if (ring->use_doorbell)
return adev->wb.wb[ring->wptr_offs];
else
- return RREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR);
+ return RREG32_SOC15(VCN, ring->me, mmUVD_RB_WPTR);
} else {
if (ring->use_doorbell)
return adev->wb.wb[ring->wptr_offs];
else
- return RREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR2);
+ return RREG32_SOC15(VCN, ring->me, mmUVD_RB_WPTR2);
}
}
@@ -1618,14 +1618,14 @@ static void vcn_v2_5_enc_ring_set_wptr(struct amdgpu_ring *ring)
adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr);
WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr));
} else {
- WREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, ring->me, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
}
} else {
if (ring->use_doorbell) {
adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr);
WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr));
} else {
- WREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
+ WREG32_SOC15(VCN, ring->me, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
}
}
}
--
2.25.4
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH 2/8] drm/amdgpu: check SMU NULL ptr on gfx hw init
2020-04-30 18:37 [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5 Alex Deucher
@ 2020-04-30 18:37 ` Alex Deucher
2020-04-30 18:37 ` [PATCH 3/8] drm/amdgpu: update the method to set kcq queue mask Alex Deucher
` (5 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Alex Deucher @ 2020-04-30 18:37 UTC (permalink / raw)
To: amd-gfx; +Cc: Alex Deucher, Likun Gao, Hawking Zhang
From: Likun Gao <Likun.Gao@amd.com>
Check SMU NULL ptr before load smu fw.
Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 16 +++++++++-------
1 file changed, 9 insertions(+), 7 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
index 63ac4308b9c4..20258c5d335f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
@@ -6540,14 +6540,16 @@ static int gfx_v10_0_hw_init(void *handle)
* loaded firstly, so in direct type, it has to load smc ucode
* here before rlc.
*/
- r = smu_load_microcode(&adev->smu);
- if (r)
- return r;
+ if (adev->smu.ppt_funcs != NULL) {
+ r = smu_load_microcode(&adev->smu);
+ if (r)
+ return r;
- r = smu_check_fw_status(&adev->smu);
- if (r) {
- pr_err("SMC firmware status is not correct\n");
- return r;
+ r = smu_check_fw_status(&adev->smu);
+ if (r) {
+ pr_err("SMC firmware status is not correct\n");
+ return r;
+ }
}
}
--
2.25.4
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH 3/8] drm/amdgpu: update the method to set kcq queue mask
2020-04-30 18:37 [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5 Alex Deucher
2020-04-30 18:37 ` [PATCH 2/8] drm/amdgpu: check SMU NULL ptr on gfx hw init Alex Deucher
@ 2020-04-30 18:37 ` Alex Deucher
2020-04-30 18:37 ` [PATCH 4/8] drm/amdgpu: re-structue members for ip discovery Alex Deucher
` (4 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Alex Deucher @ 2020-04-30 18:37 UTC (permalink / raw)
To: amd-gfx; +Cc: Alex Deucher, Likun Gao, Huang Rui
From: Likun Gao <Likun.Gao@amd.com>
Use a common method to set queue mask before set kiq resource.
The value of queue mask must suitablt for the designated form.
Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 15 ++++++++++++++-
1 file changed, 14 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
index 0103acc57474..ed212c070e8e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
@@ -485,6 +485,19 @@ int amdgpu_gfx_disable_kcq(struct amdgpu_device *adev)
return amdgpu_ring_test_helper(kiq_ring);
}
+int amdgpu_gfx_kcq_queue_mask_transform(struct amdgpu_device *adev,
+ int queue_bit)
+{
+ int mec, pipe, queue;
+ int queue_kcq_bit = 0;
+
+ amdgpu_gfx_bit_to_mec_queue(adev, queue_bit, &mec, &pipe, &queue);
+
+ queue_kcq_bit = mec * 4 * 8 + pipe * 8 + queue;
+
+ return queue_kcq_bit;
+}
+
int amdgpu_gfx_enable_kcq(struct amdgpu_device *adev)
{
struct amdgpu_kiq *kiq = &adev->gfx.kiq;
@@ -507,7 +520,7 @@ int amdgpu_gfx_enable_kcq(struct amdgpu_device *adev)
break;
}
- queue_mask |= (1ull << i);
+ queue_mask |= (1ull << amdgpu_gfx_kcq_queue_mask_transform(adev, i));
}
DRM_INFO("kiq ring mec %d pipe %d q %d\n", kiq_ring->me, kiq_ring->pipe,
--
2.25.4
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH 4/8] drm/amdgpu: re-structue members for ip discovery
2020-04-30 18:37 [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5 Alex Deucher
2020-04-30 18:37 ` [PATCH 2/8] drm/amdgpu: check SMU NULL ptr on gfx hw init Alex Deucher
2020-04-30 18:37 ` [PATCH 3/8] drm/amdgpu: update the method to set kcq queue mask Alex Deucher
@ 2020-04-30 18:37 ` Alex Deucher
2020-04-30 18:37 ` [PATCH 5/8] drm/amdgpu: skip reservation of discovery tmr region in pre-Navi Alex Deucher
` (3 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Alex Deucher @ 2020-04-30 18:37 UTC (permalink / raw)
To: amd-gfx; +Cc: Alex Deucher, Likun Gao, Hawking Zhang
From: Hawking Zhang <Hawking.Zhang@amd.com>
This is to prepare for initializing discovery tmr size per
ASIC type
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Likun Gao <Likun.Gao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu.h | 5 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 56 ++++++++++---------
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 12 ++--
3 files changed, 38 insertions(+), 35 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index bc1e0fd71a09..6048c5902fad 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -768,7 +768,6 @@ struct amdgpu_device {
uint8_t *bios;
uint32_t bios_size;
struct amdgpu_bo *stolen_vga_memory;
- struct amdgpu_bo *discovery_memory;
uint32_t bios_scratch_reg_offset;
uint32_t bios_scratch[AMDGPU_BIOS_NUM_SCRATCH];
@@ -921,7 +920,9 @@ struct amdgpu_device {
struct amdgpu_display_manager dm;
/* discovery */
- uint8_t *discovery;
+ uint8_t *discovery_bin;
+ uint32_t discovery_tmr_size;
+ struct amdgpu_bo *discovery_memory;
/* mes */
bool enable_mes;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
index 43bb22ad8add..b5d6274952a5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c
@@ -133,9 +133,10 @@ static int hw_id_map[MAX_HWIP] = {
static int amdgpu_discovery_read_binary(struct amdgpu_device *adev, uint8_t *binary)
{
uint64_t vram_size = (uint64_t)RREG32(mmRCC_CONFIG_MEMSIZE) << 20;
- uint64_t pos = vram_size - DISCOVERY_TMR_SIZE;
+ uint64_t pos = vram_size - adev->discovery_tmr_size;
- amdgpu_device_vram_access(adev, pos, (uint32_t *)binary, DISCOVERY_TMR_SIZE, false);
+ amdgpu_device_vram_access(adev, pos, (uint32_t *)binary,
+ adev->discovery_tmr_size, false);
return 0;
}
@@ -167,17 +168,18 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev)
uint16_t checksum;
int r;
- adev->discovery = kzalloc(DISCOVERY_TMR_SIZE, GFP_KERNEL);
- if (!adev->discovery)
+ adev->discovery_tmr_size = DISCOVERY_TMR_SIZE;
+ adev->discovery_bin = kzalloc(adev->discovery_tmr_size, GFP_KERNEL);
+ if (!adev->discovery_bin)
return -ENOMEM;
- r = amdgpu_discovery_read_binary(adev, adev->discovery);
+ r = amdgpu_discovery_read_binary(adev, adev->discovery_bin);
if (r) {
DRM_ERROR("failed to read ip discovery binary\n");
goto out;
}
- bhdr = (struct binary_header *)adev->discovery;
+ bhdr = (struct binary_header *)adev->discovery_bin;
if (le32_to_cpu(bhdr->binary_signature) != BINARY_SIGNATURE) {
DRM_ERROR("invalid ip discovery binary signature\n");
@@ -190,7 +192,7 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev)
size = bhdr->binary_size - offset;
checksum = bhdr->binary_checksum;
- if (!amdgpu_discovery_verify_checksum(adev->discovery + offset,
+ if (!amdgpu_discovery_verify_checksum(adev->discovery_bin + offset,
size, checksum)) {
DRM_ERROR("invalid ip discovery binary checksum\n");
r = -EINVAL;
@@ -200,7 +202,7 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev)
info = &bhdr->table_list[IP_DISCOVERY];
offset = le16_to_cpu(info->offset);
checksum = le16_to_cpu(info->checksum);
- ihdr = (struct ip_discovery_header *)(adev->discovery + offset);
+ ihdr = (struct ip_discovery_header *)(adev->discovery_bin + offset);
if (le32_to_cpu(ihdr->signature) != DISCOVERY_TABLE_SIGNATURE) {
DRM_ERROR("invalid ip discovery data table signature\n");
@@ -208,7 +210,7 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev)
goto out;
}
- if (!amdgpu_discovery_verify_checksum(adev->discovery + offset,
+ if (!amdgpu_discovery_verify_checksum(adev->discovery_bin + offset,
ihdr->size, checksum)) {
DRM_ERROR("invalid ip discovery data table checksum\n");
r = -EINVAL;
@@ -218,9 +220,9 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev)
info = &bhdr->table_list[GC];
offset = le16_to_cpu(info->offset);
checksum = le16_to_cpu(info->checksum);
- ghdr = (struct gpu_info_header *)(adev->discovery + offset);
+ ghdr = (struct gpu_info_header *)(adev->discovery_bin + offset);
- if (!amdgpu_discovery_verify_checksum(adev->discovery + offset,
+ if (!amdgpu_discovery_verify_checksum(adev->discovery_bin + offset,
ghdr->size, checksum)) {
DRM_ERROR("invalid gc data table checksum\n");
r = -EINVAL;
@@ -230,16 +232,16 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev)
return 0;
out:
- kfree(adev->discovery);
- adev->discovery = NULL;
+ kfree(adev->discovery_bin);
+ adev->discovery_bin = NULL;
return r;
}
void amdgpu_discovery_fini(struct amdgpu_device *adev)
{
- kfree(adev->discovery);
- adev->discovery = NULL;
+ kfree(adev->discovery_bin);
+ adev->discovery_bin = NULL;
}
int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev)
@@ -263,8 +265,8 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev)
return r;
}
- bhdr = (struct binary_header *)adev->discovery;
- ihdr = (struct ip_discovery_header *)(adev->discovery +
+ bhdr = (struct binary_header *)adev->discovery_bin;
+ ihdr = (struct ip_discovery_header *)(adev->discovery_bin +
le16_to_cpu(bhdr->table_list[IP_DISCOVERY].offset));
num_dies = le16_to_cpu(ihdr->num_dies);
@@ -272,7 +274,7 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev)
for (i = 0; i < num_dies; i++) {
die_offset = le16_to_cpu(ihdr->die_info[i].die_offset);
- dhdr = (struct die_header *)(adev->discovery + die_offset);
+ dhdr = (struct die_header *)(adev->discovery_bin + die_offset);
num_ips = le16_to_cpu(dhdr->num_ips);
ip_offset = die_offset + sizeof(*dhdr);
@@ -286,7 +288,7 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev)
le16_to_cpu(dhdr->die_id), num_ips);
for (j = 0; j < num_ips; j++) {
- ip = (struct ip *)(adev->discovery + ip_offset);
+ ip = (struct ip *)(adev->discovery_bin + ip_offset);
num_base_address = ip->num_base_address;
DRM_DEBUG("%s(%d) #%d v%d.%d.%d:\n",
@@ -335,24 +337,24 @@ int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id,
uint16_t num_ips;
int i, j;
- if (!adev->discovery) {
+ if (!adev->discovery_bin) {
DRM_ERROR("ip discovery uninitialized\n");
return -EINVAL;
}
- bhdr = (struct binary_header *)adev->discovery;
- ihdr = (struct ip_discovery_header *)(adev->discovery +
+ bhdr = (struct binary_header *)adev->discovery_bin;
+ ihdr = (struct ip_discovery_header *)(adev->discovery_bin +
le16_to_cpu(bhdr->table_list[IP_DISCOVERY].offset));
num_dies = le16_to_cpu(ihdr->num_dies);
for (i = 0; i < num_dies; i++) {
die_offset = le16_to_cpu(ihdr->die_info[i].die_offset);
- dhdr = (struct die_header *)(adev->discovery + die_offset);
+ dhdr = (struct die_header *)(adev->discovery_bin + die_offset);
num_ips = le16_to_cpu(dhdr->num_ips);
ip_offset = die_offset + sizeof(*dhdr);
for (j = 0; j < num_ips; j++) {
- ip = (struct ip *)(adev->discovery + ip_offset);
+ ip = (struct ip *)(adev->discovery_bin + ip_offset);
if (le16_to_cpu(ip->hw_id) == hw_id) {
if (major)
@@ -375,13 +377,13 @@ int amdgpu_discovery_get_gfx_info(struct amdgpu_device *adev)
struct binary_header *bhdr;
struct gc_info_v1_0 *gc_info;
- if (!adev->discovery) {
+ if (!adev->discovery_bin) {
DRM_ERROR("ip discovery uninitialized\n");
return -EINVAL;
}
- bhdr = (struct binary_header *)adev->discovery;
- gc_info = (struct gc_info_v1_0 *)(adev->discovery +
+ bhdr = (struct binary_header *)adev->discovery_bin;
+ gc_info = (struct gc_info_v1_0 *)(adev->discovery_bin +
le16_to_cpu(bhdr->table_list[GC].offset));
adev->gfx.config.max_shader_engines = le32_to_cpu(gc_info->gc_num_se);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index ea0199a8f9c9..085d7c238163 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -1958,15 +1958,15 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
return r;
/*
- * reserve one TMR (64K) memory at the top of VRAM which holds
+ * reserve TMR memory at the top of VRAM which holds
* IP Discovery data and is protected by PSP.
*/
r = amdgpu_bo_create_kernel_at(adev,
- adev->gmc.real_vram_size - DISCOVERY_TMR_SIZE,
- DISCOVERY_TMR_SIZE,
- AMDGPU_GEM_DOMAIN_VRAM,
- &adev->discovery_memory,
- NULL);
+ adev->gmc.real_vram_size - adev->discovery_tmr_size,
+ adev->discovery_tmr_size,
+ AMDGPU_GEM_DOMAIN_VRAM,
+ &adev->discovery_memory,
+ NULL);
if (r)
return r;
--
2.25.4
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH 5/8] drm/amdgpu: skip reservation of discovery tmr region in pre-Navi
2020-04-30 18:37 [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5 Alex Deucher
` (2 preceding siblings ...)
2020-04-30 18:37 ` [PATCH 4/8] drm/amdgpu: re-structue members for ip discovery Alex Deucher
@ 2020-04-30 18:37 ` Alex Deucher
2020-04-30 18:37 ` [PATCH 6/8] drm/amdgpu: Rename amdgpu_gfx_kcq_queue_mask_transform() Alex Deucher
` (2 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Alex Deucher @ 2020-04-30 18:37 UTC (permalink / raw)
To: amd-gfx; +Cc: Alex Deucher, Likun Gao, Hawking Zhang
From: Hawking Zhang <Hawking.Zhang@amd.com>
IP discovery is only supported in Navi series and onwards.
There is no need to reserve a portion of vram as discovery
tmr region for pre-Navi adapters.
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Likun Gao <Likun.Gao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 085d7c238163..93d057f92dcd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -1961,14 +1961,16 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
* reserve TMR memory at the top of VRAM which holds
* IP Discovery data and is protected by PSP.
*/
- r = amdgpu_bo_create_kernel_at(adev,
+ if (adev->discovery_tmr_size > 0) {
+ r = amdgpu_bo_create_kernel_at(adev,
adev->gmc.real_vram_size - adev->discovery_tmr_size,
adev->discovery_tmr_size,
AMDGPU_GEM_DOMAIN_VRAM,
&adev->discovery_memory,
NULL);
- if (r)
- return r;
+ if (r)
+ return r;
+ }
DRM_INFO("amdgpu: %uM of VRAM memory ready\n",
(unsigned) (adev->gmc.real_vram_size / (1024 * 1024)));
--
2.25.4
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH 6/8] drm/amdgpu: Rename amdgpu_gfx_kcq_queue_mask_transform()
2020-04-30 18:37 [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5 Alex Deucher
` (3 preceding siblings ...)
2020-04-30 18:37 ` [PATCH 5/8] drm/amdgpu: skip reservation of discovery tmr region in pre-Navi Alex Deucher
@ 2020-04-30 18:37 ` Alex Deucher
2020-04-30 18:37 ` [PATCH 7/8] drm/amdkfd: Use a systematic method to calculate queue mask bit Alex Deucher
2020-04-30 18:37 ` [PATCH 8/8] amd/amdgpu: Limit rlcg write registers only for nv12 Alex Deucher
6 siblings, 0 replies; 8+ messages in thread
From: Alex Deucher @ 2020-04-30 18:37 UTC (permalink / raw)
To: amd-gfx; +Cc: Alex Deucher, Yong Zhao
From: Yong Zhao <Yong.Zhao@amd.com>
Rename it to amdgpu_queue_mask_bit_to_set_resource_bit() to be more
specific about its functionality. KFD will use it later.
Signed-off-by: Yong Zhao <Yong.Zhao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 16 ++++++++--------
drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h | 2 +-
2 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
index ed212c070e8e..d612033a23ac 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
@@ -48,7 +48,7 @@ int amdgpu_gfx_mec_queue_to_bit(struct amdgpu_device *adev, int mec,
return bit;
}
-void amdgpu_gfx_bit_to_mec_queue(struct amdgpu_device *adev, int bit,
+void amdgpu_queue_mask_bit_to_mec_queue(struct amdgpu_device *adev, int bit,
int *mec, int *pipe, int *queue)
{
*queue = bit % adev->gfx.mec.num_queue_per_pipe;
@@ -274,7 +274,7 @@ static int amdgpu_gfx_kiq_acquire(struct amdgpu_device *adev,
if (test_bit(queue_bit, adev->gfx.mec.queue_bitmap))
continue;
- amdgpu_gfx_bit_to_mec_queue(adev, queue_bit, &mec, &pipe, &queue);
+ amdgpu_queue_mask_bit_to_mec_queue(adev, queue_bit, &mec, &pipe, &queue);
/*
* 1. Using pipes 2/3 from MEC 2 seems cause problems.
@@ -485,17 +485,17 @@ int amdgpu_gfx_disable_kcq(struct amdgpu_device *adev)
return amdgpu_ring_test_helper(kiq_ring);
}
-int amdgpu_gfx_kcq_queue_mask_transform(struct amdgpu_device *adev,
+int amdgpu_queue_mask_bit_to_set_resource_bit(struct amdgpu_device *adev,
int queue_bit)
{
int mec, pipe, queue;
- int queue_kcq_bit = 0;
+ int set_resource_bit = 0;
- amdgpu_gfx_bit_to_mec_queue(adev, queue_bit, &mec, &pipe, &queue);
+ amdgpu_queue_mask_bit_to_mec_queue(adev, queue_bit, &mec, &pipe, &queue);
- queue_kcq_bit = mec * 4 * 8 + pipe * 8 + queue;
+ set_resource_bit = mec * 4 * 8 + pipe * 8 + queue;
- return queue_kcq_bit;
+ return set_resource_bit;
}
int amdgpu_gfx_enable_kcq(struct amdgpu_device *adev)
@@ -520,7 +520,7 @@ int amdgpu_gfx_enable_kcq(struct amdgpu_device *adev)
break;
}
- queue_mask |= (1ull << amdgpu_gfx_kcq_queue_mask_transform(adev, i));
+ queue_mask |= (1ull << amdgpu_queue_mask_bit_to_set_resource_bit(adev, i));
}
DRM_INFO("kiq ring mec %d pipe %d q %d\n", kiq_ring->me, kiq_ring->pipe,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
index ee698f0246d8..d43c11671a38 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
@@ -364,7 +364,7 @@ void amdgpu_gfx_graphics_queue_acquire(struct amdgpu_device *adev);
int amdgpu_gfx_mec_queue_to_bit(struct amdgpu_device *adev, int mec,
int pipe, int queue);
-void amdgpu_gfx_bit_to_mec_queue(struct amdgpu_device *adev, int bit,
+void amdgpu_queue_mask_bit_to_mec_queue(struct amdgpu_device *adev, int bit,
int *mec, int *pipe, int *queue);
bool amdgpu_gfx_is_mec_queue_enabled(struct amdgpu_device *adev, int mec,
int pipe, int queue);
--
2.25.4
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH 7/8] drm/amdkfd: Use a systematic method to calculate queue mask bit
2020-04-30 18:37 [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5 Alex Deucher
` (4 preceding siblings ...)
2020-04-30 18:37 ` [PATCH 6/8] drm/amdgpu: Rename amdgpu_gfx_kcq_queue_mask_transform() Alex Deucher
@ 2020-04-30 18:37 ` Alex Deucher
2020-04-30 18:37 ` [PATCH 8/8] amd/amdgpu: Limit rlcg write registers only for nv12 Alex Deucher
6 siblings, 0 replies; 8+ messages in thread
From: Alex Deucher @ 2020-04-30 18:37 UTC (permalink / raw)
To: amd-gfx; +Cc: Alex Deucher, Yong Zhao
From: Yong Zhao <Yong.Zhao@amd.com>
The queue mask used for set_resources always assumes the queue number
per pipe is 8, so KFD needs to align with that by using function
amdgpu_queue_mask_bit_to_set_resource_bit().
Signed-off-by: Yong Zhao <Yong.Zhao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 3 +++
drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 4 +++-
2 files changed, 6 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index a501026e829c..3f2b695cf19e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -149,6 +149,9 @@ int amdgpu_amdkfd_post_reset(struct amdgpu_device *adev);
void amdgpu_amdkfd_gpu_reset(struct kgd_dev *kgd);
+int amdgpu_queue_mask_bit_to_set_resource_bit(struct amdgpu_device *adev,
+ int queue_bit);
+
/* Shared API */
int amdgpu_amdkfd_alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
void **mem_obj, uint64_t *gpu_addr,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index ae9547791813..e9c4867abeff 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -1089,7 +1089,9 @@ static int set_sched_resources(struct device_queue_manager *dqm)
break;
}
- res.queue_mask |= (1ull << i);
+ res.queue_mask |= 1ull
+ << amdgpu_queue_mask_bit_to_set_resource_bit(
+ (struct amdgpu_device *)dqm->dev->kgd, i);
}
res.gws_mask = ~0ull;
res.oac_mask = res.gds_heap_base = res.gds_heap_size = 0;
--
2.25.4
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH 8/8] amd/amdgpu: Limit rlcg write registers only for nv12
2020-04-30 18:37 [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5 Alex Deucher
` (5 preceding siblings ...)
2020-04-30 18:37 ` [PATCH 7/8] drm/amdkfd: Use a systematic method to calculate queue mask bit Alex Deucher
@ 2020-04-30 18:37 ` Alex Deucher
6 siblings, 0 replies; 8+ messages in thread
From: Alex Deucher @ 2020-04-30 18:37 UTC (permalink / raw)
To: amd-gfx; +Cc: Alex Deucher, shaoyunl
From: shaoyunl <shaoyun.liu@amd.com>
Create gfx_v10_0_rlc_funcs_sriov for nv12 with rlcg_write function pointers be initialized
so driver can use RLCG to write aceess CSIB and CP_ME_CNTL registers when nv12 in sriov mode
Signed-off-by: shaoyunl <shaoyun.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 32 ++++++++++++++++++++------
1 file changed, 25 insertions(+), 7 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
index 20258c5d335f..7ff7e516b72f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
@@ -4577,11 +4577,13 @@ static int gfx_v10_0_init_csb(struct amdgpu_device *adev)
adev->gfx.rlc.funcs->get_csb_buffer(adev, adev->gfx.rlc.cs_ptr);
/* csib */
- WREG32_SOC15_RLC(GC, 0, mmRLC_CSIB_ADDR_HI,
- adev->gfx.rlc.clear_state_gpu_addr >> 32);
- WREG32_SOC15_RLC(GC, 0, mmRLC_CSIB_ADDR_LO,
- adev->gfx.rlc.clear_state_gpu_addr & 0xfffffffc);
- WREG32_SOC15_RLC(GC, 0, mmRLC_CSIB_LENGTH, adev->gfx.rlc.clear_state_size);
+ /* amdgpu_mm_wreg_mmio_rlc will fall back to mmio if doesn't support rlcg_write */
+ amdgpu_mm_wreg_mmio_rlc(adev, SOC15_REG_OFFSET(GC, 0, mmRLC_CSIB_ADDR_HI),
+ adev->gfx.rlc.clear_state_gpu_addr >> 32, 0);
+ amdgpu_mm_wreg_mmio_rlc(adev, SOC15_REG_OFFSET(GC, 0, mmRLC_CSIB_ADDR_LO),
+ adev->gfx.rlc.clear_state_gpu_addr & 0xfffffffc, 0);
+ amdgpu_mm_wreg_mmio_rlc(adev, SOC15_REG_OFFSET(GC, 0, mmRLC_CSIB_LENGTH),
+ adev->gfx.rlc.clear_state_size, 0);
return 0;
}
@@ -5190,7 +5192,7 @@ static int gfx_v10_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable)
tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_HALT, enable ? 0 : 1);
tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_HALT, enable ? 0 : 1);
tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_HALT, enable ? 0 : 1);
- WREG32_SOC15_RLC(GC, 0, mmCP_ME_CNTL, tmp);
+ amdgpu_mm_wreg_mmio_rlc(adev, SOC15_REG_OFFSET(GC, 0, mmCP_ME_CNTL), tmp, 0);
for (i = 0; i < adev->usec_timeout; i++) {
if (RREG32_SOC15(GC, 0, mmCP_STAT) == 0)
@@ -7088,6 +7090,20 @@ static const struct amdgpu_rlc_funcs gfx_v10_0_rlc_funcs = {
.reset = gfx_v10_0_rlc_reset,
.start = gfx_v10_0_rlc_start,
.update_spm_vmid = gfx_v10_0_update_spm_vmid,
+};
+
+static const struct amdgpu_rlc_funcs gfx_v10_0_rlc_funcs_sriov = {
+ .is_rlc_enabled = gfx_v10_0_is_rlc_enabled,
+ .set_safe_mode = gfx_v10_0_set_safe_mode,
+ .unset_safe_mode = gfx_v10_0_unset_safe_mode,
+ .init = gfx_v10_0_rlc_init,
+ .get_csb_size = gfx_v10_0_get_csb_size,
+ .get_csb_buffer = gfx_v10_0_get_csb_buffer,
+ .resume = gfx_v10_0_rlc_resume,
+ .stop = gfx_v10_0_rlc_stop,
+ .reset = gfx_v10_0_rlc_reset,
+ .start = gfx_v10_0_rlc_start,
+ .update_spm_vmid = gfx_v10_0_update_spm_vmid,
.rlcg_wreg = gfx_v10_rlcg_wreg,
.is_rlcg_access_range = gfx_v10_0_is_rlcg_access_range,
};
@@ -8211,9 +8227,11 @@ static void gfx_v10_0_set_rlc_funcs(struct amdgpu_device *adev)
switch (adev->asic_type) {
case CHIP_NAVI10:
case CHIP_NAVI14:
- case CHIP_NAVI12:
adev->gfx.rlc.funcs = &gfx_v10_0_rlc_funcs;
break;
+ case CHIP_NAVI12:
+ adev->gfx.rlc.funcs = &gfx_v10_0_rlc_funcs_sriov;
+ break;
default:
break;
}
--
2.25.4
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 8+ messages in thread
end of thread, other threads:[~2020-04-30 18:37 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-04-30 18:37 [PATCH 1/8] drm/amdgpu: change HWIP from UVD to VCN for VCN2.5 Alex Deucher
2020-04-30 18:37 ` [PATCH 2/8] drm/amdgpu: check SMU NULL ptr on gfx hw init Alex Deucher
2020-04-30 18:37 ` [PATCH 3/8] drm/amdgpu: update the method to set kcq queue mask Alex Deucher
2020-04-30 18:37 ` [PATCH 4/8] drm/amdgpu: re-structue members for ip discovery Alex Deucher
2020-04-30 18:37 ` [PATCH 5/8] drm/amdgpu: skip reservation of discovery tmr region in pre-Navi Alex Deucher
2020-04-30 18:37 ` [PATCH 6/8] drm/amdgpu: Rename amdgpu_gfx_kcq_queue_mask_transform() Alex Deucher
2020-04-30 18:37 ` [PATCH 7/8] drm/amdkfd: Use a systematic method to calculate queue mask bit Alex Deucher
2020-04-30 18:37 ` [PATCH 8/8] amd/amdgpu: Limit rlcg write registers only for nv12 Alex Deucher
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.