drm/amd: Introduce `AMDGPU_PP_SENSOR_GPU_INPUT_POWER`

Some GPUs have been overloading average power values and input power
values. To disambiguate these, introduce a new
`AMDGPU_PP_SENSOR_GPU_INPUT_POWER` and the GPUs that share input
power update to use this instead of average power.

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2746Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drivers/gpu/drm/amd/include/kgd_pp_interface.h

@@ -133,6 +133,7 @@ enum amd_pp_sensors {
 	AMDGPU_PP_SENSOR_VCE_POWER,
 	AMDGPU_PP_SENSOR_UVD_POWER,
 	AMDGPU_PP_SENSOR_GPU_POWER,
+	AMDGPU_PP_SENSOR_GPU_INPUT_POWER,
 	AMDGPU_PP_SENSOR_SS_APU_SHARE,
 	AMDGPU_PP_SENSOR_SS_DGPU_SHARE,
 	AMDGPU_PP_SENSOR_STABLE_PSTATE_SCLK,

drivers/gpu/drm/amd/pm/amdgpu_pm.c

@@ -2809,7 +2809,7 @@ static ssize_t amdgpu_hwmon_show_power_input(struct device *dev,
 {
 	unsigned int val;
 
-	val = amdgpu_hwmon_get_power(dev, AMDGPU_PP_SENSOR_GPU_POWER);
+	val = amdgpu_hwmon_get_power(dev, AMDGPU_PP_SENSOR_GPU_INPUT_POWER);
 	if (val < 0)
 		return val;
 

drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c

@@ -4039,7 +4039,7 @@ static int smu7_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*((uint32_t *)value) = data->vce_power_gated ? 0 : 1;
 		*size = 4;
 		return 0;
-	case AMDGPU_PP_SENSOR_GPU_POWER:
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 		return smu7_get_gpu_power(hwmgr, (uint32_t *)value);
 	case AMDGPU_PP_SENSOR_VDDGFX:
 		if ((data->vr_config & VRCONF_VDDGFX_MASK) ==

drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c

@@ -3966,7 +3966,7 @@ static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*((uint32_t *)value) = data->vce_power_gated ? 0 : 1;
 		*size = 4;
 		break;
-	case AMDGPU_PP_SENSOR_GPU_POWER:
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 		ret = vega10_get_gpu_power(hwmgr, (uint32_t *)value);
 		break;
 	case AMDGPU_PP_SENSOR_VDDGFX:

drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega12_hwmgr.c

@@ -1529,7 +1529,7 @@ static int vega12_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*((uint32_t *)value) = data->vce_power_gated ? 0 : 1;
 		*size = 4;
 		break;
-	case AMDGPU_PP_SENSOR_GPU_POWER:
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 		ret = vega12_get_gpu_power(hwmgr, (uint32_t *)value);
 		if (!ret)
 			*size = 4;

drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_hwmgr.c

@@ -2253,7 +2253,7 @@ static int vega20_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*((uint32_t *)value) = data->vce_power_gated ? 0 : 1;
 		*size = 4;
 		break;
-	case AMDGPU_PP_SENSOR_GPU_POWER:
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 		*size = 16;
 		ret = vega20_get_gpu_power(hwmgr, (uint32_t *)value);
 		break;

drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h

@@ -1398,6 +1398,7 @@ typedef enum {
 	METRICS_PCIE_RATE,
 	METRICS_PCIE_WIDTH,
 	METRICS_CURR_FANPWM,
+	METRICS_CURR_SOCKETPOWER,
 } MetricsMember_t;
 
 enum smu_cmn2asic_mapping_type {

drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c

@@ -1169,6 +1169,7 @@ static int arcturus_read_sensor(struct smu_context *smu,
 		ret = smu_v11_0_get_gfx_vdd(smu, (uint32_t *)data);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 	default:
 		ret = -EOPNOTSUPP;
 		break;

drivers/gpu/drm/amd/pm/swsmu/smu11/cyan_skillfish_ppt.c

@@ -154,10 +154,14 @@ cyan_skillfish_get_smu_metrics_data(struct smu_context *smu,
 	case METRICS_CURR_UCLK:
 		*value = metrics->Current.MemclkFrequency;
 		break;
-	case METRICS_AVERAGE_SOCKETPOWER:
+	case METRICS_CURR_SOCKETPOWER:
 		*value = (metrics->Current.CurrentSocketPower << 8) /
 				1000;
 		break;
+	case METRICS_AVERAGE_SOCKETPOWER:
+		*value = (metrics->Average.CurrentSocketPower << 8) /
+				1000;
+		break;
 	case METRICS_TEMPERATURE_EDGE:
 		*value = metrics->Current.GfxTemperature / 100 *
 				SMU_TEMPERATURE_UNITS_PER_CENTIGRADES;
@@ -214,6 +218,12 @@ static int cyan_skillfish_read_sensor(struct smu_context *smu,
 						   (uint32_t *)data);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
+		ret = cyan_skillfish_get_smu_metrics_data(smu,
+						   METRICS_CURR_SOCKETPOWER,
+						   (uint32_t *)data);
+		*size = 4;
+		break;
 	case AMDGPU_PP_SENSOR_HOTSPOT_TEMP:
 		ret = cyan_skillfish_get_smu_metrics_data(smu,
 						   METRICS_TEMPERATURE_HOTSPOT,

drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c

@@ -2240,6 +2240,7 @@ static int navi10_read_sensor(struct smu_context *smu,
 		ret = smu_v11_0_get_gfx_vdd(smu, (uint32_t *)data);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 	default:
 		ret = -EOPNOTSUPP;
 		break;

drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c

@@ -1962,6 +1962,7 @@ static int sienna_cichlid_read_sensor(struct smu_context *smu,
 			ret = -EOPNOTSUPP;
 		}
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 	default:
 		ret = -EOPNOTSUPP;
 		break;

drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c

@@ -390,6 +390,10 @@ static int vangogh_get_smu_metrics_data(struct smu_context *smu,
 		*value = metrics->Current.UvdActivity;
 		break;
 	case METRICS_AVERAGE_SOCKETPOWER:
+		*value = (metrics->Average.CurrentSocketPower << 8) /
+		1000;
+		break;
+	case METRICS_CURR_SOCKETPOWER:
 		*value = (metrics->Current.CurrentSocketPower << 8) /
 		1000;
 		break;
@@ -1542,6 +1546,12 @@ static int vangogh_read_sensor(struct smu_context *smu,
 						   (uint32_t *)data);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
+		ret = vangogh_common_get_smu_metrics_data(smu,
+						   METRICS_CURR_SOCKETPOWER,
+						   (uint32_t *)data);
+		*size = 4;
+		break;
 	case AMDGPU_PP_SENSOR_EDGE_TEMP:
 		ret = vangogh_common_get_smu_metrics_data(smu,
 						   METRICS_TEMPERATURE_EDGE,

drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c

@@ -1197,7 +1197,7 @@ static int renoir_get_smu_metrics_data(struct smu_context *smu,
 	case METRICS_AVERAGE_VCNACTIVITY:
 		*value = metrics->AverageUvdActivity / 100;
 		break;
-	case METRICS_AVERAGE_SOCKETPOWER:
+	case METRICS_CURR_SOCKETPOWER:
 		if (((adev->ip_versions[MP1_HWIP][0] == IP_VERSION(12, 0, 1)) && (adev->pm.fw_version >= 0x40000f)) ||
 		((adev->ip_versions[MP1_HWIP][0] == IP_VERSION(12, 0, 0)) && (adev->pm.fw_version >= 0x373200)))
 			*value = metrics->CurrentSocketPower << 8;
@@ -1297,9 +1297,9 @@ static int renoir_read_sensor(struct smu_context *smu,
 						  (uint32_t *)data);
 		*size = 4;
 		break;
-	case AMDGPU_PP_SENSOR_GPU_POWER:
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 		ret = renoir_get_smu_metrics_data(smu,
-						  METRICS_AVERAGE_SOCKETPOWER,
+						  METRICS_CURR_SOCKETPOWER,
 						  (uint32_t *)data);
 		*size = 4;
 		break;

drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c

@@ -1183,6 +1183,7 @@ static int aldebaran_read_sensor(struct smu_context *smu,
 		ret = smu_v13_0_get_gfx_vdd(smu, (uint32_t *)data);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 	default:
 		ret = -EOPNOTSUPP;
 		break;

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c

@@ -997,6 +997,7 @@ static int smu_v13_0_0_read_sensor(struct smu_context *smu,
 						       (uint32_t *)data);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 	default:
 		ret = -EOPNOTSUPP;
 		break;

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c

@@ -321,6 +321,9 @@ static int smu_v13_0_4_get_smu_metrics_data(struct smu_context *smu,
 		*value = metrics->UvdActivity;
 		break;
 	case METRICS_AVERAGE_SOCKETPOWER:
+		*value = (metrics->AverageSocketPower << 8) / 1000;
+		break;
+	case METRICS_CURR_SOCKETPOWER:
 		*value = (metrics->CurrentSocketPower << 8) / 1000;
 		break;
 	case METRICS_TEMPERATURE_EDGE:
@@ -575,6 +578,12 @@ static int smu_v13_0_4_read_sensor(struct smu_context *smu,
 						       (uint32_t *)data);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
+		ret = smu_v13_0_4_get_smu_metrics_data(smu,
+						       METRICS_CURR_SOCKETPOWER,
+						       (uint32_t *)data);
+		*size = 4;
+		break;
 	case AMDGPU_PP_SENSOR_EDGE_TEMP:
 		ret = smu_v13_0_4_get_smu_metrics_data(smu,
 						       METRICS_TEMPERATURE_EDGE,

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c

@@ -288,7 +288,7 @@ static int smu_v13_0_5_get_smu_metrics_data(struct smu_context *smu,
 	case METRICS_AVERAGE_VCNACTIVITY:
 		*value = metrics->UvdActivity;
 		break;
-	case METRICS_AVERAGE_SOCKETPOWER:
+	case METRICS_CURR_SOCKETPOWER:
 		*value = (metrics->CurrentSocketPower << 8) / 1000;
 		break;
 	case METRICS_TEMPERATURE_EDGE:
@@ -332,9 +332,9 @@ static int smu_v13_0_5_read_sensor(struct smu_context *smu,
 								(uint32_t *)data);
 		*size = 4;
 		break;
-	case AMDGPU_PP_SENSOR_GPU_POWER:
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 		ret = smu_v13_0_5_get_smu_metrics_data(smu,
-								METRICS_AVERAGE_SOCKETPOWER,
+								METRICS_CURR_SOCKETPOWER,
 								(uint32_t *)data);
 		*size = 4;
 		break;

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c

@@ -714,7 +714,7 @@ static int smu_v13_0_6_get_smu_metrics_data(struct smu_context *smu,
 	case METRICS_AVERAGE_MEMACTIVITY:
 		*value = SMUQ10_TO_UINT(metrics->DramBandwidthUtilization);
 		break;
-	case METRICS_AVERAGE_SOCKETPOWER:
+	case METRICS_CURR_SOCKETPOWER:
 		*value = SMUQ10_TO_UINT(metrics->SocketPower) << 8;
 		break;
 	case METRICS_TEMPERATURE_HOTSPOT:
@@ -1139,15 +1139,6 @@ static int smu_v13_0_6_get_current_activity_percent(struct smu_context *smu,
 	return ret;
 }
 
-static int smu_v13_0_6_get_gpu_power(struct smu_context *smu, uint32_t *value)
-{
-	if (!value)
-		return -EINVAL;
-
-	return smu_v13_0_6_get_smu_metrics_data(smu, METRICS_AVERAGE_SOCKETPOWER,
-					       value);
-}
-
 static int smu_v13_0_6_thermal_get_temperature(struct smu_context *smu,
 					       enum amd_pp_sensors sensor,
 					       uint32_t *value)
@@ -1193,8 +1184,10 @@ static int smu_v13_0_6_read_sensor(struct smu_context *smu,
 							       (uint32_t *)data);
 		*size = 4;
 		break;
-	case AMDGPU_PP_SENSOR_GPU_POWER:
-		ret = smu_v13_0_6_get_gpu_power(smu, (uint32_t *)data);
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
+		ret = smu_v13_0_6_get_smu_metrics_data(smu,
+						       METRICS_CURR_SOCKETPOWER,
+						       (uint32_t *)data);
 		*size = 4;
 		break;
 	case AMDGPU_PP_SENSOR_HOTSPOT_TEMP:

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c

@@ -978,6 +978,7 @@ static int smu_v13_0_7_read_sensor(struct smu_context *smu,
 						       (uint32_t *)data);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 	default:
 		ret = -EOPNOTSUPP;
 		break;

drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c

@@ -365,7 +365,7 @@ static int yellow_carp_get_smu_metrics_data(struct smu_context *smu,
 	case METRICS_AVERAGE_VCNACTIVITY:
 		*value = metrics->UvdActivity;
 		break;
-	case METRICS_AVERAGE_SOCKETPOWER:
+	case METRICS_CURR_SOCKETPOWER:
 		*value = (metrics->CurrentSocketPower << 8) / 1000;
 		break;
 	case METRICS_TEMPERATURE_EDGE:
@@ -423,9 +423,9 @@ static int yellow_carp_read_sensor(struct smu_context *smu,
 								(uint32_t *)data);
 		*size = 4;
 		break;
-	case AMDGPU_PP_SENSOR_GPU_POWER:
+	case AMDGPU_PP_SENSOR_GPU_INPUT_POWER:
 		ret = yellow_carp_get_smu_metrics_data(smu,
-								METRICS_AVERAGE_SOCKETPOWER,
+								METRICS_CURR_SOCKETPOWER,
 								(uint32_t *)data);
 		*size = 4;
 		break;