[PATCH 1/5] drm/amd/powerplay: support hotspot/memory critical limit values

[PATCH 1/5] drm/amd/powerplay: support hotspot/memory critical limit values

[Evan Quan]

These new interfaces(temp2_crit, temp2_crit_hyst, temp3_crit,
temp3_crit_hyst) are supported on SOC15 dGPUs only.

Change-Id: Ia87e3f6ad816b51d6680eb74c8f755d6c2b0a6ae
Signed-off-by: Evan Quan <evan.quan@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h       |  8 +++
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c        | 55 ++++++++++++++++++-
 .../drm/amd/powerplay/hwmgr/hardwaremanager.c | 12 +++-
 .../drm/amd/powerplay/hwmgr/vega10_hwmgr.c    | 10 +++-
 .../drm/amd/powerplay/hwmgr/vega12_hwmgr.c    | 11 +++-
 .../drm/amd/powerplay/hwmgr/vega20_hwmgr.c    | 11 +++-
 .../gpu/drm/amd/powerplay/inc/power_state.h   |  4 ++
 .../gpu/drm/amd/powerplay/inc/pp_thermal.h    |  8 +--
 8 files changed, 103 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
index dca35407879d..8df54443ec78 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -75,6 +75,14 @@ struct amdgpu_dpm_thermal {
 	int                min_temp;
 	/* high temperature threshold */
 	int                max_temp;
+	/* hotspot low temperature threshold */
+	int                min_hotspot_temp;
+	/* hotspot high temperature critical threshold */
+	int                max_hotspot_crit_temp;
+	/* memory low temperature threshold */
+	int                min_mem_temp;
+	/* memory high temperature critical threshold */
+	int                max_mem_crit_temp;
 	/* was last interrupt low to high or high to low */
 	bool               high_to_low;
 	/* interrupt source */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 552127b74f78..f427e4a64453 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -1457,6 +1457,38 @@ static ssize_t amdgpu_hwmon_show_temp_thresh(struct device *dev,
 	return snprintf(buf, PAGE_SIZE, "%d\n", temp);
 }
 
+static ssize_t amdgpu_hwmon_show_hotspot_temp_thresh(struct device *dev,
+					     struct device_attribute *attr,
+					     char *buf)
+{
+	struct amdgpu_device *adev = dev_get_drvdata(dev);
+	int hyst = to_sensor_dev_attr(attr)->index;
+	int temp;
+
+	if (hyst)
+		temp = adev->pm.dpm.thermal.min_hotspot_temp;
+	else
+		temp = adev->pm.dpm.thermal.max_hotspot_crit_temp;
+
+	return snprintf(buf, PAGE_SIZE, "%d\n", temp);
+}
+
+static ssize_t amdgpu_hwmon_show_mem_temp_thresh(struct device *dev,
+					     struct device_attribute *attr,
+					     char *buf)
+{
+	struct amdgpu_device *adev = dev_get_drvdata(dev);
+	int hyst = to_sensor_dev_attr(attr)->index;
+	int temp;
+
+	if (hyst)
+		temp = adev->pm.dpm.thermal.min_mem_temp;
+	else
+		temp = adev->pm.dpm.thermal.max_mem_crit_temp;
+
+	return snprintf(buf, PAGE_SIZE, "%d\n", temp);
+}
+
 static ssize_t amdgpu_hwmon_get_pwm1_enable(struct device *dev,
 					    struct device_attribute *attr,
 					    char *buf)
@@ -2028,9 +2060,11 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
  *
  * - temp1_input: the on die GPU temperature in millidegrees Celsius
  *
- * - temp1_crit: temperature critical max value in millidegrees Celsius
+ * - temp[1-3]_crit: temperature critical max value in millidegrees Celsius
+ *   - temp2_crit and temp3_crit are supported on SOC15 dGPUs only
  *
- * - temp1_crit_hyst: temperature hysteresis for critical limit in millidegrees Celsius
+ * - temp[1-3]_crit_hyst: temperature hysteresis for critical limit in millidegrees Celsius
+ *   - temp2_crit_hyst and temp3_crit_hyst are supported on SOC15 dGPUs only
  *
  * hwmon interfaces for GPU voltage:
  *
@@ -2081,6 +2115,10 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
 static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit_hyst, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 1);
+static SENSOR_DEVICE_ATTR(temp2_crit, S_IRUGO, amdgpu_hwmon_show_hotspot_temp_thresh, NULL, 0);
+static SENSOR_DEVICE_ATTR(temp2_crit_hyst, S_IRUGO, amdgpu_hwmon_show_hotspot_temp_thresh, NULL, 1);
+static SENSOR_DEVICE_ATTR(temp3_crit, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 0);
+static SENSOR_DEVICE_ATTR(temp3_crit_hyst, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 1);
 static SENSOR_DEVICE_ATTR(pwm1, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1, amdgpu_hwmon_set_pwm1, 0);
 static SENSOR_DEVICE_ATTR(pwm1_enable, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1_enable, amdgpu_hwmon_set_pwm1_enable, 0);
 static SENSOR_DEVICE_ATTR(pwm1_min, S_IRUGO, amdgpu_hwmon_get_pwm1_min, NULL, 0);
@@ -2107,6 +2145,10 @@ static struct attribute *hwmon_attributes[] = {
 	&sensor_dev_attr_temp1_input.dev_attr.attr,
 	&sensor_dev_attr_temp1_crit.dev_attr.attr,
 	&sensor_dev_attr_temp1_crit_hyst.dev_attr.attr,
+	&sensor_dev_attr_temp2_crit.dev_attr.attr,
+	&sensor_dev_attr_temp2_crit_hyst.dev_attr.attr,
+	&sensor_dev_attr_temp3_crit.dev_attr.attr,
+	&sensor_dev_attr_temp3_crit_hyst.dev_attr.attr,
 	&sensor_dev_attr_pwm1.dev_attr.attr,
 	&sensor_dev_attr_pwm1_enable.dev_attr.attr,
 	&sensor_dev_attr_pwm1_min.dev_attr.attr,
@@ -2229,6 +2271,15 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 	     attr == &sensor_dev_attr_freq2_label.dev_attr.attr))
 		return 0;
 
+	/* only SOC15 dGPUs support hotspot and mem temperatures */
+	if (((adev->flags & AMD_IS_APU) ||
+	     adev->asic_type < CHIP_VEGA10) &&
+	    (attr == &sensor_dev_attr_temp2_crit.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp2_crit_hyst.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp3_crit.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp3_crit_hyst.dev_attr.attr))
+		return 0;
+
 	return effective_mode;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
index 70f7f47a2fcf..af6ab04130ef 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
@@ -225,7 +225,13 @@ int phm_register_irq_handlers(struct pp_hwmgr *hwmgr)
 int phm_start_thermal_controller(struct pp_hwmgr *hwmgr)
 {
 	int ret = 0;
-	struct PP_TemperatureRange range = {TEMP_RANGE_MIN, TEMP_RANGE_MAX};
+	struct PP_TemperatureRange range = {
+		TEMP_RANGE_MIN,
+		TEMP_RANGE_MAX,
+		TEMP_RANGE_MIN,
+		TEMP_RANGE_MAX,
+		TEMP_RANGE_MIN,
+		TEMP_RANGE_MAX};
 	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (hwmgr->hwmgr_func->get_thermal_temperature_range)
@@ -239,6 +245,10 @@ int phm_start_thermal_controller(struct pp_hwmgr *hwmgr)
 
 	adev->pm.dpm.thermal.min_temp = range.min;
 	adev->pm.dpm.thermal.max_temp = range.max;
+	adev->pm.dpm.thermal.min_hotspot_temp = range.hotspot_min;
+	adev->pm.dpm.thermal.max_hotspot_crit_temp = range.hotspot_crit_max;
+	adev->pm.dpm.thermal.min_mem_temp = range.mem_min;
+	adev->pm.dpm.thermal.max_mem_crit_temp = range.mem_crit_max;
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 384c37875cd0..4e1df44f094b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -4852,12 +4852,16 @@ static int vega10_notify_cac_buffer_info(struct pp_hwmgr *hwmgr,
 static int vega10_get_thermal_temperature_range(struct pp_hwmgr *hwmgr,
 		struct PP_TemperatureRange *thermal_data)
 {
-	struct phm_ppt_v2_information *table_info =
-			(struct phm_ppt_v2_information *)hwmgr->pptable;
+	struct vega10_hwmgr *data = hwmgr->backend;
+	PPTable_t *pp_table = &(data->smc_state_table.pp_table);
 
 	memcpy(thermal_data, &SMU7ThermalWithDelayPolicy[0], sizeof(struct PP_TemperatureRange));
 
-	thermal_data->max = table_info->tdp_table->usSoftwareShutdownTemp *
+	thermal_data->max = pp_table->TedgeLimit *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->hotspot_crit_max = pp_table->ThotspotLimit *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->mem_crit_max = pp_table->ThbmLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index 707cd4b0357f..4f63570ea257 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -2526,12 +2526,17 @@ static int vega12_notify_cac_buffer_info(struct pp_hwmgr *hwmgr,
 static int vega12_get_thermal_temperature_range(struct pp_hwmgr *hwmgr,
 		struct PP_TemperatureRange *thermal_data)
 {
-	struct phm_ppt_v3_information *pptable_information =
-		(struct phm_ppt_v3_information *)hwmgr->pptable;
+	struct vega12_hwmgr *data =
+			(struct vega12_hwmgr *)(hwmgr->backend);
+	PPTable_t *pp_table = &(data->smc_state_table.pp_table);
 
 	memcpy(thermal_data, &SMU7ThermalWithDelayPolicy[0], sizeof(struct PP_TemperatureRange));
 
-	thermal_data->max = pptable_information->us_software_shutdown_temp *
+	thermal_data->max = pp_table->TedgeLimit *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->hotspot_crit_max = pp_table->ThotspotLimit *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->mem_crit_max = pp_table->ThbmLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
index 39a547084e90..129ffa781b89 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
@@ -3963,12 +3963,17 @@ static int vega20_notify_cac_buffer_info(struct pp_hwmgr *hwmgr,
 static int vega20_get_thermal_temperature_range(struct pp_hwmgr *hwmgr,
 		struct PP_TemperatureRange *thermal_data)
 {
-	struct phm_ppt_v3_information *pptable_information =
-		(struct phm_ppt_v3_information *)hwmgr->pptable;
+	struct vega20_hwmgr *data =
+			(struct vega20_hwmgr *)(hwmgr->backend);
+	PPTable_t *pp_table = &(data->smc_state_table.pp_table);
 
 	memcpy(thermal_data, &SMU7ThermalWithDelayPolicy[0], sizeof(struct PP_TemperatureRange));
 
-	thermal_data->max = pptable_information->us_software_shutdown_temp *
+	thermal_data->max = pp_table->TedgeLimit *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->hotspot_crit_max = pp_table->ThotspotLimit *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->mem_crit_max = pp_table->ThbmLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/powerplay/inc/power_state.h b/drivers/gpu/drm/amd/powerplay/inc/power_state.h
index a99b5cbb113e..a8988e7d58c6 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/power_state.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/power_state.h
@@ -124,6 +124,10 @@ struct PP_StateSoftwareAlgorithmBlock {
 struct PP_TemperatureRange {
 	int min;
 	int max;
+	int hotspot_min;
+	int hotspot_crit_max;
+	int mem_min;
+	int mem_crit_max;
 };
 
 struct PP_StateValidationBlock {
diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_thermal.h b/drivers/gpu/drm/amd/powerplay/inc/pp_thermal.h
index 201d2b6329ab..75a0a2f8bea2 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/pp_thermal.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/pp_thermal.h
@@ -27,14 +27,14 @@
 
 static const struct PP_TemperatureRange SMU7ThermalWithDelayPolicy[] =
 {
-	{-273150,  99000},
-	{ 120000, 120000},
+	{-273150,  99000, -273150, 99000, -273150, 99000},
+	{ 120000, 120000, 120000, 120000, 120000, 120000},
 };
 
 static const struct PP_TemperatureRange SMU7ThermalPolicy[] =
 {
-	{-273150,  99000},
-	{ 120000, 120000},
+	{-273150,  99000, -273150, 99000, -273150, 99000},
+	{ 120000, 120000, 120000, 120000, 120000, 120000},
 };
 
 #endif
-- 
2.21.0

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

[PATCH 2/5] drm/amd/powerplay: support temperature emergency max values

[Evan Quan]

These new interfaces(temp1_emergency, temp2_emergency,
temp3_emergency) are supported on SOC15 dGPUs only.

Change-Id: I2552df63f9c8c50294b3940bb2a402217673c2bc
Signed-off-by: Evan Quan <evan.quan@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h       |  6 +++
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c        | 40 ++++++++++++++++++-
 .../drm/amd/powerplay/hwmgr/hardwaremanager.c |  6 +++
 .../drm/amd/powerplay/hwmgr/vega10_hwmgr.c    |  6 +++
 .../drm/amd/powerplay/hwmgr/vega12_hwmgr.c    |  6 +++
 .../drm/amd/powerplay/hwmgr/vega20_hwmgr.c    |  6 +++
 .../gpu/drm/amd/powerplay/inc/power_state.h   |  3 ++
 .../gpu/drm/amd/powerplay/inc/pp_thermal.h    | 12 ++++--
 8 files changed, 80 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
index 8df54443ec78..521dbd0d9af8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -75,14 +75,20 @@ struct amdgpu_dpm_thermal {
 	int                min_temp;
 	/* high temperature threshold */
 	int                max_temp;
+	/* edge max emergency(shutdown) temp */
+	int                max_edge_emergency_temp;
 	/* hotspot low temperature threshold */
 	int                min_hotspot_temp;
 	/* hotspot high temperature critical threshold */
 	int                max_hotspot_crit_temp;
+	/* hotspot max emergency(shutdown) temp */
+	int                max_hotspot_emergency_temp;
 	/* memory low temperature threshold */
 	int                min_mem_temp;
 	/* memory high temperature critical threshold */
 	int                max_mem_crit_temp;
+	/* memory max emergency(shutdown) temp */
+	int                max_mem_emergency_temp;
 	/* was last interrupt low to high or high to low */
 	bool               high_to_low;
 	/* interrupt source */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index f427e4a64453..77de010a4ceb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -1489,6 +1489,32 @@ static ssize_t amdgpu_hwmon_show_mem_temp_thresh(struct device *dev,
 	return snprintf(buf, PAGE_SIZE, "%d\n", temp);
 }
 
+static ssize_t amdgpu_hwmon_show_temp_emergency(struct device *dev,
+					     struct device_attribute *attr,
+					     char *buf)
+{
+	struct amdgpu_device *adev = dev_get_drvdata(dev);
+	int channel = to_sensor_dev_attr(attr)->index;
+	int temp;
+
+	if (channel >= PP_TEMP_MAX)
+		return -EINVAL;
+
+	switch (channel) {
+	case PP_TEMP_JUNCTION:
+		temp = adev->pm.dpm.thermal.max_hotspot_emergency_temp;
+		break;
+	case PP_TEMP_EDGE:
+		temp = adev->pm.dpm.thermal.max_edge_emergency_temp;
+		break;
+	case PP_TEMP_MEM:
+		temp = adev->pm.dpm.thermal.max_mem_emergency_temp;
+		break;
+	}
+
+	return snprintf(buf, PAGE_SIZE, "%d\n", temp);
+}
+
 static ssize_t amdgpu_hwmon_get_pwm1_enable(struct device *dev,
 					    struct device_attribute *attr,
 					    char *buf)
@@ -2066,6 +2092,9 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
  * - temp[1-3]_crit_hyst: temperature hysteresis for critical limit in millidegrees Celsius
  *   - temp2_crit_hyst and temp3_crit_hyst are supported on SOC15 dGPUs only
  *
+ * - temp[1-3]_emergency: temperature emergency max value(asic shutdown) in millidegrees Celsius
+ *   - these are supported on SOC15 dGPUs only
+ *
  * hwmon interfaces for GPU voltage:
  *
  * - in0_input: the voltage on the GPU in millivolts
@@ -2115,10 +2144,13 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
 static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit_hyst, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 1);
+static SENSOR_DEVICE_ATTR(temp1_emergency, S_IRUGO, amdgpu_hwmon_show_temp_emergency, NULL, PP_TEMP_EDGE);
 static SENSOR_DEVICE_ATTR(temp2_crit, S_IRUGO, amdgpu_hwmon_show_hotspot_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp2_crit_hyst, S_IRUGO, amdgpu_hwmon_show_hotspot_temp_thresh, NULL, 1);
+static SENSOR_DEVICE_ATTR(temp2_emergency, S_IRUGO, amdgpu_hwmon_show_temp_emergency, NULL, PP_TEMP_JUNCTION);
 static SENSOR_DEVICE_ATTR(temp3_crit, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp3_crit_hyst, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 1);
+static SENSOR_DEVICE_ATTR(temp3_emergency, S_IRUGO, amdgpu_hwmon_show_temp_emergency, NULL, PP_TEMP_MEM);
 static SENSOR_DEVICE_ATTR(pwm1, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1, amdgpu_hwmon_set_pwm1, 0);
 static SENSOR_DEVICE_ATTR(pwm1_enable, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1_enable, amdgpu_hwmon_set_pwm1_enable, 0);
 static SENSOR_DEVICE_ATTR(pwm1_min, S_IRUGO, amdgpu_hwmon_get_pwm1_min, NULL, 0);
@@ -2149,6 +2181,9 @@ static struct attribute *hwmon_attributes[] = {
 	&sensor_dev_attr_temp2_crit_hyst.dev_attr.attr,
 	&sensor_dev_attr_temp3_crit.dev_attr.attr,
 	&sensor_dev_attr_temp3_crit_hyst.dev_attr.attr,
+	&sensor_dev_attr_temp1_emergency.dev_attr.attr,
+	&sensor_dev_attr_temp2_emergency.dev_attr.attr,
+	&sensor_dev_attr_temp3_emergency.dev_attr.attr,
 	&sensor_dev_attr_pwm1.dev_attr.attr,
 	&sensor_dev_attr_pwm1_enable.dev_attr.attr,
 	&sensor_dev_attr_pwm1_min.dev_attr.attr,
@@ -2277,7 +2312,10 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 	    (attr == &sensor_dev_attr_temp2_crit.dev_attr.attr ||
 	     attr == &sensor_dev_attr_temp2_crit_hyst.dev_attr.attr ||
 	     attr == &sensor_dev_attr_temp3_crit.dev_attr.attr ||
-	     attr == &sensor_dev_attr_temp3_crit_hyst.dev_attr.attr))
+	     attr == &sensor_dev_attr_temp3_crit_hyst.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp1_emergency.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp2_emergency.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp3_emergency.dev_attr.attr))
 		return 0;
 
 	return effective_mode;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
index af6ab04130ef..cc57fb953e62 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
@@ -228,9 +228,12 @@ int phm_start_thermal_controller(struct pp_hwmgr *hwmgr)
 	struct PP_TemperatureRange range = {
 		TEMP_RANGE_MIN,
 		TEMP_RANGE_MAX,
+		TEMP_RANGE_MAX,
 		TEMP_RANGE_MIN,
 		TEMP_RANGE_MAX,
+		TEMP_RANGE_MAX,
 		TEMP_RANGE_MIN,
+		TEMP_RANGE_MAX,
 		TEMP_RANGE_MAX};
 	struct amdgpu_device *adev = hwmgr->adev;
 
@@ -245,10 +248,13 @@ int phm_start_thermal_controller(struct pp_hwmgr *hwmgr)
 
 	adev->pm.dpm.thermal.min_temp = range.min;
 	adev->pm.dpm.thermal.max_temp = range.max;
+	adev->pm.dpm.thermal.max_edge_emergency_temp = range.edge_emergency_max;
 	adev->pm.dpm.thermal.min_hotspot_temp = range.hotspot_min;
 	adev->pm.dpm.thermal.max_hotspot_crit_temp = range.hotspot_crit_max;
+	adev->pm.dpm.thermal.max_hotspot_emergency_temp = range.hotspot_emergency_max;
 	adev->pm.dpm.thermal.min_mem_temp = range.mem_min;
 	adev->pm.dpm.thermal.max_mem_crit_temp = range.mem_crit_max;
+	adev->pm.dpm.thermal.max_mem_emergency_temp = range.mem_emergency_max;
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 4e1df44f094b..1422bc4e45d1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -4859,10 +4859,16 @@ static int vega10_get_thermal_temperature_range(struct pp_hwmgr *hwmgr,
 
 	thermal_data->max = pp_table->TedgeLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->edge_emergency_max = (pp_table->TedgeLimit + CTF_OFFSET_EDGE) *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	thermal_data->hotspot_crit_max = pp_table->ThotspotLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->hotspot_emergency_max = (pp_table->ThotspotLimit + CTF_OFFSET_HOTSPOT) *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	thermal_data->mem_crit_max = pp_table->ThbmLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->mem_emergency_max = (pp_table->ThbmLimit + CTF_OFFSET_HBM)*
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index 4f63570ea257..60c9f9502e65 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -2534,10 +2534,16 @@ static int vega12_get_thermal_temperature_range(struct pp_hwmgr *hwmgr,
 
 	thermal_data->max = pp_table->TedgeLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->edge_emergency_max = (pp_table->TedgeLimit + CTF_OFFSET_EDGE) *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	thermal_data->hotspot_crit_max = pp_table->ThotspotLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->hotspot_emergency_max = (pp_table->ThotspotLimit + CTF_OFFSET_HOTSPOT) *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	thermal_data->mem_crit_max = pp_table->ThbmLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->mem_emergency_max = (pp_table->ThbmLimit + CTF_OFFSET_HBM)*
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
index 129ffa781b89..6fcdefcfa930 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
@@ -3971,10 +3971,16 @@ static int vega20_get_thermal_temperature_range(struct pp_hwmgr *hwmgr,
 
 	thermal_data->max = pp_table->TedgeLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->edge_emergency_max = (pp_table->TedgeLimit + CTF_OFFSET_EDGE) *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	thermal_data->hotspot_crit_max = pp_table->ThotspotLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->hotspot_emergency_max = (pp_table->ThotspotLimit + CTF_OFFSET_HOTSPOT) *
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	thermal_data->mem_crit_max = pp_table->ThbmLimit *
 		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+	thermal_data->mem_emergency_max = (pp_table->ThbmLimit + CTF_OFFSET_HBM)*
+		PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/inc/power_state.h b/drivers/gpu/drm/amd/powerplay/inc/power_state.h
index a8988e7d58c6..a5f2227a3971 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/power_state.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/power_state.h
@@ -124,10 +124,13 @@ struct PP_StateSoftwareAlgorithmBlock {
 struct PP_TemperatureRange {
 	int min;
 	int max;
+	int edge_emergency_max;
 	int hotspot_min;
 	int hotspot_crit_max;
+	int hotspot_emergency_max;
 	int mem_min;
 	int mem_crit_max;
+	int mem_emergency_max;
 };
 
 struct PP_StateValidationBlock {
diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_thermal.h b/drivers/gpu/drm/amd/powerplay/inc/pp_thermal.h
index 75a0a2f8bea2..3e30768f9e1c 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/pp_thermal.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/pp_thermal.h
@@ -27,14 +27,18 @@
 
 static const struct PP_TemperatureRange SMU7ThermalWithDelayPolicy[] =
 {
-	{-273150,  99000, -273150, 99000, -273150, 99000},
-	{ 120000, 120000, 120000, 120000, 120000, 120000},
+	{-273150,  99000, 99000, -273150, 99000, 99000, -273150, 99000, 99000},
+	{ 120000, 120000, 120000, 120000, 120000, 120000, 120000, 120000, 120000},
 };
 
 static const struct PP_TemperatureRange SMU7ThermalPolicy[] =
 {
-	{-273150,  99000, -273150, 99000, -273150, 99000},
-	{ 120000, 120000, 120000, 120000, 120000, 120000},
+	{-273150,  99000, 99000, -273150, 99000, 99000, -273150, 99000, 99000},
+	{ 120000, 120000, 120000, 120000, 120000, 120000, 120000, 120000, 120000},
 };
 
+#define CTF_OFFSET_EDGE			5
+#define CTF_OFFSET_HOTSPOT		5
+#define CTF_OFFSET_HBM			5
+
 #endif
-- 
2.21.0

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

[PATCH 3/5] drm/amd/powerplay: support SMU metrics table on Vega12

[Evan Quan]

That should provide some necessary sensor information.

Change-Id: I898371cef06795c5369a14c4dd3fe8717959d81a
Signed-off-by: Evan Quan <evan.quan@amd.com>
---
 .../drm/amd/powerplay/hwmgr/vega12_hwmgr.c    | 21 +++++++++++++++++++
 .../drm/amd/powerplay/hwmgr/vega12_hwmgr.h    |  3 +++
 .../drm/amd/powerplay/smumgr/vega12_smumgr.c  | 21 +++++++++++++++++++
 3 files changed, 45 insertions(+)

diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index 60c9f9502e65..aeeeaa79056c 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -1237,6 +1237,27 @@ static uint32_t vega12_dpm_get_mclk(struct pp_hwmgr *hwmgr, bool low)
 	return (mem_clk * 100);
 }
 
+static int vega12_get_metrics_table(struct pp_hwmgr *hwmgr, SmuMetrics_t *metrics_table)
+{
+	struct vega12_hwmgr *data =
+			(struct vega12_hwmgr *)(hwmgr->backend);
+	int ret = 0;
+
+	if (!data->metrics_time || time_after(jiffies, data->metrics_time + HZ / 2)) {
+		ret = smum_smc_table_manager(hwmgr, (uint8_t *)metrics_table,
+				TABLE_SMU_METRICS, true);
+		if (ret) {
+			pr_info("Failed to export SMU metrics table!\n");
+			return ret;
+		}
+		memcpy(&data->metrics_table, metrics_table, sizeof(SmuMetrics_t));
+		data->metrics_time = jiffies;
+	} else
+		memcpy(metrics_table, &data->metrics_table, sizeof(SmuMetrics_t));
+
+	return ret;
+}
+
 static int vega12_get_gpu_power(struct pp_hwmgr *hwmgr, uint32_t *query)
 {
 #if 0
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
index b3e424d28994..73875399666a 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
@@ -396,6 +396,9 @@ struct vega12_hwmgr {
 
 	/* ---- Gfxoff ---- */
 	bool                           gfxoff_controlled_by_driver;
+
+	unsigned long                  metrics_time;
+	SmuMetrics_t                   metrics_table;
 };
 
 #define VEGA12_DPM2_NEAR_TDP_DEC                      10
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
index ddb801517667..1eaf0fa28ef7 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
@@ -287,8 +287,26 @@ static int vega12_smu_init(struct pp_hwmgr *hwmgr)
 	priv->smu_tables.entry[TABLE_OVERDRIVE].version = 0x01;
 	priv->smu_tables.entry[TABLE_OVERDRIVE].size = sizeof(OverDriveTable_t);
 
+	/* allocate space for SMU_METRICS table */
+	ret = amdgpu_bo_create_kernel((struct amdgpu_device *)hwmgr->adev,
+				      sizeof(SmuMetrics_t),
+				      PAGE_SIZE,
+				      AMDGPU_GEM_DOMAIN_VRAM,
+				      &priv->smu_tables.entry[TABLE_SMU_METRICS].handle,
+				      &priv->smu_tables.entry[TABLE_SMU_METRICS].mc_addr,
+				      &priv->smu_tables.entry[TABLE_SMU_METRICS].table);
+	if (ret)
+		goto err4;
+
+	priv->smu_tables.entry[TABLE_SMU_METRICS].version = 0x01;
+	priv->smu_tables.entry[TABLE_SMU_METRICS].size = sizeof(SmuMetrics_t);
+
 	return 0;
 
+err4:
+	amdgpu_bo_free_kernel(&priv->smu_tables.entry[TABLE_OVERDRIVE].handle,
+				&priv->smu_tables.entry[TABLE_OVERDRIVE].mc_addr,
+				&priv->smu_tables.entry[TABLE_OVERDRIVE].table);
 err3:
 	amdgpu_bo_free_kernel(&priv->smu_tables.entry[TABLE_AVFS_FUSE_OVERRIDE].handle,
 				&priv->smu_tables.entry[TABLE_AVFS_FUSE_OVERRIDE].mc_addr,
@@ -334,6 +352,9 @@ static int vega12_smu_fini(struct pp_hwmgr *hwmgr)
 		amdgpu_bo_free_kernel(&priv->smu_tables.entry[TABLE_OVERDRIVE].handle,
 				      &priv->smu_tables.entry[TABLE_OVERDRIVE].mc_addr,
 				      &priv->smu_tables.entry[TABLE_OVERDRIVE].table);
+		amdgpu_bo_free_kernel(&priv->smu_tables.entry[TABLE_SMU_METRICS].handle,
+				      &priv->smu_tables.entry[TABLE_SMU_METRICS].mc_addr,
+				      &priv->smu_tables.entry[TABLE_SMU_METRICS].table);
 		kfree(hwmgr->smu_backend);
 		hwmgr->smu_backend = NULL;
 	}
-- 
2.21.0

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

[PATCH 4/5] drm/amd/powerplay: expose current hotspot and memory temperatures V2

[Evan Quan]

Two new hwmon interfaces(temp2_input and temp3_input) are added.
They are supported on SOC15 dGPUs only.

- V2: correct thermal sensor output

Change-Id: I935c512bd38e080fb8b6e3164c5e5294baff4e91
Signed-off-by: Evan Quan <evan.quan@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c        | 45 +++++++++++++++----
 .../gpu/drm/amd/include/kgd_pp_interface.h    |  3 ++
 .../drm/amd/powerplay/hwmgr/vega10_hwmgr.c    | 12 +++++
 .../drm/amd/powerplay/hwmgr/vega12_hwmgr.c    | 19 ++++++++
 .../drm/amd/powerplay/hwmgr/vega20_hwmgr.c    | 20 ++++++++-
 5 files changed, 90 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 77de010a4ceb..291c06f40dba 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -1425,6 +1425,7 @@ static ssize_t amdgpu_hwmon_show_temp(struct device *dev,
 {
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	struct drm_device *ddev = adev->ddev;
+	int channel = to_sensor_dev_attr(attr)->index;
 	int r, temp, size = sizeof(temp);
 
 	/* Can't get temperature when the card is off */
@@ -1432,11 +1433,32 @@ static ssize_t amdgpu_hwmon_show_temp(struct device *dev,
 	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;
 
-	/* get the temperature */
-	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP,
-				   (void *)&temp, &size);
-	if (r)
-		return r;
+	if (channel >= PP_TEMP_MAX)
+		return -EINVAL;
+
+	switch (channel) {
+	case PP_TEMP_JUNCTION:
+		/* get current junction temperature */
+		r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_HOTSPOT_TEMP,
+					   (void *)&temp, &size);
+		if (r)
+			return r;
+		break;
+	case PP_TEMP_EDGE:
+		/* get current edge temperature */
+		r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_EDGE_TEMP,
+					   (void *)&temp, &size);
+		if (r)
+			return r;
+		break;
+	case PP_TEMP_MEM:
+		/* get current memory temperature */
+		r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MEM_TEMP,
+					   (void *)&temp, &size);
+		if (r)
+			return r;
+		break;
+	}
 
 	return snprintf(buf, PAGE_SIZE, "%d\n", temp);
 }
@@ -2084,7 +2106,8 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
  *
  * hwmon interfaces for GPU temperature:
  *
- * - temp1_input: the on die GPU temperature in millidegrees Celsius
+ * - temp[1-3]_input: the on die GPU temperature in millidegrees Celsius
+ *   - temp2_input and temp3_input are supported on SOC15 dGPUs only
  *
  * - temp[1-3]_crit: temperature critical max value in millidegrees Celsius
  *   - temp2_crit and temp3_crit are supported on SOC15 dGPUs only
@@ -2141,13 +2164,15 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
  *
  */
 
-static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, 0);
+static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, PP_TEMP_EDGE);
 static SENSOR_DEVICE_ATTR(temp1_crit, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit_hyst, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 1);
 static SENSOR_DEVICE_ATTR(temp1_emergency, S_IRUGO, amdgpu_hwmon_show_temp_emergency, NULL, PP_TEMP_EDGE);
+static SENSOR_DEVICE_ATTR(temp2_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, PP_TEMP_JUNCTION);
 static SENSOR_DEVICE_ATTR(temp2_crit, S_IRUGO, amdgpu_hwmon_show_hotspot_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp2_crit_hyst, S_IRUGO, amdgpu_hwmon_show_hotspot_temp_thresh, NULL, 1);
 static SENSOR_DEVICE_ATTR(temp2_emergency, S_IRUGO, amdgpu_hwmon_show_temp_emergency, NULL, PP_TEMP_JUNCTION);
+static SENSOR_DEVICE_ATTR(temp3_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, PP_TEMP_MEM);
 static SENSOR_DEVICE_ATTR(temp3_crit, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp3_crit_hyst, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 1);
 static SENSOR_DEVICE_ATTR(temp3_emergency, S_IRUGO, amdgpu_hwmon_show_temp_emergency, NULL, PP_TEMP_MEM);
@@ -2177,8 +2202,10 @@ static struct attribute *hwmon_attributes[] = {
 	&sensor_dev_attr_temp1_input.dev_attr.attr,
 	&sensor_dev_attr_temp1_crit.dev_attr.attr,
 	&sensor_dev_attr_temp1_crit_hyst.dev_attr.attr,
+	&sensor_dev_attr_temp2_input.dev_attr.attr,
 	&sensor_dev_attr_temp2_crit.dev_attr.attr,
 	&sensor_dev_attr_temp2_crit_hyst.dev_attr.attr,
+	&sensor_dev_attr_temp3_input.dev_attr.attr,
 	&sensor_dev_attr_temp3_crit.dev_attr.attr,
 	&sensor_dev_attr_temp3_crit_hyst.dev_attr.attr,
 	&sensor_dev_attr_temp1_emergency.dev_attr.attr,
@@ -2315,7 +2342,9 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 	     attr == &sensor_dev_attr_temp3_crit_hyst.dev_attr.attr ||
 	     attr == &sensor_dev_attr_temp1_emergency.dev_attr.attr ||
 	     attr == &sensor_dev_attr_temp2_emergency.dev_attr.attr ||
-	     attr == &sensor_dev_attr_temp3_emergency.dev_attr.attr))
+	     attr == &sensor_dev_attr_temp3_emergency.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp2_input.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp3_input.dev_attr.attr))
 		return 0;
 
 	return effective_mode;
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index 2b579ba9b685..a8bf8e90ceeb 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -111,6 +111,9 @@ enum amd_pp_sensors {
 	AMDGPU_PP_SENSOR_GPU_LOAD,
 	AMDGPU_PP_SENSOR_GFX_MCLK,
 	AMDGPU_PP_SENSOR_GPU_TEMP,
+	AMDGPU_PP_SENSOR_EDGE_TEMP = AMDGPU_PP_SENSOR_GPU_TEMP,
+	AMDGPU_PP_SENSOR_HOTSPOT_TEMP,
+	AMDGPU_PP_SENSOR_MEM_TEMP,
 	AMDGPU_PP_SENSOR_VCE_POWER,
 	AMDGPU_PP_SENSOR_UVD_POWER,
 	AMDGPU_PP_SENSOR_GPU_POWER,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 1422bc4e45d1..d5d0db456021 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -3785,6 +3785,18 @@ static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*((uint32_t *)value) = vega10_thermal_get_temperature(hwmgr);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_HOTSPOT_TEMP:
+		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetTemperatureHotspot);
+		*((uint32_t *)value) = smum_get_argument(hwmgr) *
+			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+		*size = 4;
+		break;
+	case AMDGPU_PP_SENSOR_MEM_TEMP:
+		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetTemperatureHBM);
+		*((uint32_t *)value) = smum_get_argument(hwmgr) *
+			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+		*size = 4;
+		break;
 	case AMDGPU_PP_SENSOR_UVD_POWER:
 		*((uint32_t *)value) = data->uvd_power_gated ? 0 : 1;
 		*size = 4;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index aeeeaa79056c..8d2865b72c7b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -1338,6 +1338,7 @@ static int vega12_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 			      void *value, int *size)
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
+	SmuMetrics_t metrics_table;
 	int ret = 0;
 
 	switch (idx) {
@@ -1360,6 +1361,24 @@ static int vega12_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*((uint32_t *)value) = vega12_thermal_get_temperature(hwmgr);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_HOTSPOT_TEMP:
+		ret = vega12_get_metrics_table(hwmgr, &metrics_table);
+		if (ret)
+			return ret;
+
+		*((uint32_t *)value) = metrics_table.TemperatureHotspot *
+			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+		*size = 4;
+		break;
+	case AMDGPU_PP_SENSOR_MEM_TEMP:
+		ret = vega12_get_metrics_table(hwmgr, &metrics_table);
+		if (ret)
+			return ret;
+
+		*((uint32_t *)value) = metrics_table.TemperatureHBM *
+			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+		*size = 4;
+		break;
 	case AMDGPU_PP_SENSOR_UVD_POWER:
 		*((uint32_t *)value) = data->uvd_power_gated ? 0 : 1;
 		*size = 4;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
index 6fcdefcfa930..abd3df9a0211 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
@@ -2138,10 +2138,28 @@ static int vega20_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		if (!ret)
 			*size = 4;
 		break;
-	case AMDGPU_PP_SENSOR_GPU_TEMP:
+	case AMDGPU_PP_SENSOR_HOTSPOT_TEMP:
 		*((uint32_t *)value) = vega20_thermal_get_temperature(hwmgr);
 		*size = 4;
 		break;
+	case AMDGPU_PP_SENSOR_EDGE_TEMP:
+		ret = vega20_get_metrics_table(hwmgr, &metrics_table);
+		if (ret)
+			return ret;
+
+		*((uint32_t *)value) = metrics_table.TemperatureEdge *
+			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+		*size = 4;
+		break;
+	case AMDGPU_PP_SENSOR_MEM_TEMP:
+		ret = vega20_get_metrics_table(hwmgr, &metrics_table);
+		if (ret)
+			return ret;
+
+		*((uint32_t *)value) = metrics_table.TemperatureHBM *
+			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+		*size = 4;
+		break;
 	case AMDGPU_PP_SENSOR_UVD_POWER:
 		*((uint32_t *)value) = data->uvd_power_gated ? 0 : 1;
 		*size = 4;
-- 
2.21.0

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

[PATCH 5/5] drm/amd/powerplay: support hwmon temperature channel labels V2

[Evan Quan]

Expose temp[1-3]_label hwmon interfaces. While temp2_label
and temp3_label are visible for SOC15 dGPUs only.

- V2: correct temp1_label as "edge"

Change-Id: I7f1e10c52ec21d272027554cdf6da97103e0be58
Signed-off-by: Evan Quan <evan.quan@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c        | 35 ++++++++++++++++++-
 .../gpu/drm/amd/include/kgd_pp_interface.h    |  7 ++++
 2 files changed, 41 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 291c06f40dba..4fe1b58892a2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -120,6 +120,15 @@ static const struct cg_flag_name clocks[] = {
 	{0, NULL},
 };
 
+static const struct hwmon_temp_label {
+	enum PP_HWMON_TEMP channel;
+	const char *label;
+} temp_label[] = {
+	{PP_TEMP_EDGE, "edge"},
+	{PP_TEMP_JUNCTION, "junction"},
+	{PP_TEMP_MEM, "mem"},
+};
+
 void amdgpu_pm_acpi_event_handler(struct amdgpu_device *adev)
 {
 	if (adev->pm.dpm_enabled) {
@@ -1511,6 +1520,19 @@ static ssize_t amdgpu_hwmon_show_mem_temp_thresh(struct device *dev,
 	return snprintf(buf, PAGE_SIZE, "%d\n", temp);
 }
 
+static ssize_t amdgpu_hwmon_show_temp_label(struct device *dev,
+					     struct device_attribute *attr,
+					     char *buf)
+{
+	struct amdgpu_device *adev = dev_get_drvdata(dev);
+	int channel = to_sensor_dev_attr(attr)->index;
+
+	if (channel >= PP_TEMP_MAX)
+		return -EINVAL;
+
+	return snprintf(buf, PAGE_SIZE, "%s\n", temp_label[channel].label);
+}
+
 static ssize_t amdgpu_hwmon_show_temp_emergency(struct device *dev,
 					     struct device_attribute *attr,
 					     char *buf)
@@ -2109,6 +2131,9 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
  * - temp[1-3]_input: the on die GPU temperature in millidegrees Celsius
  *   - temp2_input and temp3_input are supported on SOC15 dGPUs only
  *
+ * - temp[1-3]_label: temperature channel label
+ *   - temp2_label and temp3_label are supported on SOC15 dGPUs only
+ *
  * - temp[1-3]_crit: temperature critical max value in millidegrees Celsius
  *   - temp2_crit and temp3_crit are supported on SOC15 dGPUs only
  *
@@ -2176,6 +2201,9 @@ static SENSOR_DEVICE_ATTR(temp3_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, PP
 static SENSOR_DEVICE_ATTR(temp3_crit, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp3_crit_hyst, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 1);
 static SENSOR_DEVICE_ATTR(temp3_emergency, S_IRUGO, amdgpu_hwmon_show_temp_emergency, NULL, PP_TEMP_MEM);
+static SENSOR_DEVICE_ATTR(temp1_label, S_IRUGO, amdgpu_hwmon_show_temp_label, NULL, PP_TEMP_EDGE);
+static SENSOR_DEVICE_ATTR(temp2_label, S_IRUGO, amdgpu_hwmon_show_temp_label, NULL, PP_TEMP_JUNCTION);
+static SENSOR_DEVICE_ATTR(temp3_label, S_IRUGO, amdgpu_hwmon_show_temp_label, NULL, PP_TEMP_MEM);
 static SENSOR_DEVICE_ATTR(pwm1, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1, amdgpu_hwmon_set_pwm1, 0);
 static SENSOR_DEVICE_ATTR(pwm1_enable, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1_enable, amdgpu_hwmon_set_pwm1_enable, 0);
 static SENSOR_DEVICE_ATTR(pwm1_min, S_IRUGO, amdgpu_hwmon_get_pwm1_min, NULL, 0);
@@ -2211,6 +2239,9 @@ static struct attribute *hwmon_attributes[] = {
 	&sensor_dev_attr_temp1_emergency.dev_attr.attr,
 	&sensor_dev_attr_temp2_emergency.dev_attr.attr,
 	&sensor_dev_attr_temp3_emergency.dev_attr.attr,
+	&sensor_dev_attr_temp1_label.dev_attr.attr,
+	&sensor_dev_attr_temp2_label.dev_attr.attr,
+	&sensor_dev_attr_temp3_label.dev_attr.attr,
 	&sensor_dev_attr_pwm1.dev_attr.attr,
 	&sensor_dev_attr_pwm1_enable.dev_attr.attr,
 	&sensor_dev_attr_pwm1_min.dev_attr.attr,
@@ -2344,7 +2375,9 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 	     attr == &sensor_dev_attr_temp2_emergency.dev_attr.attr ||
 	     attr == &sensor_dev_attr_temp3_emergency.dev_attr.attr ||
 	     attr == &sensor_dev_attr_temp2_input.dev_attr.attr ||
-	     attr == &sensor_dev_attr_temp3_input.dev_attr.attr))
+	     attr == &sensor_dev_attr_temp3_input.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp2_label.dev_attr.attr ||
+	     attr == &sensor_dev_attr_temp3_label.dev_attr.attr))
 		return 0;
 
 	return effective_mode;
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index a8bf8e90ceeb..30788d510576 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -162,6 +162,13 @@ struct pp_states_info {
 	uint32_t states[16];
 };
 
+enum PP_HWMON_TEMP {
+	PP_TEMP_EDGE = 0,
+	PP_TEMP_JUNCTION,
+	PP_TEMP_MEM,
+	PP_TEMP_MAX
+};
+
 #define PP_GROUP_MASK        0xF0000000
 #define PP_GROUP_SHIFT       28
 
-- 
2.21.0

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

Re: [PATCH 5/5] drm/amd/powerplay: support hwmon temperature channel labels V2

[Deucher, Alexander]

[-- Attachment #1.1: Type: text/plain, Size: 5943 bytes --]

Series is:
Reviewed-by: Alex Deucher <alexander.deucher-5C7GfCeVMHo@public.gmane.org>
________________________________
From: amd-gfx <amd-gfx-bounces-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org> on behalf of Evan Quan <evan.quan-5C7GfCeVMHo@public.gmane.org>
Sent: Tuesday, April 23, 2019 4:37 AM
To: amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
Cc: Deucher, Alexander; Quan, Evan
Subject: [PATCH 5/5] drm/amd/powerplay: support hwmon temperature channel labels V2

Expose temp[1-3]_label hwmon interfaces. While temp2_label
and temp3_label are visible for SOC15 dGPUs only.

- V2: correct temp1_label as "edge"

Change-Id: I7f1e10c52ec21d272027554cdf6da97103e0be58
Signed-off-by: Evan Quan <evan.quan-5C7GfCeVMHo@public.gmane.org>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c        | 35 ++++++++++++++++++-
 .../gpu/drm/amd/include/kgd_pp_interface.h    |  7 ++++
 2 files changed, 41 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 291c06f40dba..4fe1b58892a2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -120,6 +120,15 @@ static const struct cg_flag_name clocks[] = {
         {0, NULL},
 };

+static const struct hwmon_temp_label {
+       enum PP_HWMON_TEMP channel;
+       const char *label;
+} temp_label[] = {
+       {PP_TEMP_EDGE, "edge"},
+       {PP_TEMP_JUNCTION, "junction"},
+       {PP_TEMP_MEM, "mem"},
+};
+
 void amdgpu_pm_acpi_event_handler(struct amdgpu_device *adev)
 {
         if (adev->pm.dpm_enabled) {
@@ -1511,6 +1520,19 @@ static ssize_t amdgpu_hwmon_show_mem_temp_thresh(struct device *dev,
         return snprintf(buf, PAGE_SIZE, "%d\n", temp);
 }

+static ssize_t amdgpu_hwmon_show_temp_label(struct device *dev,
+                                            struct device_attribute *attr,
+                                            char *buf)
+{
+       struct amdgpu_device *adev = dev_get_drvdata(dev);
+       int channel = to_sensor_dev_attr(attr)->index;
+
+       if (channel >= PP_TEMP_MAX)
+               return -EINVAL;
+
+       return snprintf(buf, PAGE_SIZE, "%s\n", temp_label[channel].label);
+}
+
 static ssize_t amdgpu_hwmon_show_temp_emergency(struct device *dev,
                                              struct device_attribute *attr,
                                              char *buf)
@@ -2109,6 +2131,9 @@ static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
  * - temp[1-3]_input: the on die GPU temperature in millidegrees Celsius
  *   - temp2_input and temp3_input are supported on SOC15 dGPUs only
  *
+ * - temp[1-3]_label: temperature channel label
+ *   - temp2_label and temp3_label are supported on SOC15 dGPUs only
+ *
  * - temp[1-3]_crit: temperature critical max value in millidegrees Celsius
  *   - temp2_crit and temp3_crit are supported on SOC15 dGPUs only
  *
@@ -2176,6 +2201,9 @@ static SENSOR_DEVICE_ATTR(temp3_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, PP
 static SENSOR_DEVICE_ATTR(temp3_crit, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp3_crit_hyst, S_IRUGO, amdgpu_hwmon_show_mem_temp_thresh, NULL, 1);
 static SENSOR_DEVICE_ATTR(temp3_emergency, S_IRUGO, amdgpu_hwmon_show_temp_emergency, NULL, PP_TEMP_MEM);
+static SENSOR_DEVICE_ATTR(temp1_label, S_IRUGO, amdgpu_hwmon_show_temp_label, NULL, PP_TEMP_EDGE);
+static SENSOR_DEVICE_ATTR(temp2_label, S_IRUGO, amdgpu_hwmon_show_temp_label, NULL, PP_TEMP_JUNCTION);
+static SENSOR_DEVICE_ATTR(temp3_label, S_IRUGO, amdgpu_hwmon_show_temp_label, NULL, PP_TEMP_MEM);
 static SENSOR_DEVICE_ATTR(pwm1, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1, amdgpu_hwmon_set_pwm1, 0);
 static SENSOR_DEVICE_ATTR(pwm1_enable, S_IRUGO | S_IWUSR, amdgpu_hwmon_get_pwm1_enable, amdgpu_hwmon_set_pwm1_enable, 0);
 static SENSOR_DEVICE_ATTR(pwm1_min, S_IRUGO, amdgpu_hwmon_get_pwm1_min, NULL, 0);
@@ -2211,6 +2239,9 @@ static struct attribute *hwmon_attributes[] = {
         &sensor_dev_attr_temp1_emergency.dev_attr.attr,
         &sensor_dev_attr_temp2_emergency.dev_attr.attr,
         &sensor_dev_attr_temp3_emergency.dev_attr.attr,
+       &sensor_dev_attr_temp1_label.dev_attr.attr,
+       &sensor_dev_attr_temp2_label.dev_attr.attr,
+       &sensor_dev_attr_temp3_label.dev_attr.attr,
         &sensor_dev_attr_pwm1.dev_attr.attr,
         &sensor_dev_attr_pwm1_enable.dev_attr.attr,
         &sensor_dev_attr_pwm1_min.dev_attr.attr,
@@ -2344,7 +2375,9 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
              attr == &sensor_dev_attr_temp2_emergency.dev_attr.attr ||
              attr == &sensor_dev_attr_temp3_emergency.dev_attr.attr ||
              attr == &sensor_dev_attr_temp2_input.dev_attr.attr ||
-            attr == &sensor_dev_attr_temp3_input.dev_attr.attr))
+            attr == &sensor_dev_attr_temp3_input.dev_attr.attr ||
+            attr == &sensor_dev_attr_temp2_label.dev_attr.attr ||
+            attr == &sensor_dev_attr_temp3_label.dev_attr.attr))
                 return 0;

         return effective_mode;
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index a8bf8e90ceeb..30788d510576 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -162,6 +162,13 @@ struct pp_states_info {
         uint32_t states[16];
 };

+enum PP_HWMON_TEMP {
+       PP_TEMP_EDGE = 0,
+       PP_TEMP_JUNCTION,
+       PP_TEMP_MEM,
+       PP_TEMP_MAX
+};
+
 #define PP_GROUP_MASK        0xF0000000
 #define PP_GROUP_SHIFT       28

--
2.21.0

_______________________________________________
amd-gfx mailing list
amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

[-- Attachment #1.2: Type: text/html, Size: 10405 bytes --]

[-- Attachment #2: Type: text/plain, Size: 153 bytes --]

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx