path: root/common/recipes-kernel/linux/linux-yocto-4.14.71/5245-drm-amd-powerplay-added-vega20-overdrive-support-V3.patch

From 74b0a922b04d1b411801ff7e08fc8c0d8798e89b Mon Sep 17 00:00:00 2001
From: Evan Quan <evan.quan@amd.com>
Date: Wed, 29 Aug 2018 14:38:50 +0800
Subject: [PATCH 5245/5725] drm/amd/powerplay: added vega20 overdrive support
 V3

Added vega20 overdrive support based on existing OD sysfs
APIs. However, the OD logics are simplified on vega20. So,
the behavior will be a little different and works only on
some limited levels.

V2: fix typo
    fix commit description
    revise error logs
    add support for clock OD

V3: separate clock from voltage OD settings

Change-Id: I403cb38a95863db664cf06d030ac42a19bff6b33
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c             |  45 ++++
 drivers/gpu/drm/amd/include/kgd_pp_interface.h     |   2 +
 drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c | 289 ++++++++++++++++++++-
 3 files changed, 335 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index f5404f2..e63b0c2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -474,6 +474,8 @@ static ssize_t amdgpu_set_pp_table(struct device *dev,
  * in each power level within a power state.  The pp_od_clk_voltage is used for
  * this.
  *
+ * < For Vega10 and previous ASICs >
+ *
  * Reading the file will display:
  *
  * - a list of engine clock levels and voltages labeled OD_SCLK
@@ -491,6 +493,44 @@ static ssize_t amdgpu_set_pp_table(struct device *dev,
  * "c" (commit) to the file to commit your changes.  If you want to reset to the
  * default power levels, write "r" (reset) to the file to reset them.
  *
+ *
+ * < For Vega20 >
+ *
+ * Reading the file will display:
+ *
+ * - minimum and maximum engine clock labeled OD_SCLK
+ *
+ * - maximum memory clock labeled OD_MCLK
+ *
+ * - three <frequency, voltage offset> points labeled OD_VDDC_CURVE.
+ *   They can be used to calibrate the sclk voltage curve.
+ *
+ * - a list of valid ranges for sclk, mclk, and voltage curve points
+ *   labeled OD_RANGE
+ *
+ * To manually adjust these settings:
+ *
+ * - First select manual using power_dpm_force_performance_level
+ *
+ * - For clock frequency setting, enter a new value by writing a
+ *   string that contains "s/m index clock" to the file. The index
+ *   should be 0 if to set minimum clock. And 1 if to set maximum
+ *   clock. E.g., "s 0 500" will update minimum sclk to be 500 MHz.
+ *   "m 1 800" will update maximum mclk to be 800Mhz.
+ *
+ *   For sclk voltage curve, enter the new values by writing a
+ *   string that contains "vc point clock voff" to the file. The
+ *   points are indexed by 0, 1 and 2. E.g., "vc 0 300 10" will
+ *   update point1 with clock set as 300Mhz and voltage increased
+ *   by 10mV. "vc 2 1000 -10" will update point3 with clock set
+ *   as 1000Mhz and voltage drop by 10mV.
+ *
+ * - When you have edited all of the states as needed, write "c" (commit)
+ *   to the file to commit your changes
+ *
+ * - If you want to reset to the default power levels, write "r" (reset)
+ *   to the file to reset them
+ *
  */
 
 static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev,
@@ -520,6 +560,8 @@ static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev,
 		type = PP_OD_RESTORE_DEFAULT_TABLE;
 	else if (*buf == 'c')
 		type = PP_OD_COMMIT_DPM_TABLE;
+	else if (!strncmp(buf, "vc", 2))
+		type = PP_OD_EDIT_VDDC_CURVE;
 	else
 		return -EINVAL;
 
@@ -527,6 +569,8 @@ static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev,
 
 	tmp_str = buf_cpy;
 
+	if (type == PP_OD_EDIT_VDDC_CURVE)
+		tmp_str++;
 	while (isspace(*++tmp_str));
 
 	while (tmp_str[0]) {
@@ -570,6 +614,7 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
 	if (adev->powerplay.pp_funcs->print_clock_levels) {
 		size = amdgpu_dpm_print_clock_levels(adev, OD_SCLK, buf);
 		size += amdgpu_dpm_print_clock_levels(adev, OD_MCLK, buf+size);
+		size += amdgpu_dpm_print_clock_levels(adev, OD_VDDC_CURVE, buf+size);
 		size += amdgpu_dpm_print_clock_levels(adev, OD_RANGE, buf+size);
 		return size;
 	} else {
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index 6a41b81..448dee4 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -94,6 +94,7 @@ enum pp_clock_type {
 	PP_PCIE,
 	OD_SCLK,
 	OD_MCLK,
+	OD_VDDC_CURVE,
 	OD_RANGE,
 };
 
@@ -141,6 +142,7 @@ enum {
 enum PP_OD_DPM_TABLE_COMMAND {
 	PP_OD_EDIT_SCLK_VDDC_TABLE,
 	PP_OD_EDIT_MCLK_VDDC_TABLE,
+	PP_OD_EDIT_VDDC_CURVE,
 	PP_OD_RESTORE_DEFAULT_TABLE,
 	PP_OD_COMMIT_DPM_TABLE
 };
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
index fb32b28..3efd59e 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
@@ -2325,11 +2325,207 @@ static int vega20_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
+static int vega20_odn_edit_dpm_table(struct pp_hwmgr *hwmgr,
+					enum PP_OD_DPM_TABLE_COMMAND type,
+					long *input, uint32_t size)
+{
+	struct vega20_hwmgr *data =
+			(struct vega20_hwmgr *)(hwmgr->backend);
+	struct vega20_od8_single_setting *od8_settings =
+			data->od8_settings.od8_settings_array;
+	OverDriveTable_t *od_table =
+			&(data->smc_state_table.overdrive_table);
+	struct pp_clock_levels_with_latency clocks;
+	int32_t input_index, input_clk, input_vol, i;
+	int ret;
+
+	PP_ASSERT_WITH_CODE(input, "NULL user input for clock and voltage",
+				return -EINVAL);
+
+	switch (type) {
+	case PP_OD_EDIT_SCLK_VDDC_TABLE:
+		if (!(od8_settings[OD8_SETTING_GFXCLK_FMIN].feature_id &&
+		      od8_settings[OD8_SETTING_GFXCLK_FMAX].feature_id)) {
+			pr_info("Sclk min/max frequency overdrive not supported\n");
+			return -EOPNOTSUPP;
+		}
+
+		for (i = 0; i < size; i += 2) {
+			if (i + 2 > size) {
+				pr_info("invalid number of input parameters %d\n",
+					size);
+				return -EINVAL;
+			}
+
+			input_index = input[i];
+			input_clk = input[i + 1];
+
+			if (input_index != 0 && input_index != 1) {
+				pr_info("Invalid index %d\n", input_index);
+				pr_info("Support min/max sclk frequency setting only which index by 0/1\n");
+				return -EINVAL;
+			}
+
+			if (input_clk < od8_settings[OD8_SETTING_GFXCLK_FMIN].min_value ||
+			    input_clk > od8_settings[OD8_SETTING_GFXCLK_FMAX].max_value) {
+				pr_info("clock freq %d is not within allowed range [%d - %d]\n",
+					input_clk,
+					od8_settings[OD8_SETTING_GFXCLK_FMIN].min_value,
+					od8_settings[OD8_SETTING_GFXCLK_FMAX].max_value);
+				return -EINVAL;
+			}
+
+			if (input_index == 0)
+				od_table->GfxclkFmin = input_clk;
+			else
+				od_table->GfxclkFmax = input_clk;
+		}
+
+		break;
+
+	case PP_OD_EDIT_MCLK_VDDC_TABLE:
+		if (!od8_settings[OD8_SETTING_UCLK_FMAX].feature_id) {
+			pr_info("Mclk max frequency overdrive not supported\n");
+			return -EOPNOTSUPP;
+		}
+
+		ret = vega20_get_memclocks(hwmgr, &clocks);
+		PP_ASSERT_WITH_CODE(!ret,
+				"Attempt to get memory clk levels failed!",
+				return ret);
+
+		for (i = 0; i < size; i += 2) {
+			if (i + 2 > size) {
+				pr_info("invalid number of input parameters %d\n",
+					size);
+				return -EINVAL;
+			}
+
+			input_index = input[i];
+			input_clk = input[i + 1];
+
+			if (input_index != 1) {
+				pr_info("Invalid index %d\n", input_index);
+				pr_info("Support max Mclk frequency setting only which index by 1\n");
+				return -EINVAL;
+			}
+
+			if (input_clk < clocks.data[0].clocks_in_khz / 100 ||
+			    input_clk > od8_settings[OD8_SETTING_UCLK_FMAX].max_value) {
+				pr_info("clock freq %d is not within allowed range [%d - %d]\n",
+					input_clk,
+					clocks.data[0].clocks_in_khz / 100,
+					od8_settings[OD8_SETTING_UCLK_FMAX].max_value);
+				return -EINVAL;
+			}
+
+			od_table->UclkFmax = input_clk;
+		}
+
+		break;
+
+	case PP_OD_EDIT_VDDC_CURVE:
+		if (!(od8_settings[OD8_SETTING_GFXCLK_FREQ1].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_FREQ2].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_FREQ3].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE2].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE3].feature_id)) {
+			pr_info("Voltage curve calibrate not supported\n");
+			return -EOPNOTSUPP;
+		}
+
+		for (i = 0; i < size; i += 3) {
+			if (i + 3 > size) {
+				pr_info("invalid number of input parameters %d\n",
+					size);
+				return -EINVAL;
+			}
+
+			input_index = input[i];
+			input_clk = input[i + 1];
+			input_vol = input[i + 2];
+
+			if (input_index > 2) {
+				pr_info("Setting for point %d is not supported\n",
+						input_index + 1);
+				pr_info("Three supported points index by 0, 1, 2\n");
+				return -EINVAL;
+			}
+
+			if (input_clk < od8_settings[OD8_SETTING_GFXCLK_FMIN].min_value ||
+			    input_clk > od8_settings[OD8_SETTING_GFXCLK_FMAX].max_value) {
+				pr_info("clock freq %d is not within allowed range [%d - %d]\n",
+					input_clk,
+					od8_settings[OD8_SETTING_GFXCLK_FMIN].min_value,
+					od8_settings[OD8_SETTING_GFXCLK_FMAX].max_value);
+				return -EINVAL;
+			}
+
+			/* TODO: suppose voltage1/2/3 has the same min/max value */
+			if (input_vol < od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].min_value ||
+			    input_vol > od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].max_value) {
+				pr_info("clock voltage offset %d is not within allowed range [%d - %d]\n",
+					input_vol,
+					od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].min_value,
+					od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].max_value);
+				return -EINVAL;
+			}
+
+			switch (input_index) {
+			case 0:
+				od_table->GfxclkFreq1 = input_clk;
+				od_table->GfxclkOffsetVolt1 = input_vol;
+				break;
+			case 1:
+				od_table->GfxclkFreq2 = input_clk;
+				od_table->GfxclkOffsetVolt2 = input_vol;
+				break;
+			case 2:
+				od_table->GfxclkFreq3 = input_clk;
+				od_table->GfxclkOffsetVolt3 = input_vol;
+				break;
+			}
+		}
+		break;
+
+	case PP_OD_RESTORE_DEFAULT_TABLE:
+		ret = vega20_copy_table_from_smc(hwmgr,
+				(uint8_t *)od_table,
+				TABLE_OVERDRIVE);
+		PP_ASSERT_WITH_CODE(!ret,
+				"Failed to export overdrive table!",
+				return ret);
+		break;
+
+	case PP_OD_COMMIT_DPM_TABLE:
+		ret = vega20_copy_table_to_smc(hwmgr,
+				(uint8_t *)od_table,
+				TABLE_OVERDRIVE);
+		PP_ASSERT_WITH_CODE(!ret,
+				"Failed to import overdrive table!",
+				return ret);
+
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static int vega20_print_clock_levels(struct pp_hwmgr *hwmgr,
 		enum pp_clock_type type, char *buf)
 {
-	int i, now, size = 0;
+	struct vega20_hwmgr *data =
+			(struct vega20_hwmgr *)(hwmgr->backend);
+	struct vega20_od8_single_setting *od8_settings =
+			data->od8_settings.od8_settings_array;
+	OverDriveTable_t *od_table =
+			&(data->smc_state_table.overdrive_table);
 	struct pp_clock_levels_with_latency clocks;
+	int i, now, size = 0;
 	int ret = 0;
 
 	switch (type) {
@@ -2370,6 +2566,95 @@ static int vega20_print_clock_levels(struct pp_hwmgr *hwmgr,
 	case PP_PCIE:
 		break;
 
+	case OD_SCLK:
+		if (od8_settings[OD8_SETTING_GFXCLK_FMIN].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_FMAX].feature_id) {
+			size = sprintf(buf, "%s:\n", "OD_SCLK");
+			size += sprintf(buf + size, "0: %10uMhz\n",
+				od_table->GfxclkFmin);
+			size += sprintf(buf + size, "1: %10uMhz\n",
+				od_table->GfxclkFmax);
+		}
+		break;
+
+	case OD_MCLK:
+		if (od8_settings[OD8_SETTING_UCLK_FMAX].feature_id) {
+			size = sprintf(buf, "%s:\n", "OD_MCLK");
+			size += sprintf(buf + size, "1: %10uMhz\n",
+				od_table->UclkFmax);
+		}
+
+		break;
+
+	case OD_VDDC_CURVE:
+		if (od8_settings[OD8_SETTING_GFXCLK_FREQ1].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_FREQ2].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_FREQ3].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE2].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE3].feature_id) {
+			size = sprintf(buf, "%s:\n", "OD_VDDC_CURVE");
+			size += sprintf(buf + size, "0: %10uMhz %10dmV\n",
+				od_table->GfxclkFreq1,
+				od_table->GfxclkOffsetVolt1);
+			size += sprintf(buf + size, "1: %10uMhz %10dmV\n",
+				od_table->GfxclkFreq2,
+				od_table->GfxclkOffsetVolt2);
+			size += sprintf(buf + size, "2: %10uMhz %10dmV\n",
+				od_table->GfxclkFreq3,
+				od_table->GfxclkOffsetVolt3);
+		}
+
+		break;
+
+	case OD_RANGE:
+		size = sprintf(buf, "%s:\n", "OD_RANGE");
+
+		if (od8_settings[OD8_SETTING_GFXCLK_FMIN].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_FMAX].feature_id) {
+			size += sprintf(buf + size, "SCLK: %7uMhz %10uMhz\n",
+				od8_settings[OD8_SETTING_GFXCLK_FMIN].min_value,
+				od8_settings[OD8_SETTING_GFXCLK_FMAX].max_value);
+		}
+
+		if (od8_settings[OD8_SETTING_UCLK_FMAX].feature_id) {
+			ret = vega20_get_memclocks(hwmgr, &clocks);
+			PP_ASSERT_WITH_CODE(!ret,
+					"Fail to get memory clk levels!",
+					return ret);
+
+			size += sprintf(buf + size, "MCLK: %7uMhz %10uMhz\n",
+				clocks.data[0].clocks_in_khz / 100,
+				od8_settings[OD8_SETTING_UCLK_FMAX].max_value);
+		}
+
+		if (od8_settings[OD8_SETTING_GFXCLK_FREQ1].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_FREQ2].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_FREQ3].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE2].feature_id &&
+		    od8_settings[OD8_SETTING_GFXCLK_VOLTAGE3].feature_id) {
+			size += sprintf(buf + size, "VDDC_CURVE_SCLK[0]: %7uMhz %10uMhz\n",
+				od8_settings[OD8_SETTING_GFXCLK_FREQ1].min_value,
+				od8_settings[OD8_SETTING_GFXCLK_FREQ1].max_value);
+			size += sprintf(buf + size, "VDDC_CURVE_VOFF[0]: %7dmV %11dmV\n",
+				od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].min_value,
+				od8_settings[OD8_SETTING_GFXCLK_VOLTAGE1].max_value);
+			size += sprintf(buf + size, "VDDC_CURVE_SCLK[1]: %7uMhz %10uMhz\n",
+				od8_settings[OD8_SETTING_GFXCLK_FREQ2].min_value,
+				od8_settings[OD8_SETTING_GFXCLK_FREQ2].max_value);
+			size += sprintf(buf + size, "VDDC_CURVE_VOFF[1]: %7dmV %11dmV\n",
+				od8_settings[OD8_SETTING_GFXCLK_VOLTAGE2].min_value,
+				od8_settings[OD8_SETTING_GFXCLK_VOLTAGE2].max_value);
+			size += sprintf(buf + size, "VDDC_CURVE_SCLK[2]: %7uMhz %10uMhz\n",
+				od8_settings[OD8_SETTING_GFXCLK_FREQ3].min_value,
+				od8_settings[OD8_SETTING_GFXCLK_FREQ3].max_value);
+			size += sprintf(buf + size, "VDDC_CURVE_VOFF[2]: %7dmV %11dmV\n",
+				od8_settings[OD8_SETTING_GFXCLK_VOLTAGE3].min_value,
+				od8_settings[OD8_SETTING_GFXCLK_VOLTAGE3].max_value);
+		}
+
+		break;
 	default:
 		break;
 	}
@@ -2977,6 +3262,8 @@ static const struct pp_hwmgr_func vega20_hwmgr_funcs = {
 		vega20_get_mclk_od,
 	.set_mclk_od =
 		vega20_set_mclk_od,
+	.odn_edit_dpm_table =
+		vega20_odn_edit_dpm_table,
 	/* for sysfs to retrive/set gfxclk/memclk */
 	.force_clock_level =
 		vega20_force_clock_level,
-- 
2.7.4