Backmerge tag 'v4.7' into drm-next

Linux 4.7

As requested by Daniel Vetter as the conflicts were getting messy.
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index e19520c..d9c88d13 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -1106,6 +1106,10 @@
 	if (fences == 0 && handles == 0) {
 		if (adev->pm.dpm_enabled) {
 			amdgpu_dpm_enable_uvd(adev, false);
+			/* just work around for uvd clock remain high even
+			 * when uvd dpm disabled on Polaris10 */
+			if (adev->asic_type == CHIP_POLARIS10)
+				amdgpu_asic_set_uvd_clocks(adev, 0, 0);
 		} else {
 			amdgpu_asic_set_uvd_clocks(adev, 0, 0);
 		}
diff --git a/drivers/gpu/drm/amd/amdgpu/atombios_i2c.c b/drivers/gpu/drm/amd/amdgpu/atombios_i2c.c
index 13cdb01..bc56c8a 100644
--- a/drivers/gpu/drm/amd/amdgpu/atombios_i2c.c
+++ b/drivers/gpu/drm/amd/amdgpu/atombios_i2c.c
@@ -156,3 +156,18 @@
 	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
 }
 
+void amdgpu_atombios_i2c_channel_trans(struct amdgpu_device* adev, u8 slave_addr, u8 line_number, u8 offset, u8 data)
+{
+	PROCESS_I2C_CHANNEL_TRANSACTION_PS_ALLOCATION args;
+	int index = GetIndexIntoMasterTable(COMMAND, ProcessI2cChannelTransaction);
+
+	args.ucRegIndex = offset;
+	args.lpI2CDataOut = data;
+	args.ucFlag = 1;
+	args.ucI2CSpeed = TARGET_HW_I2C_CLOCK;
+	args.ucTransBytes = 1;
+	args.ucSlaveAddr = slave_addr;
+	args.ucLineNumber = line_number;
+
+	amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/atombios_i2c.h b/drivers/gpu/drm/amd/amdgpu/atombios_i2c.h
index d6128d9d..251aaf4 100644
--- a/drivers/gpu/drm/amd/amdgpu/atombios_i2c.h
+++ b/drivers/gpu/drm/amd/amdgpu/atombios_i2c.h
@@ -27,5 +27,7 @@
 int amdgpu_atombios_i2c_xfer(struct i2c_adapter *i2c_adap,
 		      struct i2c_msg *msgs, int num);
 u32 amdgpu_atombios_i2c_func(struct i2c_adapter *adap);
+void amdgpu_atombios_i2c_channel_trans(struct amdgpu_device* adev,
+		u8 slave_addr, u8 line_number, u8 offset, u8 data);
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index c30b6ac..aabaf2f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -28,6 +28,7 @@
 #include "vid.h"
 #include "amdgpu_ucode.h"
 #include "amdgpu_atombios.h"
+#include "atombios_i2c.h"
 #include "clearstate_vi.h"
 
 #include "gmc/gmc_8_2_d.h"
@@ -47,6 +48,8 @@
 #include "dce/dce_10_0_d.h"
 #include "dce/dce_10_0_sh_mask.h"
 
+#include "smu/smu_7_1_3_d.h"
+
 #define GFX8_NUM_GFX_RINGS     1
 #define GFX8_NUM_COMPUTE_RINGS 8
 
@@ -282,6 +285,7 @@
 	mmTCP_ADDR_CONFIG, 0x000003ff, 0x000000f3,
 	mmTCP_CHAN_STEER_HI, 0xffffffff, 0x00000000,
 	mmTCP_CHAN_STEER_LO, 0xffffffff, 0x00003210,
+	mmVGT_RESET_DEBUG, 0x00000004, 0x00000004,
 };
 
 static const u32 polaris11_golden_common_all[] =
@@ -312,6 +316,7 @@
 	mmTCC_CTRL, 0x00100000, 0xf31fff7f,
 	mmTCP_ADDR_CONFIG, 0x000003ff, 0x000000f7,
 	mmTCP_CHAN_STEER_HI, 0xffffffff, 0x00000000,
+	mmVGT_RESET_DEBUG, 0x00000004, 0x00000004,
 };
 
 static const u32 polaris10_golden_common_all[] =
@@ -693,6 +698,11 @@
 		amdgpu_program_register_sequence(adev,
 						 polaris10_golden_common_all,
 						 (const u32)ARRAY_SIZE(polaris10_golden_common_all));
+		WREG32_SMC(ixCG_ACLK_CNTL, 0x0000001C);
+		if (adev->pdev->revision == 0xc7) {
+			amdgpu_atombios_i2c_channel_trans(adev, 0x10, 0x96, 0x1E, 0xDD);
+			amdgpu_atombios_i2c_channel_trans(adev, 0x10, 0x96, 0x1F, 0xD0);
+		}
 		break;
 	case CHIP_CARRIZO:
 		amdgpu_program_register_sequence(adev,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c
index 9d764c4..a6c9b42 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c
@@ -780,7 +780,7 @@
 			table->Smio[level] |=
 				data->mvdd_voltage_table.entries[level].smio_low;
 		}
-		table->SmioMask2 = data->vddci_voltage_table.mask_low;
+		table->SmioMask2 = data->mvdd_voltage_table.mask_low;
 
 		table->MvddLevelCount = (uint32_t) PP_HOST_TO_SMC_UL(count);
 	}
@@ -1470,22 +1470,19 @@
 
 	table->ACPILevel.Flags &= ~PPSMC_SWSTATE_FLAG_DC;
 
-	if (!data->sclk_dpm_key_disabled) {
-		/* Get MinVoltage and Frequency from DPM0,
-		 * already converted to SMC_UL */
-		sclk_frequency = data->dpm_table.sclk_table.dpm_levels[0].value;
-		result = polaris10_get_dependency_volt_by_clk(hwmgr,
-				table_info->vdd_dep_on_sclk,
-				table->ACPILevel.SclkFrequency,
-				&table->ACPILevel.MinVoltage, &mvdd);
-		PP_ASSERT_WITH_CODE((0 == result),
-				"Cannot find ACPI VDDC voltage value "
-				"in Clock Dependency Table", );
-	} else {
-		sclk_frequency = data->vbios_boot_state.sclk_bootup_value;
-		table->ACPILevel.MinVoltage =
-				data->vbios_boot_state.vddc_bootup_value * VOLTAGE_SCALE;
-	}
+
+	/* Get MinVoltage and Frequency from DPM0,
+	 * already converted to SMC_UL */
+	sclk_frequency = data->dpm_table.sclk_table.dpm_levels[0].value;
+	result = polaris10_get_dependency_volt_by_clk(hwmgr,
+			table_info->vdd_dep_on_sclk,
+			sclk_frequency,
+			&table->ACPILevel.MinVoltage, &mvdd);
+	PP_ASSERT_WITH_CODE((0 == result),
+			"Cannot find ACPI VDDC voltage value "
+			"in Clock Dependency Table",
+			);
+
 
 	result = polaris10_calculate_sclk_params(hwmgr, sclk_frequency,  &(table->ACPILevel.SclkSetting));
 	PP_ASSERT_WITH_CODE(result == 0, "Error retrieving Engine Clock dividers from VBIOS.", return result);
@@ -1510,24 +1507,18 @@
 	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw1_frac);
 	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Sclk_ss_slew_rate);
 
-	if (!data->mclk_dpm_key_disabled) {
-		/* Get MinVoltage and Frequency from DPM0, already converted to SMC_UL */
-		table->MemoryACPILevel.MclkFrequency =
-				data->dpm_table.mclk_table.dpm_levels[0].value;
-		result = polaris10_get_dependency_volt_by_clk(hwmgr,
-				table_info->vdd_dep_on_mclk,
-				table->MemoryACPILevel.MclkFrequency,
-				&table->MemoryACPILevel.MinVoltage, &mvdd);
-		PP_ASSERT_WITH_CODE((0 == result),
-				"Cannot find ACPI VDDCI voltage value "
-				"in Clock Dependency Table",
-				);
-	} else {
-		table->MemoryACPILevel.MclkFrequency =
-				data->vbios_boot_state.mclk_bootup_value;
-		table->MemoryACPILevel.MinVoltage =
-				data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE;
-	}
+
+	/* Get MinVoltage and Frequency from DPM0, already converted to SMC_UL */
+	table->MemoryACPILevel.MclkFrequency =
+			data->dpm_table.mclk_table.dpm_levels[0].value;
+	result = polaris10_get_dependency_volt_by_clk(hwmgr,
+			table_info->vdd_dep_on_mclk,
+			table->MemoryACPILevel.MclkFrequency,
+			&table->MemoryACPILevel.MinVoltage, &mvdd);
+	PP_ASSERT_WITH_CODE((0 == result),
+			"Cannot find ACPI VDDCI voltage value "
+			"in Clock Dependency Table",
+			);
 
 	us_mvdd = 0;
 	if ((POLARIS10_VOLTAGE_CONTROL_NONE == data->mvdd_control) ||
@@ -1572,6 +1563,7 @@
 	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
 			table_info->mm_dep_table;
 	struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
+	uint32_t vddci;
 
 	table->VceLevelCount = (uint8_t)(mm_table->count);
 	table->VceBootLevel = 0;
@@ -1581,9 +1573,18 @@
 		table->VceLevel[count].MinVoltage = 0;
 		table->VceLevel[count].MinVoltage |=
 				(mm_table->entries[count].vddc * VOLTAGE_SCALE) << VDDC_SHIFT;
+
+		if (POLARIS10_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
+			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
+						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
+		else if (POLARIS10_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
+			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
+		else
+			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+
 		table->VceLevel[count].MinVoltage |=
-				((mm_table->entries[count].vddc - data->vddc_vddci_delta) *
-						VOLTAGE_SCALE) << VDDCI_SHIFT;
+				(vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
 		table->VceLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
 
 		/*retrieve divider value for VBIOS */
@@ -1612,6 +1613,7 @@
 	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
 			table_info->mm_dep_table;
 	struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
+	uint32_t vddci;
 
 	table->SamuBootLevel = 0;
 	table->SamuLevelCount = (uint8_t)(mm_table->count);
@@ -1622,8 +1624,16 @@
 		table->SamuLevel[count].Frequency = mm_table->entries[count].samclock;
 		table->SamuLevel[count].MinVoltage |= (mm_table->entries[count].vddc *
 				VOLTAGE_SCALE) << VDDC_SHIFT;
-		table->SamuLevel[count].MinVoltage |= ((mm_table->entries[count].vddc -
-				data->vddc_vddci_delta) * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+		if (POLARIS10_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
+			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
+						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
+		else if (POLARIS10_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
+			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
+		else
+			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+		table->SamuLevel[count].MinVoltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
 		table->SamuLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
 
 		/* retrieve divider value for VBIOS */
@@ -1706,6 +1716,7 @@
 	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
 			table_info->mm_dep_table;
 	struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
+	uint32_t vddci;
 
 	table->UvdLevelCount = (uint8_t)(mm_table->count);
 	table->UvdBootLevel = 0;
@@ -1716,8 +1727,16 @@
 		table->UvdLevel[count].DclkFrequency = mm_table->entries[count].dclk;
 		table->UvdLevel[count].MinVoltage |= (mm_table->entries[count].vddc *
 				VOLTAGE_SCALE) << VDDC_SHIFT;
-		table->UvdLevel[count].MinVoltage |= ((mm_table->entries[count].vddc -
-				data->vddc_vddci_delta) * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+		if (POLARIS10_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
+			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
+						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
+		else if (POLARIS10_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
+			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
+		else
+			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+		table->UvdLevel[count].MinVoltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
 		table->UvdLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
 
 		/* retrieve divider value for VBIOS */
@@ -1738,8 +1757,8 @@
 		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].VclkFrequency);
 		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].DclkFrequency);
 		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].MinVoltage);
-
 	}
+
 	return result;
 }
 
@@ -1839,20 +1858,26 @@
 	for (i = 0; i < sclk_table->count; i++) {
 		data->smc_state_table.Sclk_CKS_masterEn0_7 |=
 				sclk_table->entries[i].cks_enable << i;
-
-		volt_without_cks =  (uint32_t)(((ro - 40) * 1000 - 2753594 - sclk_table->entries[i].clk/100 * 136418 /1000) / \
-					(sclk_table->entries[i].clk/100 * 1132925 /10000 - 242418)/100);
-
-		volt_with_cks = (uint32_t)((ro * 1000 -2396351 - sclk_table->entries[i].clk/100 * 329021/1000) / \
-				(sclk_table->entries[i].clk/10000 * 649434 /1000  - 18005)/10);
+		if (hwmgr->chip_id == CHIP_POLARIS10) {
+			volt_without_cks = (uint32_t)((2753594000U + (sclk_table->entries[i].clk/100) * 136418 -(ro - 70) * 1000000) / \
+						(2424180 - (sclk_table->entries[i].clk/100) * 1132925/1000));
+			volt_with_cks = (uint32_t)((2797202000U + sclk_table->entries[i].clk/100 * 3232 - (ro - 65) * 1000000) / \
+					(2522480 - sclk_table->entries[i].clk/100 * 115764/100));
+		} else {
+			volt_without_cks = (uint32_t)((2416794800U + (sclk_table->entries[i].clk/100) * 1476925/10 -(ro - 50) * 1000000) / \
+						(2625416 - (sclk_table->entries[i].clk/100) * (12586807/10000)));
+			volt_with_cks = (uint32_t)((2999656000U - sclk_table->entries[i].clk/100 * 392803 - (ro - 44) * 1000000) / \
+					(3422454 - sclk_table->entries[i].clk/100 * (18886376/10000)));
+		}
 
 		if (volt_without_cks >= volt_with_cks)
 			volt_offset = (uint8_t)(((volt_without_cks - volt_with_cks +
-					sclk_table->entries[i].cks_voffset) * 100 / 625) + 1);
+					sclk_table->entries[i].cks_voffset) * 100 + 624) / 625);
 
 		data->smc_state_table.Sclk_voltageOffset[i] = volt_offset;
 	}
 
+	data->smc_state_table.LdoRefSel = (table_info->cac_dtp_table->ucCKS_LDO_REFSEL != 0) ? table_info->cac_dtp_table->ucCKS_LDO_REFSEL : 6;
 	/* Populate CKS Lookup Table */
 	if (stretch_amount == 1 || stretch_amount == 2 || stretch_amount == 5)
 		stretch_amount2 = 0;
@@ -2629,6 +2654,8 @@
 	PP_ASSERT_WITH_CODE((0 == tmp_result),
 			"Failed to enable VR hot GPIO interrupt!", result = tmp_result);
 
+	smum_send_msg_to_smc(hwmgr->smumgr, (PPSMC_Msg)PPSMC_HasDisplay);
+
 	tmp_result = polaris10_enable_sclk_control(hwmgr);
 	PP_ASSERT_WITH_CODE((0 == tmp_result),
 			"Failed to enable SCLK control!", result = tmp_result);
@@ -2847,7 +2874,7 @@
 {
 	struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
 	uint16_t vv_id;
-	uint16_t vddc = 0;
+	uint32_t vddc = 0;
 	uint16_t i, j;
 	uint32_t sclk = 0;
 	struct phm_ppt_v1_information *table_info =
@@ -2878,8 +2905,9 @@
 				continue;
 			}
 
-			/* need to make sure vddc is less than 2v or else, it could burn the ASIC. */
-			PP_ASSERT_WITH_CODE((vddc < 2000 && vddc != 0),
+			/* need to make sure vddc is less than 2v or else, it could burn the ASIC.
+			 * real voltage level in unit of 0.01mv */
+			PP_ASSERT_WITH_CODE((vddc < 200000 && vddc != 0),
 					"Invalid VDDC value", result = -EINVAL;);
 
 			/* the voltage should not be zero nor equal to leakage ID */
@@ -3105,6 +3133,31 @@
 	return 0;
 }
 
+int polaris10_patch_voltage_workaround(struct pp_hwmgr *hwmgr)
+{
+	struct phm_ppt_v1_information *table_info =
+		       (struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_mclk_table =
+			table_info->vdd_dep_on_mclk;
+	struct phm_ppt_v1_voltage_lookup_table *lookup_table =
+			table_info->vddc_lookup_table;
+	uint32_t i;
+
+	if (hwmgr->chip_id == CHIP_POLARIS10 && hwmgr->hw_revision == 0xC7) {
+		if (lookup_table->entries[dep_mclk_table->entries[dep_mclk_table->count-1].vddInd].us_vdd >= 1000)
+			return 0;
+
+		for (i = 0; i < lookup_table->count; i++) {
+			if (lookup_table->entries[i].us_vdd < 0xff01 && lookup_table->entries[i].us_vdd >= 1000) {
+				dep_mclk_table->entries[dep_mclk_table->count-1].vddInd = (uint8_t) i;
+				return 0;
+			}
+		}
+	}
+	return 0;
+}
+
+
 int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
 {
 	struct polaris10_hwmgr *data;
@@ -3188,6 +3241,7 @@
 
 	polaris10_set_features_platform_caps(hwmgr);
 
+	polaris10_patch_voltage_workaround(hwmgr);
 	polaris10_init_dpm_defaults(hwmgr);
 
 	/* Get leakage voltage based on leakage ID. */
@@ -4557,6 +4611,15 @@
 	return 0;
 }
 
+static int polaris10_notify_smc_display(struct pp_hwmgr *hwmgr)
+{
+	struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
+
+	smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+		(PPSMC_Msg)PPSMC_MSG_SetVBITimeout, data->frame_time_x2);
+	return (smum_send_msg_to_smc(hwmgr->smumgr, (PPSMC_Msg)PPSMC_HasDisplay) == 0) ?  0 : -EINVAL;
+}
+
 static int polaris10_set_power_state_tasks(struct pp_hwmgr *hwmgr, const void *input)
 {
 	int tmp_result, result = 0;
@@ -4605,6 +4668,11 @@
 			"Failed to program memory timing parameters!",
 			result = tmp_result);
 
+	tmp_result = polaris10_notify_smc_display(hwmgr);
+	PP_ASSERT_WITH_CODE((0 == tmp_result),
+			"Failed to notify smc display settings!",
+			result = tmp_result);
+
 	tmp_result = polaris10_unfreeze_sclk_mclk_dpm(hwmgr);
 	PP_ASSERT_WITH_CODE((0 == tmp_result),
 			"Failed to unfreeze SCLK MCLK DPM!",
@@ -4639,6 +4707,7 @@
 			PPSMC_MSG_SetFanPwmMax, us_max_fan_pwm);
 }
 
+
 int polaris10_notify_smc_display_change(struct pp_hwmgr *hwmgr, bool has_display)
 {
 	PPSMC_Msg msg = has_display ? (PPSMC_Msg)PPSMC_HasDisplay : (PPSMC_Msg)PPSMC_NoDisplay;
@@ -4658,8 +4727,6 @@
 
 	if (num_active_displays > 1)  /* to do && (pHwMgr->pPECI->displayConfiguration.bMultiMonitorInSync != TRUE)) */
 		polaris10_notify_smc_display_change(hwmgr, false);
-	else
-		polaris10_notify_smc_display_change(hwmgr, true);
 
 	return 0;
 }
@@ -4700,6 +4767,8 @@
 	frame_time_in_us = 1000000 / refresh_rate;
 
 	pre_vbi_time_in_us = frame_time_in_us - 200 - mode_info.vblank_time_us;
+	data->frame_time_x2 = frame_time_in_us * 2 / 100;
+
 	display_gap2 = pre_vbi_time_in_us * (ref_clock / 100);
 
 	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL2, display_gap2);
@@ -4708,8 +4777,6 @@
 
 	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, data->soft_regs_start + offsetof(SMU74_SoftRegisters, VBlankTimeout), (frame_time_in_us - pre_vbi_time_in_us));
 
-	polaris10_notify_smc_display_change(hwmgr, num_active_displays != 0);
-
 	return 0;
 }
 
@@ -4821,7 +4888,7 @@
 		return 0;
 	}
 
-	data->need_long_memory_training = true;
+	data->need_long_memory_training = false;
 
 /*
  *	PPMCME_FirmwareDescriptorEntry *pfd = NULL;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h
index fd38b0d..dbc6d9b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h
@@ -311,6 +311,7 @@
 	bool apply_optimized_settings;
 	uint32_t                              avfs_vdroop_override_setting;
 	bool                                  apply_avfs_cks_off_voltage;
+	uint32_t                              frame_time_x2;
 };
 
 /* To convert to Q8.8 format for firmware */
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
index 5d70e2c..a64db7b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
@@ -1256,7 +1256,7 @@
 }
 
 int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
-				uint32_t sclk, uint16_t virtual_voltage_Id, uint16_t *voltage)
+				uint32_t sclk, uint16_t virtual_voltage_Id, uint32_t *voltage)
 {
 
 	int result;
@@ -1274,7 +1274,7 @@
 	if (0 != result)
 		return result;
 
-	*voltage = get_voltage_info_param_space.usVoltageLevel;
+	*voltage = ((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_3 *)(&get_voltage_info_param_space))->ulVoltageLevel;
 
 	return result;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
index 248c5db..1e35a96 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
@@ -305,7 +305,7 @@
 extern int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock,
 								uint8_t level);
 extern int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
-				uint32_t sclk, uint16_t virtual_voltage_Id, uint16_t *voltage);
+				uint32_t sclk, uint16_t virtual_voltage_Id, uint32_t *voltage);
 extern int atomctrl_get_smc_sclk_range_table(struct pp_hwmgr *hwmgr, struct pp_atom_ctrl_sclk_range_table *table);
 
 extern int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr, struct pp_atom_ctrl__avfs_parameters *param);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c
index 6c4553c..072b3b6 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c
@@ -1302,7 +1302,7 @@
 			table->Smio[count] |=
 				data->mvdd_voltage_table.entries[count].smio_low;
 		}
-		table->SmioMask2 = data->vddci_voltage_table.mask_low;
+		table->SmioMask2 = data->mvdd_voltage_table.mask_low;
 
 		CONVERT_FROM_HOST_TO_SMC_UL(table->MvddLevelCount);
 	}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c
index 94d6b47..4662d3d 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c
@@ -302,7 +302,7 @@
 			(((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usPPMTableOffset));
 
 		if (0 != powerplay_table->usPPMTableOffset) {
-			if (1 == get_platform_power_management_table(hwmgr, atom_ppm_table)) {
+			if (get_platform_power_management_table(hwmgr, atom_ppm_table) == 0) {
 				phm_cap_set(hwmgr->platform_descriptor.platformCaps,
 					PHM_PlatformCaps_EnablePlatformPowerManagement);
 			}
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 0bbc42a..bf0d2ac 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -415,6 +415,8 @@
 	uint8_t  ucVr_I2C_Line;
 	uint8_t  ucPlx_I2C_address;
 	uint8_t  ucPlx_I2C_Line;
+	uint32_t usBoostPowerLimit;
+	uint8_t  ucCKS_LDO_REFSEL;
 };
 
 struct phm_ppm_table {
diff --git a/drivers/gpu/drm/amd/powerplay/inc/polaris10_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/polaris10_ppsmc.h
index d41d37a..b8f4b73 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/polaris10_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/polaris10_ppsmc.h
@@ -392,6 +392,8 @@
 #define PPSMC_MSG_SetGpuPllDfsForSclk         ((uint16_t) 0x300)
 #define PPSMC_MSG_Didt_Block_Function		  ((uint16_t) 0x301)
 
+#define PPSMC_MSG_SetVBITimeout               ((uint16_t) 0x306)
+
 #define PPSMC_MSG_SecureSRBMWrite             ((uint16_t) 0x600)
 #define PPSMC_MSG_SecureSRBMRead              ((uint16_t) 0x601)
 #define PPSMC_MSG_SetAddress                  ((uint16_t) 0x800)
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu74_discrete.h b/drivers/gpu/drm/amd/powerplay/inc/smu74_discrete.h
index b85ff54..899d6d8 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu74_discrete.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu74_discrete.h
@@ -270,7 +270,8 @@
 	uint8_t                             BootPhases;
 
 	uint8_t                             VRHotLevel;
-	uint8_t                             Reserved1[3];
+	uint8_t                             LdoRefSel;
+	uint8_t                             Reserved1[2];
 	uint16_t                            FanStartTemperature;
 	uint16_t                            FanStopTemperature;
 	uint16_t                            MaxVoltage;
diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c
index 067632a..6f10b42 100644
--- a/drivers/gpu/drm/i915/i915_gem_shrinker.c
+++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c
@@ -40,7 +40,7 @@
 	if (!mutex_is_locked(mutex))
 		return false;
 
-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES)
+#if defined(CONFIG_DEBUG_MUTEXES) || defined(CONFIG_MUTEX_SPIN_ON_OWNER)
 	return mutex->owner == task;
 #else
 	/* Since UP may be pre-empted, we cannot assume that we own the lock */
diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c
index c3b33a1..3edb958 100644
--- a/drivers/gpu/drm/i915/intel_csr.c
+++ b/drivers/gpu/drm/i915/intel_csr.c
@@ -41,15 +41,15 @@
  * be moved to FW_FAILED.
  */
 
-#define I915_CSR_KBL "i915/kbl_dmc_ver1_01.bin"
+#define I915_CSR_KBL "i915/kbl_dmc_ver1.bin"
 MODULE_FIRMWARE(I915_CSR_KBL);
 #define KBL_CSR_VERSION_REQUIRED	CSR_VERSION(1, 1)
 
-#define I915_CSR_SKL "i915/skl_dmc_ver1_26.bin"
+#define I915_CSR_SKL "i915/skl_dmc_ver1.bin"
 MODULE_FIRMWARE(I915_CSR_SKL);
-#define SKL_CSR_VERSION_REQUIRED	CSR_VERSION(1, 26)
+#define SKL_CSR_VERSION_REQUIRED	CSR_VERSION(1, 23)
 
-#define I915_CSR_BXT "i915/bxt_dmc_ver1_07.bin"
+#define I915_CSR_BXT "i915/bxt_dmc_ver1.bin"
 MODULE_FIRMWARE(I915_CSR_BXT);
 #define BXT_CSR_VERSION_REQUIRED	CSR_VERSION(1, 7)
 
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index be3b2ca..70555c5 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -12016,6 +12016,12 @@
 		ret = intel_color_check(crtc, crtc_state);
 		if (ret)
 			return ret;
+
+		/*
+		 * Changing color management on Intel hardware is
+		 * handled as part of planes update.
+		 */
+		crtc_state->planes_changed = true;
 	}
 
 	ret = 0;
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 0c5ba34..21b04c3 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -4336,7 +4336,7 @@
 
 	intel_dp->detect_done = false;
 
-	if (intel_connector->detect_edid)
+	if (is_edp(intel_dp) || intel_connector->detect_edid)
 		return connector_status_connected;
 	else
 		return connector_status_disconnected;
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 70c6990..414ddda 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1097,6 +1097,7 @@
 						uint32_t *const batch,
 						uint32_t index)
 {
+	struct drm_i915_private *dev_priv = engine->i915;
 	uint32_t l3sqc4_flush = (0x40400000 | GEN8_LQSC_FLUSH_COHERENT_LINES);
 
 	/*
@@ -1105,8 +1106,8 @@
 	 * this batch updates GEN8_L3SQCREG4 with default value we need to
 	 * set this bit here to retain the WA during flush.
 	 */
-	if (IS_SKL_REVID(engine->i915, 0, SKL_REVID_E0) ||
-	    IS_KBL_REVID(engine->i915, 0, KBL_REVID_E0))
+	if (IS_SKL_REVID(dev_priv, 0, SKL_REVID_E0) ||
+	    IS_KBL_REVID(dev_priv, 0, KBL_REVID_E0))
 		l3sqc4_flush |= GEN8_LQSC_RO_PERF_DIS;
 
 	wa_ctx_emit(batch, index, (MI_STORE_REGISTER_MEM_GEN8 |
@@ -1267,11 +1268,12 @@
 				    uint32_t *offset)
 {
 	int ret;
+	struct drm_i915_private *dev_priv = engine->i915;
 	uint32_t index = wa_ctx_start(wa_ctx, *offset, CACHELINE_DWORDS);
 
 	/* WaDisableCtxRestoreArbitration:skl,bxt */
-	if (IS_SKL_REVID(engine->i915, 0, SKL_REVID_D0) ||
-	    IS_BXT_REVID(engine->i915, 0, BXT_REVID_A1))
+	if (IS_SKL_REVID(dev_priv, 0, SKL_REVID_D0) ||
+	    IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1))
 		wa_ctx_emit(batch, index, MI_ARB_ON_OFF | MI_ARB_DISABLE);
 
 	/* WaFlushCoherentL3CacheLinesAtContextSwitch:skl,bxt */
@@ -1282,7 +1284,7 @@
 
 	/* WaClearSlmSpaceAtContextSwitch:kbl */
 	/* Actual scratch location is at 128 bytes offset */
-	if (IS_KBL_REVID(engine->i915, 0, KBL_REVID_A0)) {
+	if (IS_KBL_REVID(dev_priv, 0, KBL_REVID_A0)) {
 		uint32_t scratch_addr
 			= engine->scratch.gtt_offset + 2*CACHELINE_BYTES;
 
diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c
index c27d5eb..adca262 100644
--- a/drivers/gpu/drm/i915/intel_opregion.c
+++ b/drivers/gpu/drm/i915/intel_opregion.c
@@ -1072,5 +1072,16 @@
 		return -ENODEV;
 	}
 
+	/*
+	 * FIXME On Dell XPS 13 9350 the OpRegion panel type (0) gives us
+	 * low vswing for eDP, whereas the VBT panel type (2) gives us normal
+	 * vswing instead. Low vswing results in some display flickers, so
+	 * let's simply ignore the OpRegion panel type on SKL for now.
+	 */
+	if (IS_SKYLAKE(dev_priv)) {
+		DRM_DEBUG_KMS("Ignoring OpRegion panel type (%d)\n", ret - 1);
+		return -ENODEV;
+	}
+
 	return ret - 1;
 }
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 5a8ee0c..f4f3fcc 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -57,7 +57,7 @@
 
 static void gen9_init_clock_gating(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	/* See Bspec note for PSR2_CTL bit 31, Wa#828:skl,bxt,kbl */
 	I915_WRITE(CHICKEN_PAR1_1,
@@ -7046,7 +7046,7 @@
 
 static void kabylake_init_clock_gating(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	gen9_init_clock_gating(dev);
 
@@ -7067,7 +7067,7 @@
 
 static void skylake_init_clock_gating(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	gen9_init_clock_gating(dev);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
index 22706c0..49bd5da 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
@@ -40,7 +40,8 @@
 gf119_sor_dp_pattern(struct nvkm_output_dp *outp, int pattern)
 {
 	struct nvkm_device *device = outp->base.disp->engine.subdev.device;
-	nvkm_mask(device, 0x61c110, 0x0f0f0f0f, 0x01010101 * pattern);
+	const u32 soff = gf119_sor_soff(outp);
+	nvkm_mask(device, 0x61c110 + soff, 0x0f0f0f0f, 0x01010101 * pattern);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/sun4i/sun4i_crtc.c b/drivers/gpu/drm/sun4i/sun4i_crtc.c
index f628b6d..4a19221 100644
--- a/drivers/gpu/drm/sun4i/sun4i_crtc.c
+++ b/drivers/gpu/drm/sun4i/sun4i_crtc.c
@@ -77,6 +77,14 @@
 	DRM_DEBUG_DRIVER("Disabling the CRTC\n");
 
 	sun4i_tcon_disable(drv->tcon);
+
+	if (crtc->state->event && !crtc->state->active) {
+		spin_lock_irq(&crtc->dev->event_lock);
+		drm_crtc_send_vblank_event(crtc, crtc->state->event);
+		spin_unlock_irq(&crtc->dev->event_lock);
+
+		crtc->state->event = NULL;
+	}
 }
 
 static void sun4i_crtc_enable(struct drm_crtc *crtc)
diff --git a/drivers/gpu/drm/sun4i/sun4i_drv.c b/drivers/gpu/drm/sun4i/sun4i_drv.c
index 5b89940..4dc543e 100644
--- a/drivers/gpu/drm/sun4i/sun4i_drv.c
+++ b/drivers/gpu/drm/sun4i/sun4i_drv.c
@@ -92,7 +92,7 @@
 	/* Frame Buffer Operations */
 
 	/* VBlank Operations */
-	.get_vblank_counter	= drm_vblank_count,
+	.get_vblank_counter	= drm_vblank_no_hw_counter,
 	.enable_vblank		= sun4i_drv_enable_vblank,
 	.disable_vblank		= sun4i_drv_disable_vblank,
 };
@@ -300,6 +300,7 @@
 
 		count += sun4i_drv_add_endpoints(&pdev->dev, &match,
 						pipeline);
+		of_node_put(pipeline);
 
 		DRM_DEBUG_DRIVER("Queued %d outputs on pipeline %d\n",
 				 count, i);
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index e340d0d6..3452748 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -1044,9 +1044,9 @@
 	return ret;
 }
 
-static bool ttm_bo_mem_compat(struct ttm_placement *placement,
-			      struct ttm_mem_reg *mem,
-			      uint32_t *new_flags)
+bool ttm_bo_mem_compat(struct ttm_placement *placement,
+		       struct ttm_mem_reg *mem,
+		       uint32_t *new_flags)
 {
 	int i;
 
@@ -1078,6 +1078,7 @@
 
 	return false;
 }
+EXPORT_SYMBOL(ttm_bo_mem_compat);
 
 int ttm_bo_validate(struct ttm_buffer_object *bo,
 			struct ttm_placement *placement,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
index 9b078a4..0cd8890 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
@@ -49,6 +49,7 @@
 {
 	struct ttm_buffer_object *bo = &buf->base;
 	int ret;
+	uint32_t new_flags;
 
 	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
 	if (unlikely(ret != 0))
@@ -60,7 +61,12 @@
 	if (unlikely(ret != 0))
 		goto err;
 
-	ret = ttm_bo_validate(bo, placement, interruptible, false);
+	if (buf->pin_count > 0)
+		ret = ttm_bo_mem_compat(placement, &bo->mem,
+					&new_flags) == true ? 0 : -EINVAL;
+	else
+		ret = ttm_bo_validate(bo, placement, interruptible, false);
+
 	if (!ret)
 		vmw_bo_pin_reserved(buf, true);
 
@@ -91,6 +97,7 @@
 {
 	struct ttm_buffer_object *bo = &buf->base;
 	int ret;
+	uint32_t new_flags;
 
 	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
 	if (unlikely(ret != 0))
@@ -102,6 +109,12 @@
 	if (unlikely(ret != 0))
 		goto err;
 
+	if (buf->pin_count > 0) {
+		ret = ttm_bo_mem_compat(&vmw_vram_gmr_placement, &bo->mem,
+					&new_flags) == true ? 0 : -EINVAL;
+		goto out_unreserve;
+	}
+
 	ret = ttm_bo_validate(bo, &vmw_vram_gmr_placement, interruptible,
 			      false);
 	if (likely(ret == 0) || ret == -ERESTARTSYS)
@@ -161,6 +174,7 @@
 	struct ttm_placement placement;
 	struct ttm_place place;
 	int ret = 0;
+	uint32_t new_flags;
 
 	place = vmw_vram_placement.placement[0];
 	place.lpfn = bo->num_pages;
@@ -185,10 +199,15 @@
 	 */
 	if (bo->mem.mem_type == TTM_PL_VRAM &&
 	    bo->mem.start < bo->num_pages &&
-	    bo->mem.start > 0)
+	    bo->mem.start > 0 &&
+	    buf->pin_count == 0)
 		(void) ttm_bo_validate(bo, &vmw_sys_placement, false, false);
 
-	ret = ttm_bo_validate(bo, &placement, interruptible, false);
+	if (buf->pin_count > 0)
+		ret = ttm_bo_mem_compat(&placement, &bo->mem,
+					&new_flags) == true ? 0 : -EINVAL;
+	else
+		ret = ttm_bo_validate(bo, &placement, interruptible, false);
 
 	/* For some reason we didn't end up at the start of vram */
 	WARN_ON(ret == 0 && bo->offset != 0);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
index 5d5c951..e8ae3dc 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
@@ -233,6 +233,7 @@
 static int vmw_restrict_iommu;
 static int vmw_force_coherent;
 static int vmw_restrict_dma_mask;
+static int vmw_assume_16bpp;
 
 static int vmw_probe(struct pci_dev *, const struct pci_device_id *);
 static void vmw_master_init(struct vmw_master *);
@@ -249,6 +250,8 @@
 module_param_named(force_coherent, vmw_force_coherent, int, 0600);
 MODULE_PARM_DESC(restrict_dma_mask, "Restrict DMA mask to 44 bits with IOMMU");
 module_param_named(restrict_dma_mask, vmw_restrict_dma_mask, int, 0600);
+MODULE_PARM_DESC(assume_16bpp, "Assume 16-bpp when filtering modes");
+module_param_named(assume_16bpp, vmw_assume_16bpp, int, 0600);
 
 
 static void vmw_print_capabilities(uint32_t capabilities)
@@ -660,6 +663,8 @@
 	dev_priv->vram_start = pci_resource_start(dev->pdev, 1);
 	dev_priv->mmio_start = pci_resource_start(dev->pdev, 2);
 
+	dev_priv->assume_16bpp = !!vmw_assume_16bpp;
+
 	dev_priv->enable_fb = enable_fbdev;
 
 	vmw_write(dev_priv, SVGA_REG_ID, SVGA_ID_2);
@@ -706,6 +711,13 @@
 			vmw_read(dev_priv,
 				 SVGA_REG_SUGGESTED_GBOBJECT_MEM_SIZE_KB);
 
+		/*
+		 * Workaround for low memory 2D VMs to compensate for the
+		 * allocation taken by fbdev
+		 */
+		if (!(dev_priv->capabilities & SVGA_CAP_3D))
+			mem_size *= 2;
+
 		dev_priv->max_mob_pages = mem_size * 1024 / PAGE_SIZE;
 		dev_priv->prim_bb_mem =
 			vmw_read(dev_priv,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 9a90f82..74304b0 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -387,6 +387,7 @@
 	spinlock_t hw_lock;
 	spinlock_t cap_lock;
 	bool has_dx;
+	bool assume_16bpp;
 
 	/*
 	 * VGA registers.
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
index 679a4cb..d2d9395 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
@@ -517,28 +517,6 @@
 
 	par->set_fb = &vfb->base;
 
-	if (!par->bo_ptr) {
-		/*
-		 * Pin before mapping. Since we don't know in what placement
-		 * to pin, call into KMS to do it for us.
-		 */
-		ret = vfb->pin(vfb);
-		if (ret) {
-			DRM_ERROR("Could not pin the fbdev framebuffer.\n");
-			return ret;
-		}
-
-		ret = ttm_bo_kmap(&par->vmw_bo->base, 0,
-				  par->vmw_bo->base.num_pages, &par->map);
-		if (ret) {
-			vfb->unpin(vfb);
-			DRM_ERROR("Could not map the fbdev framebuffer.\n");
-			return ret;
-		}
-
-		par->bo_ptr = ttm_kmap_obj_virtual(&par->map, &par->bo_iowrite);
-	}
-
 	return 0;
 }
 
@@ -601,6 +579,31 @@
 	if (ret)
 		goto out_unlock;
 
+	if (!par->bo_ptr) {
+		struct vmw_framebuffer *vfb = vmw_framebuffer_to_vfb(set.fb);
+
+		/*
+		 * Pin before mapping. Since we don't know in what placement
+		 * to pin, call into KMS to do it for us.
+		 */
+		ret = vfb->pin(vfb);
+		if (ret) {
+			DRM_ERROR("Could not pin the fbdev framebuffer.\n");
+			goto out_unlock;
+		}
+
+		ret = ttm_bo_kmap(&par->vmw_bo->base, 0,
+				  par->vmw_bo->base.num_pages, &par->map);
+		if (ret) {
+			vfb->unpin(vfb);
+			DRM_ERROR("Could not map the fbdev framebuffer.\n");
+			goto out_unlock;
+		}
+
+		par->bo_ptr = ttm_kmap_obj_virtual(&par->map, &par->bo_iowrite);
+	}
+
+
 	vmw_fb_dirty_mark(par, par->fb_x, par->fb_y,
 			  par->set_fb->width, par->set_fb->height);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
index 8a69d4d..bf28ccc 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
@@ -1555,14 +1555,10 @@
 		DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_PVSYNC)
 	};
 	int i;
-	u32 assumed_bpp = 2;
+	u32 assumed_bpp = 4;
 
-	/*
-	 * If using screen objects, then assume 32-bpp because that's what the
-	 * SVGA device is assuming
-	 */
-	if (dev_priv->active_display_unit == vmw_du_screen_object)
-		assumed_bpp = 4;
+	if (dev_priv->assume_16bpp)
+		assumed_bpp = 2;
 
 	if (dev_priv->active_display_unit == vmw_du_screen_target) {
 		max_width  = min(max_width,  dev_priv->stdu_max_width);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
index f0374f9..e57a0ba 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
@@ -300,6 +300,9 @@
 		break;
 	}
 
+	if (retries == RETRIES)
+		return -EINVAL;
+
 	*msg_len = reply_len;
 	*msg     = reply;
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c b/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
index 9ca818f..41932a7 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
@@ -399,8 +399,10 @@
 
 	WARN_ON_ONCE(!stdu->defined);
 
-	if (!vfb->dmabuf && new_fb->width == mode->hdisplay &&
-	    new_fb->height == mode->vdisplay)
+	new_vfbs = (vfb->dmabuf) ? NULL : vmw_framebuffer_to_vfbs(new_fb);
+
+	if (new_vfbs && new_vfbs->surface->base_size.width == mode->hdisplay &&
+	    new_vfbs->surface->base_size.height == mode->vdisplay)
 		new_content_type = SAME_AS_DISPLAY;
 	else if (vfb->dmabuf)
 		new_content_type = SEPARATE_DMA;
@@ -444,7 +446,6 @@
 			content_srf.mip_levels[0]     = 1;
 			content_srf.multisample_count = 0;
 		} else {
-			new_vfbs = vmw_framebuffer_to_vfbs(new_fb);
 			content_srf = *new_vfbs->surface;
 		}
 
@@ -464,7 +465,6 @@
 			return ret;
 		}
 	} else if (new_content_type == SAME_AS_DISPLAY) {
-		new_vfbs = vmw_framebuffer_to_vfbs(new_fb);
 		new_display_srf = vmw_surface_reference(new_vfbs->surface);
 	}