sync code with last fixes and improvements from OpenBSD

sys/dev/pci/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c

@@ -2737,6 +2737,9 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence **ef)
 			if (!attachment->is_mapped)
 				continue;
 
+			if (attachment->bo_va->base.bo->tbo.pin_count)
+				continue;
+
 			kfd_mem_dmaunmap_attachment(mem, attachment);
 			ret = update_gpuvm_pte(mem, attachment, &sync_obj);
 			if (ret) {

sys/dev/pci/drm/amd/amdgpu/amdgpu_vm.c

@@ -1730,18 +1730,30 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 
 	/* Insert partial mapping before the range */
 	if (!list_empty(&before->list)) {
+		struct amdgpu_bo *bo = before->bo_va->base.bo;
+
 		amdgpu_vm_it_insert(before, &vm->va);
 		if (before->flags & AMDGPU_PTE_PRT)
 			amdgpu_vm_prt_get(adev);
+
+		if (bo && bo->tbo.base.resv == vm->root.bo->tbo.base.resv &&
+		    !before->bo_va->base.moved)
+			amdgpu_vm_bo_moved(&before->bo_va->base);
 	} else {
 		kfree(before);
 	}
 
 	/* Insert partial mapping after the range */
 	if (!list_empty(&after->list)) {
+		struct amdgpu_bo *bo = after->bo_va->base.bo;
+
 		amdgpu_vm_it_insert(after, &vm->va);
 		if (after->flags & AMDGPU_PTE_PRT)
 			amdgpu_vm_prt_get(adev);
+
+		if (bo && bo->tbo.base.resv == vm->root.bo->tbo.base.resv &&
+		    !after->bo_va->base.moved)
+			amdgpu_vm_bo_moved(&after->bo_va->base);
 	} else {
 		kfree(after);
 	}

sys/dev/pci/drm/amd/amdgpu/nbio_v2_3.c

@@ -346,7 +346,7 @@ static void nbio_v2_3_init_registers(struct amdgpu_device *adev)
 
 #define NAVI10_PCIE__LC_L0S_INACTIVITY_DEFAULT		0x00000000 // off by default, no gains over L1
 #define NAVI10_PCIE__LC_L1_INACTIVITY_DEFAULT		0x00000009 // 1=1us, 9=1ms
-#define NAVI10_PCIE__LC_L1_INACTIVITY_TBT_DEFAULT	0x0000000E // 4ms
+#define NAVI10_PCIE__LC_L1_INACTIVITY_TBT_DEFAULT	0x0000000E // 400ms
 
 static void nbio_v2_3_enable_aspm(struct amdgpu_device *adev,
 				  bool enable)
@@ -479,9 +479,12 @@ static void nbio_v2_3_program_aspm(struct amdgpu_device *adev)
 		WREG32_SOC15(NBIO, 0, mmRCC_BIF_STRAP5, data);
 
 	def = data = RREG32_PCIE(smnPCIE_LC_CNTL);
-	data &= ~PCIE_LC_CNTL__LC_L0S_INACTIVITY_MASK;
-	data |= 0x9 << PCIE_LC_CNTL__LC_L1_INACTIVITY__SHIFT;
-	data |= 0x1 << PCIE_LC_CNTL__LC_PMI_TO_L1_DIS__SHIFT;
+	data |= NAVI10_PCIE__LC_L0S_INACTIVITY_DEFAULT << PCIE_LC_CNTL__LC_L0S_INACTIVITY__SHIFT;
+	if (pci_is_thunderbolt_attached(adev->pdev))
+		data |= NAVI10_PCIE__LC_L1_INACTIVITY_TBT_DEFAULT  << PCIE_LC_CNTL__LC_L1_INACTIVITY__SHIFT;
+	else
+		data |= NAVI10_PCIE__LC_L1_INACTIVITY_DEFAULT << PCIE_LC_CNTL__LC_L1_INACTIVITY__SHIFT;
+	data &= ~PCIE_LC_CNTL__LC_PMI_TO_L1_DIS_MASK;
 	if (def != data)
 		WREG32_PCIE(smnPCIE_LC_CNTL, data);
 

sys/dev/pci/drm/amd/amdgpu/sdma_v4_0.c

@@ -2330,7 +2330,7 @@ const struct amd_ip_funcs sdma_v4_0_ip_funcs = {
 
 static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs = {
 	.type = AMDGPU_RING_TYPE_SDMA,
-	.align_mask = 0xf,
+	.align_mask = 0xff,
 	.nop = SDMA_PKT_NOP_HEADER_OP(SDMA_OP_NOP),
 	.support_64bit_ptrs = true,
 	.secure_submission_supported = true,
@@ -2400,7 +2400,7 @@ static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs_2nd_mmhub = {
 
 static const struct amdgpu_ring_funcs sdma_v4_0_page_ring_funcs = {
 	.type = AMDGPU_RING_TYPE_SDMA,
-	.align_mask = 0xf,
+	.align_mask = 0xff,
 	.nop = SDMA_PKT_NOP_HEADER_OP(SDMA_OP_NOP),
 	.support_64bit_ptrs = true,
 	.secure_submission_supported = true,

sys/dev/pci/drm/amd/display/amdgpu_dm/amdgpu_dm.c

@@ -6974,7 +6974,13 @@ static int amdgpu_dm_connector_get_modes(struct drm_connector *connector)
 				drm_add_modes_noedid(connector, 640, 480);
 	} else {
 		amdgpu_dm_connector_ddc_get_modes(connector, edid);
-		amdgpu_dm_connector_add_common_modes(encoder, connector);
+		/* most eDP supports only timings from its edid,
+		 * usually only detailed timings are available
+		 * from eDP edid. timings which are not from edid
+		 * may damage eDP
+		 */
+		if (connector->connector_type != DRM_MODE_CONNECTOR_eDP)
+			amdgpu_dm_connector_add_common_modes(encoder, connector);
 		amdgpu_dm_connector_add_freesync_modes(connector, edid);
 	}
 	amdgpu_dm_fbc_init(connector);

sys/dev/pci/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c

@@ -42,6 +42,30 @@
 #include "dm_helpers.h"
 #include "ddc_service_types.h"
 
+static u32 edid_extract_panel_id(struct edid *edid)
+{
+	return (u32)edid->mfg_id[0] << 24   |
+	       (u32)edid->mfg_id[1] << 16   |
+	       (u32)EDID_PRODUCT_ID(edid);
+}
+
+static void apply_edid_quirks(struct edid *edid, struct dc_edid_caps *edid_caps)
+{
+	uint32_t panel_id = edid_extract_panel_id(edid);
+
+	switch (panel_id) {
+	/* Workaround for some monitors which does not work well with FAMS */
+	case drm_edid_encode_panel_id('S', 'A', 'M', 0x0E5E):
+	case drm_edid_encode_panel_id('S', 'A', 'M', 0x7053):
+	case drm_edid_encode_panel_id('S', 'A', 'M', 0x71AC):
+		DRM_DEBUG_DRIVER("Disabling FAMS on monitor with panel id %X\n", panel_id);
+		edid_caps->panel_patch.disable_fams = true;
+		break;
+	default:
+		return;
+	}
+}
+
 /* dm_helpers_parse_edid_caps
  *
  * Parse edid caps
@@ -113,6 +137,8 @@ enum dc_edid_status dm_helpers_parse_edid_caps(
 	else
 		edid_caps->speaker_flags = DEFAULT_SPEAKER_LOCATION;
 
+	apply_edid_quirks(edid_buf, edid_caps);
+
 	kfree(sads);
 	kfree(sadb);
 

sys/dev/pci/drm/amd/display/dc/core/amdgpu_dc.c

@@ -1539,6 +1539,9 @@ bool dc_validate_boot_timing(const struct dc *dc,
 		return false;
 	}
 
+	if (dc->debug.force_odm_combine)
+		return false;
+
 	/* Check for enabled DIG to identify enabled display */
 	if (!link->link_enc->funcs->is_dig_enabled(link->link_enc))
 		return false;

sys/dev/pci/drm/amd/display/dc/dce112/dce112_resource.c

@@ -970,10 +970,12 @@ enum dc_status resource_map_phy_clock_resources(
 		|| dc_is_virtual_signal(pipe_ctx->stream->signal))
 		pipe_ctx->clock_source =
 				dc->res_pool->dp_clock_source;
-	else
-		pipe_ctx->clock_source = find_matching_pll(
-			&context->res_ctx, dc->res_pool,
-			stream);
+	else {
+		if (stream && stream->link && stream->link->link_enc)
+			pipe_ctx->clock_source = find_matching_pll(
+				&context->res_ctx, dc->res_pool,
+				stream);
+	}
 
 	if (pipe_ctx->clock_source == NULL)
 		return DC_NO_CLOCK_SOURCE_RESOURCE;

sys/dev/pci/drm/amd/display/dc/dcn20/dcn20_hwseq.c

@@ -1678,6 +1678,17 @@ static void dcn20_program_pipe(
 
 		if (hws->funcs.setup_vupdate_interrupt)
 			hws->funcs.setup_vupdate_interrupt(dc, pipe_ctx);
+
+		if (hws->funcs.calculate_dccg_k1_k2_values && dc->res_pool->dccg->funcs->set_pixel_rate_div) {
+			unsigned int k1_div, k2_div;
+
+			hws->funcs.calculate_dccg_k1_k2_values(pipe_ctx, &k1_div, &k2_div);
+
+			dc->res_pool->dccg->funcs->set_pixel_rate_div(
+				dc->res_pool->dccg,
+				pipe_ctx->stream_res.tg->inst,
+				k1_div, k2_div);
+		}
 	}
 
 	if (pipe_ctx->update_flags.bits.odm)

sys/dev/pci/drm/amd/display/dc/dcn32/dcn32_hwseq.c

@@ -1165,10 +1165,6 @@ unsigned int dcn32_calculate_dccg_k1_k2_values(struct pipe_ctx *pipe_ctx, unsign
 	unsigned int odm_combine_factor = 0;
 	bool two_pix_per_container = false;
 
-	// For phantom pipes, use the same programming as the main pipes
-	if (pipe_ctx->stream->mall_stream_config.type == SUBVP_PHANTOM) {
-		stream = pipe_ctx->stream->mall_stream_config.paired_stream;
-	}
 	two_pix_per_container = optc2_is_two_pixels_per_containter(&stream->timing);
 	odm_combine_factor = get_odm_config(pipe_ctx, NULL);
 

sys/dev/pci/drm/amd/display/dc/dcn32/dcn32_optc.c

@@ -98,7 +98,7 @@ static void optc32_set_odm_combine(struct timing_generator *optc, int *opp_id, i
 	optc1->opp_count = opp_cnt;
 }
 
-static void optc32_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode)
+void optc32_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 

sys/dev/pci/drm/amd/display/dc/dcn32/dcn32_optc.h

@@ -250,5 +250,6 @@
 	SF(OTG0_OTG_DRR_CONTROL, OTG_V_TOTAL_LAST_USED_BY_DRR, mask_sh)
 
 void dcn32_timing_generator_init(struct optc *optc1);
+void optc32_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode);
 
 #endif /* __DC_OPTC_DCN32_H__ */

sys/dev/pci/drm/amd/display/dmub/dmub_srv.h

@@ -471,7 +471,7 @@ struct dmub_notification {
  * of a firmware to know if feature or functionality is supported or present.
  */
 #define DMUB_FW_VERSION(major, minor, revision) \
-	((((major) & 0xFF) << 24) | (((minor) & 0xFF) << 16) | ((revision) & 0xFFFF))
+	((((major) & 0xFF) << 24) | (((minor) & 0xFF) << 16) | (((revision) & 0xFF) << 8))
 
 /**
  * dmub_srv_create() - creates the DMUB service.

sys/dev/pci/drm/amd/pm/swsmu/inc/amdgpu_smu.h

@@ -168,6 +168,7 @@ struct smu_temperature_range {
 	int mem_crit_max;
 	int mem_emergency_max;
 	int software_shutdown_temp;
+	int software_shutdown_temp_offset;
 };
 
 struct smu_state_validation_block {

sys/dev/pci/drm/amd/pm/swsmu/inc/smu_v13_0.h

@@ -297,5 +297,9 @@ int smu_v13_0_get_pptable_from_firmware(struct smu_context *smu,
 					uint32_t *size,
 					uint32_t pptable_id);
 
+int smu_v13_0_update_pcie_parameters(struct smu_context *smu,
+				     uint32_t pcie_gen_cap,
+				     uint32_t pcie_width_cap);
+
 #endif
 #endif

sys/dev/pci/drm/amd/pm/swsmu/smu13/smu_v13_0.c

@@ -1381,6 +1381,7 @@ static int smu_v13_0_irq_process(struct amdgpu_device *adev,
 	 */
 	uint32_t ctxid = entry->src_data[0];
 	uint32_t data;
+	uint32_t high;
 
 	if (client_id == SOC15_IH_CLIENTID_THM) {
 		switch (src_id) {
@@ -1437,6 +1438,36 @@ static int smu_v13_0_irq_process(struct amdgpu_device *adev,
 					schedule_work(&smu->throttling_logging_work);
 
 				break;
+			case 0x8:
+				high = smu->thermal_range.software_shutdown_temp +
+					smu->thermal_range.software_shutdown_temp_offset;
+				high = min_t(typeof(high),
+					     SMU_THERMAL_MAXIMUM_ALERT_TEMP,
+					     high);
+				dev_emerg(adev->dev, "Reduce soft CTF limit to %d (by an offset %d)\n",
+							high,
+							smu->thermal_range.software_shutdown_temp_offset);
+
+				data = RREG32_SOC15(THM, 0, regTHM_THERMAL_INT_CTRL);
+				data = REG_SET_FIELD(data, THM_THERMAL_INT_CTRL,
+							DIG_THERM_INTH,
+							(high & 0xff));
+				data = data & (~THM_THERMAL_INT_CTRL__THERM_TRIGGER_MASK_MASK);
+				WREG32_SOC15(THM, 0, regTHM_THERMAL_INT_CTRL, data);
+				break;
+			case 0x9:
+				high = min_t(typeof(high),
+					     SMU_THERMAL_MAXIMUM_ALERT_TEMP,
+					     smu->thermal_range.software_shutdown_temp);
+				dev_emerg(adev->dev, "Recover soft CTF limit to %d\n", high);
+
+				data = RREG32_SOC15(THM, 0, regTHM_THERMAL_INT_CTRL);
+				data = REG_SET_FIELD(data, THM_THERMAL_INT_CTRL,
+							DIG_THERM_INTH,
+							(high & 0xff));
+				data = data & (~THM_THERMAL_INT_CTRL__THERM_TRIGGER_MASK_MASK);
+				WREG32_SOC15(THM, 0, regTHM_THERMAL_INT_CTRL, data);
+				break;
 			}
 		}
 	}
@@ -2458,3 +2489,74 @@ int smu_v13_0_mode1_reset(struct smu_context *smu)
 
 	return ret;
 }
+
+/*
+ * Intel hosts such as Raptor Lake and Sapphire Rapids don't support dynamic
+ * speed switching. Until we have confirmation from Intel that a specific host
+ * supports it, it's safer that we keep it disabled for all.
+ *
+ * https://edc.intel.com/content/www/us/en/design/products/platforms/details/raptor-lake-s/13th-generation-core-processors-datasheet-volume-1-of-2/005/pci-express-support/
+ * https://gitlab.freedesktop.org/drm/amd/-/issues/2663
+ */
+static bool smu_v13_0_is_pcie_dynamic_switching_supported(void)
+{
+#if IS_ENABLED(CONFIG_X86)
+#ifdef __linux__
+	struct cpuinfo_x86 *c = &cpu_data(0);
+
+	if (c->x86_vendor == X86_VENDOR_INTEL)
+#else
+	if (strcmp(cpu_vendor, "GenuineIntel") == 0)
+#endif
+		return false;
+#endif
+	return true;
+}
+
+int smu_v13_0_update_pcie_parameters(struct smu_context *smu,
+				     uint32_t pcie_gen_cap,
+				     uint32_t pcie_width_cap)
+{
+	struct smu_13_0_dpm_context *dpm_context = smu->smu_dpm.dpm_context;
+	struct smu_13_0_pcie_table *pcie_table =
+				&dpm_context->dpm_tables.pcie_table;
+	int num_of_levels = pcie_table->num_of_link_levels;
+	uint32_t smu_pcie_arg;
+	int ret, i;
+
+	if (!smu_v13_0_is_pcie_dynamic_switching_supported()) {
+		if (pcie_table->pcie_gen[num_of_levels - 1] < pcie_gen_cap)
+			pcie_gen_cap = pcie_table->pcie_gen[num_of_levels - 1];
+
+		if (pcie_table->pcie_lane[num_of_levels - 1] < pcie_width_cap)
+			pcie_width_cap = pcie_table->pcie_lane[num_of_levels - 1];
+
+		/* Force all levels to use the same settings */
+		for (i = 0; i < num_of_levels; i++) {
+			pcie_table->pcie_gen[i] = pcie_gen_cap;
+			pcie_table->pcie_lane[i] = pcie_width_cap;
+		}
+	} else {
+		for (i = 0; i < num_of_levels; i++) {
+			if (pcie_table->pcie_gen[i] > pcie_gen_cap)
+				pcie_table->pcie_gen[i] = pcie_gen_cap;
+			if (pcie_table->pcie_lane[i] > pcie_width_cap)
+				pcie_table->pcie_lane[i] = pcie_width_cap;
+		}
+	}
+
+	for (i = 0; i < num_of_levels; i++) {
+		smu_pcie_arg = i << 16;
+		smu_pcie_arg |= pcie_table->pcie_gen[i] << 8;
+		smu_pcie_arg |= pcie_table->pcie_lane[i];
+
+		ret = smu_cmn_send_smc_msg_with_param(smu,
+						      SMU_MSG_OverridePcieParameters,
+						      smu_pcie_arg,
+						      NULL);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}

sys/dev/pci/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c

@@ -1216,37 +1216,6 @@ static int smu_v13_0_0_force_clk_levels(struct smu_context *smu,
 	return ret;
 }
 
-static int smu_v13_0_0_update_pcie_parameters(struct smu_context *smu,
-					      uint32_t pcie_gen_cap,
-					      uint32_t pcie_width_cap)
-{
-	struct smu_13_0_dpm_context *dpm_context = smu->smu_dpm.dpm_context;
-	struct smu_13_0_pcie_table *pcie_table =
-				&dpm_context->dpm_tables.pcie_table;
-	uint32_t smu_pcie_arg;
-	int ret, i;
-
-	for (i = 0; i < pcie_table->num_of_link_levels; i++) {
-		if (pcie_table->pcie_gen[i] > pcie_gen_cap)
-			pcie_table->pcie_gen[i] = pcie_gen_cap;
-		if (pcie_table->pcie_lane[i] > pcie_width_cap)
-			pcie_table->pcie_lane[i] = pcie_width_cap;
-
-		smu_pcie_arg = i << 16;
-		smu_pcie_arg |= pcie_table->pcie_gen[i] << 8;
-		smu_pcie_arg |= pcie_table->pcie_lane[i];
-
-		ret = smu_cmn_send_smc_msg_with_param(smu,
-						      SMU_MSG_OverridePcieParameters,
-						      smu_pcie_arg,
-						      NULL);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
 static const struct smu_temperature_range smu13_thermal_policy[] = {
 	{-273150,  99000, 99000, -273150, 99000, 99000, -273150, 99000, 99000},
 	{ 120000, 120000, 120000, 120000, 120000, 120000, 120000, 120000, 120000},
@@ -1281,6 +1250,7 @@ static int smu_v13_0_0_get_thermal_temperature_range(struct smu_context *smu,
 	range->mem_emergency_max = (pptable->SkuTable.TemperatureLimit[TEMP_MEM] + CTF_OFFSET_MEM)*
 		SMU_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	range->software_shutdown_temp = powerplay_table->software_shutdown_temp;
+	range->software_shutdown_temp_offset = pptable->SkuTable.FanAbnormalTempLimitOffset;
 
 	return 0;
 }
@@ -2036,7 +2006,7 @@ static const struct pptable_funcs smu_v13_0_0_ppt_funcs = {
 	.feature_is_enabled = smu_cmn_feature_is_enabled,
 	.print_clk_levels = smu_v13_0_0_print_clk_levels,
 	.force_clk_levels = smu_v13_0_0_force_clk_levels,
-	.update_pcie_parameters = smu_v13_0_0_update_pcie_parameters,
+	.update_pcie_parameters = smu_v13_0_update_pcie_parameters,
 	.get_thermal_temperature_range = smu_v13_0_0_get_thermal_temperature_range,
 	.register_irq_handler = smu_v13_0_register_irq_handler,
 	.enable_thermal_alert = smu_v13_0_enable_thermal_alert,

sys/dev/pci/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c

@@ -1225,37 +1225,6 @@ static int smu_v13_0_7_force_clk_levels(struct smu_context *smu,
 	return ret;
 }
 
-static int smu_v13_0_7_update_pcie_parameters(struct smu_context *smu,
-					      uint32_t pcie_gen_cap,
-					      uint32_t pcie_width_cap)
-{
-	struct smu_13_0_dpm_context *dpm_context = smu->smu_dpm.dpm_context;
-	struct smu_13_0_pcie_table *pcie_table =
-				&dpm_context->dpm_tables.pcie_table;
-	uint32_t smu_pcie_arg;
-	int ret, i;
-
-	for (i = 0; i < pcie_table->num_of_link_levels; i++) {
-		if (pcie_table->pcie_gen[i] > pcie_gen_cap)
-			pcie_table->pcie_gen[i] = pcie_gen_cap;
-		if (pcie_table->pcie_lane[i] > pcie_width_cap)
-			pcie_table->pcie_lane[i] = pcie_width_cap;
-
-		smu_pcie_arg = i << 16;
-		smu_pcie_arg |= pcie_table->pcie_gen[i] << 8;
-		smu_pcie_arg |= pcie_table->pcie_lane[i];
-
-		ret = smu_cmn_send_smc_msg_with_param(smu,
-						      SMU_MSG_OverridePcieParameters,
-						      smu_pcie_arg,
-						      NULL);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
 static const struct smu_temperature_range smu13_thermal_policy[] =
 {
 	{-273150,  99000, 99000, -273150, 99000, 99000, -273150, 99000, 99000},
@@ -1288,6 +1257,7 @@ static int smu_v13_0_7_get_thermal_temperature_range(struct smu_context *smu,
 	range->mem_emergency_max = (pptable->SkuTable.TemperatureLimit[TEMP_MEM] + CTF_OFFSET_MEM)*
 		SMU_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	range->software_shutdown_temp = powerplay_table->software_shutdown_temp;
+	range->software_shutdown_temp_offset = pptable->SkuTable.FanAbnormalTempLimitOffset;
 
 	return 0;
 }
@@ -1751,7 +1721,7 @@ static const struct pptable_funcs smu_v13_0_7_ppt_funcs = {
 	.feature_is_enabled = smu_cmn_feature_is_enabled,
 	.print_clk_levels = smu_v13_0_7_print_clk_levels,
 	.force_clk_levels = smu_v13_0_7_force_clk_levels,
-	.update_pcie_parameters = smu_v13_0_7_update_pcie_parameters,
+	.update_pcie_parameters = smu_v13_0_update_pcie_parameters,
 	.get_thermal_temperature_range = smu_v13_0_7_get_thermal_temperature_range,
 	.register_irq_handler = smu_v13_0_register_irq_handler,
 	.enable_thermal_alert = smu_v13_0_enable_thermal_alert,

sys/dev/pci/drm/drm_atomic.c

@@ -140,6 +140,12 @@ drm_atomic_state_init(struct drm_device *dev, struct drm_atomic_state *state)
 	if (!state->planes)
 		goto fail;
 
+	/*
+	 * Because drm_atomic_state can be committed asynchronously we need our
+	 * own reference and cannot rely on the on implied by drm_file in the
+	 * ioctl call.
+	 */
+	drm_dev_get(dev);
 	state->dev = dev;
 
 	drm_dbg_atomic(dev, "Allocated atomic state %p\n", state);
@@ -299,7 +305,8 @@ EXPORT_SYMBOL(drm_atomic_state_clear);
 void __drm_atomic_state_free(struct kref *ref)
 {
 	struct drm_atomic_state *state = container_of(ref, typeof(*state), ref);
-	struct drm_mode_config *config = &state->dev->mode_config;
+	struct drm_device *dev = state->dev;
+	struct drm_mode_config *config = &dev->mode_config;
 
 	drm_atomic_state_clear(state);
 
@@ -311,6 +318,8 @@ void __drm_atomic_state_free(struct kref *ref)
 		drm_atomic_state_default_release(state);
 		kfree(state);
 	}
+
+	drm_dev_put(dev);
 }
 EXPORT_SYMBOL(__drm_atomic_state_free);
 

sys/dev/pci/drm/drm_atomic_helper.c

@@ -1231,7 +1231,16 @@ disable_outputs(struct drm_device *dev, struct drm_atomic_state *old_state)
 			continue;
 
 		ret = drm_crtc_vblank_get(crtc);
-		WARN_ONCE(ret != -EINVAL, "driver forgot to call drm_crtc_vblank_off()\n");
+		/*
+		 * Self-refresh is not a true "disable"; ensure vblank remains
+		 * enabled.
+		 */
+		if (new_crtc_state->self_refresh_active)
+			WARN_ONCE(ret != 0,
+				  "driver disabled vblank in self-refresh\n");
+		else
+			WARN_ONCE(ret != -EINVAL,
+				  "driver forgot to call drm_crtc_vblank_off()\n");
 		if (ret == 0)
 			drm_crtc_vblank_put(crtc);
 	}

sys/dev/pci/drm/drm_client.c

@@ -122,13 +122,34 @@ EXPORT_SYMBOL(drm_client_init);
  * drm_client_register() it is no longer permissible to call drm_client_release()
  * directly (outside the unregister callback), instead cleanup will happen
  * automatically on driver unload.
+ *
+ * Registering a client generates a hotplug event that allows the client
+ * to set up its display from pre-existing outputs. The client must have
+ * initialized its state to able to handle the hotplug event successfully.
  */
 void drm_client_register(struct drm_client_dev *client)
 {
 	struct drm_device *dev = client->dev;
+	int ret;
 
 	mutex_lock(&dev->clientlist_mutex);
 	list_add(&client->list, &dev->clientlist);
+
+	if (client->funcs && client->funcs->hotplug) {
+		/*
+		 * Perform an initial hotplug event to pick up the
+		 * display configuration for the client. This step
+		 * has to be performed *after* registering the client
+		 * in the list of clients, or a concurrent hotplug
+		 * event might be lost; leaving the display off.
+		 *
+		 * Hold the clientlist_mutex as for a regular hotplug
+		 * event.
+		 */
+		ret = client->funcs->hotplug(client);
+		if (ret)
+			drm_dbg_kms(dev, "client hotplug ret=%d\n", ret);
+	}
 	mutex_unlock(&dev->clientlist_mutex);
 }
 EXPORT_SYMBOL(drm_client_register);

sys/dev/pci/drm/drm_drv.c

@@ -1309,6 +1309,7 @@ drm_attach(struct device *parent, struct device *self, void *aux)
 
 	sc->sc_drm = dev;
 
+	kref_init(&dev->ref);
 	dev->dev = self;
 	dev->dev_private = parent;
 	dev->driver = da->driver;

sys/dev/pci/drm/drm_fb_helper.c

@@ -2717,10 +2717,6 @@ void drm_fbdev_generic_setup(struct drm_device *dev,
 		preferred_bpp = 32;
 	fb_helper->preferred_bpp = preferred_bpp;
 
-	ret = drm_fbdev_client_hotplug(&fb_helper->client);
-	if (ret)
-		drm_dbg_kms(dev, "client hotplug ret=%d\n", ret);
-
 	drm_client_register(&fb_helper->client);
 }
 EXPORT_SYMBOL(drm_fbdev_generic_setup);

sys/dev/pci/drm/i915/display/intel_display.c

@@ -5133,7 +5133,6 @@ copy_bigjoiner_crtc_state_modeset(struct intel_atomic_state *state,
 	saved_state->uapi = slave_crtc_state->uapi;
 	saved_state->scaler_state = slave_crtc_state->scaler_state;
 	saved_state->shared_dpll = slave_crtc_state->shared_dpll;
-	saved_state->dpll_hw_state = slave_crtc_state->dpll_hw_state;
 	saved_state->crc_enabled = slave_crtc_state->crc_enabled;
 
 	intel_crtc_free_hw_state(slave_crtc_state);

sys/dev/pci/drm/i915/gt/intel_gtt.c

@@ -611,7 +611,7 @@ __vm_create_scratch_for_read(struct i915_address_space *vm, unsigned long size)
 	if (IS_ERR(obj))
 		return ERR_CAST(obj);
 
-	i915_gem_object_set_cache_coherency(obj, I915_CACHING_CACHED);
+	i915_gem_object_set_cache_coherency(obj, I915_CACHE_LLC);
 
 	vma = i915_vma_instance(obj, vm, NULL);
 	if (IS_ERR(vma)) {

sys/dev/pci/drm/ttm/ttm_bo.c

@@ -1191,6 +1191,7 @@ int ttm_bo_swapout(struct ttm_buffer_object *bo, struct ttm_operation_ctx *ctx,
 		ret = ttm_bo_handle_move_mem(bo, evict_mem, true, &ctx, &hop);
 		if (unlikely(ret != 0)) {
 			WARN(ret == -EMULTIHOP, "Unexpected multihop in swaput - likely driver bug.\n");
+			ttm_resource_free(bo, &evict_mem);
 			goto out;
 		}
 	}