Merge tag 'amd-drm-fixes-6.3-2023-03-09' of...

Merge tag 'amd-drm-fixes-6.3-2023-03-09' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.3-2023-03-09:

amdgpu:
- Misc display fixes
- UMC 8.10 fixes
- Driver unload fixes
- NBIO 7.3.0 fix
- Error checking fixes for soc15, nv, soc21 read register interface
- Fix video cap query for VCN 4.0.4

amdkfd:
- Fix return check in doorbell handling
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230310031314.1296929-1-alexander.deucher@amd.com

drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c

@@ -543,6 +543,7 @@ static void amdgpu_discovery_read_from_harvest_table(struct amdgpu_device *adev,
 	struct harvest_table *harvest_info;
 	u16 offset;
 	int i;
+	uint32_t umc_harvest_config = 0;
 
 	bhdr = (struct binary_header *)adev->mman.discovery_bin;
 	offset = le16_to_cpu(bhdr->table_list[HARVEST_INFO].offset);
@@ -570,12 +571,17 @@ static void amdgpu_discovery_read_from_harvest_table(struct amdgpu_device *adev,
 			adev->harvest_ip_mask |= AMD_HARVEST_IP_DMU_MASK;
 			break;
 		case UMC_HWID:
+			umc_harvest_config |=
+				1 << (le16_to_cpu(harvest_info->list[i].number_instance));
 			(*umc_harvest_count)++;
 			break;
 		default:
 			break;
 		}
 	}
+
+	adev->umc.active_mask = ((1 << adev->umc.node_inst_num) - 1) &
+				~umc_harvest_config;
 }
 
 /* ================================================== */
@@ -1156,8 +1162,10 @@ static int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev)
 						AMDGPU_MAX_SDMA_INSTANCES);
 			}
 
-			if (le16_to_cpu(ip->hw_id) == UMC_HWID)
+			if (le16_to_cpu(ip->hw_id) == UMC_HWID) {
 				adev->gmc.num_umc++;
+				adev->umc.node_inst_num++;
+			}
 
 			for (k = 0; k < num_base_address; k++) {
 				/*

drivers/gpu/drm/amd/amdgpu/amdgpu_object.c

@@ -1315,7 +1315,7 @@ void amdgpu_bo_release_notify(struct ttm_buffer_object *bo)
 
 	if (!bo->resource || bo->resource->mem_type != TTM_PL_VRAM ||
 	    !(abo->flags & AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE) ||
-	    adev->in_suspend || adev->shutdown)
+	    adev->in_suspend || drm_dev_is_unplugged(adev_to_drm(adev)))
 		return;
 
 	if (WARN_ON_ONCE(!dma_resv_trylock(bo->base.resv)))

drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c

@@ -602,27 +602,14 @@ psp_cmd_submit_buf(struct psp_context *psp,
 		   struct psp_gfx_cmd_resp *cmd, uint64_t fence_mc_addr)
 {
 	int ret;
-	int index, idx;
+	int index;
 	int timeout = 20000;
 	bool ras_intr = false;
 	bool skip_unsupport = false;
-	bool dev_entered;
 
 	if (psp->adev->no_hw_access)
 		return 0;
 
-	dev_entered = drm_dev_enter(adev_to_drm(psp->adev), &idx);
-	/*
-	 * We allow sending PSP messages LOAD_ASD and UNLOAD_TA without acquiring
-	 * a lock in drm_dev_enter during driver unload because we must call
-	 * drm_dev_unplug as the beginning  of unload driver sequence . It is very
-	 * crucial that userspace can't access device instances anymore.
-	 */
-	if (!dev_entered)
-		WARN_ON(psp->cmd_buf_mem->cmd_id != GFX_CMD_ID_LOAD_ASD &&
-			psp->cmd_buf_mem->cmd_id != GFX_CMD_ID_UNLOAD_TA &&
-			psp->cmd_buf_mem->cmd_id != GFX_CMD_ID_INVOKE_CMD);
-
 	memset(psp->cmd_buf_mem, 0, PSP_CMD_BUFFER_SIZE);
 
 	memcpy(psp->cmd_buf_mem, cmd, sizeof(struct psp_gfx_cmd_resp));
@@ -686,8 +673,6 @@ psp_cmd_submit_buf(struct psp_context *psp,
 	}
 
 exit:
-	if (dev_entered)
-		drm_dev_exit(idx);
 	return ret;
 }
 

drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h

@@ -42,7 +42,7 @@
 #define LOOP_UMC_INST_AND_CH(umc_inst, ch_inst) LOOP_UMC_INST((umc_inst)) LOOP_UMC_CH_INST((ch_inst))
 
 #define LOOP_UMC_NODE_INST(node_inst) \
-		for ((node_inst) = 0; (node_inst) < adev->umc.node_inst_num; (node_inst)++)
+		for_each_set_bit((node_inst), &(adev->umc.active_mask), adev->umc.node_inst_num)
 
 #define LOOP_UMC_EACH_NODE_INST_AND_CH(node_inst, umc_inst, ch_inst) \
 		LOOP_UMC_NODE_INST((node_inst)) LOOP_UMC_INST_AND_CH((umc_inst), (ch_inst))
@@ -69,7 +69,7 @@ struct amdgpu_umc {
 	/* number of umc instance with memory map register access */
 	uint32_t umc_inst_num;
 
-	/*number of umc node instance with memory map register access*/
+	/* Total number of umc node instance including harvest one */
 	uint32_t node_inst_num;
 
 	/* UMC regiser per channel offset */
@@ -82,6 +82,9 @@ struct amdgpu_umc {
 
 	const struct amdgpu_umc_funcs *funcs;
 	struct amdgpu_umc_ras *ras;
+
+	/* active mask for umc node instance */
+	unsigned long active_mask;
 };
 
 int amdgpu_umc_ras_late_init(struct amdgpu_device *adev, struct ras_common_if *ras_block);

drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c

@@ -567,7 +567,6 @@ static void gmc_v11_0_set_umc_funcs(struct amdgpu_device *adev)
 	case IP_VERSION(8, 10, 0):
 		adev->umc.channel_inst_num = UMC_V8_10_CHANNEL_INSTANCE_NUM;
 		adev->umc.umc_inst_num = UMC_V8_10_UMC_INSTANCE_NUM;
-		adev->umc.node_inst_num = adev->gmc.num_umc;
 		adev->umc.max_ras_err_cnt_per_query = UMC_V8_10_TOTAL_CHANNEL_NUM(adev);
 		adev->umc.channel_offs = UMC_V8_10_PER_CHANNEL_OFFSET;
 		adev->umc.retire_unit = UMC_V8_10_NA_COL_2BITS_POWER_OF_2_NUM;

drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c

@@ -382,11 +382,6 @@ static void nbio_v7_2_init_registers(struct amdgpu_device *adev)
 		if (def != data)
 			WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regBIF1_PCIE_MST_CTRL_3), data);
 		break;
-	case IP_VERSION(7, 5, 1):
-		data = RREG32_SOC15(NBIO, 0, regRCC_DEV2_EPF0_STRAP2);
-		data &= ~RCC_DEV2_EPF0_STRAP2__STRAP_NO_SOFT_RESET_DEV2_F0_MASK;
-		WREG32_SOC15(NBIO, 0, regRCC_DEV2_EPF0_STRAP2, data);
-		fallthrough;
 	default:
 		def = data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CONFIG_CNTL));
 		data = REG_SET_FIELD(data, PCIE_CONFIG_CNTL,
@@ -399,6 +394,15 @@ static void nbio_v7_2_init_registers(struct amdgpu_device *adev)
 		break;
 	}
 
+	switch (adev->ip_versions[NBIO_HWIP][0]) {
+	case IP_VERSION(7, 3, 0):
+	case IP_VERSION(7, 5, 1):
+		data = RREG32_SOC15(NBIO, 0, regRCC_DEV2_EPF0_STRAP2);
+		data &= ~RCC_DEV2_EPF0_STRAP2__STRAP_NO_SOFT_RESET_DEV2_F0_MASK;
+		WREG32_SOC15(NBIO, 0, regRCC_DEV2_EPF0_STRAP2, data);
+		break;
+	}
+
 	if (amdgpu_sriov_vf(adev))
 		adev->rmmio_remap.reg_offset = SOC15_REG_OFFSET(NBIO, 0,
 			regBIF_BX_PF0_HDP_MEM_COHERENCY_FLUSH_CNTL) << 2;

drivers/gpu/drm/amd/amdgpu/nv.c

@@ -444,9 +444,10 @@ static int nv_read_register(struct amdgpu_device *adev, u32 se_num,
 	*value = 0;
 	for (i = 0; i < ARRAY_SIZE(nv_allowed_read_registers); i++) {
 		en = &nv_allowed_read_registers[i];
-		if (adev->reg_offset[en->hwip][en->inst] &&
-		    reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg]
-				   + en->reg_offset))
+		if (!adev->reg_offset[en->hwip][en->inst])
+			continue;
+		else if (reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg]
+					+ en->reg_offset))
 			continue;
 
 		*value = nv_get_register_value(adev,

drivers/gpu/drm/amd/amdgpu/soc15.c

@@ -439,8 +439,9 @@ static int soc15_read_register(struct amdgpu_device *adev, u32 se_num,
 	*value = 0;
 	for (i = 0; i < ARRAY_SIZE(soc15_allowed_read_registers); i++) {
 		en = &soc15_allowed_read_registers[i];
-		if (adev->reg_offset[en->hwip][en->inst] &&
-			reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg]
+		if (!adev->reg_offset[en->hwip][en->inst])
+			continue;
+		else if (reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg]
 					+ en->reg_offset))
 			continue;
 

drivers/gpu/drm/amd/amdgpu/soc21.c

@@ -111,6 +111,7 @@ static int soc21_query_video_codecs(struct amdgpu_device *adev, bool encode,
 	switch (adev->ip_versions[UVD_HWIP][0]) {
 	case IP_VERSION(4, 0, 0):
 	case IP_VERSION(4, 0, 2):
+	case IP_VERSION(4, 0, 4):
 		if (adev->vcn.harvest_config & AMDGPU_VCN_HARVEST_VCN0) {
 			if (encode)
 				*codecs = &vcn_4_0_0_video_codecs_encode_vcn1;
@@ -291,9 +292,10 @@ static int soc21_read_register(struct amdgpu_device *adev, u32 se_num,
 	*value = 0;
 	for (i = 0; i < ARRAY_SIZE(soc21_allowed_read_registers); i++) {
 		en = &soc21_allowed_read_registers[i];
-		if (adev->reg_offset[en->hwip][en->inst] &&
-		    reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg]
-				   + en->reg_offset))
+		if (!adev->reg_offset[en->hwip][en->inst])
+			continue;
+		else if (reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg]
+					+ en->reg_offset))
 			continue;
 
 		*value = soc21_get_register_value(adev,

drivers/gpu/drm/amd/amdgpu/umc_v8_10.h

@@ -31,9 +31,9 @@
 /* number of umc instance with memory map register access */
 #define UMC_V8_10_UMC_INSTANCE_NUM		2
 
-/* Total channel instances for all umc nodes */
+/* Total channel instances for all available umc nodes */
 #define UMC_V8_10_TOTAL_CHANNEL_NUM(adev) \
-	(UMC_V8_10_CHANNEL_INSTANCE_NUM * UMC_V8_10_UMC_INSTANCE_NUM * (adev)->umc.node_inst_num)
+	(UMC_V8_10_CHANNEL_INSTANCE_NUM * UMC_V8_10_UMC_INSTANCE_NUM * (adev)->gmc.num_umc)
 
 /* UMC regiser per channel offset */
 #define UMC_V8_10_PER_CHANNEL_OFFSET	0x400

drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c

@@ -280,7 +280,7 @@ phys_addr_t kfd_get_process_doorbells(struct kfd_process_device *pdd)
 	if (!pdd->doorbell_index) {
 		int r = kfd_alloc_process_doorbells(pdd->dev,
 						    &pdd->doorbell_index);
-		if (r)
+		if (r < 0)
 			return 0;
 	}
 

drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c

@@ -529,6 +529,19 @@ static struct clk_bw_params vg_bw_params = {
 
 };
 
+static uint32_t find_max_clk_value(const uint32_t clocks[], uint32_t num_clocks)
+{
+	uint32_t max = 0;
+	int i;
+
+	for (i = 0; i < num_clocks; ++i) {
+		if (clocks[i] > max)
+			max = clocks[i];
+	}
+
+	return max;
+}
+
 static unsigned int find_dcfclk_for_voltage(const struct vg_dpm_clocks *clock_table,
 		unsigned int voltage)
 {
@@ -572,12 +585,16 @@ static void vg_clk_mgr_helper_populate_bw_params(
 
 	bw_params->clk_table.num_entries = j + 1;
 
-	for (i = 0; i < bw_params->clk_table.num_entries; i++, j--) {
+	for (i = 0; i < bw_params->clk_table.num_entries - 1; i++, j--) {
 		bw_params->clk_table.entries[i].fclk_mhz = clock_table->DfPstateTable[j].fclk;
 		bw_params->clk_table.entries[i].memclk_mhz = clock_table->DfPstateTable[j].memclk;
 		bw_params->clk_table.entries[i].voltage = clock_table->DfPstateTable[j].voltage;
 		bw_params->clk_table.entries[i].dcfclk_mhz = find_dcfclk_for_voltage(clock_table, clock_table->DfPstateTable[j].voltage);
 	}
+	bw_params->clk_table.entries[i].fclk_mhz = clock_table->DfPstateTable[j].fclk;
+	bw_params->clk_table.entries[i].memclk_mhz = clock_table->DfPstateTable[j].memclk;
+	bw_params->clk_table.entries[i].voltage = clock_table->DfPstateTable[j].voltage;
+	bw_params->clk_table.entries[i].dcfclk_mhz = find_max_clk_value(clock_table->DcfClocks, VG_NUM_DCFCLK_DPM_LEVELS);
 
 	bw_params->vram_type = bios_info->memory_type;
 	bw_params->num_channels = bios_info->ma_channel_number;

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

@@ -46,6 +46,7 @@
 #include "asic_reg/mp/mp_13_0_0_sh_mask.h"
 #include "smu_cmn.h"
 #include "amdgpu_ras.h"
+#include "umc_v8_10.h"
 
 /*
  * DO NOT use these for err/warn/info/debug messages.
@@ -90,6 +91,12 @@
 
 #define DEBUGSMC_MSG_Mode1Reset	2
 
+/*
+ * SMU_v13_0_10 supports ECCTABLE since version 80.34.0,
+ * use this to check ECCTABLE feature whether support
+ */
+#define SUPPORT_ECCTABLE_SMU_13_0_10_VERSION 0x00502200
+
 static struct cmn2asic_msg_mapping smu_v13_0_0_message_map[SMU_MSG_MAX_COUNT] = {
 	MSG_MAP(TestMessage,			PPSMC_MSG_TestMessage,                 1),
 	MSG_MAP(GetSmuVersion,			PPSMC_MSG_GetSmuVersion,               1),
@@ -229,6 +236,7 @@ static struct cmn2asic_mapping smu_v13_0_0_table_map[SMU_TABLE_COUNT] = {
 	TAB_MAP(ACTIVITY_MONITOR_COEFF),
 	[SMU_TABLE_COMBO_PPTABLE] = {1, TABLE_COMBO_PPTABLE},
 	TAB_MAP(I2C_COMMANDS),
+	TAB_MAP(ECCINFO),
 };
 
 static struct cmn2asic_mapping smu_v13_0_0_pwr_src_map[SMU_POWER_SOURCE_COUNT] = {
@@ -462,6 +470,8 @@ static int smu_v13_0_0_tables_init(struct smu_context *smu)
 		       AMDGPU_GEM_DOMAIN_VRAM);
 	SMU_TABLE_INIT(tables, SMU_TABLE_COMBO_PPTABLE, MP0_MP1_DATA_REGION_SIZE_COMBOPPTABLE,
 			PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM);
+	SMU_TABLE_INIT(tables, SMU_TABLE_ECCINFO, sizeof(EccInfoTable_t),
+			PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM);
 
 	smu_table->metrics_table = kzalloc(sizeof(SmuMetricsExternal_t), GFP_KERNEL);
 	if (!smu_table->metrics_table)
@@ -477,8 +487,14 @@ static int smu_v13_0_0_tables_init(struct smu_context *smu)
 	if (!smu_table->watermarks_table)
 		goto err2_out;
 
+	smu_table->ecc_table = kzalloc(tables[SMU_TABLE_ECCINFO].size, GFP_KERNEL);
+	if (!smu_table->ecc_table)
+		goto err3_out;
+
 	return 0;
 
+err3_out:
+	kfree(smu_table->watermarks_table);
 err2_out:
 	kfree(smu_table->gpu_metrics_table);
 err1_out:
@@ -2036,6 +2052,64 @@ static int smu_v13_0_0_send_bad_mem_channel_flag(struct smu_context *smu,
 	return ret;
 }
 
+static int smu_v13_0_0_check_ecc_table_support(struct smu_context *smu)
+{
+	struct amdgpu_device *adev = smu->adev;
+	uint32_t if_version = 0xff, smu_version = 0xff;
+	int ret = 0;
+
+	ret = smu_cmn_get_smc_version(smu, &if_version, &smu_version);
+	if (ret)
+		return -EOPNOTSUPP;
+
+	if ((adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 10)) &&
+		(smu_version >= SUPPORT_ECCTABLE_SMU_13_0_10_VERSION))
+		return ret;
+	else
+		return -EOPNOTSUPP;
+}
+
+static ssize_t smu_v13_0_0_get_ecc_info(struct smu_context *smu,
+									void *table)
+{
+	struct smu_table_context *smu_table = &smu->smu_table;
+	struct amdgpu_device *adev = smu->adev;
+	EccInfoTable_t *ecc_table = NULL;
+	struct ecc_info_per_ch *ecc_info_per_channel = NULL;
+	int i, ret = 0;
+	struct umc_ecc_info *eccinfo = (struct umc_ecc_info *)table;
+
+	ret = smu_v13_0_0_check_ecc_table_support(smu);
+	if (ret)
+		return ret;
+
+	ret = smu_cmn_update_table(smu,
+					SMU_TABLE_ECCINFO,
+					0,
+					smu_table->ecc_table,
+					false);
+	if (ret) {
+		dev_info(adev->dev, "Failed to export SMU ecc table!\n");
+		return ret;
+	}
+
+	ecc_table = (EccInfoTable_t *)smu_table->ecc_table;
+
+	for (i = 0; i < UMC_V8_10_TOTAL_CHANNEL_NUM(adev); i++) {
+		ecc_info_per_channel = &(eccinfo->ecc[i]);
+		ecc_info_per_channel->ce_count_lo_chip =
+				ecc_table->EccInfo[i].ce_count_lo_chip;
+		ecc_info_per_channel->ce_count_hi_chip =
+				ecc_table->EccInfo[i].ce_count_hi_chip;
+		ecc_info_per_channel->mca_umc_status =
+				ecc_table->EccInfo[i].mca_umc_status;
+		ecc_info_per_channel->mca_umc_addr =
+				ecc_table->EccInfo[i].mca_umc_addr;
+	}
+
+	return ret;
+}
+
 static const struct pptable_funcs smu_v13_0_0_ppt_funcs = {
 	.get_allowed_feature_mask = smu_v13_0_0_get_allowed_feature_mask,
 	.set_default_dpm_table = smu_v13_0_0_set_default_dpm_table,
@@ -2111,6 +2185,7 @@ static const struct pptable_funcs smu_v13_0_0_ppt_funcs = {
 	.send_hbm_bad_pages_num = smu_v13_0_0_smu_send_bad_mem_page_num,
 	.send_hbm_bad_channel_flag = smu_v13_0_0_send_bad_mem_channel_flag,
 	.gpo_control = smu_v13_0_gpo_control,
+	.get_ecc_info = smu_v13_0_0_get_ecc_info,
 };
 
 void smu_v13_0_0_set_ppt_funcs(struct smu_context *smu)

drivers/gpu/drm/display/drm_hdmi_helper.c

@@ -44,10 +44,8 @@ int drm_hdmi_infoframe_set_hdr_metadata(struct hdmi_drm_infoframe *frame,
 
 	/* Sink EOTF is Bit map while infoframe is absolute values */
 	if (!is_eotf_supported(hdr_metadata->hdmi_metadata_type1.eotf,
-	    connector->hdr_sink_metadata.hdmi_type1.eotf)) {
-		DRM_DEBUG_KMS("EOTF Not Supported\n");
-		return -EINVAL;
-	}
+	    connector->hdr_sink_metadata.hdmi_type1.eotf))
+		DRM_DEBUG_KMS("Unknown EOTF %d\n", hdr_metadata->hdmi_metadata_type1.eotf);
 
 	err = hdmi_drm_infoframe_init(frame);
 	if (err < 0)

drivers/gpu/drm/drm_atomic.c

@@ -1070,6 +1070,7 @@ static void drm_atomic_connector_print_state(struct drm_printer *p,
 	drm_printf(p, "connector[%u]: %s\n", connector->base.id, connector->name);
 	drm_printf(p, "\tcrtc=%s\n", state->crtc ? state->crtc->name : "(null)");
 	drm_printf(p, "\tself_refresh_aware=%d\n", state->self_refresh_aware);
+	drm_printf(p, "\tmax_requested_bpc=%d\n", state->max_requested_bpc);
 
 	if (connector->connector_type == DRM_MODE_CONNECTOR_WRITEBACK)
 		if (state->writeback_job && state->writeback_job->fb)