1 files changed, 32 insertions, 36 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 2dad7aa9a03b..ad490c1e2f57 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -1267,7 +1267,7 @@ static ssize_t amdgpu_ras_sysfs_features_read(struct device *dev,
 	struct amdgpu_ras *con =
 		container_of(attr, struct amdgpu_ras, features_attr);
 
-	return scnprintf(buf, PAGE_SIZE, "feature mask: 0x%x\n", con->features);
+	return sysfs_emit(buf, "feature mask: 0x%x\n", con->features);
 }
 
 static void amdgpu_ras_sysfs_remove_bad_page_node(struct amdgpu_device *adev)
@@ -1561,7 +1561,6 @@ static void amdgpu_ras_interrupt_poison_consumption_handler(struct ras_manager *
 {
 	bool poison_stat = false;
 	struct amdgpu_device *adev = obj->adev;
-	struct ras_err_data err_data = {0, 0, 0, NULL};
 	struct amdgpu_ras_block_object *block_obj =
 		amdgpu_ras_get_ras_block(adev, obj->head.block, 0);
 
@@ -1584,7 +1583,7 @@ static void amdgpu_ras_interrupt_poison_consumption_handler(struct ras_manager *
 	}
 
 	if (!adev->gmc.xgmi.connected_to_cpu)
-		amdgpu_umc_poison_handler(adev, &err_data, false);
+		amdgpu_umc_poison_handler(adev, false);
 
 	if (block_obj->hw_ops->handle_poison_consumption)
 		poison_stat = block_obj->hw_ops->handle_poison_consumption(adev);
@@ -1949,8 +1948,12 @@ static void amdgpu_ras_do_recovery(struct work_struct *work)
 
 		reset_context.method = AMD_RESET_METHOD_NONE;
 		reset_context.reset_req_dev = adev;
-		clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
-		clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
+
+		/* Perform full reset in fatal error mode */
+		if (!amdgpu_ras_is_poison_mode_supported(ras->adev))
+			set_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
+		else
+			clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
 
 		amdgpu_device_gpu_recover(ras->adev, NULL, &reset_context);
 	}
@@ -2268,6 +2271,25 @@ static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev)
 
 static bool amdgpu_ras_asic_supported(struct amdgpu_device *adev)
 {
+	if (amdgpu_sriov_vf(adev)) {
+		switch (adev->ip_versions[MP0_HWIP][0]) {
+		case IP_VERSION(13, 0, 2):
+			return true;
+		default:
+			return false;
+		}
+	}
+
+	if (adev->asic_type == CHIP_IP_DISCOVERY) {
+		switch (adev->ip_versions[MP0_HWIP][0]) {
+		case IP_VERSION(13, 0, 0):
+		case IP_VERSION(13, 0, 10):
+			return true;
+		default:
+			return false;
+		}
+	}
+
 	return adev->asic_type == CHIP_VEGA10 ||
 		adev->asic_type == CHIP_VEGA20 ||
 		adev->asic_type == CHIP_ARCTURUS ||
@@ -2311,11 +2333,6 @@ static void amdgpu_ras_check_supported(struct amdgpu_device *adev)
 	    !amdgpu_ras_asic_supported(adev))
 		return;
 
-	/* If driver run on sriov guest side, only enable ras for aldebaran */
-	if (amdgpu_sriov_vf(adev) &&
-		adev->ip_versions[MP1_HWIP][0] != IP_VERSION(13, 0, 2))
-		return;
-
 	if (!adev->gmc.xgmi.connected_to_cpu) {
 		if (amdgpu_atomfirmware_mem_ecc_supported(adev)) {
 			dev_info(adev->dev, "MEM ECC is active.\n");
@@ -2331,7 +2348,8 @@ static void amdgpu_ras_check_supported(struct amdgpu_device *adev)
 				adev->ras_hw_enabled |= ~(1 << AMDGPU_RAS_BLOCK__UMC |
 							    1 << AMDGPU_RAS_BLOCK__DF);
 
-				if (adev->ip_versions[VCN_HWIP][0] == IP_VERSION(2, 6, 0))
+				if (adev->ip_versions[VCN_HWIP][0] == IP_VERSION(2, 6, 0) ||
+				    adev->ip_versions[VCN_HWIP][0] == IP_VERSION(4, 0, 0))
 					adev->ras_hw_enabled |= (1 << AMDGPU_RAS_BLOCK__VCN |
 							1 << AMDGPU_RAS_BLOCK__JPEG);
 				else
@@ -2835,7 +2853,6 @@ static int amdgpu_bad_page_notifier(struct notifier_block *nb,
 	struct amdgpu_device *adev = NULL;
 	uint32_t gpu_id = 0;
 	uint32_t umc_inst = 0, ch_inst = 0;
-	struct ras_err_data err_data = {0, 0, 0, NULL};
 
 	/*
 	 * If the error was generated in UMC_V2, which belongs to GPU UMCs,
@@ -2874,31 +2891,10 @@ static int amdgpu_bad_page_notifier(struct notifier_block *nb,
 	dev_info(adev->dev, "Uncorrectable error detected in UMC inst: %d, chan_idx: %d",
 			     umc_inst, ch_inst);
 
-	err_data.err_addr =
-		kcalloc(adev->umc.max_ras_err_cnt_per_query,
-			sizeof(struct eeprom_table_record), GFP_KERNEL);
-	if (!err_data.err_addr) {
-		dev_warn(adev->dev,
-			"Failed to alloc memory for umc error record in mca notifier!\n");
+	if (!amdgpu_umc_page_retirement_mca(adev, m->addr, ch_inst, umc_inst))
+		return NOTIFY_OK;
+	else
 		return NOTIFY_DONE;
-	}
-
-	/*
-	 * Translate UMC channel address to Physical address
-	 */
-	if (adev->umc.ras &&
-	    adev->umc.ras->convert_ras_error_address)
-		adev->umc.ras->convert_ras_error_address(adev,
-			&err_data, m->addr, ch_inst, umc_inst);
-
-	if (amdgpu_bad_page_threshold != 0) {
-		amdgpu_ras_add_bad_pages(adev, err_data.err_addr,
-						err_data.err_addr_cnt);
-		amdgpu_ras_save_bad_pages(adev);
-	}
-
-	kfree(err_data.err_addr);
-	return NOTIFY_OK;
 }
 
 static struct notifier_block amdgpu_bad_page_nb = {