drm/amdgpu: add ras event id support

add amdgpu ras event id support to better distinguish different
error information sources in dmesg logs.

the following log will be identify by event id:
{event_id} interrupt to inform RAS event
{event_id} ACA logs
{event_id} errors statistic since from current injection/error query
{event_id} errors statistic since from gpu load
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

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

@@ -210,22 +210,26 @@ int amdgpu_mca_smu_set_debug_mode(struct amdgpu_device *adev, bool enable)
 	return -EOPNOTSUPP;
 }
 
-static void amdgpu_mca_smu_mca_bank_dump(struct amdgpu_device *adev, int idx, struct mca_bank_entry *entry)
+static void amdgpu_mca_smu_mca_bank_dump(struct amdgpu_device *adev, int idx, struct mca_bank_entry *entry,
+					 struct ras_query_context *qctx)
 {
-	dev_info(adev->dev, HW_ERR "Accelerator Check Architecture events logged\n");
-	dev_info(adev->dev, HW_ERR "aca entry[%02d].STATUS=0x%016llx\n",
-		 idx, entry->regs[MCA_REG_IDX_STATUS]);
-	dev_info(adev->dev, HW_ERR "aca entry[%02d].ADDR=0x%016llx\n",
-		 idx, entry->regs[MCA_REG_IDX_ADDR]);
-	dev_info(adev->dev, HW_ERR "aca entry[%02d].MISC0=0x%016llx\n",
-		 idx, entry->regs[MCA_REG_IDX_MISC0]);
-	dev_info(adev->dev, HW_ERR "aca entry[%02d].IPID=0x%016llx\n",
-		 idx, entry->regs[MCA_REG_IDX_IPID]);
-	dev_info(adev->dev, HW_ERR "aca entry[%02d].SYND=0x%016llx\n",
-		 idx, entry->regs[MCA_REG_IDX_SYND]);
+	u64 event_id = qctx->event_id;
+
+	RAS_EVENT_LOG(adev, event_id, HW_ERR "Accelerator Check Architecture events logged\n");
+	RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].STATUS=0x%016llx\n",
+		      idx, entry->regs[MCA_REG_IDX_STATUS]);
+	RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].ADDR=0x%016llx\n",
+		      idx, entry->regs[MCA_REG_IDX_ADDR]);
+	RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].MISC0=0x%016llx\n",
+		      idx, entry->regs[MCA_REG_IDX_MISC0]);
+	RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].IPID=0x%016llx\n",
+		      idx, entry->regs[MCA_REG_IDX_IPID]);
+	RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].SYND=0x%016llx\n",
+		      idx, entry->regs[MCA_REG_IDX_SYND]);
 }
 
-int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type, struct ras_err_data *err_data)
+int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type,
+				 struct ras_err_data *err_data, struct ras_query_context *qctx)
 {
 	struct amdgpu_smuio_mcm_config_info mcm_info;
 	struct ras_err_addr err_addr = {0};
@@ -244,7 +248,7 @@ int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_blo
 	list_for_each_entry(node, &mca_set.list, node) {
 		entry = &node->entry;
 
-		amdgpu_mca_smu_mca_bank_dump(adev, i++, entry);
+		amdgpu_mca_smu_mca_bank_dump(adev, i++, entry, qctx);
 
 		count = 0;
 		ret = amdgpu_mca_smu_parse_mca_error_count(adev, blk, type, entry, &count);

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

@@ -169,6 +169,7 @@ void amdgpu_mca_smu_debugfs_init(struct amdgpu_device *adev, struct dentry *root
 void amdgpu_mca_bank_set_init(struct mca_bank_set *mca_set);
 int amdgpu_mca_bank_set_add_entry(struct mca_bank_set *mca_set, struct mca_bank_entry *entry);
 void amdgpu_mca_bank_set_release(struct mca_bank_set *mca_set);
-int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type, struct ras_err_data *err_data);
+int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type,
+				 struct ras_err_data *err_data, struct ras_query_context *qctx);
 
 #endif

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

@@ -64,6 +64,14 @@ struct amdgpu_iv_entry;
 /* The high three bits indicates socketid */
 #define AMDGPU_RAS_GET_FEATURES(val)  ((val) & ~AMDGPU_RAS_FEATURES_SOCKETID_MASK)
 
+#define RAS_EVENT_LOG(_adev, _id, _fmt, ...)				\
+do {									\
+	if (amdgpu_ras_event_id_is_valid((_adev), (_id)))			\
+	    dev_info((_adev)->dev, "{%llu}" _fmt, (_id), ##__VA_ARGS__);	\
+	else								\
+	    dev_info((_adev)->dev, _fmt, ##__VA_ARGS__);			\
+} while (0)
+
 enum amdgpu_ras_block {
 	AMDGPU_RAS_BLOCK__UMC = 0,
 	AMDGPU_RAS_BLOCK__SDMA,
@@ -419,6 +427,21 @@ struct umc_ecc_info {
 	int record_ce_addr_supported;
 };
 
+enum ras_event_type {
+	RAS_EVENT_TYPE_INVALID = -1,
+	RAS_EVENT_TYPE_ISR = 0,
+	RAS_EVENT_TYPE_COUNT,
+};
+
+struct ras_event_manager {
+	atomic64_t seqnos[RAS_EVENT_TYPE_COUNT];
+};
+
+struct ras_query_context {
+	enum ras_event_type type;
+	u64 event_id;
+};
+
 struct amdgpu_ras {
 	/* ras infrastructure */
 	/* for ras itself. */
@@ -479,6 +502,11 @@ struct amdgpu_ras {
 	atomic_t page_retirement_req_cnt;
 	/* Fatal error detected flag */
 	atomic_t fed;
+
+	/* RAS event manager */
+	struct ras_event_manager __event_mgr;
+	struct ras_event_manager *event_mgr;
+
 };
 
 struct ras_fs_data {
@@ -879,4 +907,6 @@ void amdgpu_ras_del_mca_err_addr(struct ras_err_info *err_info,
 void amdgpu_ras_set_fed(struct amdgpu_device *adev, bool status);
 bool amdgpu_ras_get_fed_status(struct amdgpu_device *adev);
 
+bool amdgpu_ras_event_id_is_valid(struct amdgpu_device *adev, u64 id);
+u64 amdgpu_ras_acquire_event_id(struct amdgpu_device *adev, enum ras_event_type type);
 #endif

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

@@ -44,6 +44,7 @@ struct amdgpu_hive_info {
 
 	struct amdgpu_reset_domain *reset_domain;
 	atomic_t ras_recovery;
+	struct ras_event_manager event_mgr;
 };
 
 struct amdgpu_pcs_ras_field {

drivers/gpu/drm/amd/amdgpu/umc_v12_0.c

@@ -404,10 +404,16 @@ static int umc_v12_0_err_cnt_init_per_channel(struct amdgpu_device *adev,
 static void umc_v12_0_ecc_info_query_ras_error_count(struct amdgpu_device *adev,
 					void *ras_error_status)
 {
+	struct ras_query_context qctx;
+
+	memset(&qctx, 0, sizeof(qctx));
+	qctx.event_id = amdgpu_ras_acquire_event_id(adev, amdgpu_ras_intr_triggered() ?
+						    RAS_EVENT_TYPE_ISR : RAS_EVENT_TYPE_INVALID);
+
 	amdgpu_mca_smu_log_ras_error(adev,
-		AMDGPU_RAS_BLOCK__UMC, AMDGPU_MCA_ERROR_TYPE_CE, ras_error_status);
+		AMDGPU_RAS_BLOCK__UMC, AMDGPU_MCA_ERROR_TYPE_CE, ras_error_status, &qctx);
 	amdgpu_mca_smu_log_ras_error(adev,
-		AMDGPU_RAS_BLOCK__UMC, AMDGPU_MCA_ERROR_TYPE_UE, ras_error_status);
+		AMDGPU_RAS_BLOCK__UMC, AMDGPU_MCA_ERROR_TYPE_UE, ras_error_status, &qctx);
 }
 
 static void umc_v12_0_ecc_info_query_ras_error_address(struct amdgpu_device *adev,