drm/amdgpu: add interface to check mca umc status

Add interface to check mca umc status.
Signed-off-by: YiPeng Chai <YiPeng.Chai@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

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

@@ -27,6 +27,16 @@
 #include "umc/umc_6_7_0_offset.h"
 #include "umc/umc_6_7_0_sh_mask.h"
 
+static bool amdgpu_mca_is_deferred_error(struct amdgpu_device *adev,
+					uint64_t mc_status)
+{
+	if (adev->umc.ras->check_ecc_err_status)
+		return adev->umc.ras->check_ecc_err_status(adev,
+				AMDGPU_MCA_ERROR_TYPE_DE, &mc_status);
+
+	return false;
+}
+
 void amdgpu_mca_query_correctable_error_count(struct amdgpu_device *adev,
 					      uint64_t mc_status_addr,
 					      unsigned long *error_count)
@@ -257,7 +267,7 @@ int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_blo
 			amdgpu_ras_error_statistic_ue_count(err_data,
 				&mcm_info, &err_addr, (uint64_t)count);
 		else {
-			if (!!(MCA_REG__STATUS__DEFERRED(entry->regs[MCA_REG_IDX_STATUS])))
+			if (amdgpu_mca_is_deferred_error(adev, entry->regs[MCA_REG_IDX_STATUS]))
 				amdgpu_ras_error_statistic_de_count(err_data,
 					&mcm_info, &err_addr, (uint64_t)count);
 			else

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

@@ -65,6 +65,7 @@ enum amdgpu_mca_ip {
 enum amdgpu_mca_error_type {
 	AMDGPU_MCA_ERROR_TYPE_UE = 0,
 	AMDGPU_MCA_ERROR_TYPE_CE,
+	AMDGPU_MCA_ERROR_TYPE_DE,
 };
 
 struct amdgpu_mca_ras_block {

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

@@ -21,7 +21,7 @@
 #ifndef __AMDGPU_UMC_H__
 #define __AMDGPU_UMC_H__
 #include "amdgpu_ras.h"
-
+#include "amdgpu_mca.h"
 /*
  * (addr / 256) * 4096, the higher 26 bits in ErrorAddr
  * is the index of 4KB block
@@ -64,6 +64,8 @@ struct amdgpu_umc_ras {
 				      void *ras_error_status);
 	void (*ecc_info_query_ras_error_address)(struct amdgpu_device *adev,
 					void *ras_error_status);
+	bool (*check_ecc_err_status)(struct amdgpu_device *adev,
+			enum amdgpu_mca_error_type type, void *ras_error_status);
 	/* support different eeprom table version for different asic */
 	void (*set_eeprom_table_version)(struct amdgpu_ras_eeprom_table_header *hdr);
 };

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

@@ -422,6 +422,25 @@ static void umc_v12_0_ecc_info_query_ras_error_address(struct amdgpu_device *ade
 	}
 }
 
+static bool umc_v12_0_check_ecc_err_status(struct amdgpu_device *adev,
+			enum amdgpu_mca_error_type type, void *ras_error_status)
+{
+	uint64_t mc_umc_status = *(uint64_t *)ras_error_status;
+
+	switch (type) {
+	case AMDGPU_MCA_ERROR_TYPE_UE:
+		return umc_v12_0_is_uncorrectable_error(adev, mc_umc_status);
+	case AMDGPU_MCA_ERROR_TYPE_CE:
+		return umc_v12_0_is_correctable_error(adev, mc_umc_status);
+	case AMDGPU_MCA_ERROR_TYPE_DE:
+		return umc_v12_0_is_deferred_error(adev, mc_umc_status);
+	default:
+		return false;
+	}
+
+	return false;
+}
+
 static void umc_v12_0_err_cnt_init(struct amdgpu_device *adev)
 {
 	amdgpu_umc_loop_channels(adev,
@@ -507,5 +526,6 @@ struct amdgpu_umc_ras umc_v12_0_ras = {
 	.query_ras_poison_mode = umc_v12_0_query_ras_poison_mode,
 	.ecc_info_query_ras_error_count = umc_v12_0_ecc_info_query_ras_error_count,
 	.ecc_info_query_ras_error_address = umc_v12_0_ecc_info_query_ras_error_address,
+	.check_ecc_err_status = umc_v12_0_check_ecc_err_status,
 };
 

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c

@@ -2557,9 +2557,9 @@ static int mca_umc_mca_get_err_count(const struct mca_ras_info *mca_ras, struct
 		return 0;
 	}
 
-	if ((type == AMDGPU_MCA_ERROR_TYPE_UE && umc_v12_0_is_uncorrectable_error(adev, status0)) ||
-	    (type == AMDGPU_MCA_ERROR_TYPE_CE && (umc_v12_0_is_correctable_error(adev, status0) ||
-	     umc_v12_0_is_deferred_error(adev, status0))))
+	if (umc_v12_0_is_deferred_error(adev, status0) ||
+	    umc_v12_0_is_uncorrectable_error(adev, status0) ||
+	    umc_v12_0_is_correctable_error(adev, status0))
 		*count = 1;
 
 	return 0;