torvalds/linux
torvalds/linux
2
0
Fork 0
mirror of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git synced 2025-09-04 20:19:47 +08:00
Code

drm/amdgpu: add aca deferred error type support

Browse Source 
add aca deferred error type support

Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Yang Wang 2024-02-21 15:07:30 +08:00 committed by Alex Deucher
parent 70dfaa3c02
commit 865d339763
2 changed files with 9 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c
View File

@ -434,6 +434,8 @@ static int aca_log_aca_error_data(struct aca_bank_error *bank_error, enum aca_er
amdgpu_ras_error_statistic_ce_count(err_data, &mcm_info, NULL, count); amdgpu_ras_error_statistic_ce_count(err_data, &mcm_info, NULL, count);
break; break;
case ACA_ERROR_TYPE_DEFERRED: case ACA_ERROR_TYPE_DEFERRED:
amdgpu_ras_error_statistic_de_count(err_data, &mcm_info, NULL, count);
break;
default: default:
break; break;
} }
@ -474,6 +476,7 @@ static int __aca_get_error_data(struct amdgpu_device *adev, struct aca_handle *h
smu_type = ACA_SMU_TYPE_UE; smu_type = ACA_SMU_TYPE_UE;
break; break;
case ACA_ERROR_TYPE_CE: case ACA_ERROR_TYPE_CE:
case ACA_ERROR_TYPE_DEFERRED:
smu_type = ACA_SMU_TYPE_CE; smu_type = ACA_SMU_TYPE_CE;
break; break;
default: default:

8
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
View File

@ -1291,8 +1291,8 @@ ssize_t amdgpu_ras_aca_sysfs_read(struct device *dev, struct device_attribute *a
if (amdgpu_ras_query_error_status(obj->adev, &info)) if (amdgpu_ras_query_error_status(obj->adev, &info))
return -EINVAL; return -EINVAL;
return sysfs_emit(buf, "%s: %lu\n%s: %lu\n", "ue", info.ue_count, return sysfs_emit(buf, "%s: %lu\n%s: %lu\n%s: %lu\n", "ue", info.ue_count,
"ce", info.ce_count); "ce", info.ce_count, "de", info.ue_count);
} }
static int amdgpu_ras_query_error_status_helper(struct amdgpu_device *adev, static int amdgpu_ras_query_error_status_helper(struct amdgpu_device *adev,
@ -1341,6 +1341,10 @@ static int amdgpu_ras_query_error_status_helper(struct amdgpu_device *adev,
ret = amdgpu_aca_log_ras_error_data(adev, blk, ACA_ERROR_TYPE_CE, err_data); ret = amdgpu_aca_log_ras_error_data(adev, blk, ACA_ERROR_TYPE_CE, err_data);
if (ret) if (ret)
return ret; return ret;
ret = amdgpu_aca_log_ras_error_data(adev, blk, ACA_ERROR_TYPE_DEFERRED, err_data);
if (ret)
return ret;
} else { } else {
/* FIXME: add code to check return value later */ /* FIXME: add code to check return value later */
amdgpu_mca_smu_log_ras_error(adev, blk, AMDGPU_MCA_ERROR_TYPE_UE, err_data, qctx); amdgpu_mca_smu_log_ras_error(adev, blk, AMDGPU_MCA_ERROR_TYPE_UE, err_data, qctx);