iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

drm/amdgpu: add RAS is_rma flag

Browse Source 
Set the flag to true if bad page number reaches threshold.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Tao Zhou 2024-05-23 11:23:20 +08:00 committed by Alex Deucher
parent 15c2990e0f
commit b95fa494d6
4 changed files with 12 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
View File

@ -2926,7 +2926,6 @@ int amdgpu_ras_recovery_init(struct amdgpu_device *adev)
struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
struct ras_err_handler_data **data;
u32 max_eeprom_records_count = 0;
bool exc_err_limit = false;
int ret;
if (!con || amdgpu_sriov_vf(adev))
@ -2963,12 +2962,12 @@ int amdgpu_ras_recovery_init(struct amdgpu_device *adev)
*/
if (adev->gmc.xgmi.pending_reset)
return 0;
ret = amdgpu_ras_eeprom_init(&con->eeprom_control, &exc_err_limit);
ret = amdgpu_ras_eeprom_init(&con->eeprom_control);
/*
* This calling fails when exc_err_limit is true or
* This calling fails when is_rma is true or
* ret != 0.
*/
if (exc_err_limit || ret)
if (con->is_rma || ret)
goto free;
if (con->eeprom_control.ras_num_recs) {
@ -3016,7 +3015,7 @@ out:
* Except error threshold exceeding case, other failure cases in this
* function would not fail amdgpu driver init.
*/
if (!exc_err_limit)
if (!con->is_rma)
ret = 0;
else
ret = -EINVAL;

1
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
View File

@ -522,6 +522,7 @@ struct amdgpu_ras {
bool update_channel_flag;
/* Record status of smu mca debug mode */
bool is_aca_debug_mode;
bool is_rma;
/* Record special requirements of gpu reset caller */
uint32_t gpu_reset_flags;

10
drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
View File

@ -750,6 +750,9 @@ amdgpu_ras_eeprom_update_header(struct amdgpu_ras_eeprom_control *control)
control->tbl_rai.health_percent = 0;
}
if (amdgpu_bad_page_threshold != -1)
ras->is_rma = true;
/* ignore the -ENOTSUPP return value */
amdgpu_dpm_send_rma_reason(adev);
}
@ -1321,8 +1324,7 @@ Out:
return res == RAS_TABLE_V2_1_INFO_SIZE ? 0 : res;
}
int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control,
bool *exceed_err_limit)
int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
{
struct amdgpu_device *adev = to_amdgpu_device(control);
unsigned char buf[RAS_TABLE_HEADER_SIZE] = { 0 };
@ -1330,7 +1332,7 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control,
struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
int res;
*exceed_err_limit = false;
ras->is_rma = false;
if (!__is_ras_eeprom_supported(adev))
return 0;
@ -1422,7 +1424,7 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control,
dev_warn(adev->dev, "GPU will be initialized due to bad_page_threshold = -1.");
res = 0;
} else {
*exceed_err_limit = true;
ras->is_rma = true;
dev_err(adev->dev,
"RAS records:%d exceed threshold:%d, "
"GPU will not be initialized. Replace this GPU or increase the threshold",

3
drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h
View File

@ -129,8 +129,7 @@ struct eeprom_table_record {
unsigned char mcumc_id;
} __packed;
int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control,
bool *exceed_err_limit);
int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control);
int amdgpu_ras_eeprom_reset_table(struct amdgpu_ras_eeprom_control *control);