[PATCH 12/23] drm/amdgpu: store only one RAS bad page record for all pages in one row
Tao Zhou
tao.zhou1 at amd.com
Fri Nov 8 11:14:12 UTC 2024
So eeprom space can be saved, compatible with legacy way.
Signed-off-by: Tao Zhou <tao.zhou1 at amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 35 +++++++++++++++++++------
1 file changed, 27 insertions(+), 8 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index f31085c3a12d..b45344fe8cc8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -2789,7 +2789,7 @@ int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev,
struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
struct ras_err_handler_data *data;
struct amdgpu_ras_eeprom_control *control;
- int save_count;
+ int save_count, unit_num, bad_page_num, i;
if (!con || !con->eh_data) {
if (new_cnt)
@@ -2801,19 +2801,38 @@ int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev,
mutex_lock(&con->recovery_lock);
control = &con->eeprom_control;
data = con->eh_data;
- save_count = data->count - control->ras_num_recs;
+ bad_page_num = control->ras_num_recs;
+ /* one record on eeprom stands for all pages in one memory row
+ * in this mode
+ */
+ if (control->rec_type == AMDGPU_RAS_EEPROM_REC_MCA)
+ bad_page_num = control->ras_num_recs * adev->umc.retire_unit;
+
+ save_count = data->count - bad_page_num;
mutex_unlock(&con->recovery_lock);
+ unit_num = save_count / adev->umc.retire_unit;
if (new_cnt)
- *new_cnt = save_count / adev->umc.retire_unit;
+ *new_cnt = unit_num;
/* only new entries are saved */
if (save_count > 0) {
- if (amdgpu_ras_eeprom_append(control,
- &data->bps[control->ras_num_recs],
- save_count)) {
- dev_err(adev->dev, "Failed to save EEPROM table data!");
- return -EIO;
+ if (control->rec_type == AMDGPU_RAS_EEPROM_REC_PA) {
+ if (amdgpu_ras_eeprom_append(control,
+ &data->bps[control->ras_num_recs],
+ save_count)) {
+ dev_err(adev->dev, "Failed to save EEPROM table data!");
+ return -EIO;
+ }
+ } else {
+ for (i = 0; i < unit_num; i++) {
+ if (amdgpu_ras_eeprom_append(control,
+ &data->bps[bad_page_num + i * adev->umc.retire_unit],
+ 1)) {
+ dev_err(adev->dev, "Failed to save EEPROM table data!");
+ return -EIO;
+ }
+ }
}
dev_info(adev->dev, "Saved %d pages to EEPROM table.\n", save_count);
--
2.34.1
More information about the amd-gfx
mailing list