From 3bdf8dd84e55e9ea0806eb62bd4437c8c7746521 Mon Sep 17 00:00:00 2001 From: Lijo Lazar Date: Wed, 4 Jun 2025 09:36:09 +0530 Subject: drm/amdgpu: Clear reset flags from ras context Once RAS errors are cleared with appropriate recovery mechanism, clear reset flags also from RAS context. Otherwise, stale flag values could affect the subsequent RAS reset handling on the device. Signed-off-by: Lijo Lazar Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index de0944947eaf..e91d4d37b5b3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -4414,8 +4414,10 @@ void amdgpu_ras_clear_err_state(struct amdgpu_device *adev) struct amdgpu_ras *ras; ras = amdgpu_ras_get_context(adev); - if (ras) + if (ras) { ras->ras_err_state = 0; + ras->gpu_reset_flags = 0; + } } void amdgpu_ras_set_err_poison(struct amdgpu_device *adev, -- cgit From f43411978d1367af6e983ae09ab4c2f7a811574a Mon Sep 17 00:00:00 2001 From: Xiang Liu Date: Fri, 6 Jun 2025 11:10:40 +0800 Subject: drm/amdgpu: Add debug mask to disable CE logs Add debug mask to disable kernel logs of RAS correctable errors, including both ACA and CE error counter kernel messages. Signed-off-by: Xiang Liu Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index e91d4d37b5b3..69b8733a2b2d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -1107,6 +1107,9 @@ static void amdgpu_ras_error_print_error_data(struct amdgpu_device *adev, err_info->de_count, blk_name); } } else { + if (adev->debug_disable_ce_logs) + return; + for_each_ras_error(err_node, err_data) { err_info = &err_node->err_info; mcm_info = &err_info->mcm_info; -- cgit From cfce8f4fa7686657e119392e03964f401efea765 Mon Sep 17 00:00:00 2001 From: ganglxie Date: Thu, 26 Jun 2025 17:00:45 +0800 Subject: drm/amdgpu: refine ras error injection when eeprom initialization failed when eeprom initialization failed, we still support ras error injection, and reserve bad pages, but do not save bad pages to eeprom Signed-off-by: ganglxie Reviewed-by: Tao Zhou Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index 69b8733a2b2d..662cddc35fd3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -3006,6 +3006,15 @@ int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev, return 0; } + if (!con->eeprom_control.is_eeprom_valid) { + dev_warn(adev->dev, + "Failed to save EEPROM table data because of EEPROM data corruption!"); + if (new_cnt) + *new_cnt = 0; + + return 0; + } + mutex_lock(&con->recovery_lock); control = &con->eeprom_control; data = con->eh_data; @@ -3491,8 +3500,7 @@ int amdgpu_ras_init_badpage_info(struct amdgpu_device *adev) control = &con->eeprom_control; ret = amdgpu_ras_eeprom_init(control); - if (ret) - return ret; + control->is_eeprom_valid = !ret; if (!adev->umc.ras || !adev->umc.ras->convert_ras_err_addr) control->ras_num_pa_recs = control->ras_num_recs; @@ -3501,10 +3509,12 @@ int amdgpu_ras_init_badpage_info(struct amdgpu_device *adev) adev->umc.ras->get_retire_flip_bits) adev->umc.ras->get_retire_flip_bits(adev); - if (control->ras_num_recs) { + if (control->ras_num_recs && control->is_eeprom_valid) { ret = amdgpu_ras_load_bad_pages(adev); - if (ret) - return ret; + if (ret) { + control->is_eeprom_valid = false; + return 0; + } amdgpu_dpm_send_hbm_bad_pages_num( adev, control->ras_num_bad_pages); @@ -3523,7 +3533,7 @@ int amdgpu_ras_init_badpage_info(struct amdgpu_device *adev) dev_warn(adev->dev, "Failed to format RAS EEPROM data in V3 version!\n"); } - return ret; + return 0; } int amdgpu_ras_recovery_init(struct amdgpu_device *adev, bool init_bp_info) -- cgit From a6d6a86e94bfdf298662ac4bd9d6c92b77ac9abb Mon Sep 17 00:00:00 2001 From: YiPeng Chai Date: Fri, 27 Jun 2025 16:49:24 +0800 Subject: drm/amdgpu: Remove useless timeout error message The timeout is only used to interrupt polling and not need to print a error message. Signed-off-by: YiPeng Chai Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index 662cddc35fd3..c508697a3412 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -3306,7 +3306,6 @@ static int amdgpu_ras_poison_creation_handler(struct amdgpu_device *adev, uint64_t de_queried_count; uint32_t new_detect_count, total_detect_count; uint32_t need_query_count = poison_creation_count; - bool query_data_timeout = false; enum ras_event_type type = RAS_EVENT_TYPE_POISON_CREATION; memset(&info, 0, sizeof(info)); @@ -3335,21 +3334,13 @@ static int amdgpu_ras_poison_creation_handler(struct amdgpu_device *adev, timeout = MAX_UMC_POISON_POLLING_TIME_ASYNC; if (timeout) { - if (!--timeout) { - query_data_timeout = true; + if (!--timeout) break; - } msleep(1); } } } while (total_detect_count < need_query_count); - if (query_data_timeout) { - dev_warn(adev->dev, "Can't find deferred error! count: %u\n", - (need_query_count - total_detect_count)); - return -ENOENT; - } - if (total_detect_count) schedule_delayed_work(&ras->page_retirement_dwork, 0); -- cgit