diff --git a/include/amd_smi/amdsmi.h b/include/amd_smi/amdsmi.h index 3ff5e536e9..18a0039545 100644 --- a/include/amd_smi/amdsmi.h +++ b/include/amd_smi/amdsmi.h @@ -2730,7 +2730,7 @@ amdsmi_status_t amdsmi_dev_get_gpu_ecc_enabled(amdsmi_processor_handle processo * * @return ::amdsmi_status_t | ::AMDSMI_STATUS_SUCCESS on success, non-zero on fail */ -amdsmi_status_t amdsmi_dev_get_ecc_status(amdsmi_processor_handle processor_handle, amdsmi_gpu_block_t block, +amdsmi_status_t amdsmi_dev_get_gpu_ecc_status(amdsmi_processor_handle processor_handle, amdsmi_gpu_block_t block, amdsmi_ras_err_state_t *state); /** diff --git a/py-interface/README.md b/py-interface/README.md index 9ca68d834e..f04e7ec1ad 100644 --- a/py-interface/README.md +++ b/py-interface/README.md @@ -2691,7 +2691,7 @@ except AmdSmiException as e: print(e) ``` -## amdsmi_dev_get_ecc_status +## amdsmi_dev_get_gpu_ecc_status Description: Retrieve the ECC status for a GPU block Input parameters: @@ -2700,7 +2700,7 @@ Input parameters: Output: ECC status for a requested GPU block -Exceptions that can be thrown by ` amdsmi_dev_get_ecc_status` function: +Exceptions that can be thrown by ` amdsmi_dev_get_gpu_ecc_status` function: * `AmdSmiLibraryException` * `AmdSmiRetryException` * `AmdSmiParameterException` @@ -2713,7 +2713,7 @@ try: print("No GPUs on machine") else: for device in devices: - status = amdsmi_dev_get_ecc_status(device, AmdSmiGpuBlock.UMC) + status = amdsmi_dev_get_gpu_ecc_status(device, AmdSmiGpuBlock.UMC) print(status) except AmdSmiException as e: print(e) diff --git a/py-interface/__init__.py b/py-interface/__init__.py index b111102b29..2521bb3218 100644 --- a/py-interface/__init__.py +++ b/py-interface/__init__.py @@ -126,7 +126,7 @@ from .amdsmi_interface import amdsmi_counter_get_available_counters # # Error Query from .amdsmi_interface import amdsmi_dev_get_gpu_ecc_count from .amdsmi_interface import amdsmi_dev_get_gpu_ecc_enabled -from .amdsmi_interface import amdsmi_dev_get_ecc_status +from .amdsmi_interface import amdsmi_dev_get_gpu_ecc_status from .amdsmi_interface import amdsmi_status_string # # System Information Query diff --git a/py-interface/amdsmi_interface.py b/py-interface/amdsmi_interface.py index 10721fbe5e..d016069b68 100644 --- a/py-interface/amdsmi_interface.py +++ b/py-interface/amdsmi_interface.py @@ -2502,7 +2502,7 @@ def amdsmi_dev_get_gpu_ecc_enabled( return blocks.value -def amdsmi_dev_get_ecc_status( +def amdsmi_dev_get_gpu_ecc_status( processor_handle: amdsmi_wrapper.amdsmi_processor_handle, block: AmdSmiGpuBlock ) -> AmdSmiRasErrState: if not isinstance(processor_handle, amdsmi_wrapper.amdsmi_processor_handle): @@ -2515,7 +2515,7 @@ def amdsmi_dev_get_ecc_status( state = amdsmi_wrapper.amdsmi_ras_err_state_t() _check_res( - amdsmi_wrapper. amdsmi_dev_get_ecc_status( + amdsmi_wrapper. amdsmi_dev_get_gpu_ecc_status( processor_handle, block, ctypes.byref(state) ) ) diff --git a/py-interface/amdsmi_wrapper.py b/py-interface/amdsmi_wrapper.py index 7560192c51..9e88b5d512 100644 --- a/py-interface/amdsmi_wrapper.py +++ b/py-interface/amdsmi_wrapper.py @@ -1583,9 +1583,9 @@ amdsmi_dev_get_gpu_ecc_count.argtypes = [amdsmi_processor_handle, amdsmi_gpu_blo amdsmi_dev_get_gpu_ecc_enabled = _libraries['libamd_smi.so'].amdsmi_dev_get_gpu_ecc_enabled amdsmi_dev_get_gpu_ecc_enabled.restype = amdsmi_status_t amdsmi_dev_get_gpu_ecc_enabled.argtypes = [amdsmi_processor_handle, ctypes.POINTER(ctypes.c_uint64)] -amdsmi_dev_get_ecc_status = _libraries['libamd_smi.so'].amdsmi_dev_get_ecc_status -amdsmi_dev_get_ecc_status.restype = amdsmi_status_t -amdsmi_dev_get_ecc_status.argtypes = [amdsmi_processor_handle, amdsmi_gpu_block_t, ctypes.POINTER(c__EA_amdsmi_ras_err_state_t)] +amdsmi_dev_get_gpu_ecc_status = _libraries['libamd_smi.so'].amdsmi_dev_get_gpu_ecc_status +amdsmi_dev_get_gpu_ecc_status.restype = amdsmi_status_t +amdsmi_dev_get_gpu_ecc_status.argtypes = [amdsmi_processor_handle, amdsmi_gpu_block_t, ctypes.POINTER(c__EA_amdsmi_ras_err_state_t)] amdsmi_status_string = _libraries['libamd_smi.so'].amdsmi_status_string amdsmi_status_string.restype = amdsmi_status_t amdsmi_status_string.argtypes = [amdsmi_status_t, ctypes.POINTER(ctypes.POINTER(ctypes.c_char))] @@ -1853,7 +1853,7 @@ __all__ = \ 'amdsmi_dev_counter_group_supported', 'amdsmi_dev_create_counter', 'amdsmi_dev_destroy_counter', 'amdsmi_get_busy_percent', 'amdsmi_get_gpu_drm_render_minor', 'amdsmi_dev_get_gpu_ecc_count', - 'amdsmi_dev_get_gpu_ecc_enabled', 'amdsmi_dev_get_ecc_status', + 'amdsmi_dev_get_gpu_ecc_enabled', 'amdsmi_dev_get_gpu_ecc_status', 'amdsmi_get_energy_count', 'amdsmi_get_gpu_fan_rpms', 'amdsmi_get_gpu_fan_speed', 'amdsmi_get_gpu_fan_speed_max', 'amdsmi_get_clk_freq', 'amdsmi_get_gpu_metrics_info', diff --git a/py-interface/rocm_smi_tool.py b/py-interface/rocm_smi_tool.py index 1fea97ba02..e645f4fb00 100644 --- a/py-interface/rocm_smi_tool.py +++ b/py-interface/rocm_smi_tool.py @@ -323,7 +323,7 @@ class Formatter: | """ + self.style.text("34 Get version string. Api: amdsmi_get_version_str ") + """ | | """ + self.style.text("35 Get device ecc counter. Api: amdsmi_dev_get_gpu_ecc_count ") + """ | | """ + self.style.text("36 Get device ecc enable. Api: amdsmi_dev_get_gpu_ecc_enabled ") + """ | - | """ + self.style.text("37 Get device ecc status. Api: amdsmi_dev_get_ecc_status ") + """ | + | """ + self.style.text("37 Get device ecc status. Api: amdsmi_dev_get_gpu_ecc_status ") + """ | | """ + self.style.text("38 Get status string. Api: amdsmi_status_string ") + """ | | """ + self.style.text("39 Get compute process info. Api: amdsmi_get_compute_process_info ") + """ | | """ + self.style.text("40 Get compute process info by pid. Api: amdsmi_get_compute_process_info_by_pid ") + """ | @@ -520,7 +520,7 @@ def amdsmi_tool_dev_ecc_status_get(dev): result = {} for gpu_block in smi_api.AmdSmiGpuBlock: try: - value = smi_api. amdsmi_dev_get_ecc_status(dev, gpu_block) + value = smi_api. amdsmi_dev_get_gpu_ecc_status(dev, gpu_block) result.update({gpu_block.name: value}) except smi_api.AmdSmiException as e: print("{}:\t{}".format(gpu_block.name, e)) diff --git a/src/amd_smi/amd_smi.cc b/src/amd_smi/amd_smi.cc index d6f28ac345..a558318876 100644 --- a/src/amd_smi/amd_smi.cc +++ b/src/amd_smi/amd_smi.cc @@ -922,7 +922,7 @@ amdsmi_get_func_iter_value(amdsmi_func_id_iter_handle_t handle, {"rsmi_dev_od_volt_info_set", " amdsmi_dev_set_od_volt_info"}, {"rsmi_dev_od_volt_curve_regions_get", " amdsmi_dev_get_od_volt_curve_regions"}, {"rsmi_dev_ecc_enabled_get", " amdsmi_dev_get_gpu_ecc_enabled"}, - {"rsmi_dev_ecc_status_get", " amdsmi_dev_get_ecc_status"}, + {"rsmi_dev_ecc_status_get", " amdsmi_dev_get_gpu_ecc_status"}, {"rsmi_dev_counter_group_supported", "amdsmi_dev_counter_group_supported"}, {"rsmi_dev_counter_create", "amdsmi_dev_create_counter"}, {"rsmi_dev_xgmi_error_status", "amdsmi_dev_xgmi_error_status"}, @@ -1027,7 +1027,7 @@ amdsmi_status_t amdsmi_dev_get_gpu_ecc_enabled(amdsmi_processor_handle processo return rsmi_wrapper(rsmi_dev_ecc_enabled_get, processor_handle, enabled_blocks); } -amdsmi_status_t amdsmi_dev_get_ecc_status(amdsmi_processor_handle processor_handle, +amdsmi_status_t amdsmi_dev_get_gpu_ecc_status(amdsmi_processor_handle processor_handle, amdsmi_gpu_block_t block, amdsmi_ras_err_state_t *state) { AMDSMI_CHECK_INIT(); diff --git a/tests/amd_smi_test/functional/err_cnt_read.cc b/tests/amd_smi_test/functional/err_cnt_read.cc index 82ee0d6a48..c43771ef1c 100755 --- a/tests/amd_smi_test/functional/err_cnt_read.cc +++ b/tests/amd_smi_test/functional/err_cnt_read.cc @@ -126,7 +126,7 @@ void TestErrCntRead::Run(void) { } for (uint32_t b = AMDSMI_GPU_BLOCK_FIRST; b <= AMDSMI_GPU_BLOCK_LAST; b = b*2) { - err = amdsmi_dev_get_ecc_status(processor_handles_[i], static_cast(b), + err = amdsmi_dev_get_gpu_ecc_status(processor_handles_[i], static_cast(b), &err_state); CHK_ERR_ASRT(err) IF_VERB(STANDARD) { @@ -135,7 +135,7 @@ void TestErrCntRead::Run(void) { " block: " << GetErrStateNameStr(err_state) << std::endl; } // Verify api support checking functionality is working - err = amdsmi_dev_get_ecc_status(processor_handles_[i], static_cast(b), + err = amdsmi_dev_get_gpu_ecc_status(processor_handles_[i], static_cast(b), nullptr); ASSERT_EQ(err, AMDSMI_STATUS_INVAL); diff --git a/tests/amd_smi_test/functional/mutual_exclusion.cc b/tests/amd_smi_test/functional/mutual_exclusion.cc index b64114b09e..b3fdaca149 100755 --- a/tests/amd_smi_test/functional/mutual_exclusion.cc +++ b/tests/amd_smi_test/functional/mutual_exclusion.cc @@ -253,7 +253,7 @@ void TestMutualExclusion::Run(void) { CHECK_RET(ret, AMDSMI_STATUS_BUSY); ret = amdsmi_dev_get_gpu_ecc_enabled(processor_handles_[0], &dmy_ui64); CHECK_RET(ret, AMDSMI_STATUS_BUSY); - ret = amdsmi_dev_get_ecc_status(processor_handles_[0], AMDSMI_GPU_BLOCK_UMC, &dmy_ras_err_st); + ret = amdsmi_dev_get_gpu_ecc_status(processor_handles_[0], AMDSMI_GPU_BLOCK_UMC, &dmy_ras_err_st); CHECK_RET(ret, AMDSMI_STATUS_BUSY); /* Other functions holding device mutexes. Listed for reference.