Max Weiss

[Issue]: rocmi-smi and rsmi_compute_process_info_by_pid_get() don't show values for processes that don't have access to all installed GPUs (Linux)

6

### Problem Description The example host has four GPUs: ``` $ rocm-smi ========================= ROCm System Management Interface ========================= =================================== Concise Info =================================== GPU Temp (DieEdge) AvgPwr SCLK MCLK Fan Perf...

[Issue]: string returned by amdsmi_status_code_to_string() contains RSMI_STATUS_*

1

### Problem Description Not a big issue, but the string returned by amdsmi_status_code_to_string() contains the RSMI_STATUS_* code. For example: ``` amdsmi_status_code_to_string(AMDSMI_STATUS_SUCCESS, &err) -> "RSMI_STATUS_SUCCESS: The function has been executed successfully."...

[Issue]: rocm-smi 6.4.1 shows "N/A" for most partition related values (MI300A, MI300X)

4

### Problem Description ROCm 6.4.1 was installed with this installer: https://repo.radeon.com/amdgpu-install/6.4.1/ubuntu/noble/amdgpu-install_6.4.60401-1_all.deb. The CLI shows "N/A" for most of the partition related values. The rocm-smi API seems to have the same...