ROCm / rocm_smi_lib

ROCm SMI LIB
https://rocm.docs.amd.com/projects/rocm_smi_lib/en/latest/
MIT License
116 stars 49 forks source link

Unable to get full info dump from -a #86

Closed Velkas closed 1 month ago

Velkas commented 3 years ago
======================= ROCm System Management Interface =======================
========================= Version of System Component ==========================
Driver version: 5.10.0-7-amd64
================================================================================
====================================== ID ======================================
GPU[0]      : GPU ID: 0x687f
================================================================================
================================== Unique ID ===================================
GPU[0]      : Unique ID: 0x213fab3ebd840c4
================================================================================
==================================== VBIOS =====================================
GPU[0]      : VBIOS version: 113-D0500100-102
================================================================================
================================= Temperature ==================================
GPU[0]      : Temperature (Sensor edge) (C): 59.0
GPU[0]      : Temperature (Sensor junction) (C): 59.0
GPU[0]      : Temperature (Sensor memory) (C): 60.0
================================================================================
========================== Current clock frequencies ===========================
GPU[0]      : dcefclk clock level: 0: (600Mhz)
ERROR: GPU[0]       : fclk clock is unsupported
GPU[0]      : mclk clock level: 3: (945Mhz)
GPU[0]      : sclk clock level: 0: (852Mhz)
GPU[0]      : socclk clock level: 5: (960Mhz)
python3: /home/<username>/Downloads/rocm_smi_lib-rocm-4.1.0/src/rocm_smi.cc:898: rsmi_status_t get_frequencies(amd::smi::DevInfoTypes, uint32_t, rsmi_frequencies_t*, uint32_t*): Assertion `f->current == RSMI_MAX_NUM_FREQUENCIES + 1' failed.
Aborted

Having issues just getting the info from all this. Not sure what all I need to provide for help.

Card: Vega64 OS: SparkyLinux 6 (Po-Tolo) [Debian testing] with 5.10.0-7-amd64 drivers.

I can do some manual control stuff like setting fans, but things like setting memclock say successful then don't actually change.

HoorayJorge commented 3 years ago

Having a similar issue. Ubuntu 20.04.3, amdspu-pro 20.30, 6600XT


========================= Supported clock frequencies ==========================
GPU[0]      : Supported dcefclk frequencies on GPU0
GPU[0]      : 0: 417Mhz
GPU[0]      : 1: 480Mhz *
GPU[0]      : 2: 1200Mhz
GPU[0]      : 
GPU[0]      : Supported fclk frequencies on GPU0
GPU[0]      : 0: 500Mhz
GPU[0]      : 1: 1248Mhz *
GPU[0]      : 2: 1801Mhz
GPU[0]      : 
GPU[0]      : Supported mclk frequencies on GPU0
GPU[0]      : 0: 96Mhz *
GPU[0]      : 1: 541Mhz
GPU[0]      : 2: 675Mhz
GPU[0]      : 3: 1000Mhz
GPU[0]      : 
python3: /home/farmer/rocm_smi_lib/src/rocm_smi.cc:913: rsmi_status_t get_frequencies(amd::smi::DevInfoTypes, uint32_t, rsmi_frequencies_t*, uint32_t*): Assertion `f->frequency[i-1] <= f->frequency[i]' failed.```
ppanchad-amd commented 1 month ago

@Velkas Apologies for the lack of response. Vega64 is no longer supported as of the latest ROCm 6.2. Closing ticket. Thanks!