RmInitAdapter failed!

[robang74]

NVIDIA Open GPU Kernel Modules Version

515.43.04

Does this happen with the proprietary driver (of the same version) as well?

No

Operating System and Version

Ubuntu 20.04.4 LTS

Kernel Release

Linux 5.4.0-109-generic #123-Ubuntu SMP Fri Apr 8 09:10:54 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux

Hardware: GPU

Unable to determine the device handle for gpu 0000:01:00.0: Not Found

Describe the bug

The nvidia-smi fails if used in combination with the Open-Source driver but not with the proprietary. I suspect a problem with the loading of the gsp.bin. Thanks, R

To Reproduce

wget https://download.nvidia.com/XFree86/Linux-x86_64/515.43.04/NVIDIA-Linux-x86_64-515.43.04-no-compat32.run

bash ./NVIDIA-Linux-x86_64-515.43.04-no-compat32.run --no-kernel-module -q -s -Z 

git clone https://github.com/NVIDIA/open-gpu-kernel-modules

cd open-gpu-kernel-modules

BUILD_EXCLUSIVE_KERNEL=$(uname -r) IGNORE_PREEMPT_RT_PRESENCE=1 make modules -j$(nproc)
BUILD_EXCLUSIVE_KERNEL=$(uname -r) IGNORE_PREEMPT_RT_PRESENCE=1 make modules_install -j$(nproc)

depmod -a; update-initramfs -u; reboot

lsmod | grep nvidia

nvidia_drm             61440  0
nvidia_modeset       1069056  1 nvidia_drm
nvidia               4558848  1 nvidia_modeset
drm_kms_helper        278528  2 nvidia_drm,i915
drm                   622592  5 drm_kms_helper,nvidia,nvidia_drm,i915

cat /proc/driver/nvidia/version

NVRM version: NVIDIA UNIX Open Kernel Module for x86_64  515.43.04  Release Build  on  Tue 17 May 2022 03:46:22 PM CEST
GCC version:  gcc version 10.2.1 20210110 (Debian 10.2.1-6)

nvidia-smi --help | grep -e "^NVIDIA"

NVIDIA System Management Interface -- v515.43.04

nvidia-smi

Unable to determine the device handle for GPU 0000:01:00.0: Not Found

nvidia-debugdump --list

Found 1 NVIDIA devices
Error: nvmlDeviceGetHandleByIndex(): Not Found
FAILED to get details on GPU (0x0): Not Found

Bug Incidence

Always

nvidia-bug-report.log.gz

nvidia-bug-report.log.gz

More Info

cat /etc/modprobe.d/nvidia-opensource-options.conf

options nvidia NVreg_EnableGpuFirmwareLogs=7
options nvidia NVreg_EnableGpuFirmware=0
options nvidia NVreg_EnablePCIeGen3=1

[robang74]

With or without device driver options, still not work. Also using this option is not going to work

options nvidia NVreg_OpenRmEnableUnsupportedGpus=1

The opensource driver that does not work with RTX A6000 works with A2.

The nVidia blog page about the OpenSource driver release claims that Ampere and Turing families are fully supported

https://developer.nvidia.com/blog/nvidia-releases-open-source-gpu-kernel-modules/

• Open kernel modules support all NVIDIA Ampere and Turing architecture GPUs

However, at the end of this page there is a table of cards that support GSP firmware which is the key of the OpenSource driver.

https://download.nvidia.com/XFree86/Linux-x86_64/515.43.04/README/gsp.html

Unfortunately the RTX A6000 is not present. This could be the reason for this malfunctioning?

[PAR2020]

Sorry for the delayed response, @robang74. Could you try the latest drop 515.48.07?
RTX A6000 is supported, DevIDs also added for clarity in the README.md file.
Thank you.

[PAR2020]

Ping @robang74...

[PAR2020]

Ping @robang74...?

[YusufKhan-gamedev]

I have a similar-ish thing, that release didnt fix it on my tu106

[robang74]

Sorry for the delayed response, @robang74. Could you try the latest drop 515.48.07? RTX A6000 is supported, DevIDs also added for clarity in the README.md file. Thank you.

@PAR2020 : I have tried the last version and it works. Thank you.