I have Suse Leap42.3 and trying to install CUDA 9.2 .
I followed the installation steps in the installation Guide Linux - v9.2.88
During the installation, it asked to install a later driver (from 390.? to 396.26).
That seems to have worked as now nvidia-smi reports:
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 396.26 Driver Version: 396.26 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce GTX 105... On | 00000000:01:00.0 Off | N/A |
| N/A 45C P8 N/A / N/A | 0MiB / 4040MiB | 0% Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+
However, I got stuck on the Power9Setup https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html#power9-setup
Particularly the NVIDIA Persistence Daemon .
I get the following:
linux-98bz:~> which nvidia-persistenced
/usr/bin/nvidia-persistenced
inux-98bz:~> systemctl status nvidia-persistenced
● nvidia-persistenced.service
Loaded: not-found (Reason: No such file or directory)
Active: inactive (dead)
I then try :
linux-98bz:~> sudo systemctl enable nvidia-persistenced
[sudo] password for root:
Failed to execute operation: No such file or directory
linux-98bz:~> sudo systemctl enable /usr/bin/nvidia-persistenced
Failed to execute operation: Invalid argument
I also tried, manually, not at boot time :
sudo /usr/bin/nvidia-persistenced --verbose
nvidia-persistenced failed to initialize. Check syslog for more details.
No idea how to use syslog. Something called nvidia-persistenced seems to be running though :
linux-98bz:~> ps -aux|grep -e nvidia
root 327 0.0 0.0 0 0 ? S 19:28 0:00 [nvidia-modeset]
root 3897 0.0 0.0 8524 1596 ? Ss 20:38 0:00 nvidia-persistenced
root 3899 0.0 0.0 0 0 ? S 20:38 0:00 [irq/129-nvidia]
root 3901 0.0 0.0 0 0 ? S 20:38 0:00 [nvidia]
enzo 4201 0.0 0.0 10548 1620 pts/1 S+ 21:08 0:00 grep --color=auto -e nvidia
What now ?
Also, not sure what to do with the next, extremely vague step. Opensuse Leap is not mentioned and I found 3 files in /lib/udev/rules.d/ with ‘hot’ in it:
linux-98bz:~> ls /lib/udev/rules.d/|grep -e hot
40-libgphoto2.rules
80-acpi-container-hotplug.rules
80-hotplug-cpu-mem.rules
The only one vaguely relevant seems to be 80-hotplug-cpu-mem.rules that contains :
linux-98bz:~> cat /lib/udev/rules.d/80-hotplug-cpu-mem.rules
# do not edit this file, it will be overwritten on update
# Hotplug physical CPU
SUBSYSTEM=="cpu", ACTION=="add", TEST=="online", ATTR{online}=="0", \
ATTR{online}="1"
# Hotplug physical memory
SUBSYSTEM=="memory", ACTION=="add", PROGRAM="/usr/bin/systemd-detect-virt", RESULT!="zvm", ATTR{state}=="offline", \
ATTR{state}="online", TAG+="tmpfs"
# See bnc#869603
TAG=="tmpfs", RUN+="/usr/lib/udev/remount-tmpfs"
Is it relevant should it be moved to /etc/udev/rules.d and changed ? How ?