You are not logged in.
I have installed the latest nvidia and cuda drivers through pacman.
I have verified that the nvidia module is being loaded properly. And I can use it to run CUDA applications.
However, some applications that require nvidia-uvm are failing because modprobe can't find nvidia-uvm. I am getting this error :
modprobe: FATAL: Module nvidia-uvm not found.
I checked in the location and nvidia-uvm is not inside the /lib/modules/$(uname) directory but there is a file called nvidia-uvm.ko.gz in a directory called extra-modules inside it.
How can I load nvidia-uvm? Is it required to remove nvidia drivers and install nvidia-dkms instead?
Last edited by niharsarangi (2016-05-16 17:15:56)
Offline
Have you installed opencl-nvidia ?
Offline
Yes. opencl-nvidia is up to date.
Offline
Please post the output of the following:
$ uname -a
$ modinfo nvidia-uvm
$ pacman -Qs nvidia
Sakura:-
Mobo: MSI MAG X570S TORPEDO MAX // Processor: AMD Ryzen 9 5950X @4.9GHz // GFX: AMD Radeon RX 5700 XT // RAM: 32GB (4x 8GB) Corsair DDR4 (@ 3000MHz) // Storage: 1x 3TB HDD, 6x 1TB SSD, 2x 120GB SSD, 1x 275GB M2 SSD
Making lemonade from lemons since 2015.
Offline
I am not sure what I did. But after a full system upgrade and reinstalling the nvidia drivers it seems to work. The logs are as follows:
uname -a
Linux JARVIS 4.5.4-1-ARCH #1 SMP PREEMPT Wed May 11 22:21:28 CEST 2016 x86_64 GNU/Linux
modinfo nvidia-uvm
filename: /lib/modules/4.5.4-1-ARCH/extramodules/nvidia-uvm.ko.gz
supported: external
license: MIT
depends: nvidia
vermagic: 4.5.3-1-ARCH SMP preempt mod_unload modversions
parm: uvm_pmm_mode:Set the mode of allocation (charp)
parm: uvm_debug_enable_push_desc:Enable push description tracking (int)
parm: uvm_page_table_location:Set the location for UVM-allocated page tables. Choices are: vid, sys. (charp)
parm: uvm_perf_fault_batch_count:uint
parm: uvm_perf_fault_replay_policy:uint
parm: uvm_enable_debug_procfs:Enable debug procfs entries in /proc/driver/nvidia-uvm (int)
parm: uvm_prefetch_stats:int
parm: uvm_prefetch_threshold:int
parm: uvm_prefetch_adaptive:int
parm: uvm_prefetch_epoch:int
parm: uvm_prefetch_sparsity_inc:int
parm: uvm_prefetch_sparsity_dec:int
parm: uvm_prefetch:int
parm: uvm_nothrashing:int
parm: uvm_thrashing_fault_count:int
parm: uvm_thrashing_lapse:int
parm: uvm_thrashing_nap:int
parm: uvm_driver_mode:Set the uvm kernel driver mode. Choices include: 8, full, lite. (charp)
parm: uvm_enable_builtin_tests:Enable the UVM built-in tests. (This is a security risk) (int)
pacman -Qs nvidia
local/cuda 7.5.18-2
NVIDIA's GPU programming toolkit
local/cudnn 5.0.4-1
NVIDIA CUDA Deep Neural Network library
local/libvdpau 1.1.1-2
Nvidia VDPAU library
local/nvidia 364.19-3
NVIDIA drivers for linux
local/nvidia-libgl 364.19-1
NVIDIA drivers libraries symlinks
local/nvidia-settings 364.19-1
Tool for configuring the NVIDIA graphics driver
local/nvidia-utils 364.19-1
NVIDIA drivers utilities
local/opencl-nvidia 364.19-1
OpenCL implemention for NVIDIA
Offline
If you have solved the problem, please mark your thread as solved by editing the first post and amending the topic title.
Everything looks fine in that output, but presumably it is from after you fixed the problem.
Sakura:-
Mobo: MSI MAG X570S TORPEDO MAX // Processor: AMD Ryzen 9 5950X @4.9GHz // GFX: AMD Radeon RX 5700 XT // RAM: 32GB (4x 8GB) Corsair DDR4 (@ 3000MHz) // Storage: 1x 3TB HDD, 6x 1TB SSD, 2x 120GB SSD, 1x 275GB M2 SSD
Making lemonade from lemons since 2015.
Offline
Thanks.
Marked as solved.
Offline