Ask Your Question
0

Graphics seems not rendered on Nvidia dGPU after kernel update

asked 2018-03-29 08:28:28 -0500

fghzxm gravatar image

I installed the proprietary Nvidia graphics driver 390.42 and CUDA driver on my Optimus-enabled laptop from the RPMFusion non-free repository when I was on kernel 4.15.10-300.fc27.x86_64, following the how-to on https://rpmfusion.org/Howto/NVIDIA and everything worked well. GNOME Settings/Details recognized the discrete graphics card ("GeForce GTX 1050/PCIe/SSE2").

Then after I upgraded to kernel 4.15.12-301, Settings no longer detects the Nvidia graphics, only showing "Intel® HD Graphics 630 (Kaby Lake GT2)". When I attempt to compile and run a CUDA sample program, as long as the program does not render graphics (vectorAdd, fp16ScalarProduct, matrixMul etc.) it runs without problem, but programs like nbody and randomFog (which render graphics) reports error 46 cudaErrorDevicesUnavailable (or similar).

$ sudo dmesg | grep -Fi nvidia
[    0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-4.15.12-301.fc27.x86_64 root=UUID=d6f184ca-812d-4d0b-9551-df4cfb12d885 ro rd.driver.blacklist=nouveau modprobe.blacklist=nouveau nvidia-drm.modeset=1 intel_iommu=on LANG=en_US.UTF-8
[    0.000000] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-4.15.12-301.fc27.x86_64 root=UUID=d6f184ca-812d-4d0b-9551-df4cfb12d885 ro rd.driver.blacklist=nouveau modprobe.blacklist=nouveau nvidia-drm.modeset=1 intel_iommu=on LANG=en_US.UTF-8
[    4.613266] nvidia: module license 'NVIDIA' taints kernel.
[    4.650558] nvidia-nvlink: Nvlink Core is being initialized, major device number 238
[    4.650737] nvidia 0000:01:00.0: enabling device (0006 -> 0007)
[    4.651615] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  390.42  Sat Mar  3 04:10:22 PST 2018 (using threaded interrupts)
[    4.699019] nvidia-uvm: Loaded the UVM driver in 8 mode, major device number 237
[    4.710348] nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms  390.42  Sat Mar  3 03:30:48 PST 2018
[    4.716639] [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
[    5.108433] nvidia-modeset: Allocated GPU:0 (GPU-5b31775c-873a-b047-811d-6a1da2804038) @ PCI:0000:01:00.0
[    5.109578] nvidia-modeset: Freed GPU:0 (GPU-5b31775c-873a-b047-811d-6a1da2804038) @ PCI:0000:01:00.0
[    5.112717] [drm] Initialized nvidia-drm 0.0.0 20160202 for 0000:01:00.0 on minor 1
edit retag flag offensive close merge delete

1 Answer

Sort by » oldest newest most voted
0

answered 2018-04-08 11:35:47 -0500

kwizart gravatar image

Please verify that both nvidia.ko and the userspace are at the same version. The Current RPM Fusion driver is at 390.48 nowadays.

We are looking to improve this interaction in this report https://bugzilla.rpmfusion.org/show_b...

For now, best is to restart automatically when a driver update occurs (or rely on gnome-software offline updates).

edit flag offensive delete link more

Question Tools

Stats

Asked: 2018-03-29 08:28:28 -0500

Seen: 662 times

Last updated: Mar 29 '18