Geforce 680 on Fedora: nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000917d:0:0

Hi,
I am using Fedora 5.0.14-300.fc30.x86_64 on my Mac Pro 5.1 with a Geforce 680
(05:00.0 VGA compatible controller: NVIDIA Corporation GK104 [GeForce GTX 680] (rev a1))

I installed the latest 418.74 driver, and all seems to be fine and loaded:
[root@localhost ~]# lsmod | grep -i nvidia
nvidia_drm 49152 3
nvidia_modeset 1093632 5 nvidia_drm
nvidia 17903616 171 nvidia_modeset
ipmi_msghandler 69632 1 nvidia
drm_kms_helper 208896 1 nvidia_drm
drm 495616 6 drm_kms_helper,nvidia_drm

I did remove the “default” driver.

Still I am only getting a black screen when starting GNOME:
[ 3318.334040] nvidia-modeset: WARNING: GPU:0: Lost display notification (0:0x00000000); continuing.
[ 3428.245261] nvidia-modeset: WARNING: GPU:0: Unable to read EDID for display device DVI-I-0
[ 3620.486502] nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000917d:0:0

This seems to be a known issue with newer drivers, but I have not seen an explanation or solution to it yet.
Any suggestions to what the root cause would be, and how to solve it?

For those asking why Fedora on a MacPro 5.1: No CUDA on Mojave, and no web drivers any more. Thought I could finally repurpose my NVIDIA stuff with Linux, but so far no luck.

Any help highly appreciated.

Older Kepler cards, especially with Mac vbios are prone to this, when this happens, no fix. Please try a v390 legacy driver, maybe at least that one works.

Thx, that seems to make sense.
I tried to install version 340, as it was supposed to be tested on Fedora.
Problem now is, I can’t seem to cleanly uninstall 418!
I did the uninstall, installed 340, lsmod shows them all loaded properly
But when I’m starting X, it appears to still have a reference to version 418 somewhere:

[ 1660.602475] NVRM: API mismatch: the client has the version 340.107, but
NVRM: this kernel module has the version 418.74. Please
NVRM: make sure that this kernel module and all NVIDIA driver
NVRM: components have the same version.

Anyone seen this and knows where this could come from?

Quick update:
340 or 390 would be working with my 680, but not with the Meltdown/Spectre fixes in Fedora f30.
Anything above is not working with my card, but would work with the kernel.

Verdict: Dead in the water. NVIDIA again managed to kill support. There is no way around fixing their stuff. Old drivers are not a work around.
Very disappointing!

You shouldn’t use the .run installer, use the driver from a repo, e.g.
https://rpmfusion.org/Configuration/
https://rpmfusion.org/Howto/NVIDIA#Legacy_GeForce_400.2F500

Thanks for the suggestion!
I went forward with it, and did as suggested.
Your method seems to be much cleaner than the .run file version. Makes sense!

Problem is, I’m exactly back to where I was with running the 418 installer directly.

Guess there is no way around the fact (as I perceive it now) that Nvidia does not provide a compatible driver for a current linux kernel with a GeForce GTX 680 EFI card on a Mac Pro 5.1. (Did not personally test other combinations, but forum suggests I’m not alone)

Reading through all related posts, this seems to be a known problem, with no motivation or statement from Nvidia to fix it.
Drivers <= 390 should work with the card, but are not compatible with the meltdown/spectre fixes
Drivers >400 are not working with the graphics card.

This is a real bummer! Combined with the PR disaster on the (lack of) Mojave drivers, my conclusion is to run from Nvidia mid to longterm. They clearly don’t care.

dmesg:
[ 394.497878] IPMI message handler: version 39.2
[ 394.503229] ipmi device interface
[ 394.528377] nvidia: loading out-of-tree module taints kernel.
[ 394.528384] nvidia: module license ‘NVIDIA’ taints kernel.
[ 394.528385] Disabling lock debugging due to kernel taint
[ 394.536878] nvidia: module verification failed: signature and/or required key missing - tainting kernel
[ 394.546726] nvidia-nvlink: Nvlink Core is being initialized, major device number 236
[ 394.546974] nvidia 0000:05:00.0: enabling device (0006 -> 0007)
[ 394.547045] nvidia 0000:05:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=none:owns=io+mem
[ 394.748112] NVRM: loading NVIDIA UNIX x86_64 Kernel Module 418.56 Fri Mar 15 12:59:26 CDT 2019
[ 394.761728] nvidia-uvm: Loaded the UVM driver in 8 mode, major device number 234
[ 394.764412] nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 418.56 Fri Mar 15 12:32:40 CDT 2019
[ 394.766855] [drm] [nvidia-drm] [GPU ID 0x00000500] Loading driver
[ 394.766857] [drm] Initialized nvidia-drm 0.0.0 20160202 for 0000:05:00.0 on minor 0
[ 407.316543] nvidia-modeset: WARNING: GPU:0: Lost display notification (0:0x00000000); continuing.
[ 664.956398] nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000917d:0:0

X.Org output:
(==) Using config directory: “/etc/X11/xorg.conf.d”
(==) Using system config directory “/usr/share/X11/xorg.conf.d”
Failed to connect to bus: No such file or directory

** (process:9194): WARNING **: 11:06:50.629: Could not make bus activated clients aware of XDG_CURRENT_DESKTOP=GNOME environment variable: Failed to execute child process ?dbus-launch? (No such file or directory)
xinit: connection to X server lost

waiting for X server to shut down …
xinit: X server slow to shut down, sending KILL signal

waiting for server to die …
xinit: X server refuses to die

There is nothing useful in the log

[root@localhost ~]# lsmod | grep nvidia
nvidia_drm 53248 0
nvidia_modeset 1089536 2 nvidia_drm
nvidia_uvm 946176 0
nvidia 17641472 58 nvidia_uvm,nvidia_modeset
drm_kms_helper 208896 1 nvidia_drm
ipmi_msghandler 69632 2 ipmi_devintf,nvidia
drm 495616 3 drm_kms_helper,nvidia_drm

What makes you think so?

https://devtalk.nvidia.com/default/topic/1042691/linux/black-screen-with-mac-version-of-gtx-680/3

Quote:
“I guess the remaining question is whether anyone here has been able to get a GTX 680 to work with the nvidia drivers beyond the 340 release series? I would assume that if such cards have become unsupported, this should be considered a regression in the drivers (or at least properly documented as no longer being supported).”

To use the 340 or the 390 driver on a 5.x kernel it requires patching. There was another post (would have to find it again) where Nvidia wrote that due to the alternate stack used in their code it won’t work on the kernel with the latest patches for Meltdown and Spectre.
I did try a patched version of the 340.107 driver (NVIDIA-Linux-x86_64-340.107-patched-kernel-5.0) as an alternative, but that would get the Xserver to crash immediatly upon startup. Simply not working.

Still looking for a way to install a driver for my EFI GeForce GTX 680 2GB on a Mac Pro 5.1 with latest Firmware (144.0.0.0) that would work. I installed the driver from the repository, which as stated in the linked discussion, is not working with that Nvidia card. (Genuine Nvidia MAC Edition Firmware)

That’s why you should use the repo drivers, those already contain the necessary patches for the distro kernels.

Clarification:
According to reports (see link above), I need the 340 driver to get this card to work.
According to rpmfusion this driver is only supported up to “up to Fedora 27 - EOL, no more nvidia updates”.
I downloaded a patched version of this driver that would load on f30, but this one causes an instant crash with X

Conclusion: Nvidia silently killed support for their Mac Edition card. Silently, because they refuse to state this in public.

You need to put on your glasses. The 304 driver is out of support, not the 340. Did you ever try the 390 repo driver?

Yes, sorry. Mixed them up when responding.
I tried the 418 run, the 430 beta run, the 430 final run, the 418 repo, the 403 repo and the 390 repo and he patched 340 run.
The 390 repo would not load correctly.
Install went fine, but when starting X I got the message in dmesg that there still was a 418 kernel module left.
Even after removing all driver artefacts, installing the stock config again, and then the 390, it would not load properly, as X claims there was a conflict with the 418, which wasn’t on the system any more.
Yes, I checked the /usr/lib64/xorg/modules/drivers/nvidia_drv.so

Please run a 418 .run installer using the --uninstall option, then try the 390 driver again.

did dnf remove with the repo driver
did NVIDIA-Linux-x86_64-418.74.run --uninstall
/usr/lib64/xorg/modules/drivers/ had no nvidia_drv.so
did install the repo 390

Result: black screen

lsmod | grep nvidia
nvidia_drm 53248 0
nvidia_modeset 1089536 2 nvidia_drm
nvidia_uvm 946176 0
nvidia 7641472 58 nvidia_uvm,nvidia_modeset
drm_kms_helper 208896 1 nvidia_drm
drm 495616 3 drm_kms_helper,nvidia_drm
ipmi_msghandler 69632 2 ipmi_devintf,nvidia

startx log: (extracts from log level 6)
[ 2637.484] (II) LoadModule: “nvidia”
[ 2637.485] (II) Loading /usr/lib64/xorg/modules/drivers/nvidia_drv.so
[ 2637.485] (II) Module nvidia: vendor=“NVIDIA Corporation”
[ 2637.485] compiled for 4.0.2, module version = 1.0.0
[ 2637.485] Module class: X.Org Video Driver

[ 2637.486] (II) NVIDIA dlloader X Driver 418.56 Fri Mar 15 12:37:35 CDT 2019
[ 2637.486] (II) NVIDIA Unified Driver for all Supported NVIDIA GPUs
[ 2637.486] (II) NVIDIA X compatibility module for ABI 24.0 built from xorg-server-1.19.99.902
[ 2637.486] (II) X log verbosity: 6

[ 2637.494] (**) NVIDIA(0): Enabling 2D acceleration
[ 2637.494] (II) Loading sub module “glxserver_nvidia”
[ 2637.494] (II) LoadModule: “glxserver_nvidia”
[ 2637.494] (II) Loading /usr/lib64/xorg/modules/extensions/libglxserver_nvidia.so
[ 2637.498] (II) Module glxserver_nvidia: vendor=“NVIDIA Corporation”
[ 2637.498] compiled for 4.0.2, module version = 1.0.0
[ 2637.498] Module class: X.Org Server Extension
[ 2637.498] (II) NVIDIA GLX Module 418.56 Fri Mar 15 12:36:35 CDT 2019
[ 2637.498] (WW) NVIDIA(0): Failed to initialize Base Mosaic! Reason: Only one GPU
[ 2637.498] (WW) NVIDIA(0): detected. Only one GPU will be used for this X screen.
[ 2638.697] (–) NVIDIA(0): Valid display device(s) on GPU-0 at PCI:5:0:0
[ 2638.697] (–) NVIDIA(0): CRT-0
[ 2638.697] (–) NVIDIA(0): DFP-0
[ 2638.697] (–) NVIDIA(0): DFP-1 (boot)
[ 2638.697] (–) NVIDIA(0): DFP-2
[ 2638.697] (–) NVIDIA(0): DFP-3
[ 2638.697] (–) NVIDIA(0): DFP-4
[ 2638.698] (II) NVIDIA(0): NVIDIA GPU GeForce GTX 680 (GK104) at PCI:5:0:0 (GPU-0)
[ 2638.699] (II) NVIDIA(0): GPU UUID: GPU-280f681b-509b-812f-d854-8fa45f8befc7
[ 2638.699] (–) NVIDIA(0): Memory: 2097152 kBytes
[ 2638.699] (II) NVIDIA(0): Board ID: 0xe10c
[ 2638.699] (II) NVIDIA(0): Chip SKU: 400
[ 2638.699] (II) NVIDIA(0): Chip SKU Mod: 0
[ 2638.699] (II) NVIDIA(0): Project: 2002
[ 2638.699] (II) NVIDIA(0): Project SKU: 0000
[ 2638.699] (II) NVIDIA(0): GPU RAM Type: GDDR5
[ 2638.699] (–) NVIDIA(0): VideoBIOS: 80.04.87.00.09
[ 2638.699] (–) NVIDIA(0): Found 4 heads on board
[ 2638.699] (II) NVIDIA(0): Supported display device(s): CRT-0, DFP-0, DFP-1, DFP-2,
[ 2638.699] (II) NVIDIA(0): DFP-3, DFP-4
[ 2638.699] (II) NVIDIA(0): Bus detected as PCI Express
[ 2638.699] (II) NVIDIA(0): Detected PCI Express Link width: 16X
[ 2638.699] (II) NVIDIA(0): SPS : 4
[ 2638.699] (II) NVIDIA(0): User configuration of GPU PowerMizer is supported

[ 2638.713] (–) NVIDIA(GPU-0): DFP-0: 330.0 MHz maximum pixel clock
[ 2638.713] (–) NVIDIA(GPU-0):
[ 2638.719] (–) NVIDIA(GPU-0): DELL 1907FP (DFP-1): connected
[ 2638.719] (–) NVIDIA(GPU-0): DELL 1907FP (DFP-1): Internal TMDS
[ 2638.719] (–) NVIDIA(GPU-0): DELL 1907FP (DFP-1): DFP is not internal to notebook
[ 2638.719] (–) NVIDIA(GPU-0): DELL 1907FP (DFP-1) Name Aliases:
[ 2638.719] (–) NVIDIA(GPU-0): DFP
[ 2638.719] (–) NVIDIA(GPU-0): DFP-1
[ 2638.719] (–) NVIDIA(GPU-0): DPY-2
[ 2638.719] (–) NVIDIA(GPU-0): HDMI-0
[ 2638.719] (–) NVIDIA(GPU-0): DPY-EDID-627e0586-9d81-52b1-e784-82f6af57925b
[ 2638.719] (–) NVIDIA(GPU-0): HDMI-0
[ 2638.719] (–) NVIDIA(GPU-0): DELL 1907FP (DFP-1): 165.0 MHz maximum pixel clock

[ 2638.754] (–) NVIDIA(GPU-0): Established Timings:
[ 2638.754] (–) NVIDIA(GPU-0): 640 x 480 @ 60 Hz
[ 2638.754] (–) NVIDIA(GPU-0): 800 x 600 @ 60 Hz
[ 2638.754] (–) NVIDIA(GPU-0): 1024 x 768 @ 60 Hz

[ 2638.755] (II) NVIDIA(0): Using HorizSync/VertRefresh ranges from the EDID for display
[ 2638.755] (II) NVIDIA(0): device DELL 1907FP (DFP-1).
[ 2638.755] (II) NVIDIA(GPU-0):
[ 2638.755] (II) NVIDIA(GPU-0): — Building ModePool for DELL 1907FP (DFP-1) —
[ 2638.755] (II) NVIDIA(GPU-0): Validating Mode “1280x1024_60”:
[ 2638.755] (II) NVIDIA(GPU-0): Mode Source: EDID
[ 2638.755] (II) NVIDIA(GPU-0): 1280 x 1024 @ 60 Hz
[ 2638.755] (II) NVIDIA(GPU-0): Pixel Clock : 108.00 MHz
[ 2638.755] (II) NVIDIA(GPU-0): HRes, HSyncStart : 1280, 1328
[ 2638.755] (II) NVIDIA(GPU-0): HSyncEnd, HTotal : 1440, 1688
[ 2638.755] (II) NVIDIA(GPU-0): VRes, VSyncStart : 1024, 1025
[ 2638.755] (II) NVIDIA(GPU-0): VSyncEnd, VTotal : 1028, 1066
[ 2638.755] (II) NVIDIA(GPU-0): Sync Polarity : +H +V
[ 2638.755] (II) NVIDIA(GPU-0): Viewport 1280x1024+0+0
[ 2638.755] (II) NVIDIA(GPU-0): Horizontal Taps 1
[ 2638.755] (II) NVIDIA(GPU-0): Vertical Taps 1
[ 2638.755] (II) NVIDIA(GPU-0): Core Depth 64
[ 2638.755] (II) NVIDIA(GPU-0): Base Depth 64
[ 2638.755] (II) NVIDIA(GPU-0): Distributed Rendering 1
[ 2638.755] (II) NVIDIA(GPU-0): Overlay Depth 32
[ 2638.755] (II) NVIDIA(GPU-0): Mode “1280x1024_60” is valid.

[ 2638.793] (II) NVIDIA(GPU-0): — Done building ModePool for Lenovo Group Limited X1 (DFP-2)
[ 2638.793] (II) NVIDIA(GPU-0): —
[ 2638.793] (II) NVIDIA(GPU-0):
[ 2638.793] (II) NVIDIA(GPU-0): Frequency information for Lenovo Group Limited X1 (DFP-2):
[ 2638.793] (II) NVIDIA(GPU-0): HorizSync : 24.000-140.000 kHz
[ 2638.793] (II) NVIDIA(GPU-0): VertRefresh : 23.000-80.000 Hz
[ 2638.793] (II) NVIDIA(GPU-0): (HorizSync from EDID)
[ 2638.793] (II) NVIDIA(GPU-0): (VertRefresh from EDID)
[ 2638.793] (II) NVIDIA(GPU-0):
[ 2638.793] (II) NVIDIA(GPU-0): — Modes in ModePool for Lenovo Group Limited X1 (DFP-2) —
[ 2638.793] (II) NVIDIA(GPU-0): “nvidia-auto-select” : 1920 x 1080 @ 60.0 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1920x1080” : 1920 x 1080 @ 60.0 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1920x1080_60” : 1920 x 1080 @ 60.0 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1920x1080_60_0” : CTA-861G:#16:1920x1080x59.940Hz/P (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1920x1080_50” : CTA-861G:#31:1920x1080x50.000Hz/P (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1920x1080_60i” : CTA-861G:#5:1920x1080x59.940Hz/I (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1920x1080_50i” : CTA-861G:#20:1920x1080x50.000Hz/I (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1680x1050” : 1680 x 1050 @ 60.0 Hz (from: X Server, EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1680x1050_60” : 1680 x 1050 @ 60.0 Hz (from: X Server, EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1600x900” : 1600 x 900 @ 60.0 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1600x900_60” : 1600 x 900 @ 60.0 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1440x900” : 1440 x 900 @ 59.9 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1440x900_60” : 1440 x 900 @ 59.9 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1280x1024” : 1280 x 1024 @ 60.0 Hz (from: X Server, EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1280x1024_60” : 1280 x 1024 @ 60.0 Hz (from: X Server, EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1280x800” : 1280 x 800 @ 59.8 Hz (from: X Server, EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1280x800_60” : 1280 x 800 @ 59.8 Hz (from: X Server, EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1280x720” : 1280 x 720 @ 60.0 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1280x720_60” : 1280 x 720 @ 60.0 Hz (from: EDID)
[ 2638.793] (II) NVIDIA(GPU-0): “1280x720_60_0” : CTA-861G:#4:1280x720x59.940Hz/P (from: EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “1280x720_50” : CTA-861G:#19:1280x720x50.000Hz/P (from: EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “1024x768” : 1024 x 768 @ 60.0 Hz (from: X Server, EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “1024x768_60” : 1024 x 768 @ 60.0 Hz (from: X Server, EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “800x600” : 800 x 600 @ 60.3 Hz (from: X Server, EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “800x600_60” : 800 x 600 @ 60.3 Hz (from: X Server, EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “720x576” : CTA-861G:#17:720x576x50.000Hz/P (from: EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “720x576_50” : CTA-861G:#17:720x576x50.000Hz/P (from: EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “720x480” : CTA-861G:#2:720x480x59.940Hz/P (from: EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “720x480_60” : CTA-861G:#2:720x480x59.940Hz/P (from: EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “640x480” : 640 x 480 @ 59.9 Hz (from: X Server, EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “640x480_60” : 640 x 480 @ 59.9 Hz (from: X Server, EDID)
[ 2638.794] (II) NVIDIA(GPU-0): “640x480_60_0” : CTA-861G:#1:640x480x59.940Hz/P (from: EDID)
[ 2638.794] (II) NVIDIA(GPU-0): — End of ModePool for Lenovo Group Limited X1 (DFP-2): —

[ 2638.799] (II) NVIDIA: Using 6144.00 MB of virtual memory for indirect memory
[ 2638.799] (II) NVIDIA: access.
[ 2638.804] (II) NVIDIA(0): kernel module enabled successfully
[ 2638.804] (II) NVIDIA(0): GPU initialized
[ 2638.805] (II) NVIDIA(0): ACPI: failed to connect to the ACPI event daemon; the daemon
[ 2638.805] (II) NVIDIA(0): may not be running or the “AcpidSocketPath” X
[ 2638.805] (II) NVIDIA(0): configuration option may not be set correctly. When the
[ 2638.805] (II) NVIDIA(0): ACPI event daemon is available, the NVIDIA X driver will
[ 2638.805] (II) NVIDIA(0): try to use it to receive ACPI event notifications. For
[ 2638.805] (II) NVIDIA(0): details, please see the “ConnectToAcpid” and
[ 2638.805] (II) NVIDIA(0): “AcpidSocketPath” X configuration options in Appendix B: X
[ 2638.805] (II) NVIDIA(0): Config Options in the README.
[ 2638.808] (==) NVIDIA(0): Using 8388608 bytes for software rendering cache
[ 2638.846] (II) NVIDIA(0): Setting mode “DFP-1:nvidia-auto-select,DFP-2:nvidia-auto-select”
[ 2677.849] (II) NVIDIA(0): First mode initialized
[ 2681.855] (II) NVIDIA(0): Initialized OpenGL Acceleration
[ 2681.856] (II) NVIDIA(0): Visuals set up
[ 2681.856] (II) NVIDIA(0): Pixmap depths set up
[ 2681.856] (II) NVIDIA(0): GLX visuals set up
[ 2681.856] (II) NVIDIA(0): Framebuffer set up
[ 2681.856] (==) NVIDIA(0): Disabling shared memory pixmaps
[ 2681.856] (II) NVIDIA(0): Initialized X Rendering Acceleration
[ 2681.856] (==) NVIDIA(0): Backing store enabled
[ 2681.856] (==) NVIDIA(0): Silken mouse enabled
[ 2681.856] (II) NVIDIA(0): Default colormap initialized.
[ 2681.856] (II) NVIDIA(0): Palette loaded
[ 2681.856] (==) NVIDIA(0): DPMS enabled
[ 2681.856] (II) NVIDIA(0): Loading NVIDIA Xinerama information

How did you install the 390 repo driver? The X log tells that you have the full 418.56 kernel/X/GL driver installed.

Well, agreed! That’s the magic!
dnf install xorg-x11-drv-nvidia-390xx akmod-nvidia-390xx
dnf install xorg-x11-drv-nvidia-390xx-cuda
dnf update -y

does
dnf list --installed nvidia
or similar list any other installed packages?

[root@localhost ~]# dnf list --installed | grep nvidia
akmod-nvidia-390xx.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree
kmod-nvidia-390xx-5.0.14-300.fc30.x86_64.x86_64 3:390.116-2.fc30 @@commandline
nvidia-persistenced.x86_64 3:418.56-1.fc30 @rpmfusion-nonfree
nvidia-settings-390xx.x86_64 390.116-2.fc30 @rpmfusion-nonfree
xorg-x11-drv-nvidia-390xx.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree
xorg-x11-drv-nvidia-390xx-cuda.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree
xorg-x11-drv-nvidia-390xx-cuda-libs.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree
xorg-x11-drv-nvidia-390xx-kmodsrc.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree
xorg-x11-drv-nvidia-390xx-libs.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree

shoot, that’s interesting! Good hint. Let’s see how this can be cleaned…

edit: the dependency is from the 390x-cuda package! That should not be, right?
Entfernen:
nvidia-persistenced x86_64 3:418.56-1.fc30 @rpmfusion-nonfree 98 k
Abhängige Pakete werden entfernt:
xorg-x11-drv-nvidia-390xx-cuda x86_64 3:390.116-2.fc30 @rpmfusion-nonfree 865 k
Nicht benötigte Abhängigkeiten werden entfernt:
xorg-x11-drv-nvidia-390xx-cuda-libs x86_64 3:390.116-2.fc30 @rpmfusion-nonfree 84 M

Ok, manually removed cuda (dnf remove …)
[root@localhost ~]# dnf list --installed | grep nvidia
akmod-nvidia-390xx.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree
kmod-nvidia-390xx-5.0.14-300.fc30.x86_64.x86_64 3:390.116-2.fc30 @@commandline
nvidia-settings-390xx.x86_64 390.116-2.fc30 @rpmfusion-nonfree
xorg-x11-drv-nvidia-390xx.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree
xorg-x11-drv-nvidia-390xx-kmodsrc.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree
xorg-x11-drv-nvidia-390xx-libs.x86_64 3:390.116-2.fc30 @rpmfusion-nonfree

No 418.56 packages left.

Still:
[ 2637.484] (II) LoadModule: “nvidia”
[ 2637.485] (II) Loading /usr/lib64/xorg/modules/drivers/nvidia_drv.so
[ 2637.485] (II) Module nvidia: vendor=“NVIDIA Corporation”
[ 2637.486] (II) NVIDIA dlloader X Driver 418.56 Fri Mar 15 12:37:35 CDT 2019
[ 2637.486] (II) NVIDIA Unified Driver for all Supported NVIDIA GPUs
[ 2637.486] (II) NVIDIA X compatibility module for ABI 24.0 built from xorg-server-1.19.99.902

When removing 418 previously I watched how /usr/lib64/xorg/modules/drivers/nvidia_drv.so disappeared.
It was gone before installing the 390 package.
Ergo: either the 390 package installed the 418, or the CUDA from 390 did overwrite it with 418. I will uninstall 390 again, (watching for the module) and install again without cuda. This will at least clarify if both packages have the wrong driver, or only the code package screwed up…

Please create a new nvidia-bug-report.log and attach.