So, right now, I installed a fresh Fedora and I did a dnf upgrade which installed 34 new packages and upgraded 759 packages with a total of 1.1 GB.
I rebooted and everything works as expected and as it was before. Wayland is running and the GPU is recognized as N117 which is the nouveau code for the GM107 code and this is the Nvidia Quadro card.
Next I am going to install the Nvidia drivers from copr.
So nope, even after the fresh install and running the installer from copr, after a reboot I am running into the black screen.
What I don’t get is still the fact, that if I boot the alternate 5.11.12-300 kernel, I don’t run into the black screen but I always get the output “Nvidia kernel module is missing. Falling back to nouveau” and then Fedora boots up.
The installation of the packages when running the installer seems to break something.
Here is the list of packages that were installed when running the installer:
akmod-nvidia-3:470.57.02-1.fc34.x86_64 akmods-0.5.6-27.fc34.noarch annobin-9.79-2.fc34.x86_64
bison-3.7.4-2.fc34.x86_64 dwz-0.14-1.fc34.x86_64 ed-1.14.2-10.fc34.x86_64
efi-srpm-macros-5-2.fc34.noarch egl-wayland-1.1.7-1.fc34.x86_64 elfutils-libelf-devel-0.185-2.fc34.x86_64
fakeroot-1.25.3-2.fc34.x86_64 fakeroot-libs-1.25.3-2.fc34.x86_64 flex-2.6.4-7.fc34.x86_64
fonts-srpm-macros-1:2.0.5-5.fc34.noarch fpc-srpm-macros-1.3-3.fc34.noarch ghc-srpm-macros-1.5.0-4.fc34.noarch
gnat-srpm-macros-4-13.fc34.noarch go-srpm-macros-3.0.10-1.fc34.noarch http-parser-2.9.4-4.fc34.x86_64
info-6.7-10.fc34.x86_64 kernel-devel-5.13.9-200.fc34.x86_64 kernel-srpm-macros-1.0-4.fc34.noarch
kmodtool-1-42.fc34.noarch koji-1.25.1-1.fc34.noarch libgit2-1.1.0-4.fc34.x86_64
libglvnd-opengl-1:1.3.3-1.fc34.x86_64 libvdpau-1.4-4.fc34.x86_64 lua-srpm-macros-1-4.fc34.noarch
m4-1.4.18-17.fc34.x86_64 nim-srpm-macros-3-4.fc34.noarch nvidia-settings-3:470.57.02-1.fc34.x86_64
ocaml-srpm-macros-6-4.fc34.noarch openblas-srpm-macros-2-9.fc34.noarch openssl-devel-1:1.1.1k-1.fc34.x86_64
patch-2.7.6-14.fc34.x86_64 perl-srpm-macros-1-39.fc34.noarch python-srpm-macros-3.9-38.fc34.noarch
python3-cached_property-1.5.2-2.fc34.noarch python3-gssapi-1.6.9-3.fc34.x86_64 python3-koji-1.25.1-1.fc34.noarch
python3-progressbar2-3.52.1-2.fc34.noarch python3-pygit2-1.5.0-1.fc34.x86_64 python3-requests-gssapi-1.2.3-1.fc34.noarch
python3-rpmautospec-0.2.5-1.fc34.noarch python3-utils-2.4.0-3.fc34.noarch qt5-srpm-macros-5.15.2-2.fc34.noarch
redhat-rpm-config-183-1.fc34.noarch rpm-build-4.16.1.3-1.fc34.x86_64 rpmautospec-rpm-macros-0.2.5-1.fc34.noarch
rpmdevtools-9.5-1.fc34.noarch rust-srpm-macros-18-1.fc34.noarch xemacs-filesystem-21.5.34-39.20200331hge2ac728aa576.fc34.noarch
xorg-x11-drv-nvidia-3:470.57.02-1.fc34.x86_64 xorg-x11-drv-nvidia-kmodsrc-3:470.57.02-1.fc34.x86_64 xorg-x11-drv-nvidia-libs-3:470.57.02-1.fc34.x86_64
zlib-devel-1.2.11-26.fc34.x86_64 zstd-1.5.0-1.fc34.x86_64
I did see in the notes at the end of the copr page that it requires secure boot to be disabled. The system will not load the nvidia driver with secure boot enabled since the module is not signed and the kernel will prevent loading an unsigned module when secure boot is active.
I don’t recall if I said anything about secure boot before, but try disabling that and boot again.
I can’t test the driver on my newly installed VM, but I did install everything and it downloaded 55 packages so that is similar to what it did for you.
Secure boot is disabled. I checked it over and over again.
And till the last kernel changes my Nvidia drivers were running good and without problems. It all began after the last kernel changes.