Yesterday everything was fine, today after kernel upgrade I just got a black screen. I've pinned the working deployment using https://www.projectatomic.io/blog/2018/05/pinning-deployments-ostree-based-systems/ as a guide, and here it is how it looks now:
State: idle AutomaticUpdates: stage; rpm-ostreed-automatic.timer: no runs since boot Deployments: ostree://fedora-workstation:fedora/29/x86_64/silverblue Version: 29.20181219.0 (2018-12-19T01:09:40Z) BaseCommit: 1850037a080228ae8bea8b4e01f971988c0e3a6b1caebffb555a231c9a3ce5a6 ├─ repo-0 (2018-12-19T00:37:19Z) └─ repo-1 (2018-10-24T22:20:15Z) Commit: 93de9f02951ff781349c8c6038429fd6e78185c7f04a11376bacb655a1e3f8d8 ├─ fedora (2018-10-24T22:20:15Z) └─ updates (2018-12-19T02:19:55Z) Staged: no StateRoot: fedora-workstation GPGSignature: 1 signature Signature made Wed 19 Dec 2018 01:09:51 GMT using RSA key ID A20AA56B429476B4 Good signature from "Fedora 29 <fedora-29@fedoraproject.org>" LayeredPackages: ansible bcc-tools chromium dconf-editor evince evolution fedora-toolbox gmp-devel gnome-photos gnome-tweak-tool kernel-tools latencytop libreoffice mesa-vulkan-drivers.i686 mozilla-ublock-origin neovim ocaml ocamldoc opam pass perf perl-Pod-Html python3-psutil ripgrep stow tig weston zsh Initramfs: regenerate ostree://fedora-workstation:fedora/29/x86_64/silverblue Version: 29.20181219.0 (2018-12-19T01:09:40Z) BaseCommit: 1850037a080228ae8bea8b4e01f971988c0e3a6b1caebffb555a231c9a3ce5a6 ├─ repo-0 (2018-12-19T00:37:19Z) └─ repo-1 (2018-10-24T22:20:15Z) Commit: ed30b4b96d2338bc7f1589e01acad4f7caca055cb4bf96ac8c490528bad2a273 ├─ fedora (2018-10-24T22:20:15Z) └─ updates (2018-12-19T02:19:55Z) StateRoot: fedora-workstation GPGSignature: 1 signature Signature made Wed 19 Dec 2018 01:09:51 GMT using RSA key ID A20AA56B429476B4 Good signature from "Fedora 29 <fedora-29@fedoraproject.org>" LayeredPackages: ansible bcc-tools chromium dconf-editor evince evolution fedora-toolbox gmp-devel gnome-photos gnome-tweak-tool kernel-tools latencytop libreoffice mesa-vulkan-drivers.i686 mozilla-ublock-origin neovim ocaml ocamldoc opam pass perf perl-Pod-Html python3-psutil ripgrep stow tig weston zsh ● ostree://fedora-workstation:fedora/29/x86_64/silverblue Version: 29.20181217.0 (2018-12-17T01:23:13Z) BaseCommit: 967d0126d5c34913e2f1d1168bfce745e0f8f02ca7f35473a406abf196291e14 ├─ repo-0 (2018-12-17T00:51:47Z) └─ repo-1 (2018-10-24T22:20:15Z) Commit: 377ba4c80d3f1db706358dfdee84e0c0156d96f07ef6311d237268f52a6f130e ├─ fedora (2018-10-24T22:20:15Z) └─ updates (2018-12-17T18:57:41Z) StateRoot: fedora-workstation GPGSignature: 1 signature Signature made Mon 17 Dec 2018 01:23:28 GMT using RSA key ID A20AA56B429476B4 Good signature from "Fedora 29 <fedora-29@fedoraproject.org>" LayeredPackages: ansible bcc-tools chromium dconf-editor evince evolution fedora-toolbox gmp-devel gnome-photos gnome-tweak-tool kernel-tools latencytop libreoffice mesa-vulkan-drivers.i686 mozilla-ublock-origin neovim ocaml ocamldoc opam pass perf perl-Pod-Html python3-psutil ripgrep stow tig weston zsh Pinned: yes
I tried enabling initramfs regeneration, but it didn't help: the file is actually missing from the filesystem:
Dec 19 19:56:06 localhost kernel: amdgpu 0000:01:00.0: Direct firmware load for amdgpu/polaris11_k_mc.bin failed with error -2 Dec 19 20:31:25 localhost kernel: amdgpu 0000:01:00.0: Direct firmware load for amdgpu/polaris11_k_mc.bin failed with error -2 Dec 19 20:46:09 localhost kernel: amdgpu 0000:01:00.0: Direct firmware load for amdgpu/polaris11_k_mc.bin failed with error -2 Dec 19 20:54:48 localhost kernel: amdgpu 0000:01:00.0: Direct firmware load for amdgpu/polaris11_k_mc.bin failed with error -2
There is only this file (which the current kernel was happy with): /usr/lib/firmware/amdgpu/polaris11_k_smc.bin
Upstream bugs: https://bugzilla.redhat.com/show_bug.cgi?id=1661024 https://bugzilla.redhat.com/show_bug.cgi?id=1659810 (this already existed prior to pushing the broken kernel to stable) https://github.com/fedora-infra/bodhi/issues/2874
Feel free to close this bug, unless you think rpm-ostree could've done some sanity checks on the presence of these files prior to rebooting?
I'm guessing this is a kernel bug. Here are the changes since 29.20181217.0:
29.20181217.0
'29.20181219.0' ostree diff commit old: onerepo:fedora/29/x86_64/silverblue^ (ccbb52c88cea9c34a9c2d1cff4cb6712ce5dda0486ee561e76eacfd1195d0ad2) ostree diff commit new: onerepo:fedora/29/x86_64/silverblue (1850037a080228ae8bea8b4e01f971988c0e3a6b1caebffb555a231c9a3ce5a6) Upgraded: libgcc 8.2.1-5.fc29 -> 8.2.1-6.fc29 libgomp 8.2.1-5.fc29 -> 8.2.1-6.fc29 libstdc++ 8.2.1-5.fc29 -> 8.2.1-6.fc29 webkit2gtk3 2.22.4-1.fc29 -> 2.22.5-1.fc29 webkit2gtk3-jsc 2.22.4-1.fc29 -> 2.22.5-1.fc29 webkit2gtk3-plugin-process-gtk2 2.22.4-1.fc29 -> 2.22.5-1.fc29 xorg-x11-server-Xorg 1.20.3-1.fc29 -> 1.20.3-2.fc29 xorg-x11-server-Xwayland 1.20.3-1.fc29 -> 1.20.3-2.fc29 xorg-x11-server-common 1.20.3-1.fc29 -> 1.20.3-2.fc29 '29.20181218.0' ostree diff commit old: onerepo:fedora/29/x86_64/silverblue^^ (967d0126d5c34913e2f1d1168bfce745e0f8f02ca7f35473a406abf196291e14) ostree diff commit new: onerepo:fedora/29/x86_64/silverblue^ (ccbb52c88cea9c34a9c2d1cff4cb6712ce5dda0486ee561e76eacfd1195d0ad2) Upgraded: flatpak 1.0.6-3.fc29 -> 1.0.6-4.fc29 flatpak-libs 1.0.6-3.fc29 -> 1.0.6-4.fc29 gnutls 3.6.4-4.fc29 -> 3.6.5-1.fc29 kernel 4.19.8-300.fc29 -> 4.19.9-300.fc29 kernel-core 4.19.8-300.fc29 -> 4.19.9-300.fc29 kernel-modules 4.19.8-300.fc29 -> 4.19.9-300.fc29 kernel-modules-extra 4.19.8-300.fc29 -> 4.19.9-300.fc29 nautilus 3.30.4-1.fc29 -> 3.30.5-1.fc29 nautilus-extensions 3.30.4-1.fc29 -> 3.30.5-1.fc29 osinfo-db 20181116-1.fc29 -> 20181214-1.fc29 podman 1:0.12.1.1-1.git66d3499.fc29 -> 1:0.12.1.2-1.git9551f6b.fc29 rpm-ostree 2018.9-3.fc29.1 -> 2018.10-1.fc29 rpm-ostree-libs 2018.9-3.fc29.1 -> 2018.10-1.fc29
so the linux-firmware package, which provides /usr/lib/firmware/amdgpu/polaris11_k_smc.bin hasn't changed, but the kernel has and you are seeing this bug.
linux-firmware
/usr/lib/firmware/amdgpu/polaris11_k_smc.bin
Can you open a bug against the kernel?
actually, bug already exists.. looks like we need to get a new firmware package into stable:
please follow: https://bugzilla.redhat.com/show_bug.cgi?id=1659810
at least your old deployment works fine :)
Metadata Update from @dustymabe: - Issue status updated to: Closed (was: Open)
Login to comment on this ticket.