[solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

archelvetic · 2021-08-05 10:55:26

My gpu doesn't work anymore after updating[0].

Thus I restore the backup and everything works again.

Now having everything working again and looking at the packages with

payman -Su

, I only see

nvidia-settings

grepping the packages to be upgraded.

Therefore I'm understanding that

nvidia-settings

must be the root of all evil.

Yet, after

pacman -Su --ignore libxnvctrl-465.31-1 nvidia-settings

(and confirming it's not listed on the packages to update anymore) I'm again facing the same problem[0].

So, how to update and have nvidia left out of this?

[0] nvidia-smi-> No devices found -> https://bbs.archlinux.org/viewtopic.php?id=268481

Last edited by archelvetic (2021-08-06 09:22:16)

V1del · 2021-08-05 11:49:22

nvidia-settings is optional and basically irrelevant to any actual usage

It is extremely unlikely to be related to your issue, unless you have a login script that explicitly invokes it. Which other packages are part of that update?

archelvetic · 2021-08-05 12:52:15

sudo pacman -Su

Packages (195) alembic-1.8.2-1  alsa-card-profiles-1:0.3.30-1  alsa-ucm-conf-1.2.5-2  alsa-utils-1.2.5-3  aom-3.1.1-1  blender-17:2.93.0-4  boost-1.76.0-1  boost-libs-1.76.0-1  ca-certificates-20210603-1
               ca-certificates-utils-20210603-1  enchant-2.2.15-3  ffmpeg-2:4.4-4  ghc-libs-8.10.5-1  git-2.32.0-1  gnuplot-5.4.2-1  gptfdisk-1.0.8-1  haskell-aeson-1.5.6.0-31  haskell-aeson-pretty-0.8.8-169
               haskell-ansi-terminal-0.11-128  haskell-appar-0.1.8-11  haskell-asn1-encoding-0.9.6-116  haskell-asn1-parse-0.9.5-116  haskell-asn1-types-0.3.4-95  haskell-assoc-1.0.2-84  haskell-async-2.2.3-45
               haskell-attoparsec-0.14.1-15  haskell-base-compat-0.11.2-5  haskell-base-compat-batteries-0.11.2-122  haskell-base-orphans-0.8.4-109  haskell-base16-bytestring-1.0.1.0-52
               haskell-base64-bytestring-1.2.0.1-59  haskell-basement-0.0.12-2  haskell-bifunctors-5.5.11-9  haskell-bitarray-0.0.1.1-21  haskell-blaze-builder-0.4.2.1-3  haskell-blaze-html-0.9.1.2-115
               haskell-blaze-markup-0.8.2.8-21  haskell-byteable-0.1.1-23  haskell-byteorder-1.0.4-22  haskell-case-insensitive-1.2.1.0-95  haskell-cereal-0.5.8.1-10  haskell-citeproc-0.4-4
               haskell-cmark-gfm-0.2.2-71  haskell-cmdargs-0.10.21-2  haskell-colour-2.3.5-186  haskell-commonmark-0.2-4  haskell-commonmark-extensions-0.2.1.2-3  haskell-commonmark-pandoc-0.2.1-6
               haskell-comonad-5.0.8-88  haskell-conduit-1.3.4.1-24  haskell-conduit-extra-1.3.5-138  haskell-connection-0.3.1-128  haskell-cookie-0.4.5-10  haskell-cryptonite-0.29-9
               haskell-data-default-0.7.1.1-151  haskell-data-default-class-0.1.2.0-22  haskell-data-default-instances-containers-0.0.1-34  haskell-data-default-instances-dlist-0.0.1-164
               haskell-data-default-instances-old-locale-0.0.1-34  haskell-data-fix-0.3.1-31  haskell-digest-0.0.1.2-23  haskell-distributive-0.6.2.1-92  haskell-dlist-1.0-87  haskell-doclayout-0.3.0.2-17
               haskell-doctemplates-0.9-56  haskell-emojis-0.1-103  haskell-erf-2.0.0.0-22  haskell-errors-2.3.0-60  haskell-file-embed-0.0.14.0-2  haskell-glob-0.10.1-84  haskell-haddock-library-1.10.0-27
               haskell-hashable-1.3.2.0-9  haskell-hourglass-0.2.12-138  haskell-hslua-1.3.0.1-42  haskell-hslua-module-path-0.1.0.1-27  haskell-hslua-module-system-0.2.2.1-98
               haskell-hslua-module-text-0.3.0.1-86  haskell-hsyaml-0.2.1.0-108  haskell-http-4000.3.16-32  haskell-http-client-0.7.8-14  haskell-http-client-tls-0.3.5.3-426  haskell-http-types-0.12.3-166
               haskell-hxt-9.3.1.22-16  haskell-hxt-charproperties-9.5.0.0-2  haskell-hxt-regex-xmlschema-9.2.0.7-3  haskell-hxt-unicode-9.0.2.4-23  haskell-indexed-traversable-0.1.1-4
               haskell-integer-logarithms-1.0.3.1-4  haskell-iproute-1.7.11-18  haskell-ipynb-0.1.0.1-209  haskell-jira-wiki-markup-1.4.0-4  haskell-juicypixels-3.3.5-94  haskell-lexer-1.1-9
               haskell-memory-0.16.0-9  haskell-mime-types-0.1.0.9-12  haskell-mono-traversable-1.0.15.1-137  haskell-network-3.1.2.1-57  haskell-network-uri-2.6.4.1-25  haskell-old-locale-1.0.0.7-28
               haskell-old-time-1.1.0.3-28  haskell-pandoc-types-1.22-87  haskell-pem-0.2.4-172  haskell-pretty-show-1.10-9  haskell-primitive-0.7.1.0-137  haskell-quickcheck-2.14.2-173  haskell-random-1.2.0-230
               haskell-regex-base-0.94.0.1-2  haskell-regex-pcre-0.95.0.0-10  haskell-resourcet-1.2.4.2-87  haskell-safe-0.3.19-6  haskell-scientific-0.3.7.0-5  haskell-sha-1.6.4.4-17
               haskell-skylighting-0.10.5.1-20  haskell-skylighting-core-0.10.5.1-20  haskell-socks-0.6.1-116  haskell-split-0.2.3.4-137  haskell-splitmix-0.1.0.3-117  haskell-streaming-commons-0.2.2.1-85
               haskell-strict-0.4.0.1-58  haskell-syb-0.7.2.1-56  haskell-tagged-0.8.6.1-3  haskell-tagsoup-0.14.8-118  haskell-temporary-1.3-269  haskell-texmath-0.12.3-11  haskell-text-conversions-0.3.1-73
               haskell-text-icu-0.7.1.0-2  haskell-th-abstraction-0.4.2.0-3  haskell-th-compat-0.1.2-18  haskell-th-lift-0.8.2-8  haskell-th-lift-instances-0.1.18-57  haskell-these-1.1.1.1-85
               haskell-time-compat-1.9.6-5  haskell-time-locale-compat-0.1.1.5-15  haskell-tls-1.5.5-32  haskell-transformers-compat-0.6.6-4  haskell-typed-process-0.2.6.0-119  haskell-unicode-collation-0.1.3-4
               haskell-unicode-transforms-0.3.7.1-69  haskell-uniplate-1.6.13-62  haskell-unliftio-core-0.2.0.1-7  haskell-unordered-containers-0.2.14.0-5  haskell-utf8-string-1.0.2-42
               haskell-uuid-types-1.0.5-9  haskell-vector-0.12.3.0-21  haskell-vector-algorithms-0.8.0.4-57  haskell-x509-1.7.5-152  haskell-x509-store-1.6.7-151  haskell-x509-system-1.6.6-219
               haskell-x509-validation-1.6.11-151  haskell-xml-1.3.14-28  haskell-xml-conduit-1.9.1.1-12  haskell-xml-types-0.3.8-6  haskell-zip-archive-0.4.1-123  haskell-zlib-0.6.2.3-25  imath-3.0.4-2
               inetutils-2.0-1  inkscape-1.1-4  isync-1.4.2-1  ledger-3.2.1-7  libcap-2.50-2  libcerf-1:1.15-1  libcmis-0.5.2-8  libedit-20210522_3.1-1  libgusb-0.3.7-1  libixion-0.16.1-6  liblouis-3.18.0-1
               liborcus-0.16.1-7  libp11-kit-0.24.0-1  libreoffice-fresh-7.1.4-1  libx11-1.7.2-1  libxcrypt-4.4.22-4  libxnvctrl-465.31-1  lyx-2.3.6.1-2  mpg123-1.28.0-1  nettle-3.7.3-1  nvidia-settings-465.31-1
               openimageio-2.2.15.1-3  openmpi-4.1.1-1  openshadinglanguage-1.11.14.0-3  openvdb-8.0.1-5  p11-kit-0.24.0-1  pambase-20210605-2  pandoc-2.14.0.1-2  perl-datetime-locale-1.32-1  perl-json-xs-4.03-1
               python-hypothesis-6.14.0-1  python-queuelib-1.6.1-1  python-websockets-9.1-1  sqlitebrowser-3.12.2-1  supercollider-3.11.2-4  terraform-1.0.0-1  util-linux-2.37-3  util-linux-libs-2.37-3
               xkeyboard-config-2.33-1  xterm-368-1

uname -r (working state)

5.10.42-1-lts

sudo pacman -Qs nvidia

local/cuda 11.3.0-2
    NVIDIA's GPU programming toolkit
local/egl-wayland 1.1.7-1
    EGLStream-based Wayland external platform
local/libvdpau 1.4-1
    Nvidia VDPAU library
local/libxnvctrl 465.27-1
    NVIDIA NV-CONTROL X extension
local/nvidia-lts 1:465.31-6
    NVIDIA drivers for linux-lts
local/nvidia-prime 1.0-4
    NVIDIA Prime Render Offload configuration and utilities
local/nvidia-settings 465.27-1
    Tool for configuring the NVIDIA graphics driver
local/nvidia-utils 465.31-1
    NVIDIA drivers utilities
local/opencl-nvidia 465.31-1
    OpenCL implemention for NVIDIA

I've appended

libxnvctrl-465.31-1

to my inital question which I forgot to mention

Last edited by archelvetic (2021-08-05 12:56:56)

seth · 2021-08-05 14:40:38

The kernel module is tied to the kernel version.
Downgrading nvidia or nvidia-lts doen't work, you've to use the 465xx version of nvidia-dkms.

archelvetic · 2021-08-05 15:19:28

@seth

At what point can I be sure the system works on dkms?

sudo pacman -U https://archive.archlinux.org/packages/n/nvidia-dkms/nvidia-dkms-465.31-1-x86_64.pkg.tar.zst

loading packages...
warning: nvidia-dkms-465.31-1 is up to date -- reinstalling
resolving dependencies...
looking for conflicting packages...

Packages (1) nvidia-dkms-465.31-1

Total Installed Size:  46.71 MiB
Net Upgrade Size:       0.00 MiB

:: Proceed with installation? [Y/n] y
(1/1) checking keys in keyring                                  [##################################] 100%(1/1) checking package integrity                                [##################################] 100%(1/1) loading package files                                     [##################################] 100%(1/1) checking for file conflicts                               [##################################] 100%(1/1) checking available disk space                             [##################################] 100%:: Running pre-transaction hooks...
(1/1) Remove upgraded DKMS modules
==> Unable to remove module nvidia/465.31 for kernel *: Not found in dkms status output.
:: Processing package changes...
(1/1) reinstalling nvidia-dkms                                  [##################################] 100%:: Running post-transaction hooks...
(1/2) Arming ConditionNeedsUpdate...
(2/2) Install DKMS modules
==> Unable to install module nvidia/465.31 for kernel *: Missing kernel headers.

Yet

pacman -Qs nvidia

lists it

local/nvidia-dkms 465.31-1
    NVIDIA drivers - module sources

However

nvidia-smi

doesn't contain lts or dkms but only 465.31

I have to admin I'm not grasping the dkms concept

seth · 2021-08-05 15:21:21

==> Unable to install module nvidia/465.31 for kernel *: Missing kernel headers.

https://archlinux.org/packages/core/x86 … x-headers/
https://archlinux.org/packages/core/x86 … s-headers/
https://wiki.archlinux.org/title/Dynami … le_Support

archelvetic · 2021-08-05 15:51:38

sudo pacman -U https://archive.archlinux.org/packages/l/linux-lts-headers/linux-lts-headers-5.10.42-1-x86_64.pkg.tar.zst 
sudo pacman -U https://archive.archlinux.org/packages/n/nvidia-dkms/nvidia-dkms-465.31-1-x86_64.pkg.tar.zst
sudo pacman -R nvidia-lts

sudo dkms status

nvidia, 465.31, 5.10.42-1-lts, x86_64: installed (original_module exists) (WARNI
NG! Diff between built and installed module!) (WARNING! Diff between built and i
nstalled module!) (WARNING! Diff between built and installed module!)
v4l2loopback, 0.12.5.r226.g81b8df7, 5.10.42-1-lts, x86_64: installed
v4l2loopback-dc, 1.7.3, 5.10.42-1-lts, x86_64: installed

Reboot

nvidia-smi

NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.

seth · 2021-08-05 15:54:39

original_module exists

get rid of the nvidia and nvidia-lts packages first, *then* trigger a dkms rebuild…

archelvetic · 2021-08-06 09:22:27

Solved. Got rid of all nvidia packages listed in pacman -Qs nvidia (even cuda) and reinstalled nvidia-dkms from the archive.

Afterwards I run

sudo pacman -Su --ignore nvidia-dkms,nvidia-utils

and tada: Updated machine and gpu work smoothly again.

Thank you so much for holding my hand during that journey

Arch Linux

#1 2021-08-05 10:55:26

[solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

#2 2021-08-05 11:49:22

Re: [solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

#3 2021-08-05 12:52:15

Re: [solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

#4 2021-08-05 14:40:38

Re: [solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

#5 2021-08-05 15:19:28

Re: [solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

#6 2021-08-05 15:21:21

Re: [solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

#7 2021-08-05 15:51:38

Re: [solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

#8 2021-08-05 15:54:39

Re: [solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

#9 2021-08-06 09:22:27

Re: [solved] Why does ignoring nvidia in pacman still cause gpu errors@upd

Board footer