Remove proprietary nvidia driver and revert back to nouveau

Kernels & Hardware, configuring network, installing services

Remove proprietary nvidia driver and revert back to nouveau

Postby tuxxie » 2017-12-16 12:15

I've downloaded and run the official nvidia driver installer "NVIDIA-Linux-x86_64-384.98.run", so now the Kernel driver in use is nvidia for my graphics card.

I would like to remove it and revert to nouveau like a fresh install..
tuxxie
 
Posts: 21
Joined: 2016-01-03 14:04

Re: Remove proprietary nvidia driver and revert back to nouv

Postby bw123 » 2017-12-16 12:39

You might want to try this, and if it doesn't work post back where it fails?

https://wiki.debian.org/NvidiaGraphicsD ... of_failure


hmm, I see you installed something from nvidia, I think it might be better to look there at the website where you downloaded the installer for a README first actually. http://www.nvidia.com/Download/driverRe ... 6577/en-us The page relies on javascript for the navigation so I can't search it for you.

good luck
User avatar
bw123
 
Posts: 3189
Joined: 2011-05-09 06:02
Location: TN_USA

Re: Remove proprietary nvidia driver and revert back to nouv

Postby tuxxie » 2017-12-16 13:01

It was easier than expected. What you have to do is:

  • Login to a terminal (for example hit Control+Alt+F2)
  • View desktop managers in use:
    Code: Select all
    sudo pgrep dm -l
  • Stop the X server and desktop environment. In my case "service sddm stop" because I'm using KDE PLASMA. For other graphical environments:
    Code: Select all
    sudo service sddm stop
    sudo service gdm stop
    sudo service lightdm stop
    sudo service kdm stop
    sudo service mdm stop
    sudo stop lightdm
    sudo /etc/init.d/gdm3 stop
  • Run the nvidia proprietary installer again with the --uninstall argument.
  • Now the /etc/X11/xorg.conf should be reverted to the original file, and if not just delete it. The nvidia uninstaller will prompt to do it for you.

Note: the usual "apt-get purge nvidia" does not work here because the official nvidia installer does not install packages, it just modifies the filesystem directly.

Code: Select all
./NVIDIA-Linux-x86_64-384.98.run --uninstall


After rebooting the kernel will load nouveau driver instead of nvidia.

More information:
Code: Select all
./NVIDIA-Linux-x86_64-384.98.run --help
./NVIDIA-Linux-x86_64-384.98.run --advanced-options

Code: Select all
This program will install the NVIDIA Accelerated Graphics Driver for
Linux-x86_64 384.98 by unpacking the embedded tarball and executing
the ./nvidia-installer  installation utility.

--info
  Print embedded info (title, default target directory) and exit.

--lsm
  Print embedded lsm entry (or no LSM) and exit.

--pkg-history
  Print the package history of this file and exit.

--list
  Print the list of files in the archive and exit.

--check
  Check integrity of the archive and exit.

-x, --extract-only
  Extract the contents of ../Descargas/NVIDIA-Linux-x86_64-384.98.run, but do not
  run 'nvidia-installer'.

--add-this-kernel
  Build a precompiled kernel interface for the currently running
  kernel and repackage the .run file to include this newly built
  precompiled kernel interface.  The new .run file will be placed
  in the current directory and the string "-custom" appended
  to its name, unless already present, to distinguish it from the
  original .run file.

--apply-patch [Patch]
  Apply the patch 'Patch' to the kernel interface files included
  in the .run file, remove any precompiled kernel interfaces
  and then repackage the .run file.  The new .run file will be
  placed in the current directory and the string "-custom"
  appended to its name, unless already present, to distinguish it
  from the original .run file.

--keep
  Do not delete target directory when done.

--target [NewDirectory]
  Extract contents in 'NewDirectory'

--extract-decompress
  Extract the embedded decompression program to stdout


The following arguments will be passed on to the ./nvidia-installer
utility:

COMMON OPTIONS:

  -a, --accept-license
      Bypass the display and prompting for acceptance of the
      NVIDIA Software License Agreement.  By passing this option
      to nvidia-installer, you indicate that you have read and
      accept the License Agreement contained in the file
      'LICENSE' (in the top level directory of the driver
      package).

  -v, --version
      Print the nvidia-installer version and exit.

  -h, --help
      Print usage information for the common commandline options
      and exit.

  -A, --advanced-options
      Print usage information for the common commandline options
      as well as the advanced options, and then exit.


ADVANCED OPTIONS:

  -a, --accept-license
      Bypass the display and prompting for acceptance of the
      NVIDIA Software License Agreement.  By passing this option
      to nvidia-installer, you indicate that you have read and
      accept the License Agreement contained in the file
      'LICENSE' (in the top level directory of the driver
      package).

  -v, --version
      Print the nvidia-installer version and exit.

  -h, --help
      Print usage information for the common commandline options
      and exit.

  -A, --advanced-options
      Print usage information for the common commandline options
      as well as the advanced options, and then exit.

  -i, --driver-info
      Print information about the currently installed NVIDIA
      driver version.

  --uninstall
      Uninstall the currently installed NVIDIA driver.

  --skip-module-unload
      When uninstalling the driver, skip unloading of the NVIDIA
      kernel module. This option is ignored when the driver is
      being installed.

  --sanity
      Perform basic sanity tests on an existing NVIDIA driver
      installation.

  -e, --expert
      Enable 'expert' installation mode; more detailed questions
      will be asked, and more verbose output will be printed;
      intended for expert users.  The questions may be suppressed
      with the '--no-questions' commandline option.

  -q, --no-questions
      Do not ask any questions; the default (normally 'yes') is
      assumed for all yes/no questions, and the default string is
      assumed in any situation where the user is prompted for
      string input.  The one question that is not bypassed by
      this option is license acceptance; the license may be
      accepted with the commandline option '--accept-license'.

  -s, --silent
      Run silently; no questions are asked and no output is
      printed, except for error messages to stderr.  This option
      implies '--ui=none --no-questions --accept-license'.

  --x-prefix=X-PREFIX
      The prefix under which the X components of the NVIDIA
      driver will be installed; the default is '/usr/X11R6'
      unless nvidia-installer detects that X.Org >= 7.0 is
      installed, in which case the default is '/usr'.  Only under
      rare circumstances should this option be used.

  --xfree86-prefix=XFREE86-PREFIX
      This is a deprecated synonym for --x-prefix.

  --x-module-path=X-MODULE-PATH
      The path under which the NVIDIA X server modules will be
      installed.  If this option is not specified,
      nvidia-installer uses the following search order and
      selects the first valid directory it finds: 1) `X
      -showDefaultModulePath`, 2) `pkg-config
      --variable=moduledir xorg-server`, or 3) the X library path
      (see the '--x-library-path' option) plus either 'modules'
      (for X servers older than X.Org 7.0) or 'xorg/modules' (for
      X.Org 7.0 or later).

  --x-library-path=X-LIBRARY-PATH
      The path under which the NVIDIA X libraries will be
      installed.  If this option is not specified,
      nvidia-installer uses the following search order and
      selects the first valid directory it finds: 1) `X
      -showDefaultLibPath`, 2) `pkg-config --variable=libdir
      xorg-server`, or 3) the X prefix (see the '--x-prefix'
      option) plus 'lib' on 32bit systems, and either 'lib64' or
      'lib' on 64bit systems, depending on the installed Linux
      distribution.

  --x-sysconfig-path=X-SYSCONFIG-PATH
      The path under which X system configuration files will be
      installed.  If this option is not specified,
      nvidia-installer uses the following search order and
      selects the first valid directory it finds: 1) `pkg-config
      --variable=sysconfigdir xorg-server`, or 2)
      /usr/share/X11/xorg.conf.d.

  --opengl-prefix=OPENGL-PREFIX
      The prefix under which the OpenGL components of the NVIDIA
      driver will be installed; the default is: '/usr'.  Only
      under rare circumstances should this option be used.  The
      Linux OpenGL ABI
      (http://oss.sgi.com/projects/ogl-sample/ABI/) mandates this
      default value.

  --opengl-libdir=OPENGL-LIBDIR
      The path relative to the OpenGL library installation prefix
      under which the NVIDIA OpenGL components will be installed.
      The default is 'lib' on 32bit systems, and 'lib64' or 'lib'
      on 64bit systems, depending on the installed Linux
      distribution.  Only under very rare circumstances should
      this option be used.

  --installer-prefix=INSTALLER-PREFIX
      The prefix under which the installer binary will be
      installed; the default is: '/usr'.  Note: please use the
      '--utility-prefix' option instead.

  --utility-prefix=UTILITY-PREFIX
      The prefix under which the NVIDIA utilities
      (nvidia-installer, nvidia-settings, nvidia-xconfig,
      nvidia-bug-report.sh) and the NVIDIA utility libraries will
      be installed; the default is: '/usr'.

  --utility-libdir=UTILITY-LIBDIR
      The path relative to the utility installation prefix under
      which the NVIDIA utility libraries will be installed.  The
      default is 'lib' on 32bit systems, and 'lib64' or 'lib' on
      64bit systems, depending on the installed Linux
      distribution.

  --documentation-prefix=DOCUMENTATION-PREFIX
      The prefix under which the documentation files for the
      NVIDIA driver will be installed.  The default is: '/usr'.

  --application-profile-path=APPLICATION-PROFILE-PATH
      The directory under which default application profiles for
      the NVIDIA driver will be installed. The default is:
      '/usr/share/nvidia'.

  --kernel-include-path=KERNEL-INCLUDE-PATH
      The directory containing the kernel include files that
      should be used when compiling the NVIDIA kernel module.
      This option is deprecated; please use
      '--kernel-source-path' instead.

  --kernel-source-path=KERNEL-SOURCE-PATH
      The directory containing the kernel source files that
      should be used when compiling the NVIDIA kernel module.
      When not specified, the installer will use
      '/lib/modules/`uname -r`/build', if that directory exists.
      Otherwise, it will use '/usr/src/linux'.

  --kernel-output-path=KERNEL-OUTPUT-PATH
      The directory containing any KBUILD output files if either
      one of the 'KBUILD_OUTPUT' or 'O' parameters were passed to
      KBUILD when building the kernel image/modules.  When not
      specified, the installer will assume that no separate
      output directory was used.

  --kernel-install-path=KERNEL-INSTALL-PATH
      The directory in which the NVIDIA kernel module should be
      installed.  The default value is either
      '/lib/modules/`uname -r`/kernel/drivers/video' (if
      '/lib/modules/`uname -r`/kernel' exists) or
      '/lib/modules/`uname -r`/video'.

  --proc-mount-point=PROC-MOUNT-POINT
      The mount point for the proc file system; if not specified,
      then this value defaults to '/proc' (which is normally
      correct).  The mount point of the proc filesystem is needed
      because the contents of '<proc filesystem>/version' is used
      when identifying if a precompiled kernel interface is
      available for the currently running kernel.  This option
      should only be needed in very rare circumstances.

  --log-file-name=LOG-FILE-NAME
      File name of the installation log file (the default is:
      '/var/log/nvidia-installer.log').

  --tmpdir=TMPDIR
      Use the specified directory as a temporary directory when
      generating transient files used by the installer; if not
      given, then the following list will be searched, and the
      first one that exists will be used: /tmp, /tmp, .,
      /home/dalvik.

  --ui=UI
      Specify what user interface to use, if available.  Valid
      values for UI are 'ncurses' (the default) or 'none'. If the
      ncurses interface fails to initialize, or 'none' is
      specified, then a simple printf/scanf interface will be
      used.

  -c, --no-ncurses-color
      Disable use of color in the ncurses user interface.

  --opengl-headers
      Normally, installation will not install NVIDIA's OpenGL
      header files; the OpenGL header files packaged by the Linux
      distribution or available from
      http://www.opengl.org/registry/ should be preferred.
      However, http://www.opengl.org/registry/ does not yet
      provide a glx.h or gl.h.  Until that is resolved, NVIDIA's
      OpenGL header files can still be chosen, through this
      installer option.

  --no-nvidia-modprobe
      Skip installation of 'nvidia-modprobe', a setuid root
      utility which nvidia-installer installs by default.
      nvidia-modprobe can be used by user-space NVIDIA driver
      components to load the NVIDIA kernel module, and create the
      NVIDIA device files, when those components run without
      sufficient privileges to do so on their own, e.g., the CUDA
      driver run within the permissions of a non-privileged user.
      This utility is only needed if other means of loading the
      NVIDIA kernel module and creating the NVIDIA device files
      are unavailable.

  --force-tls=FORCE-TLS
      NVIDIA's OpenGL libraries are compiled with one of two
      different thread local storage (TLS) mechanisms: 'classic
      tls' which is used on systems with glibc 2.2 or older, and
      'new tls' which is used on systems with tls-enabled glibc
      2.3 or newer.  nvidia-installer will select the OpenGL
      libraries appropriate for your system; however, you may use
      this option to force the installer to install one library
      type or another.  Valid values for FORCE-TLS are 'new' and
      'classic'.

  -k KERNEL-NAME, --kernel-name=KERNEL-NAME
      Build and install the NVIDIA kernel module for the
      non-running kernel specified by KERNEL-NAME (KERNEL-NAME
      should be the output of `uname -r` when the target kernel
      is actually running).  This option implies
      '--no-precompiled-interface'.  If the options
      '--kernel-install-path' and '--kernel-source-path' are not
      given, then they will be inferred from KERNEL-NAME; eg:
      '/lib/modules/KERNEL-NAME/kernel/drivers/video/' and
      '/lib/modules/KERNEL-NAME/build/', respectively.

  -n, --no-precompiled-interface
      Disable use of precompiled kernel interfaces.

  --no-abi-note
      The NVIDIA OpenGL libraries contain an OS ABI note tag,
      which identifies the minimum kernel version needed to use
      the library.  This option causes the installer to remove
      this note from the OpenGL libraries during installation.

  --no-rpms
      Normally, the installer will check for several rpms that
      conflict with the driver (specifically: NVIDIA_GLX and
      NVIDIA_kernel), and remove them if present.  This option
      disables this check.

  -b, --no-backup
      During driver installation, conflicting files are backed
      up, so that they can be restored when the driver is
      uninstalled.  This option causes the installer to simply
      delete conflicting files, rather than back them up.

  -r, --no-recursion
      Normally, nvidia-installer will recursively search for
      potentially conflicting libraries under the default OpenGL
      and X server installation locations.  With this option set,
      the installer will only search in the top-level
      directories.

  -K, --kernel-module-only
      Install a kernel module only, and do not uninstall the
      existing driver.  This is intended to be used to install
      kernel modules for additional kernels (in cases where you
      might boot between several different kernels).  To use this
      option, you must already have a driver installed, and the
      version of the installed driver must match the version of
      this kernel module.

  --no-kernel-module
      Install everything but the kernel module, and do not remove
      any existing, possibly conflicting kernel modules.  This
      can be useful in some DEBUG environments.  If you use this
      option, you must be careful to ensure that a NVIDIA kernel
      module matching this driver version is installed
      seperately.

  --no-x-check
      Do not abort the installation if nvidia-installer detects
      that an X server is running.  Only under very rare
      circumstances should this option be used.

  --precompiled-kernel-interfaces-path=PRECOMPILED-KERNEL-INTERFA
  CES-PATH
      Before searching for a precompiled kernel interface in the
      .run file, search in the specified directory.

  -z, --no-nouveau-check
      Normally, nvidia-installer aborts installation if the
      nouveau kernel driver is in use.  Use this option to
      disable this check.

  -Z, --disable-nouveau
      If the nouveau kernel module is detected by
      nvidia-installer, the installer offers to attempt to
      disable nouveau. The default action is to not attempt to
      disable nouveau; use this option to change the default
      action to attempt to disable nouveau.

  -X, --run-nvidia-xconfig
      nvidia-installer can optionally invoke the nvidia-xconfig
      utility.  This will update the system X configuration file
      so that the NVIDIA X driver is used.  The pre-existing X
      configuration file will be backed up.  At the end of
      installation, nvidia-installer will ask the user if they
      wish to run nvidia-xconfig; the default response is 'no'.
      Use this option to make the default response 'yes'.  This
      is useful with the '--no-questions' or '--silent' options,
      which assume the default values for all questions.

  --force-selinux=FORCE-SELINUX
      Linux installations using SELinux (Security-Enhanced Linux)
      require that the security type of all shared libraries be
      set to 'shlib_t' or 'textrel_shlib_t', depending on the
      distribution. nvidia-installer will detect when to set the
      security type, and set it using chcon(1) on the shared
      libraries it installs.  If the execstack(8) system utility
      is present, nvidia-installer will use it to also clear the
      executable stack flag of the libraries.  Use this option to
      override nvidia-installer's detection of when to set the
      security type.  Valid values for FORCE-SELINUX are 'yes'
      (force setting of the security type), 'no' (prevent setting
      of the security type), and 'default' (let nvidia-installer
      decide when to set the security type).

  --selinux-chcon-type=SELINUX-CHCON-TYPE
      When SELinux support is enabled, nvidia-installer will try
      to determine which chcon argument to use by first trying
      'textrel_shlib_t', then 'texrel_shlib_t', then 'shlib_t'.
      Use this option to override this detection logic.

  --no-sigwinch-workaround
      Normally, nvidia-installer ignores the SIGWINCH signal
      before it forks to execute commands, e.g. to build the
      kernel module, and restores the SIGWINCH signal handler
      after the child process has terminated.  This option
      disables this behavior.

  --no-cc-version-check
      The NVIDIA kernel module should be compiled with the same
      compiler that was used to compile the currently running
      kernel. The layout of some Linux kernel data structures may
      be dependent on the version of gcc used to compile it. The
      Linux 2.6 kernel modules are tagged with information about
      the compiler and the Linux kernel's module loader performs
      a strict version match check. nvidia-installer checks for
      mismatches prior to building the NVIDIA kernel module and
      aborts the installation in case of failures. Use this
      option to override this check.

  --no-distro-scripts
      Normally, nvidia-installer will run scripts from
      /usr/lib/nvidia before and after installing or uninstalling
      the driver.  Use this option to disable execution of these
      scripts.

  --no-opengl-files
      Do not install any of the OpenGL-related driver files.

  --kernel-module-source-prefix=KERNEL-MODULE-SOURCE-PREFIX
      Specify a path where the source directory for the kernel
      module will be installed. Default: install source directory
      at /usr/src

  --kernel-module-source-dir=KERNEL-MODULE-SOURCE-DIR
      Specify the name of the directory where the kernel module
      sources will be installed. Default: directory name is
      nvidia-VERSION

  --no-kernel-module-source
      Skip installation of the kernel module source.

  --dkms
      nvidia-installer can optionally register the NVIDIA kernel
      module sources, if installed, with DKMS, then build and
      install a kernel module using the DKMS-registered sources.
      This will allow the DKMS infrastructure to automatically
      build a new kernel module when changing kernels.  During
      installation, if DKMS is detected, nvidia-installer will
      ask the user if they wish to register the module with DKMS;
      the default response is 'no'.  This option will bypass the
      detection of DKMS, and cause the installer to attempt a
      DKMS-based installation regardless of whether DKMS is
      present.

  --module-signing-secret-key=MODULE-SIGNING-SECRET-KEY
      Specify a path to a private key to use for signing the
      NVIDIA kernel module. The corresponding public key must
      also be provided.

  --module-signing-public-key=MODULE-SIGNING-PUBLIC-KEY
      Specify a path to a public key to use for verifying the
      signature of the NVIDIA kernel module. The corresponding
      private key must also be provided.

  --module-signing-script=MODULE-SIGNING-SCRIPT
      Specify a path to a program to use for signing the NVIDIA
      kernel module. The program will be called with the
      arguments: program-name <HASH> <PRIVATEKEY> <PUBLICKEY>
      <MODULE>; if the program returns an error status, it will
      be called again with the arguments: program-name
      <PRIVATEKEY> <PUBLICKEY> <MODULE>. Default: use the
      sign-file script in the kernel source directory.

  --module-signing-key-path=MODULE-SIGNING-KEY-PATH
      Specify a path where signing keys generated by
      nvidia-installer will be installed. Default: install keys
      to '/usr/share/nvidia'.

  --module-signing-hash=MODULE-SIGNING-HASH
      Specify a cryptographic hash algorithm to use for signing
      kernel modules. This requires a module signing tool that
      allows explicit selection of the hash algorithm, and the
      hash algorithm name must be recognizable by the module
      signing tool. Default: select a hash algorithm
      automatically, based on the kernel's configuration.

  --module-signing-x509-hash=MODULE-SIGNING-X509-HASH
      Specify a cryptographic hash algorithm to use for signing
      X.509 certificates generated by nvidia-installer. The hash
      algorithm name must be one of the message digest algorithms
      recognized by the x509(1) command.

  --no-check-for-alternate-installs
      Maintainers of alternate driver installation methods can
      report the presence and/or availability of an alternate
      driver installation to nvidia-installer. Setting this
      option skips the check for alternate driver installations.

  --no-unified-memory
      Do not install the NVIDIA Unified Memory kernel module.
      This kernel module is required for CUDA on 64-bit systems,
      and if it is not installed, the CUDA driver and CUDA
      applications will not be able to run. The
      '--no-unified-memory' option should only be used to work
      around failures to build or install the Unified Memory
      kernel module on systems that do not need to run CUDA.

  --no-drm
      Do not install the nvidia-drm kernel module. This kernel
      module provides several features, including X11
      autoconfiguration, support for PRIME, and DRM-KMS. The
      latter is used to support modesetting on windowing systems
      that run independently of X11. The '--no-drm' option should
      only be used to work around failures to build or install
      the nvidia-drm kernel module on systems that do not need
      these features.

  -j CONCURRENCY-LEVEL, --concurrency-level=CONCURRENCY-LEVEL
      Set the concurrency level for operations such as building
      the kernel module which may be parallelized on SMP systems.
      By default, this will be set to the number of detected
      CPUs, or to '1', if nvidia-installer fails to detect the
      number of CPUs. Systems with a large number of CPUs will
      have the default concurrency level limited to 32; setting a
      higher level on the command line will override this limit.

  --force-libglx-indirect
      If the package includes a libglvnd-based OpenGL library,
      then always install a libGLX_indirect.so.0 symlink,
      overwriting one if it exists.

  --no-libglx-indirect
      Do not install a libGLX_indirect.so.0 symlink.

  --install-libglvnd, --no-install-libglvnd
      If the package includes a libglvnd-based OpenGL library,
      then it will try to determine whether the libglvnd
      libraries are already available, and will install them if
      they're not. Use --install-libglvnd to always install the
      libglvnd libraries, overwriting any that already exist. Use
      --no-install-libglvnd to exclude the libglvnd libraries,
      even if they appear to be missing.

  --glvnd-glx-client, --no-glvnd-glx-client
      By default, the NVIDIA GLX driver will be installed with
      the new GLVND architecture, to support coexisting with
      other GLVND-compliant GLX drivers. However, some
      applications which do not conform to the Linux OpenGL ABI
      may not be fully compatible with a GLVND-based GLX driver.
      The --no-glvnd-glx-client option will select a non-GLVND
      GLX client library (libGL.so.1), which may help to avoid
      compatibility issues with such applications.

  --glvnd-egl-config-path=GLVND-EGL-CONFIG-PATH
      If the package includes a libglvnd-based EGL library, then
      install the EGL vendor library config file to this
      directory. If the libglvnd libraries are already present,
      then by default the installer will try to determine the
      path by running `pkg-config --variable=datadir libglvnd`.
      If that fails, then it will default to
      /usr/share/glvnd/egl_vendor.d.

  --glvnd-egl-client, --no-glvnd-egl-client
      By default, the NVIDIA EGL driver will be installed with
      the new GLVND architecture, to support coexisting with
      other GLVND-compliant EGL drivers. However, some
      applications may not be fully compatible with a GLVND-based
      EGL driver. The --no-glvnd-egl-client option will select a
      non-GLVND EGL client library, which may help to avoid
      compatibility issues with such applications.

  --egl-external-platform-config-path=EGL-EXTERNAL-PLATFORM-CONFI
  G-PATH
      If the package includes an EGL external platform library,
      then install the EGL external platform library config file
      to this directory. Defaults to
      /usr/share/egl/egl_external_platform.d.
tuxxie
 
Posts: 21
Joined: 2016-01-03 14:04

Re: Remove proprietary nvidia driver and revert back to nouv

Postby dilberts_left_nut » 2017-12-16 20:10

There is also this package available:
https://packages.debian.org/stretch/nvi ... er-cleanup
AdrianTM wrote:There's no hacker in my grandma...
User avatar
dilberts_left_nut
 
Posts: 4858
Joined: 2009-10-05 07:54
Location: enzed

Re: Remove proprietary nvidia driver and revert back to nouv

Postby tuxxie » 2017-12-16 21:25

dilberts_left_nut wrote:There is also this package available:
https://packages.debian.org/stretch/nvi ... er-cleanup

It says
Furthermore using the nvidia-installer to install a new version of the drivers is prevented as long as this package is installed

Is that right, does that even prevent the nvidia official installer from working properly?
tuxxie
 
Posts: 21
Joined: 2016-01-03 14:04

Re: Remove proprietary nvidia driver and revert back to nouv

Postby sunrat » 2017-12-18 22:47

tuxxie wrote:It says
Furthermore using the nvidia-installer to install a new version of the drivers is prevented as long as this package is installed

Is that right, does that even prevent the nvidia official installer from working properly?

Most likely it forces using the nvidia-driver package from Debian repo, which is the most reliable and best supported way to install anyway.
“ computer users can be divided into 2 categories:
Those who have lost data
...and those who have not lost data YET ”
Remember to BACKUP!
User avatar
sunrat
 
Posts: 2391
Joined: 2006-08-29 09:12
Location: Melbourne, Australia

Re: Remove proprietary nvidia driver and revert back to nouv

Postby tuxxie » 2017-12-19 05:28

sunrat wrote:Most likely it forces using the nvidia-driver package from Debian repo, which is the most reliable and best supported way to install anyway.

At least on ubuntu I tried getting 3 acceleration to work work the packages from the repositories, but no luck no matter how I tried. The only way to get nvidia proprietary drivers properly working was to use the .run installer.

Furthermore, I think that Nvidia knows things that open source developers don't.
tuxxie
 
Posts: 21
Joined: 2016-01-03 14:04

Re: Remove proprietary nvidia driver and revert back to nouv

Postby dilberts_left_nut » 2017-12-19 06:16

tuxxie wrote:Furthermore, I think that Nvidia knows things that open source developers don't.

And maybe open source developers care about things that nvidia doesn't.
AdrianTM wrote:There's no hacker in my grandma...
User avatar
dilberts_left_nut
 
Posts: 4858
Joined: 2009-10-05 07:54
Location: enzed

Re: Remove proprietary nvidia driver and revert back to nouv

Postby tuxxie » 2017-12-19 06:29

dilberts_left_nut wrote:
tuxxie wrote:Furthermore, I think that Nvidia knows things that open source developers don't.

And maybe open source developers care about things that nvidia doesn't.

I know probably there is some issue with the scripts in the debs.

Furthermore, when you download official drivers from nvidia, you choose which graphics card you have and you download the proper .run installer, which one of them may be common for several cards yet there may be variants of them. What apt does is just to install the same nvidia package for every single nvidia card which I'm not sure it's the right thing to do or that it will work flawlessly as expected.
Last edited by tuxxie on 2017-12-19 07:11, edited 1 time in total.
tuxxie
 
Posts: 21
Joined: 2016-01-03 14:04

Re: Remove proprietary nvidia driver and revert back to nouv

Postby dilberts_left_nut » 2017-12-19 07:08

You're still getting the same blob, either way (given the same version numbers ... because ... stable etc).
The proprietary nvidia drivers support the "current range" of cards - older ones are supported by the *-legacy packages (rather than d/l'ing the older iteration from nvidia when you pick that card).

The difference is that the Debian packaged versions are put together by maintainers that know Debian, to be compatible with the shipped kernels and graphics stack and to play nice with the rest of the system.
AdrianTM wrote:There's no hacker in my grandma...
User avatar
dilberts_left_nut
 
Posts: 4858
Joined: 2009-10-05 07:54
Location: enzed


Re: Remove proprietary nvidia driver and revert back to nouv

Postby bw123 » 2017-12-19 10:48

tuxxie wrote:[*]Stop the X server and desktop environment. In my case "service sddm stop" because I'm using KDE PLASMA. For other graphical environments:
Code: Select all
sudo service sddm stop
sudo service gdm stop
sudo service lightdm stop
sudo service kdm stop
sudo service mdm stop
sudo stop lightdm
sudo /etc/init.d/gdm3 stop



Debian seems to be using a display-manager.service now, so the way I've been doing this is"
Code: Select all
# systemctl stop display-manager


This would be good for people who don't know what dm they are using...
User avatar
bw123
 
Posts: 3189
Joined: 2011-05-09 06:02
Location: TN_USA

Re: Remove proprietary nvidia driver and revert back to nouv

Postby sunrat » 2017-12-19 11:27

tuxxie wrote:What apt does is just to install the same nvidia package for every single nvidia card which I'm not sure it's the right thing to do or that it will work flawlessly as expected.

There are 3 separate drivers in the repo to install nvidia. You can use the package nvidia-detect to detect and tell you which one is correct for your card.

The 'nvidia-detect' script in this package checks for an NVIDIA GPU in the system and recommends one of the non-free accelerated driver meta-packages (nvidia-driver, nvidia-legacy-340xx-driver, or nvidia-legacy-304xx-driver) for installation.
“ computer users can be divided into 2 categories:
Those who have lost data
...and those who have not lost data YET ”
Remember to BACKUP!
User avatar
sunrat
 
Posts: 2391
Joined: 2006-08-29 09:12
Location: Melbourne, Australia

Re: Remove proprietary nvidia driver and revert back to nouv

Postby tuxxie » 2018-05-01 12:57

Finally could get the nvidia propriertary driver working on debian buster/sid, directly from the nouveau, just adding the non-free repository and installing nvidia-driver:
NvidiaGraphicsDrivers

Code: Select all
01:00.0 VGA compatible controller: NVIDIA Corporation GP104M [GeForce GTX 1060] (rev a1)
        Subsystem: Micro-Star International Co., Ltd. [MSI] GP104M [GeForce GTX 1060 Mobile]
        Kernel driver in use: nvidia
        Kernel modules: nvidia
tuxxie
 
Posts: 21
Joined: 2016-01-03 14:04

Re: Remove proprietary nvidia driver and revert back to nouv

Postby debiman » 2018-05-01 15:11

tuxxie wrote:I think that Nvidia knows things about Nvidia hardware that open source developers don't.

fixed that for you.
User avatar
debiman
 
Posts: 2610
Joined: 2013-03-12 07:18


Return to System configuration

Who is online

Users browsing this forum: No registered users and 8 guests

fashionable