For the open driver, the `nvidia-uvm` module does not auto-load after
`nvidia`, which makes CUDA not work.
In this case, we need to add it to `boot.kernelModules` for it to work
again.
In my experience this is required to get Plasma 6 Wayland to work on
my GTX 1080 with the proprietary driver. Otherwise, I get errors in my
`journalctl --user` about "MESA-LOADER: failed to open simpledrm" (among
many other things) and Plasma just presents a black screen with a cursor
and nothing else.
I discovered this by searching the aforementioned error and found
[this Arch Linux forum post][0] and some further clicking around led
me to [this section on the ArchWiki][1], which is where I got the
`versionAtLeast` value from.
Tested with `nvidia-x11-555.58-6.9.7`.
[0]: https://bbs.archlinux.org/viewtopic.php?id=293741
[1]: https://wiki.archlinux.org/title/NVIDIA#DRM_kernel_mode_setting
- rename hardware.opengl to hardware.graphics
- remove hardware.opengl.driSupport, which does nothing
- remove hardware.opengl.setLdLibraryPath, which should never be done
- rename hardware.opengl.driSupport32Bit to hardware.graphics.enable32Bit
- lost of small docs / formatting cleanups
This library does not actually need to match the Nvidia driver version,
so we do not need to make it available impurely.
This reverts the following commits.
9b3461e7ae4e353b67f6
these changes were generated with nixq 0.0.2, by running
nixq ">> lib.mdDoc[remove] Argument[keep]" --batchmode nixos/**.nix
nixq ">> mdDoc[remove] Argument[keep]" --batchmode nixos/**.nix
nixq ">> Inherit >> mdDoc[remove]" --batchmode nixos/**.nix
two mentions of the mdDoc function remain in nixos/, both of which
are inside of comments.
Since lib.mdDoc is already defined as just id, this commit is a no-op as
far as Nix (and the built manual) is concerned.
The NVIDIA X driver uses a UNIX domain socket to pass information to
other driver components. If unable to connect to this socket, some
driver features, such as G-Sync, may not work correctly. The socket will
be bound to a file with a name unique to the X server instance created
in the directory specified by this option. Note that on Linux, an
additional abstract socket (not associated with a file) will also be
created, with this pathname socket serving as a fallback if connecting
to the abstract socket fails.
The default, which was in effect prior to this change, was `/var/run`.
The effect of not setting this option was that GDM X sessions
(and other non-root sessions) would see this warning in the log files:
```
(WW) NVIDIA: Failed to bind sideband socket to
(WW) NVIDIA: '/var/run/nvidia-xdriver-b4f69129' Permission denied
```
I don't see any security implications of turning this on universally,
since there already was an abstract socket created according to the
docs.
Documentation:
1. [NVIDIA X Config Options](https://download.nvidia.com/XFree86/Linux-x86_64/440.82/README/xconfigoptions.html#SidebandSocketPath)
Diagnosis:
1. [Arch Linux BBS post](https://bbs.archlinux.org/viewtopic.php?pid=1909115#p1909115)
For NVLink topology systems we need fabricmanager. Fabricmanager itself is
dependent on the datacenter driver set and not the regular x11 ones, it is also
tightly tied to the driver version. Furhtermore the current cudaPackages
defaults to version 11.8, which corresponds to the 520 datacenter drivers.
Future improvement should be to switch the main nvidia datacenter driver version
on the `config.cudaVersion` since these are well known from:
> https://docs.nvidia.com/deploy/cuda-compatibility/index.html#use-the-right-compat-package
This adds nixos configuration options `hardware.nvidia.datacenter.enable` and
`hardware.nvidia.datacenter.settings` (the settings configure fabricmanager)
Other interesting external links related to this commit are:
* Fabricmanager download site:
- https://developer.download.nvidia.com/compute/cuda/redist/fabricmanager/linux-x86_64/
* Data Center drivers:
- https://www.nvidia.com/Download/driverResults.aspx/193711/en-us/
Implementation specific details:
* Fabricmanager is added as a passthru package, similar to settings and
presistenced.
* Adds `use{Settings,Persistenced,Fabricmanager}` with defaults to preserve x11
expressions.
* Utilizes mkMerge to split the `hardware.nvidia` module into three comment
delimited sections:
1. Common
2. X11/xorg
3. Data Center
* Uses asserts to make the configurations mutualy exclusive.
Notes:
* Data Center Drivers are `x86_64` only.
* Reuses the `nvidia_x11` attribute in nixpkgs on enable, e.g. doesn't change it
to `nvidia_driver` and sets that to either `nvidia_x11` or `nvidia_dc`.
* Should have a helper function which is switched on `config.cudaVersion` like
`selectHighestVersion` but rather `selectCudaCompatibleVersion`.
the conversion procedure is simple:
- find all things that look like options, ie calls to either `mkOption`
or `lib.mkOption` that take an attrset. remember the attrset as the
option
- for all options, find a `description` attribute who's value is not a
call to `mdDoc` or `lib.mdDoc`
- textually convert the entire value of the attribute to MD with a few
simple regexes (the set from mdize-module.sh)
- if the change produced a change in the manual output, discard
- if the change kept the manual unchanged, add some text to the
description to make sure we've actually found an option. if the
manual changes this time, keep the converted description
this procedure converts 80% of nixos options to markdown. around 2000
options remain to be inspected, but most of those fail the "does not
change the manual output check": currently the MD conversion process
does not faithfully convert docbook tags like <code> and <package>, so
any option using such tags will not be converted at all.
