From 0a9ac0721f0ceebce537a4ba00854665f80c4b6e Mon Sep 17 00:00:00 2001 From: modeco80 Date: Sun, 29 Sep 2024 23:42:43 -0400 Subject: [PATCH] replace readme --- README.md | 37 +++++++++++-------------------------- 1 file changed, 11 insertions(+), 26 deletions(-) diff --git a/README.md b/README.md index 1b11d64..ea5cbf7 100644 --- a/README.md +++ b/README.md @@ -1,32 +1,17 @@ -# Nvidia vGPU-Merged Packages for Arch Linux -Use Nvidia vGPU _and_ GeForce functions of consumer-grade NVIDIA GPUs. +# `nvidia-merged` -This repository contains scripts for building all components of the "merged" driver package into drop-in replacements for the Arch Linux NVIDIA packages. It adds the following additional functions by means of scripts/drop-ins: - - Wrapper for `nvidia-smi` to support recognizing consumer GPUs as vGPU-capable - - Dependency re-ordering to ensure `libvirtd`, if installed, will start _after_ the NVIDIA vGPU services. +This is NVIDIA vGPU R16 horribly hacked to build and work on up to Linux 6.11. -Compatible Linux Versions: -- Various 5.11 or older releases -- 5.12 -- 5.14 except for some of the earlier 5.14 releases -- 5.15 +# Building -## Important -These packages are not guaranteed to work out of the box (or at all), so use it at your own risk. Backups should be taken before attempting to replace stock NVIDIA drivers on any system. +`makepkg` will build with linux 6.1x patches. -## Installation -1. Build and install desired packages -2. Edit `/etc/vgpu_unlock/profile_override.toml` and set desired overrides. -3. Enable `nvidia-vgpu-mgr.service` and `nvidia-vgpud.service` -4. Reboot. +To disable them, use `_use_new_patches="false" makepkg`. This is useful if you're not using a upstream kernel -It is recommended to at least install `nvidia-merged` and `nvidia-merged-settings`, which will include the required services/kernel module sources by dependency. +**NOTICE**: the vGPU VFIO module needs CONFIG_IOMMUFD=n to function properly. No Arch kernels are built with this config option set correctly, so you will have to build your own kernel. -## Credits -This project would not be possible without all those who came before me that wrote the software and figured out how to implement it: - - [DualCoder - vgpu_unlock](https://github.com/DualCoder/vgpu_unlock) - - [rupansh - twelve.patch](https://github.com/rupansh/vgpu_unlock_5.12) - - [HiFiPhile - cvgpu.c](https://gist.github.com/HiFiPhile/b3267ce1e93f15642ce3943db6e60776/) - - [Krutav Shah - vGPU_Unlock Wiki](https://docs.google.com/document/d/1pzrWJ9h-zANCtyqRgS7Vzla0Y8Ea2-5z2HEi4X75d2Q) - - [tuh8888 - libvirt_win10_vm](https://github.com/tuh8888/libvirt_win10_vm) - - [mbilker - vgpu_unlock-rs](https://github.com/mbilker/vgpu_unlock-rs) +# Thanks + +- Erin-allison originally created the PKGBUILD I based this on. +- Snowman for vGPU-Unlock-Patcher +- vgpu-unlock_rs