I have a bit of a chicken and an egg problem. In a test program I want to start a local dbus-daemon so the test program can run in isolation, but I end up doing this too late because the NVIDIA driver has already initialized the dbus library before I get to the point where I can easily start my own dbus-daemon and set DBUS_SESSION_BUS_ADDRESS in the environment. I can work around this, but it is going to be a little cumbersome, so I was hoping there was a way, e.g via an environment variable, to instruct the NVIDIA driver not to use dbus.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Unable to tweak `nvidia-settings` on remote server | 0 | 384 | August 31, 2023 | |
| Disabling Nouveau drivers to enable use of NVIDIA drivers | 3 | 1880 | October 30, 2019 | |
| Use integrated GPU for monitors, NVIDIA for cuda calculations (linux desktop) | 0 | 713 | June 3, 2021 | |
| `nvidia-dbus.conf ` lead to high security concerns | 10 | 2378 | August 5, 2022 | |
| Cannot install NVIDIA drivers on Ubuntu 16 | 1 | 5333 | June 13, 2016 | |
| Providing Nvidia drivers in a linux distribution | 0 | 930 | June 18, 2014 | |
| NVIDIA Driver Fails to Initialize GPUs on Ubuntu 22.04: "NVRM: This PCI I/O region assigned to your NVIDIA device is invalid" | 3 | 229 | January 28, 2025 | |
| You do not appear to have an NVIDIA GPU supported "Nvidia Geforce 940MX" | 17 | 22644 | October 14, 2021 | |
| driver setup for compute only use How to install the nvidia driver for compute-only use? | 6 | 5188 | January 27, 2010 | |
| forcing the driver to install | 2 | 6978 | May 18, 2008 |