Switch between Nvidia drivers without driver removal possible?

[SamKook]

The last proprietary nvidia driver update for my RTX3080 broke VS code and brave GUI somehow(it tries to render the window but can’t, gonna have to make a post somewhere to figure that one out), everything was fine with the previous one(580xx I think, which is sadly not available anymore and none of the other fix the issue), but it also works with the open driver at least.

My problem is that I end up switching between the 2 a lot since I sometime need the proprietary driver since it works better for games and the best way I found is to leave the open driver installed and install or remove the proprietary driver as needed with mhwd and reboot.

I’m assuming we can’t just reload the graphic driver on the fly and rebooting is unavoidable, but is there a way to leave both installed(the proprietary driver has no issue with the open driver remaining installed, at least none that I noticed) and simply choose which to use on the next reboot?

[Molski]

I’m using both Brave and visual-studio-code-bin, with an RTX 3080.

I may have other issues. I have a draft almost ready to post.

But these two apps, and all my other apps work fine.

And I’m using:

pacman -Qs nvidia-open
local/linux618-nvidia-open 590.48.01-9 (linux618-extramodules)
    NVIDIA open kernel modules for linux618

What is the output of your inxi -Gxxx ?

@SamKook

I’m assuming we can’t just reload the graphic driver on the fly and rebooting is unavoidable, but is there a way to leave both installed(the proprietary driver has no issue with the open driver remaining installed, at least none that I noticed) and simply choose which to use on the next reboot?

On the fly.. Nvidia. Maybe someday, if we all dream hard enough.

I have never tried to have multiple drivers installed. Does it make you uninstall? (And not just use the last installed one?)

The Manjaro helper scripts do a lot to make it easier. Work I’d rather not do, so I would just let it uninstall what’s needed, if that’s the case.

[SamKook]

I haven’t tried the open one, the many different nvidia driver are so confusing that I needed to do more research to figure out what is what, which I haven’t spent the time to do yet.
I’ve been using mhwd-nvidia for proprietary.

This is the output(I’m currently using video-linux)

Device-1: NVIDIA GA102 [GeForce RTX 3080] vendor: EVGA driver: nouveau
    v: kernel arch: Ampere pcie: speed: 16 GT/s lanes: 16 ports:
    active: DP-2,DP-3 empty: DP-1,HDMI-A-1 bus-ID: 0e:00.0 chip-ID: 10de:2206
    class-ID: 0300
  Display: wayland server: X.org v: 1.21.1.21 with: Xwayland v: 24.1.9
    compositor: kwin_wayland driver: X: loaded: modesetting
    alternate: fbdev,vesa dri: zink gpu: nouveau d-rect: 6000x2520
    display-ID: 0
  Monitor-1: DP-2 pos: top-right model: LG (GoldStar) ULTRAWIDE
    serial: 177759 res: mode: 2560x1080 hz: 100 scale: 100% (1) dpi: 97
    size: 673x284mm (26.5x11.18") diag: 730mm (28.8") modes: max: 2560x1080
    min: 640x480
  Monitor-2: DP-3 pos: bottom-l model: LG (GoldStar) 34GN850
    serial: 007NTFA54907 res: mode: 3440x1440 hz: 100 scale: 100% (1) dpi: 109
    size: 800x335mm (31.5x13.19") diag: 867mm (34.1") modes: max: 3440x1440
    min: 640x480
  API: EGL v: 1.5 platforms: device: 0 drv: zink device: 1 drv: swrast gbm:
    drv: zink surfaceless: drv: zink wayland: drv: zink x11: drv: zink
  API: OpenGL v: 4.6 compat-v: 4.5 vendor: mesa v: 25.3.4-arch1.1 glx-v: 1.4
    direct-render: yes renderer: zink Vulkan 1.4(NVIDIA GeForce RTX 3080 (NVK
    GA102) (MESA_NVK)) device-ID: 10de:2206 display-ID: :1.0
  API: Vulkan v: 1.4.335 layers: 6 surfaces: N/A device: 0
    type: discrete-gpu driver: mesa nvk device-ID: 10de:2206
  Info: Tools: api: clinfo, eglinfo, glxinfo, vulkaninfo
    de: kscreen-console,kscreen-doctor wl: wayland-info
    x11: xdpyinfo, xprop, xrandr

It doesn’t try to uninstall video-linux, it just gives priority to the proprietary driver so if that’s installed, that’s what it’s using.
That’s why I was wondering if it was possible to switch between both, they cohabit with no problem whatsoever as far as I can tell.

[Molski]

Our generation (and Turing aka RTX 2000 series) GPUs and newer use this package now. Nvidia is moving to open source drivers. Not to be confused with the Nouveau driver.

Everything older was dropped.

@Molski

local/linux618-nvidia-open 590.48.01-9 (linux618-extramodules)

It’s using their 590 series drivers (the latest).

And it’s what the Manjaro modified, but Nvidia supplied driver assistant said:

$ nvidia-driver-assistant
Detected GPUs:
  NVIDIA GeForce RTX 3080 - (pci_id 0x2216)

Detected system:
  Manjaro Linux 

Please copy and paste the following command to install the open kernel module flavour:
  sudo pacman -S linux618-nvidia-open

[SamKook]

Using the open one doesn’t fix my problem, those 2 apps are still unhappy with the gpu driver.

Device-1: NVIDIA GA102 [GeForce RTX 3080] vendor: EVGA driver: nvidia

vs code does a bunch of these:

[3880:0203/024408.098108:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[3880:0203/024408.098176:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[3634:0203/024408.313649:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[3634:0203/024408.313664:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 9 time(s)
[3634:0203/024408.313676:FATAL:content/browser/gpu/gpu_data_manager_impl_private.cc:415] GPU process isn't usable. Goodbye.

And brave:

[4348:4348:0203/024903.688361:ERROR:content/browser/gpu/gpu_process_host.cc:1004] GPU process exited unexpectedly: exit_code=139
[4348:4348:0203/024903.915367:ERROR:content/browser/gpu/gpu_process_host.cc:1004] GPU process exited unexpectedly: exit_code=139
[4348:4348:0203/024904.143365:ERROR:content/browser/gpu/gpu_process_host.cc:1004] GPU process exited unexpectedly: exit_code=139
[4348:4348:0203/024904.143390:FATAL:content/browser/gpu/gpu_data_manager_impl_private.cc:415] GPU process isn't usable. Goodbye.

Edit: Btw, I did try to uninstall the open source driver as well, in case that was causing incompatibilities but no change.

[SamKook]

I’m trying to diagnose my nvidia driver issue with the proprietary driver and since it doesn’t seem to be an issue with the driver itself since a previously working version doesn’t work anymore(meaning the issue happens with it), I wanted to try booting the live usb iso to test if that has the same issue. I tried both 26.0.1 and 26.0.2 kde version with the same result.

When I try to boot the live usb with the nonfree option, it display the boot screen, then goes back to text loading, flashes black a few times and stops doing anything.
Looking at dmesg, it says:

[   10.989952] [drm] Initialized nova 0.0.0 for NovaCore.nova-drm.0 on minor 0
[   13.477911] igb 0000:08:00.0 enp8s0: igb: enp8s0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[   50.921723] nvidia: loading out-of-tree module taints kernel.
[   50.921734] nvidia: module verification failed: signature and/or required key missing - tainting kernel
[   51.061896] nvidia-nvlink: Nvlink Core is being initialized, major device number 510
[   51.061903] NVRM: GPU 0000:0e:00.0 is already bound to NovaCore.
[   51.065149] NVRM: The NVIDIA probe routine was not called for 1 device(s).
[   51.065152] NVRM: This can occur when another driver was loaded and 
               NVRM: obtained ownership of the NVIDIA device(s).
[   51.065153] NVRM: Try unloading the conflicting kernel module (and/or
               NVRM: reconfigure your kernel without the conflicting
               NVRM: driver(s)), then try loading the NVIDIA kernel module
               NVRM: again.
[   51.065155] NVRM: No NVIDIA devices probed.
[   51.065691] nvidia-nvlink: Unregistered Nvlink Core, major device number 510

It tries about a dozen time and that’s where it stops.

Is that a known issue and is there any way to get around that and still load the proprietary driver with a live version? I’d rather not have to install an extra drive just for a simple test.

Edit: The XFCE iso has the same behavior.

[linux-aarhus]

Do you really need the proprietary driver to boot the system?

Depending on your general system composition and the GPU at hand you may get a better result using the free driver to get a working system, then continue to apply the proprietary driver.

[SamKook]

@SamKook

I’m trying to diagnose my nvidia driver issue with the proprietary driver

^ Yes, that’s the whole point of the test. And since it’s a live iso, there’s no way to switch the driver after it’s booted as far as I know.
If there is, that would be great and it’s why I’m asking:

@SamKook

is there any way to get around that

I know my system works with both the free and non-free driver once it’s installed on a drive, but some apps don’t work with the non-free driver and I need that driver for some games.
Everything was working perfectly fine a few updates ago so I’m trying to get back to that state.

[linux-aarhus]

I am sorry - my apology for not understanding the issue at hand.

@SamKook

I know my system works with both the free and non-free driver once it’s installed on a drive

If the system work with both once installed - then I don’t understand the urge to make it work with the live system.

I mean - it is known that the live ISO may not be optimally configured when using the proprietary driver (only Nvidia) is an issue - perhaps (very old AMD as well) - and that you work around by using the free GPU option.

@SamKook

some apps don’t work with the non-free driver and I need that driver for some games

Exactly why I say like I do - you cannot play games on the live ISO - it is not designed for that task - the live ISO is designed for a preview of the desktop and subsequent installation.

@SamKook

Everything was working perfectly fine a few updates ago so I’m trying to get back to that state.

That may not be possible unless you create a system snap-shot before applying an update.

moving forward

When your goal is to troubleshoot your system’s GPU drivers, there is a better way.

In the topic Basic Rescue and Recovery, it has been described how you can use either TTY or kernel cmdline to reach a console without the need for a live ISO.

Using one of the described methods (see quotes below), you can reach a running system and perform tasks required.

When troubleshooting a GPU issue I suggest you start with disabling the display manager, usually sddm but there is also the new kid plasmalogin which must be installed by hand if desired, so example is for sddm.

sudo systemctl disable --now sddm

When troubleshooting, having the system booting to a console is - at any measure - to be preferred over booting a live ISO.

You can start the plasma desktop by using one of two

• startplasma-wayland
• startplasma-x11 (only available with plasma-x11-session)

When you are ready - re-enable the display manager (example sddm)

sudo systemctl enable --now sddm

@linux-aarhus

Using TTY console

If you can reach the login screen and sign in, but the desktop never shows, you may still be able to reach a console using the graphical console switching mode. This mode is using a hotkey combination of CtrlAlt and a function key.

On most systems TTY1, TTY2 and/or TTY7 is already occupied so the usable keys will be in range from F3 to F6.

To test if you can reach a TTY try one of the function keys e.g.

CtrlAltF4

The system should react by switching to a text based screen with a prompt.

Enter your credentials to login.

From there you have full system access and can perform all sorts of changes to the system.

@linux-aarhus

Using kernel cmdline

If your issue is a non-functional display-manager, e.g. gdm, sddm, lightdm, manifesting in a missing login screen, you may still be able to reach a working console without the need for mounting and chroot.

Assuming you are using the GRUB boot-loader

• press and hold the Shift key during boot until the GRUB menu shows.
• Press e on the highlighted entry
• Use arrow keys to navigate to line starting with linux
• Press the end key and add a space and the number three 3
• Press F10 to continue booting

When you are at the console prompt - enter your credentials to login.

From there you have full system access and can perform all sorts of changes to the system.

[SamKook]

I’m trying to identify the cause of the issue.

Using a live iso with only the default packages installed will tell me if it’s a package issue or something else, that is the only reason I’m trying to run a live version with the non-free driver.
I only have to try and launch a chrome based browser to see if it works or not to get my answer.

My system is not completely broken, I can boot fine with either driver as I said. Only certain apps and games don’t work and I’m trying to determine if it’s a new package that is causing the issues or if it’s something else.

In short, it’s just a troubleshooting step, which is why I’m trying to save myself the trouble of installing a new drive and making a full install just to test starting 1 app.

[linux-aarhus]

@SamKook

Only certain apps and games don’t work and I’m trying to determine if it’s a new package that is causing the issues or if it’s something else.

When the apps in question (I looked at topic referenced in your initial issue) is based on chromium v8 (VS is running on top of Electron, Brave is a chromium based browser) you may need to add Electron cmdline flags.

It may not be related - plasma browser integration - look into the folder ~/.var/app/ and you will find browser configurations - I don’t know if it is related at all.

[musdus]

Hi @SamKook,

From your running system:

Diagnose why your Chromium-based apps aren’t working, please run the app from the terminal and share the output. This will help identify any errors related to GPU, libraries, or driver issues.

1. Open a terminal.

2. Launch your Chromium-based app (e.g., Chrome, Chromium, Brave, Vivaldi, your apps) by typing its command. For example:

   google-chrome-stable

3. Copy and paste the full output from the terminal here, especially any lines that start with error, failed, or libGL

   Look for any red flags or warnings under “Graphics Feature Status” and share those as well.

This will help us pinpoint if the issue is related to the NVIDIA driver, missing libraries, or something else.

[philm]

@SamKook you may want to try an older ISO with older Nvidia drivers and Xorg support: Manjaro 25.0 Zetar released

[SamKook]

Basically, those apps can’t get gpu access at all, the process crashes immediately, so the logs don’t tell much, which is why I was trying to do more testing to narrow down the possible causes.

@linux-aarhus

you may need to add Electron cmdline flags.

I checked the flags, tried some I hadn’t already tried before, but none seem to be for fixing it not running at all.

@musdus

Diagnose why your Chromium-based apps aren’t working, please run the app from the terminal and share the output. This will help identify any errors related to GPU, libraries, or driver issues.

I did share the more relevant part of those output in the thread I linked in the first post, but here are the full logs for both(there are also games that stopped working so I don’t think it’s specifically an electron issue though):
brave:

[67394:67394:0220/111210.923632:ERROR:chrome/browser/ui/views/user_education/impl/browser_user_education_interface_impl.cc:154] Attempting to show IPH IPH_DiscardRing before browser initialization complete; IPH will not be shown.
[67394:67394:0220/111211.072960:ERROR:content/browser/gpu/gpu_process_host.cc:996] GPU process exited unexpectedly: exit_code=139
[67394:67394:0220/111211.308086:ERROR:content/browser/gpu/gpu_process_host.cc:996] GPU process exited unexpectedly: exit_code=139
[67394:67394:0220/111211.549590:ERROR:content/browser/gpu/gpu_process_host.cc:996] GPU process exited unexpectedly: exit_code=139
[67394:67394:0220/111211.549608:FATAL:content/browser/gpu/gpu_data_manager_impl_private.cc:416] GPU process isn't usable. Goodbye.
[0220/111211.561150:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111211.561300:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111211.561404:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111211.566214:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111211.568011:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111211.568198:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111211.568276:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111211.568668:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111211.569132:ERROR:third_party/crashpad/crashpad/util/process/process_memory_range.cc:75] read out of range
[0220/111211.569143:ERROR:third_party/crashpad/crashpad/util/process/process_memory_range.cc:75] read out of range
[0220/111211.575023:ERROR:third_party/crashpad/crashpad/util/process/process_memory_range.cc:75] read out of range
[0220/111211.575035:ERROR:third_party/crashpad/crashpad/util/process/process_memory_range.cc:75] read out of range
[0220/111211.575051:ERROR:third_party/crashpad/crashpad/util/process/process_memory_range.cc:75] read out of range
[0220/111211.575058:ERROR:third_party/crashpad/crashpad/util/process/process_memory_range.cc:75] read out of range

And vs code:

[68332:0220/111435.940796:WARNING:ui/ozone/platform/wayland/common/drm_render_node_path_finder.cc:111] Preferred drm_render_node not found, picking nvidia-drm
[68332:0220/111436.151843:WARNING:ui/ozone/platform/wayland/common/wayland_object.cc:119] Binding to wl_shm version 1 but version 2 is available.
[68332:0220/111436.151855:WARNING:ui/ozone/platform/wayland/common/wayland_object.cc:119] Binding to wl_seat version 8 but version 10 is available.
[68332:0220/111436.151867:WARNING:ui/ozone/platform/wayland/common/wayland_object.cc:119] Binding to zwp_pointer_gestures_v1 version 1 but version 3 is available.
[68332:0220/111436.151887:WARNING:ui/ozone/platform/wayland/common/wayland_object.cc:119] Binding to wp_cursor_shape_manager_v1 version 1 but version 2 is available.
[68332:0220/111436.151905:WARNING:ui/ozone/platform/wayland/common/wayland_object.cc:119] Binding to org_kde_kwin_appmenu_manager version 1 but version 2 is available.
[68332:0220/111436.151957:WARNING:ui/ozone/platform/wayland/common/wayland_object.cc:119] Binding to zwp_linux_dmabuf_v1 version 4 but version 5 is available.
[main 2026-02-20T16:14:36.270Z] PolicyConfiguration#initialize
[main 2026-02-20T16:14:36.271Z] PolicyConfiguration#updatePolicyDefinitions [
  'update.mode',
  'update.channel',
  'update.showReleaseNotes',
  'http.useLocalProxyConfiguration',
  'http.electronFetch',
  'http.proxy',
  'http.proxyStrictSSL',
  'http.proxyKerberosServicePrincipal',
  'http.noProxy',
  'http.proxyAuthorization',
  'http.proxySupport',
  'http.systemCertificates',
  'http.systemCertificatesNode',
  'http.experimental.systemCertificatesV2',
  'http.fetchAdditionalSupport',
  'http.experimental.networkInterfaceCheckInterval',
  'telemetry.telemetryLevel',
  'telemetry.feedback.enabled',
  'telemetry.enableTelemetry',
  'extensions.allowed'
]
[main 2026-02-20T16:14:36.277Z] [File Watcher (node.js)] Request to start watching: /etc/vscode/policy.json (excludes: <none>, includes: <all>, filter: <none>, correlationId: <none>),/home/sloth/.config/Code/User (excludes: <none>, includes: <all>, filter: <none>, correlationId: <none>),/home/sloth/.config/Code/User/settings.json (excludes: <none>, includes: <all>, filter: <none>, correlationId: <none>)
[main 2026-02-20T16:14:36.279Z] PolicyConfiguration#update [
  'update.mode',
  'telemetry.telemetryLevel',
  'telemetry.feedback.enabled',
  'extensions.allowed'
]
[main 2026-02-20T16:14:36.279Z] PolicyConfiguration#updatePolicyDefinitions [ 'update.enableWindowsBackgroundUpdates' ]
[main 2026-02-20T16:14:36.279Z] PolicyConfiguration#update []
[main 2026-02-20T16:14:36.279Z] [File Watcher (node.js)] ignoring a path for watching who's stat info failed to resolve: /etc/vscode/policy.json (error: Error: ENOENT: no such file or directory, stat '/etc/vscode/policy.json')
[main 2026-02-20T16:14:36.280Z] [File Watcher (node.js)] starting fs.watchFile() on /etc/vscode/policy.json (correlationId: undefined)
[main 2026-02-20T16:14:36.314Z] Starting VS Code
[main 2026-02-20T16:14:36.314Z] from: /opt/visual-studio-code/resources/app
[main 2026-02-20T16:14:36.315Z] args: {
  _: [],
  diff: false,
  merge: false,
  add: false,
  remove: false,
  goto: false,
  'new-window': false,
  'reuse-window': false,
  wait: false,
  help: false,
  'list-extensions': false,
  'show-versions': false,
  'pre-release': false,
  'update-extensions': false,
  version: false,
  verbose: true,
  status: false,
  'prof-startup': false,
  'no-cached-data': false,
  'prof-v8-extensions': false,
  'disable-extensions': false,
  'disable-lcd-text': false,
  'disable-gpu': false,
  'disable-chromium-sandbox': false,
  sandbox: false,
  telemetry: false,
  debugRenderer: false,
  'enable-smoke-test-driver': false,
  logExtensionHostCommunication: false,
  'skip-release-notes': false,
  'skip-welcome': false,
  'disable-telemetry': false,
  'disable-updates': false,
  transient: false,
  'use-inmemory-secretstorage': false,
  'disable-workspace-trust': false,
  'disable-crash-reporter': false,
  'crash-reporter-id': '7318f7ab-1fc0-4b46-b038-446b4da1fc85',
  'skip-add-to-recently-opened': false,
  'open-url': false,
  'file-write': false,
  'file-chmod': false,
  force: false,
  'do-not-sync': false,
  'do-not-include-pack-dependencies': false,
  trace: false,
  'trace-memory-infra': false,
  'preserve-env': false,
  'force-user-env': false,
  'force-disable-user-env': false,
  'open-devtools': false,
  'disable-gpu-sandbox': false,
  '__enable-file-policy': false,
  'enable-coi': false,
  'enable-rdp-display-tracking': false,
  'disable-layout-restore': false,
  'disable-experiments': false,
  'no-proxy-server': false,
  'no-sandbox': false,
  nolazy: false,
  'force-renderer-accessibility': false,
  'ignore-certificate-errors': false,
  'allow-insecure-localhost': false,
  'disable-dev-shm-usage': false,
  'profile-temp': false,
  logsPath: '/home/sloth/.config/Code/logs/20260220T111436'
}
[main 2026-02-20T16:14:36.315Z] Resolving machine identifier...
[main 2026-02-20T16:14:36.315Z] Resolving SQM identifier...
[main 2026-02-20T16:14:36.315Z] Resolving devDevice identifier...
[main 2026-02-20T16:14:36.315Z] Resolved machine identifier: 93fbd3a23e4b28332d266ebd2056d745b7f1e9f52d3d8e9c022f562805d3aaff
[main 2026-02-20T16:14:36.315Z] Resolved SQM identifier: 
[main 2026-02-20T16:14:36.315Z] Resolved devDevice identifier: bcdf4ce0-0e95-4a88-949d-5b191e961641
[main 2026-02-20T16:14:36.315Z] Main->SharedProcess#connect
[main 2026-02-20T16:14:36.317Z] PolicyConfiguration#updatePolicyDefinitions [
  'terminal.integrated.automationProfile.linux',
  'terminal.integrated.automationProfile.osx',
  'terminal.integrated.automationProfile.windows',
  'terminal.integrated.profiles.windows',
  'terminal.integrated.profiles.osx',
  'terminal.integrated.profiles.linux',
  'terminal.integrated.useWslProfiles',
  'terminal.integrated.inheritEnv',
  'terminal.integrated.persistentSessionScrollback',
  'terminal.integrated.showLinkHover',
  'terminal.integrated.ignoreProcessNames'
]
[main 2026-02-20T16:14:36.317Z] PolicyConfiguration#updatePolicyDefinitions [
  'terminal.integrated.defaultProfile.linux',
  'terminal.integrated.defaultProfile.osx',
  'terminal.integrated.defaultProfile.windows'
]
[main 2026-02-20T16:14:36.319Z] PolicyConfiguration#update []
[main 2026-02-20T16:14:36.319Z] PolicyConfiguration#update []
[main 2026-02-20T16:14:36.321Z] [File Watcher (node.js)] Started watching: '/home/sloth/.config/Code/User'
[main 2026-02-20T16:14:36.328Z] [File Watcher (node.js)] Started watching: '/home/sloth/.config/Code/User/settings.json'
[main 2026-02-20T16:14:36.329Z] StorageMainService: creating application storage
[main 2026-02-20T16:14:36.334Z] ElectronURLListener: waiting for window to be ready to handle URLs...
[main 2026-02-20T16:14:36.334Z] lifecycle (main): phase changed (value: 2)
[main 2026-02-20T16:14:36.334Z] windowsManager#open
[main 2026-02-20T16:14:36.336Z] windowsManager#open pathsToOpen [
  {
    workspace: { id: 'f81194ce3d9459af46207e7625aab580', uri: [Yr] },
    type: 2,
    exists: true
  }
]
[main 2026-02-20T16:14:36.336Z] windowsManager#doOpenFolderOrWorkspace {
  folderOrWorkspace: {
    workspace: { id: 'f81194ce3d9459af46207e7625aab580', uri: [Yr] },
    type: 2,
    exists: true
  },
  filesToOpen: undefined
}
[main 2026-02-20T16:14:36.338Z] IPC Object URL: Registered new channel vscode:2a56925d-3176-4c54-a787-f321e3179675.
[main 2026-02-20T16:14:36.338Z] window#validateWindowState: validating window state on 2 display(s) { mode: 1, x: 0, y: 0, width: 2788, height: 1260 }
[main 2026-02-20T16:14:36.338Z] window#validateWindowState: multi-monitor working area { x: 437, y: 0, width: 2560, height: 1080 }
[main 2026-02-20T16:14:36.338Z] window#ctor: using window state { mode: 1, x: 0, y: 0, width: 2788, height: 1260 }
[68332:0220/111436.346003:WARNING:ui/ozone/platform/wayland/host/wayland_surface.cc:170] Server doesn't support zcr_alpha_compositing_v1.
[68332:0220/111436.346015:WARNING:ui/ozone/platform/wayland/host/wayland_surface.cc:185] Server doesn't support overlay_prioritizer.
[main 2026-02-20T16:14:36.370Z] window#load: attempt to load window (id: 1)
[main 2026-02-20T16:14:36.373Z] StorageMainService: creating workspace storage (f81194ce3d9459af46207e7625aab580)
[main 2026-02-20T16:14:36.374Z] windowsManager#open used window count 1 (workspacesToOpen: 0, foldersToOpen: 1, emptyToRestore: 0, maybeOpenEmptyWindow: false)
[main 2026-02-20T16:14:36.374Z] lifecycle (main): phase changed (value: 3)
[main 2026-02-20T16:14:36.375Z] resolveShellEnv(): skipped (VSCODE_CLI is set)
[main 2026-02-20T16:14:36.375Z] update#setState idle
[main 2026-02-20T16:14:36.418Z] resolveShellEnv(): skipped (VSCODE_CLI is set)
[68373:0220/111436.526998:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68373:0220/111436.527092:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[0220/111436.542496:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/d618ea95-4fc9-4f2d-8492-4dec3f5a3d41: No such file or directory (2)
[68332:0220/111436.853638:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111436.853653:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 1 time(s)
[68439:0220/111437.067284:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68439:0220/111437.067682:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[main 2026-02-20T16:14:37.081Z] [File Watcher (node.js)] Request to stop watching: /etc/vscode/policy.json (excludes: <none>, includes: <all>, filter: <none>, correlationId: <none>)
[main 2026-02-20T16:14:37.081Z] [File Watcher (node.js)] stopping file watcher (/etc/vscode/policy.json (excludes: <none>, includes: <all>, filter: <none>, correlationId: <none>))
[0220/111437.081877:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/966c4c9c-cfee-4681-9d5e-5c6472b757dc: No such file or directory (2)
[68332:0220/111437.400019:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111437.400034:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 2 time(s)
[68453:0220/111437.616858:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68453:0220/111437.616941:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[0220/111437.631176:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/1437e538-e004-45dc-b66d-8a6c4c8ed34a: No such file or directory (2)
[68332:0220/111437.944824:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111437.944837:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 3 time(s)
[68467:0220/111438.083622:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68467:0220/111438.083713:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[0220/111438.098385:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/b09b0bf2-04d7-408a-b34c-3e7a2cd73ce3: No such file or directory (2)
[68332:0220/111438.601748:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111438.601759:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 4 time(s)
[68497:0220/111438.744193:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68497:0220/111438.744279:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[0220/111438.759043:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/9b86f98a-0028-47b5-84d5-ab861aa7fe72: No such file or directory (2)
[main 2026-02-20T16:14:38.877Z] lifecycle (main): phase changed (value: 4)
[68332:0220/111439.260338:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111439.260353:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 5 time(s)
[68528:0220/111439.412238:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68528:0220/111439.412333:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[0220/111439.427028:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/29819ab6-3768-495f-af60-99420467baac: No such file or directory (2)
[68332:0220/111439.918292:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111439.918303:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 6 time(s)
[68558:0220/111439.922755:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68558:0220/111439.922837:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[0220/111439.934000:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/f940aeef-e547-4c55-8e6a-86a8cbe8dce7: No such file or directory (2)
[68332:0220/111440.144213:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111440.144224:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 7 time(s)
[68568:0220/111440.148821:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68568:0220/111440.148894:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[0220/111440.239704:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/9c756c8f-07a6-45c3-a3f1-21e272085c48: No such file or directory (2)
[68332:0220/111440.375663:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111440.375675:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 8 time(s)
[68580:0220/111440.380665:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:129] Should skip nVidia device named: nvidia-drm
[68580:0220/111440.380742:WARNING:media/gpu/vaapi/vaapi_wrapper.cc:1549] Forcibly using value of --render-node-override
[0220/111440.553738:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/582e0741-9b5b-4030-9936-d5a59229aeda: No such file or directory (2)
[68332:0220/111440.618098:ERROR:content/browser/gpu/gpu_process_host.cc:999] GPU process exited unexpectedly: exit_code=139
[68332:0220/111440.618108:WARNING:content/browser/gpu/gpu_process_host.cc:1441] The GPU process has crashed 9 time(s)
[68332:0220/111440.618118:FATAL:content/browser/gpu/gpu_data_manager_impl_private.cc:415] GPU process isn't usable. Goodbye.
[0220/111440.624628:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111440.624689:ERROR:third_party/crashpad/crashpad/snapshot/elf/elf_dynamic_array_reader.h:64] tag not found
[0220/111440.875894:ERROR:third_party/crashpad/crashpad/util/file/directory_reader_posix.cc:43] opendir /home/sloth/.config/Code/Crashpad/attachments/054ca4a1-fec5-4265-b10d-ca3b1e62783e: No such file or directory (2)

@philm

you may want to try an older ISO

Thanks, trying to find how to download old versions was going to be my next step if there was no way to fix the new one.
At this point, I’m assuming launching nonfree nvidia from the iso is just broken for everyone in the current version since everyone is avoiding directly answering my question.

I’ve already tried old drivers to see if they worked with no luck and I’ve had report that those apps are not broken with the latest driver from someone with basically the same gpu I have so having an old driver in the iso wouldn’t change much(it’s actually less ideal for the test I’m trying to make), but if it boots, it’ll still a good test.

[linux-aarhus]

@SamKook

since everyone is avoiding directly answering my question

I think you mis-interpret something here.

I have not seen any direct questions here - just a lot of statements.

Further I acknowledge that

@linux-aarhus

it is known that the live ISO may not be optimally configured when using the proprietary driver (only Nvidia) is an issue - perhaps (very old AMD as well) - and that you work around by using the free GPU option.

For the sake of verifying your assumption

@SamKook

At this point, I’m assuming launching nonfree nvidia from the iso is just broken for everyone in the current version

I have started a Lenovo Legion (configured to only use discrete GPU) with Nvidia 5070 - using the minimal Plasma ISO from 260114 - it took a while, but eventually it loaded the Plasma Desktop.

@SamKook

My system is not completely broken, I can boot fine with either driver as I said. Only certain apps and games don’t work and I’m trying to determine if it’s a new package that is causing the issues or if it’s something else.

As your system is working - it would be safe to assume that it is application related somehow.

I cannot possibly guess where your problem lies - yes, it may be related to an update, but on stable branch a lot gets updated.

My last suggestion will be to switch branch to testing or unstable. Those branches is updated more frequently and troubleshooting is somewhat easier when the updates comes in smaller chunks.

I have tested both Brave browser and Visual Studio Code on the mentioned Lenovo Legion - and there is no issues with neither applications.

GPU and Driver Info

https://0x0.st/PSXQ.txt

[SamKook]

I’m not trying to be snarky or anything with the people dodging my question statement, people tend to be eager to help and focus on the bigger problem rather than answering questions directly, which tends to be what people often need and help is always appreciated.

@SamKook

Is that a known issue and is there any way to get around that and still load the proprietary driver with a live version?

Seemed like a pretty direct question to me with an exclamation mark and everything.
And yes, I did see your statement, but in my mind there’s a big difference between not optimal and not loading at all.

Thanks for testing, that means it not loading on my pc may be related to the issue so that is more helpful information acquired. It may be the fallback to nouveau kicking in, although that seemed too early in the boot process to me, that is creating that problem.

Looking at your info, it may be my lack of integrated gpu that could also be a possible cause, maybe it has issues loading the nonfree driver if the dedicated gpu is the only one available.

It started happening after one of the big system updates so I am suspecting the issue is related to one of those(I may also have switched kernel at the same time so that’s not a guarantee, but I tested switching back to 6.17 and even tried 6.19 too) and it is sadly too late for being on a more often updated branch to be helpful to narrow it down since the problem is already present.

I hadn’t narrowed down the gpu issue enough to be comfortable asking for help with it yet, which is why I made a thread about my issue booting a live version with the proprietary driver, which is the issue I’m trying to fix right now.

I certainly don’t mind people also trying to help to fix the bigger problem though.

[SamKook]

I tried the first and last release of v25 and both worked fine, but none of the v26(tried .0 as well) will boot with the proprietary driver.

Seems my pc doesn’t like something about the switch from X11 to wayland in the iso relating to the nvidia driver.

My bigger issue shouldn’t be about wayland only though since I had manually switched KDE plasma to wayland about 2 years ago on my system.

[Molski]

@SamKook

Seems my pc doesn’t like something about the switch from X11 to wayland in the iso relating to the nvidia driver.

Probably not the case since..

@SamKook

Edit: The XFCE iso has the same behavior.

Xfce wayland support doesn’t exist. (It’s expected later this year on their road map.)

[SamKook]

I guess some other behavior changed in v26 then, good to know.

[linux-aarhus]

@SamKook

my issue booting a live version with the proprietary driver, which is the issue I’m trying to fix right now

Hmm, the default live ISO uses sddm and wayland, but as you note you have no daily issues by using wayland.

@SamKook

Seemed like a pretty direct question to me with an exclamation mark and everything.

Yes it is - and I answered the first half of question.

@SamKook

Is that a known issue and is there any way to get around that and still load the proprietary driver with a live version?

As for the second half - I also answered that - but not the answer you wanted to hear

@linux-aarhus

and that you work around by using the free GPU option

Sorry if I was not clear enough.

No, there is no way around and still load the proprietary driver with a live version.
edit: 2026-02-23T06:21:00Z there may be a way below

This is because mhwd installs a driver at load time, based on the output from nvidia-driver-assistant.

Perhaps you could try the following (no idea if you can get it working) but if mhwd can do it automagically - then it could be done manually too - but I have zero knowledge on how mhwd work and what it may do in preparation, none what-so-ever

• boot the ISO to the grub menu
• edit the kernel cmdline and add 3 and continue booting
• login using the default credentials

Run nvidia-driver-assistant to see what it says.

Then use cmdline arguments to pacman and point to make pacman use the mhwd repo at /opt/mhwd/pkg

Or use pacman -U to specify the packages to install in the live environment.

When you have the driver in place, test the graphic environment with

startplasma-wayland

Or via the display manager

systemctl start sddm

ideas

None of these will help in relation to booting the ISO - that is - unless you customize an iso-profile to use a given driver version and build such ISO yourself.

first idea

If you boot from the last ISO in the 25.x series, you can copy from the live ISO /opt/mhwd/ folder, which contains the drivers used to install the system (it is also possible to extract the files without booting, requires more work, but doable).

Copy the *.pkg.tar.{zst,zst.sig} files to an USB device for safekeeping.

I don’t think they will be of much use though - as they are matched against a the specific kernel used on ISO.

second idea

This is best done using a console boot - no graphics loaded.

• Sync the kernel-headers for your kernel(s)
• Remove the kernel specific drivers
• Sync the relevant dkms driver package

 $ pamac search nvidia dkms --no-aur
nvidia-open-dkms  590.48.01-3                                      extra
    NVIDIA open kernel modules - module sources
nvidia-dkms  590.48.01-3                                           extra
    NVIDIA kernel modules - module sources
nvidia-575xx-open-dkms  575.64.05-3                                extra
    NVIDIA 575 open kernel modules - module sources
nvidia-575xx-dkms  575.64.05-3                                     extra
    NVIDIA 575 kernel modules - module sources
nvidia-570xx-open-dkms  570.211.01-1                               extra
    NVIDIA 570xx open kernel modules - module sources
nvidia-570xx-dkms  570.211.01-1                                    extra
    NVIDIA 570xx kernel modules - module sources
nvidia-470xx-dkms  470.256.02-17                                   extra
    NVIDIA drivers - module sources
nvidia-390xx-dkms  390.157-22                                      extra
    NVIDIA drivers - module sources

For some unknown reason the package nvidia-580xx-dkms can only be fetched from AUR.