[Enhancement] Auto-detect NVIDIA 580xx users & enforce DKMS + mkinitcpio MODULES during omarchy-update

[lukejmorrison]

Enhancement request

After surviving the 3.4.x kernel bump (see #4781), the root cause was crystal clear:

• nvidia-*-utils installed but no *-dkms package → mkinitcpio fails silently → emergency mode with /boot vfat unknown + missing modules.

Proposed fix (tiny & fully Oma-friendly):

• In omarchy-update: if any nvidia-*-utils is present, auto-run omarchy-pkg-install nvidia-*-dkms (or prompt).
• Auto-append the four modules (nvidia nvidia_modeset nvidia_uvm nvidia_drm) to MODULES=() in /etc/mkinitcpio.conf (with timestamped backup).
• Always force limine-mkinitcpio after any kernel change.

Zero breaking changes, respects existing hooks (omarchy_hooks.conf), and eliminates the #1 NVIDIA Pascal complaint.

Hardware example: Intel i7-5930K, ASUS X99-E, NVIDIA GTX 1070 Ti (580xx), LUKS/Btrfs/Limine.
My fix path (worked perfectly):

• omarchy-pkg-install nvidia-580xx-dkms
• Ensure MODULES line + limine-mkinitcpio
• Pre-update snapshot + reboot + omarchy-update

Happy to open a PR with the exact diff if wanted. Cross-ref: #4781 #4192 #3954.

[lukejmorrison]

Cross-link for full context: the exact repro + fix path (nvidia-580xx-dkms install + prep script + reboot) that solved the emergency-mode boot is in the original report here -> #4781 (and my earlier comment with logs).\nThis enhancement would have prevented the whole issue automatically. Happy to help test or PR the guardrail!

[lukejmorrison]

prepare-omarchy-update-nvidia-safe.sh

Field-tested helper script from this incident (shared in case useful for upstream design/testing):

/home/luke/dev/Omarchy/scripts/prepare-omarchy-update-nvidia-safe.sh

What it does:

• Creates snapshot first (omarchy-snapshot create) and prints rollback target ID.
• Ensures /etc/mkinitcpio.conf has MODULES=(nvidia nvidia_modeset nvidia_uvm nvidia_drm).
• Creates timestamped backup before editing.
• Rebuilds initramfs via limine-mkinitcpio when available (falls back to mkinitcpio -P).
• Verifies NVIDIA module files (nvidia*.ko*) exist for current kernel before allowing update.
• Hard-stops with actionable remediation if modules are missing.

Scope/safety notes:

• Uses official Omarchy flows (omarchy-snapshot, omarchy-update, Omarchy package recommendation).
• No edits outside intended config path.
• Current package remediation is tuned to Pascal/580xx (nvidia-580xx-dkms), which could be generalized by branch detection if adopted upstream.

Observed result on this hardware:

• Intel i7-5930K + ASUS X99-E + GTX 1070 Ti (Pascal), LUKS/Btrfs/Limine.
• After nvidia-580xx-dkms install + UKI rebuild, update path stabilized and completed normally.

[gregid]

Everytime there is kernel bump I dread the omarchy update because of this - and it looks I am not the only one.

[EverStarck]

I just got dropped into emergency mode with /boot showing vfat unknown and missing modules. I’m using a GTX 1080, so I think I may be running into the same issue.

Is there any way to fix this and recover my working Omarchy installation? Unfortunately, my latest snapshot also boots into emergency mode, lol

Is there another recovery option besides booting from a live USB and doing a chroot repair?