OPINIONATED OMARCHY
TROUBLESHOOTING CORPUS
790
RECORDS
790
AUDITED
0
UNCHECKED
2388
SOURCES
LIVE
omarchy-core medium rare DISPUTE

Remove the resume parameters 'omarchy hibernation remove' leaves in the kernel command line

omarchyomarchy-4hibernationliminebtrfslaptop
AUDITED Audited against its sources and confirmed accurate. · checked against Omarchy omarchy 4.0.4-1 on 2026-10-05
AUDIT NOTE

Read /usr/share/omarchy/bin/omarchy-hibernation-remove and omarchy-hibernation-setup on 4.0.4-1. Remove deletes the swapfile, the /swap subvolume, the fstab lines and /etc/mkinitcpio.conf.d/omarchy_resume.conf, then runs limine-mkinitcpio, and never touches /etc/limine-entry-tool.d/resume.conf. Setup writes RESUME_DROP_IN only inside if [[ ! -f $RESUME_DROP_IN ]]. Its only repair path fixes an empty resume_offset, and only when hibernation is already set up. #13583 and #12096 are open and describe exactly this. The fix (rm -f the drop-in, then sudo limine-mkinitcpio) is the right rebuild on Omarchy 4, and btrfs inspect-internal map-swapfile -r is the same offset command setup uses. Running omarchy hibernation setup without sudo is correct, because the script calls sudo itself. The leftover rtc-alarm.conf is harmless, as #13583 notes. This is Omarchy-specific, with no plain Arch branch needed. Not exercised: hibernation was not set up or removed here.

Symptom

After omarchy hibernation remove, cat /proc/cmdline still carries resume=/dev/mapper/root resume_offset=1929151 pointing at a swapfile that no longer exists. Worse, after setting hibernation up again later, hibernate writes an image and powers off but the next boot starts fresh and the session is lost.

Cause

omarchy-hibernation-setup writes /etc/limine-entry-tool.d/resume.conf (KERNEL_CMDLINE[default]+=" resume=... resume_offset=..."). omarchy-hibernation-remove on 4.0.4-1 deletes the swapfile, the /swap subvolume, the fstab line and /etc/mkinitcpio.conf.d/omarchy_resume.conf, then runs limine-mkinitcpio, but never deletes resume.conf, so the UKI keeps the old parameters. A later setup only writes resume.conf when it is missing (if [[ ! -f $RESUME_DROP_IN ]]), so the new swapfile's physical offset is never recorded and resume reads the wrong place. Both confirmed by reading the 4.0.4-1 scripts here.

Fix

After removing hibernation, delete the drop-in and rebuild the UKI:


sudo rm -f /etc/limine-entry-tool.d/resume.conf
sudo limine-mkinitcpio

If you are setting hibernation up again and the drop-in survived from an earlier setup, remove it first so setup writes the current offset:


sudo rm -f /etc/limine-entry-tool.d/resume.conf
omarchy hibernation setup
cat /etc/limine-entry-tool.d/resume.conf
sudo btrfs inspect-internal map-swapfile -r /swap/swapfile   # must equal resume_offset

Verify

After a reboot, grep -o 'resume[^ ]*' /proc/cmdline prints nothing when hibernation is removed, or a resume_offset equal to sudo btrfs inspect-internal map-swapfile -r /swap/swapfile when it is set up. A test systemctl hibernate restores the session.

Sources

hibernation-remove-leaves-stale-resume-offset