Summary
After a routine dnf upgrade on Qubes OS 4.3 (dom0, fc41 packages), my machine entered an endless boot loop (GRUB loads, kernel starts, hard reset before LUKS prompt, repeating) with kernel-latest 7.2.11. Older kernels from the GRUB menu booted fine.
Root cause found: dnf history shows the 7.2.11 transaction (#30, 2026-10-31) ended in Status: Started (no end time) and a follow-up transaction (#31) in Status: Error — the update was interrupted mid-transaction. Reinstalling the exact same 7.2.11 packages cleanly from the dnf cache in a single completed transaction resolved everything: kernel-latest 7.2.11 now boots normally. The kernel itself was never buggy.
Posting this mainly so that (a) other users hitting an identical boot loop after a dom0 update check dnf history for a Status other than “Ok” before assuming a kernel bug, and (b) to document that interrupted transactions can leave a state that looks exactly like a kernel regression.
Timeline (dnf history evidence)
-
#30, 2026-10-31 14:31 —
dnf upgrade: installs kernel-latest 7.2.11 + kernel 6.18.57 + xen 4.19.5→4.19.6. Status: Started, End time: 1970-01-01 (interrupted). -
#31, 2026-10-31 15:47 — retry
dnf upgrade. Status: Error. -
Machine boot-loops with 7.2.11; 7.1.10 from GRUB works.
-
#32, 2026-11-08 —
dnf remove kernel-latest-7.2.11*(partially effective: devel/modules/qubes-vm remained). -
#33+ — kernel-latest 7.2.5 installed; boots fine.
-
Final step — all four 7.2.11 packages cleanly reinstalled from
/var/cache/libdnf5/.../packages/in one completed transaction (dnf install --allow-downgrade ./kernel-latest*.rpm). Result: 7.2.11 boots fine. No kernel bug.
Environment
-
Qubes OS 4.3, dom0 packages fc41, Xen 4.19.6-2
-
Hardware: [CPU, GPU, Mainboard/Laptop-Modell]
== System ==
TUXEDO
TUXEDO InfinityBook Pro Intel Gen9
== Board ==
NB02
GXxMRXx
== CPU ==
Model name: Intel(R) Core™ Ultra 7 155H
== GPU ==
00:02.0 VGA compatible controller: Intel Corporation Meteor Lake-P [Intel Arc Graphics] (rev 08)
00:1c.5 PCI bridge: Intel Corporation Device 7e3d (rev 20)
== NIC ==
00:08.0 System peripheral: Intel Corporation Meteor Lake-P Gaussian & Neural-Network Accelerator (rev 20)
00:14.3 Network controller: Intel Corporation Meteor Lake PCH CNVi WiFi (rev 20)
2d:00.0 Ethernet controller: Motorcomm Microelectronics. YT6801 Gigabit Ethernet Controller (rev 01)
== Boot ==
UEFI
== Xen ==
xen_version : 4.19.6
Related
Suggestion
Would it be feasible for dom0 to warn more prominently when a dom0 kernel/Xen transaction did not complete (e.g. check dnf history status at boot and display a notice), since the resulting state masquerades as a kernel bug and is hard to diagnose?