Update: this same kernel (7.0.0-28-generic) also fails to hibernate
cleanly, with a different but likely related symptom, and battery-
draining consequences.

## New symptom: hibernation hangs mid-freeze, battery drains overnight,
fails to resume

Sequence of events (all on the same boot, kernel 7.0.0-28-generic):

1. ~20:50 — a hibernate cycle triggers the following kernel warning,
repeated dozens of times within the same second:

   WARNING: drivers/gpu/drm/ttm/ttm_resource.c:235 at
ttm_resource_add_bulk_move+0xf5/0x170 [ttm], CPU#12:
kworker/u64:19/78646

   Call trace shows this happening inside the device-freeze path during 
hibernation entry:
   amdgpu_device_suspend -> ttm_device_prepare_hibernation -> pci_pm_freeze -> 
dpm_run_callback -> device_suspend -> async_suspend

2. Later the same evening (~22:32), a second, intentional hibernate is
triggered (systemd-hibernate.service starts, "PM: hibernation:
hibernation entry" logged) — and the journal for that boot ends
immediately after, with no "hibernation exit" and no orderly shutdown
sequence.

3. The next boot (next day, ~16h later) is a fresh cold boot (new boot
ID), not a resume — the previous session/hibernation image was not
restored.

4. The laptop's battery was found fully drained the next morning.

## Interpretation

A genuinely successful hibernation powers the machine off almost
completely (negligible standby draw). A full battery drain overnight
strongly suggests the machine never reached that powered-off state, and
instead hung in an intermediate state (screen off, board still powered)
after the amdgpu/TTM warning above, until the battery ran out. The
hibernation image was presumably left incomplete/invalid as a result,
which is presumably why the kernel discarded it and did a normal cold
boot instead of a resume (a sensible safety fallback, but it meant
losing the whole session).

## Hardware/software (same as original report)

- Dell Inspiron 14 5425, AMD Ryzen 7 5825U, AMD Barcelo iGPU (amdgpu)
- Ubuntu 25.10, KDE Plasma 6.5, Wayland
- Kernel: 7.0.0-28-generic (7.0.0-28.28) — same kernel as the original suspend 
regression reported here
- Secure Boot disabled, hibernation resume correctly configured 
(resume=UUID=... pointing at active swap partition, /sys/power/disk shows 
[platform], swap capacity confirmed sufficient by systemd-logind at time of the 
attempt)

## Workaround

Same as for the suspend issue: reverting to linux-image-7.0.0-27-generic
avoids this — no equivalent warning or failed hibernate/resume observed
on that kernel.

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2161312

Title:
  System hangs on suspend (s2idle) with kernel 7.0.0-28-generic, works
  fine on 7.0.0-27

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2161312/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to