# Hard freeze shortly after resuming from s2idle on Intel Ice Lake (i915) with kernel 7.2.3-arch1-3

## Summary

On Omarchy 4.0.3-1 with kernel 7.2.3-arch1-3, this laptop intermittently
hard-freezes **~20-50 s after a successful s2idle resume**. Suspend and resume
both complete and log normally (`PM: suspend entry` and `PM: suspend exit`); the
machine then runs briefly (networking reconnects, the shell re-registers its QML
plugins) and stops logging entirely. No panic, oops, watchdog, MCE or lockup
message is written. Recovery requires holding the power button.

The same machine went through **42 suspend/resume cycles on the previous kernel
(7.1.9-arch1-2) without freezing right after a resume**. The regression appeared
after the in-place update to 7.2.3-arch1-3 on 2026-09-08 23:10.

This is a *different* failure mode from the other 7.2.3 suspend reports: here the
machine resumes (`PM: suspend exit` is present) and dies shortly afterward, on
Intel Ice Lake, with `s2idle` (not `deep`).

## System details

- **Omarchy**: 4.0.3-1
- **Kernel**: 7.2.3-arch1-3 (updated in place from 7.1.9-arch1-2 on 2026-09-08 23:10)
- **linux-firmware**: 20260810-2
- **Machine**: ASUSTeK ZenBook UX325JA, firmware UX325JA.312 (2022-01-17)
- **CPU/GPU**: Intel Core i7-1065G7 (Ice Lake), Iris Plus Graphics G7 `[8086:8a52]`, driver `i915`
- **`/sys/power/mem_sleep`**: `[s2idle] deep` (s2idle default; the freeze happens under s2idle)
- **Desktop**: Hyprland 0.56.2, SDDM
- **Display**: external monitor on HDMI-A-1 active, eDP-1 off
- **journald**: persistent (`/var/log/journal` present), so a stopped log is a real halt, not a lost-log artifact

## What happens

1. Suspend (lid close / idle) completes normally: `PM: suspend entry (s2idle)`.
2. Resume completes normally: `PM: suspend exit`, and userspace comes back.
3. ~20-50 s later the whole machine hard-freezes: the journal simply stops.
4. Forced power-off; next boot shows no clean-shutdown record.

### Most recent occurrence (2026-09-12, boot `fb822a61`'s predecessor)

```text
Sep 12 07:30:43 omarchy kernel: PM: suspend entry (s2idle)
Sep 12 07:30:48 omarchy kernel: PM: suspend exit
... tailscaled reconnects, NetworkManager-dispatcher, timesyncd ...
Sep 12 07:31:17 omarchy kernel: [UFW BLOCK] ...      <-- last line of the boot
Sep 12 07:34:19 omarchy kernel: Linux version 7.2.3-arch1-3   <-- next boot
```

### Previous occurrence (2026-09-10)

```text
Sep 10 07:06:19 omarchy kernel: PM: suspend entry (s2idle)
Sep 10 07:06:23 omarchy kernel: PM: suspend exit
... voxtype gdk_monitor_set_scale assertions, omarchy-shell QML plugin warnings ...
Sep 10 07:07:13 omarchy kernel: <-- last line of the boot
Sep 10 07:08:32 omarchy kernel: Linux version 7.2.3-arch1-3   <-- next boot
```

## Suspends per kernel (retained journal)

| Boot | Kernel | `PM: suspend entry` | `PM: suspend exit` | Outcome |
|------|--------|--------------------:|-------------------:|---------|
| -5 | 7.1.9 | 24 | 24 | ran to a normal end |
| -4 | 7.1.9 | 11 | 11 | ran to a normal end |
| -3 | 7.1.9 | 7 | 7 | ran to a normal end |
| -2 | 7.2.3 | 2 | 2 | froze ~50 s after the last exit |
| -1 | 7.2.3 | 3 | 3 | froze ~29 s after the last exit |

So the freeze is **not** a missing `PM: suspend exit` — resume succeeds and the
machine dies shortly afterwards.

## Not the cause

- **Not OOM**: at the 2026-09-12 death `mem_available` was ~4.4 GiB; the local
  crash ledger recorded `suspected_oom=False` and `swap_used` ~2.7 GiB.
- **No `i915` "Unclaimed access detected prior to suspending"** warning anywhere in
  the journal (count 0), unlike #11255.
- **No kernel panic/oops, no MCE, no soft lockup/hung task** — the log simply stops.
- High-ish thermals (package ~67-78 °C, `crit` 100 °C) but not at critical, and the
  same temps occur on clean-booting sessions.

## Related issues

- #11255 — Intel Raptor Lake-P, `i915`, kernel 7.2.3, *deep* suspend never resumes
  (hard hang at *entry*; here it resumes first).
- #11332 — intermittent s2idle resume failure, more frequent on 7.2.3 (AMD HP
  EliteBook; hangs at *entry*).
- #11245 — kernel 7.2.3 regression (ELAN touchpad), A/B confirmed vs linux-lts.
- #7380 — omarchy-shell crashes and relaunches on wake (FALLBACK monitor removal).
  The same shell crash is visible here during the suspend/wake path, but it
  self-heals and is separate from the hard freeze.

## Possible workarounds to test

- Pin/downgrade to kernel 7.1.9-arch1-2 (or use `linux-lts`) until a fixed kernel
  is available.
- Try `i915.enable_psr=0` / `i915.enable_dc=0`, since the failure is in the
  post-resume display re-initialisation path.
- Try `mem_sleep_default=deep` to move off s2idle.

## Diagnostics

`omarchy-debug --no-sudo --print` captured (system info, package list, cmdline,
CPU/GPU) and saved locally at `/tmp/omarchy-debug.log`.

---

Filed by deepseek-v4.1-flash via opencode.