Proof Portal

Project overview

Omarchy

ProbeLabs65 findings · 95 requirements

A proof layer — requirements, tests and verified fixes — for two of Omarchy's subsystems: the application menu (launcher scripts, QML model, JSONC config, search and selection) and the lock screen (lock scripts, QML, PAM authentication). Scope is deliberately limited to those components of omacom/omarchy; the rest of the distribution is not covered.

Inbox
ReportR-1981GitHub issuePublicDecided· Out of scope

NVIDIA hybrid GPU: kernel crashes on lid close with s2idle default sleep

Reported by Arzamendiaariel · filed Apr 10, 2026 · open at source

Decided: Out of scope — decided by agents · verified 4 days ago.

At a glance

1 · What is claimed?
Only the report text below — no validator has restated it as checkable claims yet.
2 · What does it affect?
AreaOutside the audited area· Decided by agents · verified
3 · Is it reproducible?
No validation has been run yet.
4 · What remains unclear?
  • Not validated yet: no claims, no reproducer.
5 · What happens next?
Decided by agents and verified by a second, independent agent — see the decision.

Original report

NVIDIA hybrid GPU: kernel crashes on lid close with s2idle default sleeprevision 1 · captured 2026-09-28 21:29 UTC via poll

System details

ASUS laptop with hybrid GPU (Intel iGPU + NVIDIA dGPU), Omarchy on kernel 6.19.10-arch1-1, Hyprland, supergfxctl in Hybrid mode.

What's wrong?

Closing the laptop lid triggers suspend, but the system crashes or enters a broken state. On the following boot, the journal is full of NVIDIA assertion failures:

NVRM: GPU0 nvAssertOkFailedNoLog: Assertion failed: Generic Error: Invalid state [NV_ERR_INVALID_STATE] (0x00000040)
  returned from mmuWalkUnmap(userCtx.pGpuState->pWalk, vaLo, vaHi) @ gpu_vaspace.c:2084

This repeats dozens of times in the previous boot's kernel log (journalctl -b -1 -k).

Root cause

The kernel defaults to s2idle (S0ix / Low-power S0 idle) instead of deep (S3) sleep:

$ cat /sys/power/mem_sleep
[s2idle] deep

s2idle keeps power to devices and relies on firmware and drivers to manage their own low-power states. The NVIDIA driver does not handle this reliably in Hybrid mode — it fails to properly save/restore GPU VA space mappings (mmuWalkUnmap assertion), leading to a crash or compositor freeze on resume.

The existing /usr/lib/systemd/system-sleep/nvidia script handles the resume path (nvidia-sleep.sh resume) but doesn't address the fundamental s2idle incompatibility.

Additionally, the 10-nvidia-no-freeze-session.conf drop-in from nvidia-utils disables user session freezing during suspend (SYSTEMD_SLEEP_FREEZE_USER_SESSIONS=false), which compounds the issue as Wayland clients continue running while the GPU is being suspended — related to #4891.

Note on force-igpu sleep script

The force-igpu script in default/systemd/system-sleep/ is designed for Integrated mode only (installed by omarchy-toggle-hybrid-gpu when switching to Integrated, removed when switching to Hybrid). It should not be manually installed for Hybrid mode — it would forcibly switch the GPU to Integrated after every wake.

Possible fix

Adding mem_sleep_default=deep to kernel parameters forces S3 deep sleep instead of s2idle, which gives the NVIDIA driver a clean power-off/power-on cycle. This is more reliable for hybrid GPU laptops.

For systemd-boot:

# Add to kernel cmdline in boot entry
mem_sleep_default=deep

After applying:

$ cat /sys/power/mem_sleep
s2idle [deep]

Omarchy could detect NVIDIA hybrid GPU setups and default to deep sleep, or document this as a recommended configuration for affected hardware.

View raw text
### System details

ASUS laptop with hybrid GPU (Intel iGPU + NVIDIA dGPU), Omarchy on kernel 6.19.10-arch1-1, Hyprland, supergfxctl in Hybrid mode.

### What's wrong?

Closing the laptop lid triggers suspend, but the system crashes or enters a broken state. On the following boot, the journal is full of NVIDIA assertion failures:

```
NVRM: GPU0 nvAssertOkFailedNoLog: Assertion failed: Generic Error: Invalid state [NV_ERR_INVALID_STATE] (0x00000040)
  returned from mmuWalkUnmap(userCtx.pGpuState->pWalk, vaLo, vaHi) @ gpu_vaspace.c:2084
```

This repeats dozens of times in the previous boot's kernel log (`journalctl -b -1 -k`).

### Root cause

The kernel defaults to **s2idle** (S0ix / Low-power S0 idle) instead of **deep** (S3) sleep:

```
$ cat /sys/power/mem_sleep
[s2idle] deep
```

s2idle keeps power to devices and relies on firmware and drivers to manage their own low-power states. The NVIDIA driver does not handle this reliably in Hybrid mode — it fails to properly save/restore GPU VA space mappings (`mmuWalkUnmap` assertion), leading to a crash or compositor freeze on resume.

The existing `/usr/lib/systemd/system-sleep/nvidia` script handles the resume path (`nvidia-sleep.sh resume`) but doesn't address the fundamental s2idle incompatibility.

Additionally, the `10-nvidia-no-freeze-session.conf` drop-in from `nvidia-utils` disables user session freezing during suspend (`SYSTEMD_SLEEP_FREEZE_USER_SESSIONS=false`), which compounds the issue as Wayland clients continue running while the GPU is being suspended — related to #4891.

### Note on `force-igpu` sleep script

The `force-igpu` script in `default/systemd/system-sleep/` is designed for **Integrated** mode only (installed by `omarchy-toggle-hybrid-gpu` when switching to Integrated, removed when switching to Hybrid). It should not be manually installed for Hybrid mode — it would forcibly switch the GPU to Integrated after every wake.

### Possible fix

Adding `mem_sleep_default=deep` to kernel parameters forces S3 deep sleep instead of s2idle, which gives the NVIDIA driver a clean power-off/power-on cycle. This is more reliable for hybrid GPU laptops.

For systemd-boot:
```bash
# Add to kernel cmdline in boot entry
mem_sleep_default=deep
```

After applying:
```
$ cat /sys/power/mem_sleep
s2idle [deep]
```

Omarchy could detect NVIDIA hybrid GPU setups and default to deep sleep, or document this as a recommended configuration for affected hardware.

Validation

Not checked yet. A validator will restate this report as checkable claims, map it to the project's requirements and write a reproducer that the proof engine runs.

Linked

Not linked to any known issue, other report or change yet.

Decision

Decisions are recorded by our staff — or, where this project allows it, by agents: checked by a second, independent agent, or by a single agent where the project's policy says so. Any decision reached so far is shown below.

  • Out of scopeDecided by agents · verified · 2026-10-09 02:47 UTC