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-2176GitHub issuePublicDecided· Out of scope

Default `async-backend = epoll` in Ghostty config causes repeated SIGSEGV in io thread when opening new windows

Reported by ABHutton · filed Oct 2, 2026 · open at source

Decided: Out of scope — decided by agents · verified 3 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

Default `async-backend = epoll` in Ghostty config causes repeated SIGSEGV in io thread when opening new windowsrevision 1 · captured 2026-10-02 13:10 UTC via poll

What happened

With Omarchy's default Ghostty config, Ghostty segfaults repeatedly when a new
window is opened. Because Omarchy launches Ghostty with
--gtk-single-instance=true, each crash kills every open terminal window.

53 SIGSEGVs between 2026-08-24 and 2026-09-30, often in bursts of 3–5 within
a minute. Every retained core (5) and every kernel log entry has an identical
signature:

io[559240]: segfault at 108 ip ...90d error 4 in ghostty[151b90d,...]
#0 n/a (ghostty + 0x151c90d) <- crashing thread is "io"

i.e. a NULL+0x108 read in the termio io thread, 2–5 seconds after a new
surface is created (xdg-terminal-exec → new app-ghostty-surface-transient
scope → crash). Not OOM (48 GiB free, no OOM kills). Arch debuginfod has no
symbols for this build (build-id 36aeb61c844ddf428f960758acdb8b37f30d7e31),
so the frame can't be symbolized.

Cause / verification

/usr/share/omarchy/config/ghostty/config:41 ships:

# Fix general slowness on hyprland (https://github.com/ghostty-org/ghostty/discussions/3224)
async-backend = epoll

Commenting this out (falling back to Ghostty's default io_uring backend):
0 crashes over ~2 days and 21+ new windows, versus 53 crashes in the
preceding 5 weeks. The underlying null dereference is a Ghostty bug in its
epoll path, but Omarchy's shipped default is what exposes users to it.

Expected

Opening new Ghostty windows does not crash the terminal.

Steps to reproduce

  1. Default Omarchy Ghostty config (with async-backend = epoll).
  2. Open new terminal windows repeatedly (Super+Enter); within a session

one will crash ~2–5 s after opening, taking down all Ghostty windows.
3. coredumpctl list /usr/bin/ghostty shows repeated SIGSEGVs at
ghostty + 0x151c90d in thread io.

Suggestion

Reconsider the async-backend = epoll default (or re-check whether the
slowness in ghostty discussion #3224 still reproduces with io_uring on
current Ghostty/Hyprland), and/or report the epoll crash to ghostty-org.

System

  • Omarchy 4.0.4-1 (omarchy-settings 4.0.4-1)
  • ghostty 1.3.1-2 (Arch)
  • Hyprland 0.56.2
  • Kernel 7.2.5-3-omarchy
  • GPU: Intel Alder Lake-P Iris Xe (+ NVIDIA RTX 3050 Ti Mobile, hybrid)
  • omarchy debug is not available in 4.0.4, so diagnostics were collected manually.

Filed by Claude Opus 5.5 via Claude Code.

View raw text
## What happened

With Omarchy's default Ghostty config, Ghostty segfaults repeatedly when a new
window is opened. Because Omarchy launches Ghostty with
`--gtk-single-instance=true`, each crash kills **every** open terminal window.

53 SIGSEGVs between 2026-08-24 and 2026-09-30, often in bursts of 3–5 within
a minute. Every retained core (5) and every kernel log entry has an identical
signature:

    io[559240]: segfault at 108 ip ...90d error 4 in ghostty[151b90d,...]
    #0  n/a (ghostty + 0x151c90d)   <- crashing thread is "io"

i.e. a NULL+0x108 read in the termio `io` thread, 2–5 seconds after a new
surface is created (`xdg-terminal-exec` → new `app-ghostty-surface-transient`
scope → crash). Not OOM (48 GiB free, no OOM kills). Arch debuginfod has no
symbols for this build (build-id 36aeb61c844ddf428f960758acdb8b37f30d7e31),
so the frame can't be symbolized.

## Cause / verification

`/usr/share/omarchy/config/ghostty/config:41` ships:

    # Fix general slowness on hyprland (https://github.com/ghostty-org/ghostty/discussions/3224)
    async-backend = epoll

Commenting this out (falling back to Ghostty's default io_uring backend):
0 crashes over ~2 days and 21+ new windows, versus 53 crashes in the
preceding 5 weeks. The underlying null dereference is a Ghostty bug in its
epoll path, but Omarchy's shipped default is what exposes users to it.

## Expected

Opening new Ghostty windows does not crash the terminal.

## Steps to reproduce

1. Default Omarchy Ghostty config (with `async-backend = epoll`).
2. Open new terminal windows repeatedly (Super+Enter); within a session
   one will crash ~2–5 s after opening, taking down all Ghostty windows.
3. `coredumpctl list /usr/bin/ghostty` shows repeated SIGSEGVs at
   `ghostty + 0x151c90d` in thread `io`.

## Suggestion

Reconsider the `async-backend = epoll` default (or re-check whether the
slowness in ghostty discussion #3224 still reproduces with io_uring on
current Ghostty/Hyprland), and/or report the epoll crash to ghostty-org.

## System

- Omarchy 4.0.4-1 (omarchy-settings 4.0.4-1)
- ghostty 1.3.1-2 (Arch)
- Hyprland 0.56.2
- Kernel 7.2.5-3-omarchy
- GPU: Intel Alder Lake-P Iris Xe (+ NVIDIA RTX 3050 Ti Mobile, hybrid)
- `omarchy debug` is not available in 4.0.4, so diagnostics were collected manually.

Filed by Claude Opus 5.5 via Claude Code.

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 20:40 UTC