Nomarchy/tools/monitor-fallback.nix

# Dock-transition hotplug harness — full softGL Hyprland VM with real KVM
# acceleration when available.  It uses QEMU's Virtual-1 DRM output as the
# laptop panel and creates a named DP-1 headless output as the dock monitor,
# exercises the generated transition helper and the live IPC watcher, and
# proves the safety sequence that regressed on hardware:
#
#   dock:   all internal workspaces -> external, focus external, eDP disabled,
#           low-level logind lid-switch inhibitor held
#   undock: eDP enabled before workspace restoration; inhibitor remains while
#           the simulated lid is closed, then releases when it opens
#
# Every wait is bounded.  Maintainer V2 harness (full desktop, so intentionally
# not part of the cheap default checks):
#   nix build --impure -f tools/monitor-fallback.nix --no-link -L
let
  flake = builtins.getFlake ("git+file://" + toString ../.);
  inherit (flake.inputs) nixpkgs home-manager;
  pkgs = import nixpkgs {
    system = "x86_64-linux";
    overlays = [ flake.overlays.default ];
    config.allowUnfree = true;
  };
in
pkgs.testers.runNixOSTest {
  name = "dock-transition";
  node.pkgsReadOnly = false;
  node.specialArgs = { username = "nomarchy"; };
  nodes.machine = { pkgs, ... }: {
    imports = [ flake.nixosModules.nomarchy home-manager.nixosModules.home-manager ];
    virtualisation.memorySize = 4096;
    virtualisation.cores = 4;
    hardware.graphics.enable = true;
    environment.variables.LIBGL_ALWAYS_SOFTWARE = "1";
    boot.initrd.availableKernelModules = [ "virtio_gpu" ];
    virtualisation.qemu.options = [ "-vga none" "-device virtio-gpu-pci,xres=1920,yres=1080" ];
    users.users.nomarchy = {
      isNormalUser = true;
      password = "test";
      extraGroups = [ "wheel" "video" ];
    };
    nomarchy.system.stateFile = flake + "/themes/boreal.json";
    systemd.tmpfiles.rules = [
      "d /home/nomarchy/.nomarchy 0755 nomarchy users -"
      "C /home/nomarchy/.nomarchy/state.json 0644 nomarchy users - ${flake + "/themes/boreal.json"}"
      "f /run/nomarchy-test-lid-state 0644 nomarchy users - state:_closed"
    ];
    services.greetd.settings.initial_session = {
      command = "start-hyprland";
      user = "nomarchy";
    };
    home-manager.useGlobalPkgs = true;
    home-manager.users.nomarchy = {
      imports = [ flake.homeModules.nomarchy ];
      nomarchy.stateFile = flake + "/themes/boreal.json";
      nomarchy.idle.enable = false;
      # Audio has its own generated-artifact guard; avoid restarting the VM's
      # graph during display-only hotplug assertions.
      nomarchy.dockAudio.enable = false;
      home.packages = [ pkgs.foot ];
      # QEMU's DRM output is Virtual-1 rather than eDP-*; this test-only
      # override exercises the same production lifecycle with that connector.
      wayland.windowManager.hyprland.settings.env = [
        "NOMARCHY_LID_STATE_FILE,/run/nomarchy-test-lid-state"
        "NOMARCHY_INTERNAL_OUTPUT,Virtual-1"
      ];
    };
  };
  testScript = ''
    import json

    machine.wait_for_unit("multi-user.target")
    machine.wait_for_file("/run/user/1000/hypr", 180)
    machine.sleep(15)
    machine.wait_until_succeeds(
        "pgrep -af nomarchy-display-profile-watch >/dev/null", timeout=30
    )

    env = ("XDG_RUNTIME_DIR=/run/user/1000 "
           "HYPRLAND_INSTANCE_SIGNATURE=$(ls /run/user/1000/hypr | head -1) "
           "WAYLAND_DISPLAY=wayland-1 ")
    hy = f"su - nomarchy -c '{env}hyprctl "
    user = lambda command: machine.succeed(f"su - nomarchy -c '{env}{command}'")

    def mons(all_outputs=False):
        suffix = "-j monitors all'" if all_outputs else "-j monitors'"
        return json.loads(machine.succeed(hy + suffix))

    def names(all_outputs=False):
        return [m["name"] for m in mons(all_outputs)]

    initial = names()[0]
    print(f"bootstrap output: {initial}")

    assert initial == "Virtual-1", f"test override expects Virtual-1, got {initial}"

    # Two real mapped clients keep workspaces 1 and 2 alive for handoff.
    machine.succeed(hy + "dispatch workspace 1'")
    user("foot --app-id dock-test-1 >/dev/null 2>&1 &")
    machine.wait_until_succeeds(hy + "-j clients' | jq -e 'length >= 1'", timeout=30)
    machine.succeed(hy + "dispatch workspace 2'")
    user("foot --app-id dock-test-2 >/dev/null 2>&1 &")
    machine.wait_until_succeeds(hy + "-j clients' | jq -e 'length >= 2'", timeout=30)

    machine.succeed(hy + "output create headless DP-1'")
    machine.wait_until_succeeds(
        hy + "-j monitors' | jq -e 'any(.[]; .name == \"DP-1\")'", timeout=20
    )
    machine.wait_until_succeeds(
        "systemd-inhibit --list --json=short | jq -e 'any(.[]; .what == \"handle-lid-switch\" and .why == \"Safe dock/undock display transition\" and .mode == \"block\")'",
        timeout=20,
    )
    print("external added: low-level lid inhibitor held")

    transition = "$(command -v nomarchy-display-transition)"
    user(f"{transition} dock Virtual-1 DP-1")
    machine.wait_until_succeeds(
        hy + "-j monitors' | jq -e 'length == 1 and .[0].name == \"DP-1\"'", timeout=20
    )
    ws = json.loads(machine.succeed(hy + "-j workspaces'"))
    assert {w["id"] for w in ws} >= {1, 2}, ws
    assert all(w["monitor"] == "DP-1" for w in ws if w["id"] in (1, 2)), ws
    all_state = {m["name"]: m.get("disabled", False) for m in mons(True)}
    assert all_state.get("Virtual-1") is True, all_state
    assert mons()[0].get("focused") is True, mons()
    print("dock transition: both workspaces on DP-1, internal disabled")

    # #148: every home-manager activation fires `hyprctl reload`, which
    # re-reads the generated config's catch-all `monitor = ,preferred,auto,1`
    # and would re-light Virtual-1 even though the session is still docked
    # with the (simulated) lid shut.  `enforce_dock_intent` — the tick-driven
    # invariant in the watcher — is what is supposed to put it back off.
    # Before the #148 fix it also required `settings.displayProfileAuto`
    # (the *named-profile* auto-switch toggle, unrelated to dock intent), and
    # this harness never sets that — it stays at its schema default `false`
    # — which is exactly the condition the bug shipped under: the invariant
    # silently never fired for any session that had not separately opted
    # into auto profiles.  Re-assert the lid is shut (it already is, from the
    # tmpfiles seed, but do so explicitly since this scenario depends on it),
    # reload, and prove the watcher re-disables the panel and logs the
    # re-dock action.
    machine.succeed("printf 'state: closed\\n' > /run/nomarchy-test-lid-state")
    machine.succeed(hy + "reload'")
    # A reload re-lights the *already-present* (disabled) Virtual-1 connector,
    # which emits no `monitoradded` — so the watcher only re-disables it on its
    # 1s idle tick (chatty IPC events `continue` past reconcile). Under this
    # softGL VM's load (swaync crash-loops on the absent Vulkan device, see the
    # journal) that idle tick is badly starved and the re-dock lands ~15-20s
    # out — a real, prompt re-dock on hardware, but right on the old 15s bound,
    # so this flaked. Bounded generously, in line with the 20-30s waits above.
    machine.wait_until_succeeds(
        hy + "-j monitors' | jq -e 'all(.[]; .name != \"Virtual-1\")'", timeout=45
    )
    machine.succeed("journalctl -t nomarchy-display-watch | grep -q 'action=re-dock'")
    print("reload-mid-dock: panel re-disabled by enforce_dock_intent; re-dock logged (#148)")

    # QEMU's synthetic headless backend aborts Hyprland if its last active
    # output is deleted while the DRM output is disabled.  Exercise the exact
    # production recovery primitive first, then remove DP-1: the real
    # monitorremoved event still proves that the watcher retains the inhibitor
    # until lid-open, without mistaking a backend limitation for product
    # behavior.  Physical cable-yank timing remains the bounded V3 check.
    #
    # Undocking a quiescent output is therefore all this harness can assert —
    # and that gap is exactly what shipped a black panel in rounds 6 and 7.
    # The real fault needs ZERO enabled outputs: with the panel disabled and
    # the external gone, Hyprland 0.55.4 accepts `keyword monitor` and never
    # flushes it, so the panel stays dark until some DRM event arrives (only
    # `hyprctl reload` escapes it — probed on hardware 2026-07-14).  Note the
    # QEMU abort above is that same degeneracy seen from the other side, so
    # this harness cannot reach the state without dying in it: the undock here
    # always runs while DP-1 is still enabled, i.e. the one case that never
    # needed the fix.  A pass below is NOT proof of the reload escalation.
    user(f"{transition} undock Virtual-1")
    machine.wait_until_succeeds(
        hy + "-j monitors' | jq -e 'any(.[]; .name == \"Virtual-1\")'", timeout=20
    )
    machine.wait_until_succeeds(
        hy + "-j workspaces' | jq -e 'all(.[]; (.id != 1 and .id != 2) or .monitor == \"Virtual-1\")'",
        timeout=20,
    )
    machine.succeed(hy + "output remove DP-1'")
    machine.wait_until_succeeds(
        hy + "-j monitors' | jq -e 'length == 1 and .[0].name == \"Virtual-1\"'", timeout=20
    )
    machine.succeed(
        "systemd-inhibit --list --json=short | jq -e 'any(.[]; .what == \"handle-lid-switch\" and .why == \"Safe dock/undock display transition\" and .mode == \"block\")'"
    )
    print("undock transition: eDP/workspaces restored; closed-lid inhibitor retained")

    machine.succeed("printf 'state: open\\n' > /run/nomarchy-test-lid-state")
    machine.wait_until_fails(
        "systemd-inhibit --list --json=short | jq -e 'any(.[]; .what == \"handle-lid-switch\" and .why == \"Safe dock/undock display transition\" and .mode == \"block\")'",
        timeout=20,
    )
    machine.fail("test -e /run/user/1000/nomarchy-dock-lid-inhibitor.pid")
    print("lid-open transition: inhibitor released; normal future lid policy restored")
  '';
}