Files
HermenixandClaude Opus 5 de408e4b7f the surface: CLI verbs, the module's options, and three checks the compositor votes in
Everything below the CLI was merged and green against fakes; none of it was
reachable and none of it had ever met a compositor. This wires it up and then
puts the compositor in a position to disagree.

CLI (src/hhstream/cli.py, src/hhstream_session/**)

  session container [ID], session reconcile [ID], and the seven pointer verbs
  from PLAN §9. `start` still prints the bare session id and nothing else.
  `list` gains WS / OUTPUT / POINTER, with the gap marker where a value is
  genuinely unknown rather than a blank that reads as zero -- the POINTER
  column needs a second, read-only `pointer.state` call because the `list`
  reply carries no lease state.

  The lease id is cached in `hhstream_session.lease` (under
  $XDG_RUNTIME_DIR), because PLAN §9's gesture verbs take no --lease and the
  wire requires a lease_id on every one of them; --lease and $HHSTREAM_LEASE
  override it, and a `no_lease` refusal drops the stale entry.

  Exit codes are the contract's: a gesture with no lease is 9 and never
  reaches the broker, and a compositor failure is 10 with a message that says
  the session is still running. `kill` now warns about
  `session.container_survivors` -- windows the compositor still shows -- and
  distinguishes `null` ("could not ask") from `[]`.

  `shot`/`record --output NAME` capture one container through `grim -o`; the
  frame socket publishes the main output only, so it cannot do it.

Module (nix/module.nix)

  sessions.containers.{enable,max,mode,workspaceBase},
  sessions.pointer.{enable,leaseTtl,leaseMax}, rendered into the generated
  TOML with §10's key names. hhstream-pointer.service: the operator's account,
  supplementary group hhstream, After=/BindsTo= the compositor,
  Restart=on-failure, KillMode=process, and the command line §3.4 freezes.
  tmpfiles for /run/hhstream plus RuntimeDirectoryPreserve on the broker,
  without which restarting the broker deletes the driver's socket out from
  under it. The runtime-directory length assertion is untouched.

  And ExecStartPost on the compositor, opening its IPC directory to its own
  group: Hyprland creates <runtime>/hypr with S_IRWXU, so the broker -- a
  different account by design -- could not reach the request or event socket
  at all, and degraded silently exactly as it is built to.

Checks (checks/**)

  vm-ipc: both sockets at their frozen names, measured against sun_path's 108
  bytes, and a socket2 subscriber that receives an openwindow for a window the
  test creates. The subscriber is the shipped HyprlandIPC, not a second
  implementation: this check exists to prove that this client hears this
  compositor.

  vm-containers: one hh-* output per session at the container mode, the window
  on it, a second window of the same application beside it, an unmatched
  window quarantined, grim -o isolating one container, teardown removing the
  output and freeing the slot. PLAN experiment 3 is answered: `hyprctl output
  create headless` works.

  vm-pointer: a version-2 virtual pointer, a warp and a click whose
  coordinates the *application* reports, BUSY naming the holder, an expired
  lease swept with its held button lifted, and a broker restart that leaves
  the lock free, the bookmark intact and the driver reachable.

  checks/lib gains the guest the three share, their test-script helpers (no
  assertions live there), an event-socket subscriber and a Wayland client that
  says where it was clicked -- nothing in nixpkgs both honours
  $HHSTREAM_APP_ID and reports pointer events.

Reported, not worked around: the broker reads none of the container or
pointer config (policy.py takes [defaults] and [apps] only), `start` has no
no_container field so --no-container is refused rather than silently ignored,
and nix/hyprland-conf.nix sets input:follow_mouse = 0, which sends pointer
events to the focused window only -- measured, with the cursor at the aimed
pixel and no client told. README.dev.md has the evidence for each.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WQy29x9AfHWyzNd1f7XPGK
2026-10-03 06:58:44 +00:00

220 lines
8.7 KiB
Nix

# The test-script helpers the three session checks share: how to run something
# as the session account, how to call the CLI, how to ask the compositor, and
# what to dump when an assertion fails.
#
# Helpers only. There is not one assertion in this file, deliberately: a shared
# prelude that asserted things would make each check's own script an incomplete
# account of what it proves, and the whole point of these scripts is that they
# can be read top to bottom when one of them goes red.
{
hhstream,
coreutils,
sessionUser,
compositorRuntime ? "/run/hypr-stream",
controlSocket ? "/run/hhstream/control.sock",
}:
''
import json
import time
USER = "${sessionUser}"
CONTROL = "${controlSocket}"
CONTROL_DIR = "/run/hhstream"
COMPOSITOR_RUNTIME = "${compositorRuntime}"
HHSTREAM = "${hhstream}/bin/hhstream"
TRUE = "${coreutils}/bin/true"
FALSE = "${coreutils}/bin/false"
# The module pins the compositor's XDG_RUNTIME_DIR; the instance signature
# under it is the only part not known ahead of time. Root may connect to the
# IPC socket even though it belongs to `hhstream`. Same shape as
# vm-hyprland.nix, which asks this for monitors.
HYPRCTL = (
"set -eu; "
f"sig=$(ls -1 {COMPOSITOR_RUNTIME}/hypr | head -n1); "
f"XDG_RUNTIME_DIR={COMPOSITOR_RUNTIME}"
" HYPRLAND_INSTANCE_SIGNATURE=$sig hyprctl -j"
)
#: A transient `systemd-run` unit inherits systemd's own minimal PATH, not
#: the guest's. That is an artefact of how these checks invoke things, not a
#: property of the system under test: a person's login has the system profile
#: on PATH, and `hhstream shot --output` looks `grim` up there. Measured, by
#: a check that failed with the CLI's own "grim is not on PATH, and --output
#: needs it" -- which is the right message and the wrong environment.
SYSTEM_PATH = "/run/current-system/sw/bin"
def as_user(command, env=""):
"""`command` (absolute path plus argv, never a shell) run as USER.
A transient unit rather than `su`, because it gives a clean argv with no
login shell in the middle, and `--pipe --wait` propagates the command's
own exit status -- which each check proves before anything depends on it.
"""
return (
"systemd-run --quiet --collect --pipe --wait"
f" --uid={USER} --setenv=HOME=/home/{USER}"
f" --setenv=PATH={SYSTEM_PATH} {env} -- {command}"
)
def background(unit, command, env="", user=None):
"""A transient unit that keeps running after the call returns.
For the programs a check must have running *before* the thing they
witness happens: the event subscriber, and a window that is not a
session.
No `--collect`: the unit has to stay loaded after it exits, because its
exit status is part of what is asserted, and a collected unit is
unloaded the moment it stops.
`user=None` means root, which is how the driver already reaches the
compositor's IPC socket (see HYPRCTL); pass USER for something that
should be subject to the same permissions an operator has.
"""
who = "" if user is None else f" --uid={user} --setenv=HOME=/home/{user}"
return (
f"systemd-run --quiet --unit={unit}{who}"
f" --setenv=PATH={SYSTEM_PATH} {env} -- {command}"
)
def compositor_env():
"""`--setenv` flags that point a Wayland client at this session.
The socket name is discovered, never assumed: the compositor binds the
first free `wayland-N`, and a crashed one leaves `wayland-0` behind.
"""
display = machine.succeed(
f"cd {COMPOSITOR_RUNTIME} && ls -1 wayland-* | grep -v '[.]lock$' | head -n1"
).strip()
assert display, f"no wayland socket in {COMPOSITOR_RUNTIME}"
return (
f"--setenv=XDG_RUNTIME_DIR={COMPOSITOR_RUNTIME}"
f" --setenv=WAYLAND_DISPLAY={display}"
)
def cli(args):
"""`hhstream session …` as USER, against the control socket."""
return as_user(f"{HHSTREAM} session --socket {CONTROL} {args}")
def cli_json(args, timeout=120):
"""Run a `--json` verb and parse what it printed. Fails loudly if it did not."""
out = machine.succeed(cli(args), timeout=timeout)
try:
return json.loads(out)
except json.JSONDecodeError:
raise AssertionError(f"not JSON from `session {args}`: {out!r}")
def compositor(resource):
"""`hyprctl -j <resource>`, parsed. The compositor is the authority here."""
return json.loads(machine.succeed(f"{HYPRCTL} {resource}", timeout=120))
def monitor_named(name):
for entry in compositor("monitors"):
if entry.get("name") == name:
return entry
return None
def wait_for(predicate, timeout, what, interval=1.0):
"""Poll a *Python* predicate to a deadline, and return its value.
`machine.wait_until_succeeds` can only wait on a shell command, which
forces anything with structure into jq. This is for the waits whose
condition is a parsed value -- "the client has reported a third button
event" -- where a shell pipeline would be both unreadable and quietly
wrong when `grep -c` prints 0 and exits 1.
It waits for an *event that must happen*, never for an assertion to come
good: every caller below asserts after it returns, not inside it.
"""
deadline = time.monotonic() + timeout
while True:
value = predicate()
if value:
return value
if time.monotonic() > deadline:
raise AssertionError(f"timed out after {timeout}s waiting for {what}")
time.sleep(interval)
def dump(label, session_id=None):
print(f"########## diagnostics: {label} ##########")
status, out = machine.execute("hhs-hyprland-diagnostics", timeout=300)
print(out)
print(f"----- hhs-hyprland-diagnostics exit {status} -----")
commands = [
"systemctl --no-pager --full status hhstream-sessions.service",
"journalctl --no-pager --boot --lines=200 --unit hhstream-sessions.service",
"systemctl --no-pager --full status hhstream-pointer.service",
"journalctl --no-pager --boot --lines=200 --unit hhstream-pointer.service",
f"ls -la {CONTROL_DIR} /var/lib/hhstream/sessions",
"ps -eo pid,ppid,pgid,sid,user,stat,comm,args --sort=pid",
cli("list --all"),
cli("container"),
cli("pointer state"),
]
if session_id:
commands += [
cli(f"status {session_id}"),
# Every session's captured output, not just the named one: when
# a pointer event goes to the wrong window, the wrong window's
# log is the evidence and the named session's is empty.
cli("list"),
# The application's own captured output: the only place a failed
# launch, or a pointer event that never arrived, explains itself.
cli(f"logs {session_id} --lines 200"),
]
for command in commands:
status, out = machine.execute(f"{command} 2>&1", timeout=180)
print(f"----- {command} (exit {status}) -----")
print(out)
for resource in ("monitors", "workspaces", "clients", "cursorpos"):
status, out = machine.execute(f"{HYPRCTL} {resource} 2>&1", timeout=180)
print(f"----- hyprctl {resource} (exit {status}) -----")
print(out)
print("########## end diagnostics ##########")
def bring_up(label="the session stack"):
"""Compositor, then broker, then the socket each exists to serve.
`wait_for_unit` alone is not enough for the broker and that is measured,
not assumed: the unit is `Type=simple` with no readiness protocol, so
systemd calls it active the instant it has forked. A daemon that exits
two milliseconds later and is restarted every second passes it. The
honest test of "it came up" is the socket plus a restart count of zero.
"""
try:
machine.wait_for_unit("hyprland-headless.service", timeout=300)
machine.wait_for_file(f"{COMPOSITOR_RUNTIME}/hypr", timeout=120)
machine.wait_for_unit("hhstream-sessions.service", timeout=180)
machine.wait_for_file(CONTROL, timeout=120)
machine.succeed("systemctl is-active hhstream-sessions.service")
restarts = machine.succeed(
"systemctl show -p NRestarts --value hhstream-sessions.service"
).strip()
assert restarts == "0", (
f"the broker has restarted {restarts} time(s) already: it is "
"crash-looping, not running"
)
except Exception:
dump(f"{label} did not come up")
raise
''