steam-backlog-enforcer/steam_backlog_enforcer/_engagement_probes.py
Krzysztof kuhy Rudnicki b3e629e3f0 Bill gaming time only while the user is actually playing
A running game process was treated as proof of play. On 2026-08-28 the budget
billed ~2h23m while a gatelock screen locker covered the display and the
machine could not be touched at all -- the game was simply still resident.

A tick now bills only when all three hold, evaluated together rather than
short-circuited so the audit log can answer "was the screen ALSO locked?":

  * no gatelock guard owns the screen  (_screen_hold, via /proc/locks)
  * the focused window belongs to a qualifying process or its ancestry
  * some input source has been used within idle_grace_seconds (default 300)

Any probe that cannot answer BILLS the tick and is recorded in `degraded`.
A broken detector must never become an unlimited-gaming exploit.

Three findings from measuring this machine rather than assuming:

  * /dev/input/js0 here is an *ASRock LED Controller* -- the kernel exposes
    the motherboard RGB device as a joystick. A presence check would have
    disabled the idle criterion permanently, so _controller_input watches for
    real events instead (skipping joydev's synthetic open burst).
  * Aseprite's window carries no _NET_WM_PID at all, so focus attribution
    falls back to XRes, which asks the server who owns the window rather than
    trusting the client to have announced itself.
  * /proc/locks reports gatelock's hold without taking it; gatelock's own
    read_claim_if_held detects held-ness by acquiring, which would race a
    guard every 3 seconds.

Idle pauses also refund the grace period once per walk-away, on the
engaged->paused edge only, never across a day rollover, and never on the first
tick after a restart -- the only place the counter moves backwards.

Everything is recorded to /var/log/steam-backlog-enforcer/budget.jsonl, on
change plus a heartbeat rather than one line per 3s tick. Records name the
qualifying PIDs: reconstructing this incident from journald alone was only
possible by inferring accrual from the spacing of two warning notifications.

Also: refuse to bounce Steam while a game is running. ensure_steam_debug_port
relaunches Steam when the CDP port is closed, which kills the running game --
that is how restarting the daemon to deploy this change ended a live session.
It now defers exactly as it already did for in-flight game updates.

Talks to X through ctypes over one cached connection: at a 3s tick, forking
xprintidle and xprop would cost ~40 forks a minute.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019CEsG6BzHtYR9cHcdxPZqG
2026-08-28 19:24:14 +02:00

169 lines
5.2 KiB
Python

"""Turning each raw engagement signal into a pause cause.
Split from :mod:`_playtime_engagement`, which owns the per-tick decision and the
backdate; this owns the three probes and the rule each one applies. Every probe
follows the same contract: append to *causes* when the criterion says "not
playing", append to *degraded* when the probe could not answer, and never raise.
**A probe that fails does not pause the tick.** Failure means the tick is billed.
A detector that stopped working must never become an unlimited-gaming exploit,
so the burden of proof sits on "not playing", not on "playing".
"""
from __future__ import annotations
import logging
from typing import TYPE_CHECKING
from steam_backlog_enforcer._engagement_types import (
CAUSE_FOCUS,
CAUSE_IDLE,
CAUSE_SCREEN_HELD,
PROBE_FOCUS,
PROBE_IDLE,
PROBE_SCREEN_HELD,
CauseTally,
)
from steam_backlog_enforcer._playtime_kill import _MAX_PROCESS_TREE_DEPTH, _read_ppid
from steam_backlog_enforcer._screen_hold import ScreenHoldError, screen_hold
from steam_backlog_enforcer._x_focus import focused_pid
from steam_backlog_enforcer._x_probe import XProbeError
if TYPE_CHECKING:
from pathlib import Path
from steam_backlog_enforcer._controller_input import ControllerActivity
from steam_backlog_enforcer._playtime_state import PlaytimeRules
from steam_backlog_enforcer._x_probe import XProbe
logger = logging.getLogger(__name__)
_INIT_PID = 1
def assess_screen(
holder_path: Path, tally: CauseTally
) -> tuple[bool | None, int | None]:
"""Add the screen-held cause if a gatelock guard owns the display.
Args:
holder_path: The gatelock holder lock.
tally: Accumulating causes and probe failures.
Returns:
The hold state and holding PID, both ``None`` when the probe failed.
"""
try:
hold = screen_hold(holder_path)
except ScreenHoldError as exc:
logger.warning("Screen-hold probe failed; billing this tick: %s", exc)
tally.degraded.append(PROBE_SCREEN_HELD)
return None, None
if hold.held:
tally.causes.append(CAUSE_SCREEN_HELD)
return hold.held, hold.holder_pid
def assess_focus(
probe: XProbe,
rules: PlaytimeRules,
qualifying: set[int],
tally: CauseTally,
) -> int | None:
"""Add the focus cause unless a qualifying process owns the focus.
Args:
probe: X connection.
rules: Policy for this tick.
qualifying: PIDs that would otherwise count.
tally: Accumulating causes and probe failures.
Returns:
The focused PID, or ``None``.
"""
if not rules.require_game_focus:
return None
try:
pid = focused_pid(probe.connect())
except XProbeError as exc:
logger.warning("Focus probe failed; billing this tick: %s", exc)
tally.degraded.append(PROBE_FOCUS)
probe.close()
return None
if not focus_qualifies(pid, qualifying):
tally.causes.append(CAUSE_FOCUS)
return pid
def assess_idle(
probe: XProbe,
controller: ControllerActivity,
rules: PlaytimeRules,
tally: CauseTally,
now_monotonic: float,
) -> tuple[float | None, float | None]:
"""Add the idle cause when every input source has gone quiet.
The controller is polled first and its age folded in with ``min``: the X
idle counter does not see a gamepad, and pausing during controller play
would be the gate's worst failure.
Args:
probe: X connection.
controller: Controller-activity watcher.
rules: Policy for this tick.
tally: Accumulating causes and probe failures.
now_monotonic: Monotonic timestamp for controller ages.
Returns:
The effective idle seconds and the controller idle seconds.
"""
controller.poll(now=now_monotonic)
controller_idle = controller.seconds_since_activity(now=now_monotonic)
try:
pointer_idle = probe.idle_seconds()
except XProbeError as exc:
logger.warning("Idle probe failed; billing this tick: %s", exc)
tally.degraded.append(PROBE_IDLE)
probe.close()
return None, controller_idle
idle = pointer_idle
if controller_idle is not None:
idle = min(idle, controller_idle)
if idle >= rules.idle_grace_seconds:
tally.causes.append(CAUSE_IDLE)
return idle, controller_idle
def focus_qualifies(focus_pid: int | None, qualifying: set[int]) -> bool:
"""Whether *focus_pid* is, or descends from, a qualifying process.
A game's visible window is often owned by a child of the process carrying
``SteamAppId`` — Proton in particular — so the ancestry is walked rather
than compared directly.
Args:
focus_pid: PID owning the focused window, or ``None``.
qualifying: PIDs that count as gaming.
Returns:
Whether the focused window belongs to the game.
"""
if focus_pid is None:
return False
pid = focus_pid
for _ in range(_MAX_PROCESS_TREE_DEPTH):
if pid in qualifying:
return True
if pid <= _INIT_PID:
return False
parent = _read_ppid(pid)
if parent is None:
return False
pid = parent
return False