Files
hermes-agent/tests/scripts/test_run_tests_parallel.py
ethernet 890bbbda1f Merge remote-tracking branch 'origin/main' into ethie/pm-clean
# Conflicts:
#	apps/desktop/e2e/archived-hidden-session-recoverable.spec.ts
#	apps/desktop/e2e/bot-chat-message-agent-friendly-name.spec.ts
#	apps/desktop/e2e/bot-mailbox-unreadable-ticket.spec.ts
#	apps/desktop/e2e/bot-mode-roster-localized.spec.ts
#	apps/desktop/e2e/bot-mode-row-click-mirrors-registry.spec.ts
#	apps/desktop/e2e/bot-mode-tab-shows-bot-name.spec.ts
#	apps/desktop/e2e/bot-roster-group-row-organisation.spec.ts
#	apps/desktop/e2e/bot-roster-ignores-infra-dirs.spec.ts
#	apps/desktop/e2e/bot-roster-timestamp-meta.spec.ts
#	apps/desktop/e2e/bot-roster-user-sections.spec.ts
#	apps/desktop/e2e/bot-routines-pane-narrow.spec.ts
#	apps/desktop/e2e/bot-row-open-recent-session.spec.ts
#	apps/desktop/e2e/bot-tile-ignores-ambient-composer-model.spec.ts
#	apps/desktop/e2e/group-composer-auto-grow.spec.ts
#	apps/desktop/e2e/group-create-gate-remote-roster.spec.ts
#	apps/desktop/e2e/group-prompt-renamed-primary-handle.spec.ts
#	apps/desktop/e2e/hosted-room-backend-continuity.spec.ts
#	apps/desktop/e2e/hosted-room-legacy-store-migration.spec.ts
#	apps/desktop/e2e/settings-scope-chips-bot-title.spec.ts
#	apps/desktop/e2e/worktree-branch-status.spec.ts
#	apps/desktop/electron/backend-probes.test.ts
#	apps/desktop/electron/connection-apply.test.ts
#	apps/desktop/electron/desktop-electron-pin.test.ts
#	apps/desktop/electron/desktop-uninstall.test.ts
#	apps/desktop/electron/gateway-file-download-transport.test.ts
#	apps/desktop/electron/gateway-stop-before-update.test.ts
#	apps/desktop/electron/github-api-auth.test.ts
#	apps/desktop/electron/registry-primary-profile-scope.test.ts
#	apps/desktop/electron/update-api-check.test.ts
#	apps/desktop/electron/update-handoff-marker.test.ts
#	apps/desktop/electron/venv-blocker-scan.test.ts
#	apps/desktop/scripts/after-extract.test.mjs
#	apps/desktop/scripts/local-pack-publish.test.mjs
#	apps/desktop/scripts/tasks-scroll.test.mjs
#	apps/desktop/src/app/settings/model-settings.test.tsx
#	apps/desktop/src/app/updates-overlay.blockers.test.tsx
#	apps/desktop/src/components/desktop-install-overlay.test.tsx
#	apps/desktop/src/lib/update-copy.test.ts
#	scripts/ci/check_os_marker_fakes.py
#	tests-js/desktop-mac-usage-descriptions.test.ts
#	tests-js/node-engine-alignment.test.ts
#	tests/agent/lsp/test_install_and_lint_fixes.py
#	tests/agent/test_command_token_source.py
#	tests/agent/test_compression_boundary_hook.py
#	tests/agent/test_create_openai_client_ssl_verify.py
#	tests/agent/test_custom_provider_ca_probes.py
#	tests/agent/test_endpoint_blackhole.py
#	tests/agent/test_estimator_parity.py
#	tests/agent/test_in_place_compaction.py
#	tests/agent/test_moa_loop_mode.py
#	tests/agent/test_model_metadata.py
#	tests/agent/test_skill_session_platform_gate.py
#	tests/agent/test_skill_utils.py
#	tests/agent/test_ssl_ca_guard.py
#	tests/computer_use/test_doctor.py
#	tests/cron/test_codex_execution_paths.py
#	tests/cron/test_cron_bot_chat_delivery.py
#	tests/cron/test_cron_script.py
#	tests/cron/test_media_delivery_parity.py
#	tests/cron/test_misfire_catchup.py
#	tests/cron/test_parallel_pool.py
#	tests/cron/test_recurring_eagain_redispatch.py
#	tests/gateway/test_choice_picker.py
#	tests/gateway/test_control_socket_windows_live.py
#	tests/gateway/test_dingtalk.py
#	tests/gateway/test_feishu.py
#	tests/gateway/test_feishu_onboard.py
#	tests/gateway/test_gateway_shutdown.py
#	tests/gateway/test_matrix.py
#	tests/gateway/test_model_command_custom_providers.py
#	tests/gateway/test_reasoning_command.py
#	tests/gateway/test_runtime_footer.py
#	tests/gateway/test_session.py
#	tests/gateway/test_session_hygiene.py
#	tests/gateway/test_status.py
#	tests/gateway/test_teams.py
#	tests/gateway/test_turn_lease.py
#	tests/gateway/test_whatsapp_connect.py
#	tests/hermes_cli/test_approvals_command.py
#	tests/hermes_cli/test_auth_store_lock_concurrent.py
#	tests/hermes_cli/test_backup.py
#	tests/hermes_cli/test_banner_git_state.py
#	tests/hermes_cli/test_certifi_repair.py
#	tests/hermes_cli/test_cmd_update.py
#	tests/hermes_cli/test_compat_manifest_targets.py
#	tests/hermes_cli/test_computer_use_cli.py
#	tests/hermes_cli/test_cpr_local_leak.py
#	tests/hermes_cli/test_dashboard_auth_gate.py
#	tests/hermes_cli/test_dashboard_procs_kill_grace.py
#	tests/hermes_cli/test_desktop_lifecycle_windows_live.py
#	tests/hermes_cli/test_doctor.py
#	tests/hermes_cli/test_doctor_command_install.py
#	tests/hermes_cli/test_fleet_config_migration_windows_live.py
#	tests/hermes_cli/test_gateway.py
#	tests/hermes_cli/test_gateway_platform_gating.py
#	tests/hermes_cli/test_gateway_restart_loop.py
#	tests/hermes_cli/test_gateway_task_probe.py
#	tests/hermes_cli/test_gateway_wsl.py
#	tests/hermes_cli/test_gui_command.py
#	tests/hermes_cli/test_install_cua_driver.py
#	tests/hermes_cli/test_kanban_db.py
#	tests/hermes_cli/test_lazy_command_exports.py
#	tests/hermes_cli/test_lazy_refresh_venv_repair.py
#	tests/hermes_cli/test_linux_desktop_entry.py
#	tests/hermes_cli/test_local_runtime.py
#	tests/hermes_cli/test_local_runtime_updates.py
#	tests/hermes_cli/test_managed_uv.py
#	tests/hermes_cli/test_mcp_reload_confirm_gate.py
#	tests/hermes_cli/test_nous_subscription.py
#	tests/hermes_cli/test_npm_engine.py
#	tests/hermes_cli/test_personality_none.py
#	tests/hermes_cli/test_pet_toggle.py
#	tests/hermes_cli/test_plan_reconciliation_windows_live.py
#	tests/hermes_cli/test_plugin_event_bus.py
#	tests/hermes_cli/test_plugin_manifest_v2.py
#	tests/hermes_cli/test_plugin_packs.py
#	tests/hermes_cli/test_plugins_cmd.py
#	tests/hermes_cli/test_plugins_cmd_enable_disable_nested.py
#	tests/hermes_cli/test_process_identity.py
#	tests/hermes_cli/test_profiles.py
#	tests/hermes_cli/test_profiles_sidebar_cache.py
#	tests/hermes_cli/test_pty_bridge.py
#	tests/hermes_cli/test_resolve_turn_limit.py
#	tests/hermes_cli/test_serve_runtime_inventory.py
#	tests/hermes_cli/test_session_vacuum_config.py
#	tests/hermes_cli/test_set_config_value.py
#	tests/hermes_cli/test_signal_handler_kanban_worker.py
#	tests/hermes_cli/test_slash_confirm_windows.py
#	tests/hermes_cli/test_stale_pid_guard.py
#	tests/hermes_cli/test_startup_fast_guards.py
#	tests/hermes_cli/test_status.py
#	tests/hermes_cli/test_telegram_managed_bot.py
#	tests/hermes_cli/test_tools_config.py
#	tests/hermes_cli/test_update_apply_shallow_count.py
#	tests/hermes_cli/test_update_autostash.py
#	tests/hermes_cli/test_update_concurrent_quarantine.py
#	tests/hermes_cli/test_update_fetch_failure_classifier.py
#	tests/hermes_cli/test_update_fleet_probe_resume_token.py
#	tests/hermes_cli/test_update_handoff_backend_reap.py
#	tests/hermes_cli/test_update_handoff_desktop_rebuild.py
#	tests/hermes_cli/test_update_head_moved_gate.py
#	tests/hermes_cli/test_update_host_obligation.py
#	tests/hermes_cli/test_update_import_guard.py
#	tests/hermes_cli/test_update_interrupted_recovery.py
#	tests/hermes_cli/test_update_inventory.py
#	tests/hermes_cli/test_update_launchd_unloaded_gateway.py
#	tests/hermes_cli/test_update_missing_configured_deps.py
#	tests/hermes_cli/test_update_modified_notice.py
#	tests/hermes_cli/test_update_multiplex_migration_hook.py
#	tests/hermes_cli/test_update_no_gateway_restart.py
#	tests/hermes_cli/test_update_orphan_backend_reap.py
#	tests/hermes_cli/test_update_parked_branch_guard.py
#	tests/hermes_cli/test_update_post_pull_syntax_guard.py
#	tests/hermes_cli/test_update_receipt.py
#	tests/hermes_cli/test_update_self_lock.py
#	tests/hermes_cli/test_update_shim_fail_closed.py
#	tests/hermes_cli/test_update_shim_self_lock.py
#	tests/hermes_cli/test_update_sqlite_remediation.py
#	tests/hermes_cli/test_update_stale_dashboard.py
#	tests/hermes_cli/test_update_stale_virtualenv.py
#	tests/hermes_cli/test_update_venv_health.py
#	tests/hermes_cli/test_update_venv_ownership_preflight.py
#	tests/hermes_cli/test_update_wedged_gateway.py
#	tests/hermes_cli/test_update_yes_flag.py
#	tests/hermes_cli/test_update_zip_two_phase.py
#	tests/hermes_cli/test_urllib_security.py
#	tests/hermes_cli/test_ux_messages_auth_config.py
#	tests/hermes_cli/test_ux_messages_startup.py
#	tests/hermes_cli/test_venv_holder_classifier.py
#	tests/hermes_cli/test_verify_console_scripts.py
#	tests/hermes_cli/test_verify_core_dependencies.py
#	tests/hermes_cli/test_web_server.py
#	tests/hermes_cli/test_web_server_console_ws.py
#	tests/hermes_cli/test_web_server_ws_ping.py
#	tests/hermes_cli/test_web_ui_build.py
#	tests/hermes_state/test_fts_rebuild_admission.py
#	tests/hermes_state/test_hermes_state.py
#	tests/plugins/memory/test_memory_lazy_install.py
#	tests/plugins/test_google_meet_plugin.py
#	tests/plugins/test_langfuse_plugin.py
#	tests/plugins/test_security_guidance_plugin.py
#	tests/plugins/test_transform_llm_output_hook.py
#	tests/scripts/desktop_update/test_desktop_update_windows_gateway_flag.py
#	tests/scripts/desktop_update/test_desktop_update_windows_python_handoff.py
#	tests/scripts/desktop_update/test_desktop_update_windows_timestamp.py
#	tests/scripts/install/test_install_clone_throttle_fallback.py
#	tests/scripts/install/test_install_lockfile_churn.py
#	tests/scripts/install/test_install_no_initial_commit.py
#	tests/scripts/install/test_install_sh_browser_install.py
#	tests/scripts/install/test_install_sh_node_prerelease.py
#	tests/scripts/install/test_install_sh_symlink_stomp.py
#	tests/scripts/install/test_install_sh_uv_lock_config.py
#	tests/scripts/install/test_install_unmerged_index.py
#	tests/scripts/test_contributor_map.py
#	tests/scripts/test_run_tests_parallel.py
#	tests/skills/test_competitor_news_monitor_skill.py
#	tests/skills/test_document_to_action_items_skill.py
#	tests/skills/test_google_workspace_setup.py
#	tests/skills/test_google_workspace_setup_deps.py
#	tests/skills/test_grounded_citations_skill.py
#	tests/skills/test_ip_as_logo_skill.py
#	tests/skills/test_live_dashboard_skill.py
#	tests/skills/test_mcp_oauth_remote_gateway_skill.py
#	tests/skills/test_office_document_skills.py
#	tests/skills/test_openclaw_migration.py
#	tests/skills/test_product_price_monitor_skill.py
#	tests/skills/test_scrollcraft_skill.py
#	tests/skills/test_setup_wizard_generator_skill.py
#	tests/skills/test_weekly_review_planning_skill.py
#	tests/test_engines_satisfiable.py
#	tests/test_fast_safe_load.py
#	tests/test_hermes_bootstrap.py
#	tests/test_hermes_constants.py
#	tests/test_hermes_logging.py
#	tests/test_managed_runtime_resolution.py
#	tests/test_model_tools_async_bridge.py
#	tests/test_packaging_build_guard.py
#	tests/test_packaging_metadata.py
#	tests/test_yaml_indent_consistency.py
#	tests/tools/test_approval_timeout_overflow.py
#	tests/tools/test_base_environment.py
#	tests/tools/test_bot_mode_dm.py
#	tests/tools/test_browser_chromium_check.py
#	tests/tools/test_browser_hardening.py
#	tests/tools/test_browser_homebrew_paths.py
#	tests/tools/test_browser_npx_warmup.py
#	tests/tools/test_browser_orphan_reaper.py
#	tests/tools/test_browser_real_profile.py
#	tests/tools/test_browser_use_cli.py
#	tests/tools/test_clipboard.py
#	tests/tools/test_code_execution.py
#	tests/tools/test_code_execution_modes.py
#	tests/tools/test_code_execution_windows_env.py
#	tests/tools/test_computer_use.py
#	tests/tools/test_delegate_liveness_timeout.py
#	tests/tools/test_execute_code_approval_cluster.py
#	tests/tools/test_execution_flag_detection.py
#	tests/tools/test_fal_common.py
#	tests/tools/test_file_operations.py
#	tests/tools/test_file_tools.py
#	tests/tools/test_file_tools_cwd_resolution.py
#	tests/tools/test_file_tools_live.py
#	tests/tools/test_lazy_deps.py
#	tests/tools/test_lazy_deps_durable_target.py
#	tests/tools/test_lazy_deps_managed.py
#	tests/tools/test_local_env_blocklist.py
#	tests/tools/test_local_tempdir.py
#	tests/tools/test_macos_protected_search.py
#	tests/tools/test_mcp_npx_cached_bin.py
#	tests/tools/test_oneshot_completion_linger.py
#	tests/tools/test_process_registry.py
#	tests/tools/test_read_file_schema_gating.py
#	tests/tools/test_skill_improvements.py
#	tests/tools/test_skills_sync.py
#	tests/tools/test_termux_api_detection.py
#	tests/tools/test_tirith_security.py
#	tests/tools/test_transcription_tools.py
#	tests/tools/test_tts_streaming.py
#	tests/tools/test_wake_word.py
#	tests/tui_gateway/test_compute_host_borrowed_lease.py
#	tests/tui_gateway/test_compute_host_turn_protocol.py
#	tests/tui_gateway/test_isolated_orphan_activity.py
#	tests/tui_gateway/test_protocol.py
#	tests/tui_gateway/test_slash_worker_profile_home.py
#	tests/tui_gateway/test_subprocess_encoding.py
#	tests/tui_gateway/test_tui_gateway_server.py
#	ui-tui/src/__tests__/terminalParity.test.ts
#	ui-tui/src/__tests__/termuxComposerLayout.test.ts
#	ui-tui/src/__tests__/textInputFastEcho.test.ts
2026-09-23 07:02:44 -04:00

577 lines
22 KiB
Python

"""Verify scripts/run_tests_parallel.py kills test-spawned grandchildren.
Setup
-----
A test in this file spawns a long-lived Python grandchild that writes
its PID + a nonce to a tempfile, then exits without cleaning up.
With the old ``subprocess.run`` runner, that grandchild would orphan
and outlive the test (and the whole runner). With the current Popen +
``start_new_session`` + ``_kill_tree`` runner, the grandchild gets
SIGKILL'd via process-group kill when its file's pytest exits.
The leaker test always passes — its only job is to spawn a grandchild
and walk away. The verifier runs the runner over the leaker file in a
subprocess, then waits for the grandchild PID to disappear from the
kernel's process table.
POSIX-only: Windows has its own grandchild lifecycle (no shared session,
``taskkill /F /T`` semantics). Marked accordingly.
"""
from __future__ import annotations
import json
import shutil
import os
import subprocess
import sys
import textwrap
import time
from pathlib import Path
import pytest
@pytest.fixture(autouse=True)
def isolated_probe_environment(monkeypatch):
# Probe files exercise pytest/runner mechanics, not installed third-party
# plugins. Autoloading the developer environment changes their startup cost.
monkeypatch.setenv("PYTEST_DISABLE_PLUGIN_AUTOLOAD", "1")
def _probe_root(tmp_path):
root = tmp_path / "runner-root"
scripts = root / "scripts"
(scripts / "ci").mkdir(parents=True, exist_ok=True)
real = Path(__file__).resolve().parents[2] / "scripts"
shutil.copy2(real / "run_tests_parallel.py", scripts)
# The runner shares the platforms() spec resolver with the CI lane selector.
shutil.copy2(real / "ci" / "list_os_marked_tests.py", scripts / "ci")
return root
def _pid_alive(pid: int) -> bool:
"""POSIX: send signal 0 to probe whether ``pid`` is still alive.
``os.kill(pid, 0)`` raises ``ProcessLookupError`` if the process is
gone, ``PermissionError`` if it exists but we can't signal it
(someone else's pid). We treat PermissionError as "alive" because
the process exists and that's all we need to know.
"""
if sys.platform == "win32": # pragma: no cover — POSIX-only test
# On Windows we'd use OpenProcess + GetExitCodeProcess; this
# test is skipped on Windows so the path is unreachable.
raise RuntimeError("_pid_alive POSIX-only")
try:
os.kill(pid, 0)
except ProcessLookupError:
return False
except PermissionError:
return True
return True
def test_progress_output_tolerates_legacy_stdout_encoding(tmp_path: Path) -> None:
"""Progress glyphs must not crash the runner on non-UTF-8 consoles."""
repo_root = _probe_root(tmp_path)
runner = repo_root / "scripts" / "run_tests_parallel.py"
probe_dir = tmp_path / "probe"
probe_dir.mkdir()
probe = probe_dir / "test_probe_smoke.py"
probe.write_text("def test_smoke():\n assert True\n", encoding="utf-8")
env = os.environ.copy()
env["PYTHONIOENCODING"] = "cp1252:strict"
proc = subprocess.run(
[
sys.executable,
str(runner),
"--paths",
str(probe_dir),
"-j",
"1",
"--file-timeout",
"30",
],
cwd=probe_dir,
env=env,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
text=True,
timeout=60,
)
assert proc.returncode == 0, proc.stdout
assert "UnicodeEncodeError" not in proc.stdout
assert "1 tests passed" in proc.stdout
@pytest.mark.platforms("posix")
@pytest.mark.live_system_guard_bypass
def test_grandchild_leak_is_killed_by_runner(tmp_path: Path) -> None:
"""Run the parallel runner over a probe file and verify cleanup.
1. Materialize a probe file that spawns a long-lived grandchild and
writes its PID to disk before exiting.
2. Invoke ``scripts/run_tests_parallel.py`` against the probe file.
3. Wait for the grandchild PID to vanish (poll for ~5s).
4. Assert the runner exited cleanly AND the grandchild is dead.
"""
repo_root = _probe_root(tmp_path)
runner = repo_root / "scripts" / "run_tests_parallel.py"
assert runner.exists(), f"runner missing at {runner}"
# Probe lives in a temp dir, NOT under tests/, so the regular suite
# never picks it up — only our explicit invocation does.
probe_dir = tmp_path / "probe"
probe_dir.mkdir()
probe = probe_dir / "test_probe_leaker.py"
nonce = f"{os.getpid()}-{int(time.time() * 1000)}"
handoff = tmp_path / f"grandchild-{nonce}.json"
probe_src = textwrap.dedent(f"""
import json, os, subprocess, sys, time
from pathlib import Path
HANDOFF = Path({str(handoff)!r})
def test_spawns_grandchild_and_walks_away():
# Long-lived grandchild: detached, ignores SIGTERM (we want
# SIGKILL or process-group kill to be the only thing that
# works, simulating a misbehaving server).
child = subprocess.Popen(
[
sys.executable, "-c",
"import os, signal, sys, time; "
"signal.signal(signal.SIGTERM, signal.SIG_IGN); "
"sys.stdout.write(f'gc-pgid={{os.getpgid(0)}} gc-pid={{os.getpid()}}\\\\n'); "
"sys.stdout.flush(); "
"time.sleep(600)",
],
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
# IMPORTANT: do NOT pass start_new_session here. We want
# the grandchild to inherit the pytest subprocess's
# process group, so when the runner kills the group the
# grandchild dies too.
)
# Read the first line so we can record gc's pgid in the
# handoff, then walk away — don't close the pipe (would
# signal EOF and let the child see SIGPIPE on next write).
first_line = child.stdout.readline().decode().strip()
HANDOFF.write_text(json.dumps({{
"pid": child.pid,
"diag": first_line,
"test_pid": os.getpid(),
"test_pgid": os.getpgid(0),
}}))
assert child.pid > 0
""").strip()
probe.write_text(probe_src + "\n")
# Run the parallel runner against just the probe file. The runner
# discovers under ``tests/`` by default, so we override via --paths.
proc = subprocess.run(
[
sys.executable,
str(runner),
"--paths",
str(probe_dir),
"-j",
"1",
# Tight per-file timeout: the probe finishes in <1s, no
# need for 10min.
"--file-timeout",
"30",
],
cwd=probe_dir,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
# The runner declares its stdio UTF-8 (see _make_stdio_glyph_safe);
# decode the same way so ✓-glyph assertions hold on Windows, where
# text=True alone would decode with the locale codec (cp1252).
encoding="utf-8",
errors="replace",
timeout=60,
)
assert handoff.exists(), (
f"probe never wrote handoff file; runner output:\n{proc.stdout}"
)
handoff_data = json.loads(handoff.read_text())
grandchild_pid = handoff_data["pid"]
diag = handoff_data.get("diag", "(no diag)")
test_pid = handoff_data.get("test_pid")
test_pgid = handoff_data.get("test_pgid")
handoff.unlink()
# The runner must have exited cleanly (probe test passes).
assert proc.returncode == 0, (
f"runner exited {proc.returncode}; output:\n{proc.stdout}"
)
# The grandchild must be gone. Poll for a bit because process-group
# SIGKILL + reaping isn't synchronous; on a loaded box it can take
# a beat.
deadline = time.monotonic() + 5.0
while time.monotonic() < deadline:
if not _pid_alive(grandchild_pid):
break
time.sleep(0.05)
else:
# Test cleanup: kill the leaked grandchild ourselves so a
# FAILED assertion doesn't leave a sleep(600) running.
try:
os.kill(grandchild_pid, 9)
except ProcessLookupError:
pass
pytest.fail(
f"grandchild PID {grandchild_pid} survived runner exit; "
f"diag={diag!r} test_pid={test_pid} test_pgid={test_pgid}; "
f"runner output:\n{proc.stdout}"
)
# ── Bare pytest-flag passthrough ─────────────────────────────────────────────
#
# The runner routes any token starting with ``-`` that isn't one of its own
# options (``-j``/``--jobs``, ``--paths``, ``--slice``, ``--file-timeout``,
# ``--generate-slices``, ``--files``, ``--include-integration``,
# ``--files-from``) straight
# through to each per-file pytest invocation — no ``--`` separator required.
# Before this, a bare ``-q`` errored out with "unrecognized arguments",
# forcing a retry on every run. These tests are behavior contracts, not
# snapshots: they assert that bare flags reach pytest and that value-taking
# flags (``-k expr``) keep their value instead of having it stolen by the
# positional-path discovery.
def _make_probe_dir(tmp_path: Path) -> Path:
"""Two trivial passing tests, one named test_alpha, one test_beta."""
probe_dir = tmp_path / "probe"
probe_dir.mkdir()
(probe_dir / "test_flagprobe.py").write_text(
"def test_alpha():\n assert True\n\n"
"def test_beta():\n assert True\n"
)
return probe_dir
def _run_runner(probe_dir: Path, *extra: str) -> subprocess.CompletedProcess:
repo_root = _probe_root(probe_dir.parent)
runner = repo_root / "scripts" / "run_tests_parallel.py"
return subprocess.run(
[sys.executable, str(runner), "--paths", str(probe_dir),
"-j", "1", "--file-timeout", "30", *extra],
cwd=probe_dir,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
# The runner declares its stdio UTF-8 (see _make_stdio_glyph_safe);
# decode the same way so ✓-glyph assertions hold on Windows, where
# text=True alone would decode with the locale codec (cp1252).
encoding="utf-8",
errors="replace",
timeout=60,
)
@pytest.mark.parametrize("help_flag", ["-h", "--help"])
def test_help_prints_usage_without_discovering_or_running_tests(
tmp_path: Path, help_flag: str
) -> None:
"""Runner help stays in argparse instead of becoming a pytest sweep."""
repo_root = Path(__file__).resolve().parent.parent.parent
runner = repo_root / "scripts" / "run_tests_parallel.py"
proc = subprocess.run(
[sys.executable, str(runner), "--paths", str(tmp_path / "no-tests"), help_flag],
cwd=repo_root,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
encoding="utf-8",
errors="replace",
timeout=10,
)
assert proc.returncode == 0, proc.stdout
assert "usage:" in proc.stdout
assert "Discovered" not in proc.stdout
def test_unknown_bare_flag_errors_with_usage_instead_of_sweeping(tmp_path: Path) -> None:
"""A typo'd flag fails once, up front, and never reaches a per-file pytest.
Bare tokens are checked against pytest's own option set, so real pytest
forms (attached short value ``-rA``, bare ``-x``) still pass through and
run, while ``--jbs`` is rejected with this runner's usage before discovery.
"""
probe_dir = _make_probe_dir(tmp_path)
proc = _run_runner(probe_dir, "--jbs")
assert proc.returncode == 2, proc.stdout
assert "usage:" in proc.stdout and "unrecognized arguments: --jbs" in proc.stdout
assert "Discovered" not in proc.stdout
proc = _run_runner(probe_dir, "-rA", "-x")
assert proc.returncode == 0, proc.stdout
assert "2✓" in proc.stdout or "2 passed" in proc.stdout, proc.stdout
def test_known_flag_missing_value_errors_with_usage_instead_of_sweeping(
tmp_path: Path,
) -> None:
"""``--tb`` with no value is a pytest UsageError, not a per-file sweep.
The flag itself is known, so an unknown-token check alone lets it through;
pytest's own parser must be allowed to reject it up front.
"""
probe_dir = _make_probe_dir(tmp_path)
proc = _run_runner(probe_dir, "--tb")
assert proc.returncode == 2, proc.stdout
assert "usage:" in proc.stdout and "--tb: expected one argument" in proc.stdout
assert "Discovered" not in proc.stdout
def test_file_retry_self_heals_and_prints_both_attempts(tmp_path: Path) -> None:
"""A pass-on-retry is green, loud, and retains the failing traceback."""
repo_root = _probe_root(tmp_path)
runner = repo_root / "scripts" / "run_tests_parallel.py"
marker = tmp_path / "ran-once"
probe = tmp_path / "test_flaky_probe.py"
probe.write_text(
textwrap.dedent(
f"""
from pathlib import Path
def test_flaky_once():
marker = Path({str(marker)!r})
if not marker.exists():
marker.write_text("failed once")
assert False, "simulated first-attempt flake"
assert True
"""
),
encoding="utf-8",
)
proc = subprocess.run(
[
sys.executable,
str(runner),
"--files",
str(probe),
"--file-retries",
"1",
"-j",
"1",
"-q",
],
cwd=tmp_path,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
text=True,
timeout=60,
)
assert proc.returncode == 0, proc.stdout
assert "FLAKY file" in proc.stdout
assert "simulated first-attempt flake" in proc.stdout
assert "first-attempt output" in proc.stdout
assert "retry output" in proc.stdout
# ---------------------------------------------------------------------------
# Zero-collection is not a pass; node ids are translated, not dropped.
#
# Both behaviors were real foot-guns: a run where NOTHING was collected printed
# "0 tests passed, 0 failed (100% complete)" (reads green), and a pytest node id
# (`file.py::Class::test`) was silently discarded by path discovery so the run
# ended with "No test files to run" while looking like an accepted selector.
def test_zero_collected_across_run_fails_and_says_so(tmp_path: Path) -> None:
"""A -k that matches nothing must FAIL, not report a green summary."""
probe_dir = _make_probe_dir(tmp_path)
proc = _run_runner(probe_dir, "-k", "zzz_matches_nothing")
assert proc.returncode == 1, proc.stdout
assert "NO TESTS RAN" in proc.stdout
assert "NOT a pass" in proc.stdout
@pytest.mark.parametrize("form,expected", [
("positional", ["alpha", "beta"]), ("bare-k", ["alpha"]),
("node-id", ["alpha"]), ("explicit-k", ["beta"]),
("pathsep", ["alpha", "beta", "gamma"]),
])
def test_runner_selection_records_actual_test_identity(tmp_path, form, expected):
probe = tmp_path / "probe"
other = tmp_path / "other"
probe.mkdir()
other.mkdir()
receipt = tmp_path / "witnesses"
receipt.mkdir()
for directory, filename, names in ((probe, "test_flags.py", ["alpha", "beta"]),
(other, "test_other.py", ["gamma"])):
(directory / filename).write_text("from pathlib import Path\n" + "".join(
f"def test_{name}():\n Path({str(receipt / name)!r}).touch()\n" for name in names
), encoding="utf-8")
target = str(probe / "test_flags.py")
arguments = {
"positional": [str(probe), "-q"],
"bare-k": ["--paths", str(probe), "-k", "test_alpha"],
"node-id": [target + "::test_alpha"],
"explicit-k": [target + "::test_alpha", "-k", "test_beta"],
"pathsep": ["--paths", os.pathsep.join([str(probe), str(other)])],
}[form]
runner = _probe_root(tmp_path) / "scripts/run_tests_parallel.py"
result = subprocess.run([sys.executable, str(runner), *arguments, "-j", "1", "--file-timeout", "30"],
cwd=tmp_path, capture_output=True, text=True, encoding="utf-8", timeout=60)
assert result.returncode == 0, result.stdout + result.stderr
assert sorted(path.name for path in receipt.iterdir()) == expected
@pytest.mark.platforms("posix") # POSIX signal death; Windows has no SIGSEGV exit
def test_interpreter_crash_is_reported_as_a_crash_not_as_no_tests_ran(tmp_path: Path) -> None:
"""A file whose interpreter dies by signal is classified as CRASHED (#113186).
A native fault after some tests passed leaves no pytest summary line, so
every count parses to 0. The runner used to file that under "no tests ran
(collection/import error)" beneath a summary reading ``0 failed`` — two
wrong diagnoses for one real bug. The crash must be named on the summary
line and in the failure buckets, and the run must still exit non-zero.
"""
probe_dir = tmp_path / "probe"
probe_dir.mkdir()
(probe_dir / "test_probe_crash.py").write_text(
textwrap.dedent(
"""
import os, signal
def test_before():
assert True
def test_crash():
os.kill(os.getpid(), signal.SIGSEGV)
"""
)
)
proc = _run_runner(probe_dir, "--file-retries", "0")
assert proc.returncode != 0
assert "1 file CRASHED" in proc.stdout
assert "SIGSEGV" in proc.stdout
assert "where no tests ran" not in proc.stdout
assert "NO TESTS RAN" not in proc.stdout
# ── --files-from: file-backed explicit file lists ───────────────────────────
#
# --files carries the whole list as ONE argv element, and Linux caps a
# single argument at MAX_ARG_STRLEN (128 KiB) — a much smaller limit than
# ARG_MAX. The whole-suite list (~210 KB) dies with E2BIG in execve before
# the runner's first line runs. --files-from takes the same explicit list
# from a file (or stdin via '-'), one path per line, so the list is bounded
# by the filesystem instead of one argv element.
def test_files_from_runs_exactly_the_listed_files(tmp_path: Path) -> None:
"""A newline-separated list file bypasses discovery like --files."""
probe_dir = _make_probe_dir(tmp_path)
extra = tmp_path / "probe_extra"
extra.mkdir()
(extra / "test_extra.py").write_text("def test_extra():\n assert True\n")
list_file = tmp_path / "files.txt"
list_file.write_text(
f"{probe_dir / 'test_flagprobe.py'}\n\n{extra / 'test_extra.py'}\n",
encoding="utf-8",
)
# --paths points at the probe dir too; an explicit list must win and
# NOT discover the extra file by accident.
proc = _run_runner(probe_dir, "--files-from", str(list_file), "-q")
assert proc.returncode == 0, proc.stdout
assert "Running 2 test files" in proc.stdout, proc.stdout
assert "3 passed" not in proc.stdout, proc.stdout
def test_files_from_dash_reads_the_list_from_stdin(tmp_path: Path) -> None:
"""--files-from - reads the list from stdin."""
probe_dir = _make_probe_dir(tmp_path)
repo_root = Path(__file__).resolve().parent.parent.parent
runner = repo_root / "scripts" / "run_tests_parallel.py"
proc = subprocess.run(
[sys.executable, str(runner), "--files-from", "-",
"-j", "1", "--file-timeout", "30", "-q"],
input=f"{probe_dir / 'test_flagprobe.py'}\n",
cwd=repo_root, stdout=subprocess.PIPE, stderr=subprocess.STDOUT,
encoding="utf-8", errors="replace", timeout=60,
)
assert proc.returncode == 0, proc.stdout
assert "Running 1 test files" in proc.stdout, proc.stdout
assert "✓2" in proc.stdout or "2 passed" in proc.stdout, proc.stdout
def test_scratch_root_is_per_user(tmp_path: Path, monkeypatch) -> None:
"""Two users on one host must never share the runner's scratch root.
A fixed literal in a world-writable sticky dir belongs to whoever created it first: a
root-owned root (a container or system-service run) makes every later ``mkdtemp`` there
fail with EPERM for every other user, with no non-root way back.
"""
import importlib
scripts_dir = Path(__file__).resolve().parents[2] / "scripts"
monkeypatch.syspath_prepend(str(scripts_dir))
runner = importlib.import_module("run_tests_parallel")
# Exercise the non-/var/tmp arm so the probe never mints roots in the real shared dir.
# (Narrow: on 3.14 ``Path.is_dir()`` itself goes through ``os.path.isdir``.)
monkeypatch.setattr(runner.tempfile, "gettempdir", lambda: str(tmp_path))
real_isdir = os.path.isdir
monkeypatch.setattr(runner.os.path, "isdir",
lambda path: False if str(path) == "/var/tmp" else real_isdir(path))
monkeypatch.setattr(runner.os, "getuid", lambda: 1000)
mine = Path(runner._runner_scratch_root())
monkeypatch.setattr(runner.os, "getuid", lambda: 0)
theirs = Path(runner._runner_scratch_root())
assert mine != theirs
assert mine.is_dir() and theirs.is_dir()
assert mine.parent == tmp_path and theirs.parent == tmp_path
def test_off_host_note_names_platforms_specs_that_exclude_this_host(tmp_path: Path) -> None:
"""A green local run must say which platforms() tests were skipped and where
they run; specs are resolved (posix is not off-host on Linux or macOS)."""
probe_dir = _make_probe_dir(tmp_path)
(probe_dir / "test_gated.py").write_text(
"import pytest\n\n"
"@pytest.mark.platforms('not linux')\ndef test_elsewhere():\n assert True\n\n"
"@pytest.mark.platforms('posix')\ndef test_posix():\n assert True\n\n"
"@pytest.mark.platforms('windows')\ndef test_windows():\n assert True\n",
encoding="utf-8",
)
proc = _run_runner(probe_dir)
notes = [line for line in proc.stdout.splitlines() if "SKIPPED on this host" in line]
host = {"linux": "linux", "darwin": "macos", "win32": "windows"}[sys.platform]
off_host = {"windows"} - {host}
if host == "linux":
off_host.add("not linux")
if host == "windows":
off_host.add("posix")
assert {n.split("platforms(")[1].split(")")[0].strip("'") for n in notes} == off_host, proc.stdout
assert all("they run on the" in n for n in notes), proc.stdout