fix(hooks): arm the rebuild timeout without SIGALRM on Windows

The GRAPHIFY_REBUILD_TIMEOUT watchdog in both embedded rebuild bodies was
guarded by hasattr(signal, 'SIGALRM') with no else branch, so on Windows it
never armed and a hung rebuild survived indefinitely -- the exact tail case
the #791 timeout was added to catch.

Fall back to a daemon threading.Timer that prints the same message and calls
os._exit(1). The hard exit is deliberate: the process is already stuck, so a
clean shutdown may itself be blocked, and _rebuild_lock degrades to a no-op
yield on platforms without fcntl, so there is no lock to leave stale.
This commit is contained in:
Rishet Mehra
2026-07-24 23:37:14 +01:00
committed by safishamsi
parent ce9ea7d7b4
commit ca3113ac7b
2 changed files with 44 additions and 8 deletions
+24 -8
View File
@@ -104,7 +104,7 @@ fi
# double-quote, $, backtick or backslash characters: it is carried inside a
# shell double-quoted `-c "..."` argument (see _detached_launch).
_REBUILD_BODY_COMMIT = """\
import os, signal, sys
import os, signal, sys, threading
from pathlib import Path
changed_raw = os.environ.get('GRAPHIFY_CHANGED', '')
@@ -119,9 +119,17 @@ try:
from graphify.watch import _rebuild_code, _apply_resource_limits
_apply_resource_limits()
_timeout = int(os.environ.get('GRAPHIFY_REBUILD_TIMEOUT', '600'))
if _timeout > 0 and hasattr(signal, 'SIGALRM'):
signal.signal(signal.SIGALRM, lambda *_: (_ for _ in ()).throw(TimeoutError(f'graphify rebuild exceeded {_timeout}s')))
signal.alarm(_timeout)
if _timeout > 0:
if hasattr(signal, 'SIGALRM'):
signal.signal(signal.SIGALRM, lambda *_: (_ for _ in ()).throw(TimeoutError(f'graphify rebuild exceeded {_timeout}s')))
signal.alarm(_timeout)
else:
def _bail():
print(f'[graphify hook] graphify rebuild exceeded {_timeout}s', flush=True)
os._exit(1)
_watchdog = threading.Timer(_timeout, _bail)
_watchdog.daemon = True
_watchdog.start()
_force = os.environ.get('GRAPHIFY_FORCE', '').lower() in ('1', 'true', 'yes')
_root = Path('.')
_out = os.environ.get('GRAPHIFY_OUT', 'graphify-out')
@@ -153,13 +161,21 @@ except Exception as exc:
_REBUILD_BODY_CHECKOUT = """\
from graphify.watch import _rebuild_code, _apply_resource_limits
from pathlib import Path
import os, signal, sys
import os, signal, sys, threading
try:
_apply_resource_limits()
_timeout = int(os.environ.get('GRAPHIFY_REBUILD_TIMEOUT', '600'))
if _timeout > 0 and hasattr(signal, 'SIGALRM'):
signal.signal(signal.SIGALRM, lambda *_: (_ for _ in ()).throw(TimeoutError(f'graphify rebuild exceeded {_timeout}s')))
signal.alarm(_timeout)
if _timeout > 0:
if hasattr(signal, 'SIGALRM'):
signal.signal(signal.SIGALRM, lambda *_: (_ for _ in ()).throw(TimeoutError(f'graphify rebuild exceeded {_timeout}s')))
signal.alarm(_timeout)
else:
def _bail():
print(f'[graphify hook] graphify rebuild exceeded {_timeout}s', flush=True)
os._exit(1)
_watchdog = threading.Timer(_timeout, _bail)
_watchdog.daemon = True
_watchdog.start()
_force = os.environ.get('GRAPHIFY_FORCE', '').lower() in ('1', 'true', 'yes')
# post-checkout: branch switch can touch arbitrary files; full rebuild path
# (no changed_paths) is correct here. The flock inside _rebuild_code still
+20
View File
@@ -330,6 +330,26 @@ def test_rebuild_bodies_with_graphify_root_are_valid_python():
ast.parse(body)
@pytest.mark.parametrize(
"name,body",
[("post-commit", _REBUILD_BODY_COMMIT), ("post-checkout", _REBUILD_BODY_CHECKOUT)],
)
def test_rebuild_bodies_arm_a_timeout_without_sigalrm(name, body):
"""Windows has no signal.SIGALRM, so the #791 rebuild timeout never armed
there at all (#2148). The fallback has to sit in the else-branch of the
SIGALRM check rather than merely appear somewhere in the body, so that a
watchdog firing unconditionally or on every platform still fails here."""
fallbacks = [
node.orelse
for node in ast.walk(ast.parse(body))
if isinstance(node, ast.If) and "'SIGALRM'" in ast.dump(node.test) and node.orelse
]
assert fallbacks, f"{name} has no else-branch for the missing-SIGALRM case (#2148)"
dumped = "".join(ast.dump(stmt) for stmt in fallbacks[0])
assert "attr='Timer'" in dumped, f"{name} fallback does not arm a threading.Timer (#2148)"
assert "attr='_exit'" in dumped, f"{name} fallback does not kill the stuck rebuild (#2148)"
def test_detached_launch_targets_graphify_python():
"""The launcher must run via the resolved $GRAPHIFY_PYTHON, not a bare
`python`, so it uses the same interpreter the detection block selected."""