fix(async): close unscheduled coroutines in all threadsafe bridges (#26584)

Wraps every sync->async coroutine-scheduling site in the codebase with a
new agent.async_utils.safe_schedule_threadsafe() helper that closes the
coroutine on scheduling failure (closed loop, shutdown race, etc.)
instead of leaking it as 'coroutine was never awaited' RuntimeWarnings
plus reference leaks.

22 production call sites migrated across the codebase:
- acp_adapter/events.py, acp_adapter/permissions.py
- agent/lsp/manager.py
- cron/scheduler.py (media + text delivery paths)
- gateway/platforms/feishu.py (5 sites, via existing _submit_on_loop helper
  which now delegates to safe_schedule_threadsafe)
- gateway/run.py (10 sites: telegram rename, agent:step hook, status
  callback, interim+bg-review, clarify send, exec-approval button+text,
  temp-bubble cleanup, channel-directory refresh)
- plugins/memory/hindsight, plugins/platforms/google_chat
- tools/browser_supervisor.py (3), browser_cdp_tool.py,
  computer_use/cua_backend.py, slash_confirm.py
- tools/environments/modal.py (_AsyncWorker)
- tools/mcp_tool.py (2 + 8 _run_on_mcp_loop callers converted to
  factory-style so the coroutine is never constructed on a dead loop)
- tui_gateway/ws.py

Tests: new tests/agent/test_async_utils.py covers helper behavior under
live loop, dead loop, None loop, and scheduling exceptions. Regression
tests added at three PR-original sites (acp events, acp permissions,
mcp loop runner) mirroring contributor's intent.

Live-tested end-to-end:
- Helper stress test: 1500 schedules across live/dead/race scenarios,
  zero leaked coroutines
- Race exercised: 5000 schedules with loop killed mid-flight, 100 ok /
  4900 None returns, zero leaks
- hermes chat -q with terminal tool call (exercises step_callback bridge)
- MCP probe against failing subprocess servers + factory path
- Real gateway daemon boot + SIGINT shutdown across multiple platform
  adapter inits
- WSTransport 100 live + 50 dead-loop writes
- Cron delivery path live + dead loop

Salvages PR #2657 — adopts contributor's intent over a much wider site
list and a single centralized helper instead of inline try/except at
each site. 3 of the original PR's 6 sites no longer exist on main
(environments/patches.py deleted, DingTalk refactored to native async);
the equivalent fix lives in tools/environments/modal.py instead.

Co-authored-by: JithendraNara <jithendranaidunara@gmail.com>
This commit is contained in:
Teknium 2026-05-15 14:00:01 -07:00 committed by GitHub
parent 931caf2b2d
commit 4e89c53082
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
23 changed files with 690 additions and 186 deletions

View file

@ -1,6 +1,8 @@
"""Tests for acp_adapter.events — callback factories for ACP notifications."""
import asyncio
import gc
import warnings
from concurrent.futures import Future
from unittest.mock import AsyncMock, MagicMock, patch
@ -10,6 +12,7 @@ import acp
from acp.schema import ToolCallStart, ToolCallProgress, AgentThoughtChunk, AgentMessageChunk
from acp_adapter.events import (
_send_update,
make_message_cb,
make_step_cb,
make_thinking_cb,
@ -325,3 +328,46 @@ class TestMessageCallback:
cb("")
mock_rcts.assert_not_called()
# ---------------------------------------------------------------------------
# Scheduler-failure regression
# ---------------------------------------------------------------------------
class TestSendUpdate:
def test_scheduler_failure_closes_update_coroutine(self, event_loop_fixture):
"""If run_coroutine_threadsafe raises, _send_update must close the coro."""
created = {"coro": None}
async def _session_update(session_id, update):
return None
conn = MagicMock()
def _capture_update(session_id, update):
created["coro"] = _session_update(session_id, update)
return created["coro"]
conn.session_update = _capture_update
with warnings.catch_warnings(record=True) as caught:
warnings.simplefilter("always")
with patch(
"agent.async_utils.asyncio.run_coroutine_threadsafe",
side_effect=RuntimeError("scheduler down"),
):
_send_update(conn, "session-1", event_loop_fixture, {"type": "noop"})
gc.collect()
assert created["coro"] is not None
assert created["coro"].cr_frame is None
# Only count warnings about THIS test's coroutine; other tests in the
# same xdist worker (or stdlib mock internals) may emit unrelated
# "coroutine was never awaited" warnings that bleed through.
runtime_warnings = [
w for w in caught
if issubclass(w.category, RuntimeWarning)
and "was never awaited" in str(w.message)
and "_session_update" in str(w.message)
]
assert runtime_warnings == []

View file

@ -38,7 +38,7 @@ def _invoke_callback(
scheduled["loop"] = passed_loop
return future
with patch("acp_adapter.permissions.asyncio.run_coroutine_threadsafe", side_effect=_schedule):
with patch("agent.async_utils.asyncio.run_coroutine_threadsafe", side_effect=_schedule):
cb = make_approval_callback(request_permission, loop, session_id="s1", timeout=timeout)
if use_prompt_path:
result = prompt_dangerous_approval(
@ -135,7 +135,7 @@ class TestApprovalBridge:
scheduled["loop"] = passed_loop
return future
with patch("acp_adapter.permissions.asyncio.run_coroutine_threadsafe", side_effect=_schedule):
with patch("agent.async_utils.asyncio.run_coroutine_threadsafe", side_effect=_schedule):
cb = make_approval_callback(request_permission, loop, session_id="s1", timeout=0.01)
result = cb("rm -rf /", "dangerous command")
@ -159,10 +159,53 @@ class TestApprovalBridge:
scheduled["loop"] = passed_loop
return future
with patch("acp_adapter.permissions.asyncio.run_coroutine_threadsafe", side_effect=_schedule):
with patch("agent.async_utils.asyncio.run_coroutine_threadsafe", side_effect=_schedule):
cb = make_approval_callback(request_permission, loop, session_id="s1", timeout=1.0)
result = cb("echo hi", "demo")
scheduled["coro"].close()
assert result == "deny"
# ---------------------------------------------------------------------------
# Scheduler-failure regression
# ---------------------------------------------------------------------------
import gc # noqa: E402
import warnings # noqa: E402
class TestSchedulerFailure:
def test_scheduler_failure_closes_permission_coroutine(self):
"""If run_coroutine_threadsafe raises, the coro is closed and we return 'deny'."""
loop = MagicMock(spec=asyncio.AbstractEventLoop)
created = {"coro": None}
async def _response_coro(**kwargs):
return _make_response(AllowedOutcome(option_id="allow_once", outcome="selected"))
def _request_permission(**kwargs):
created["coro"] = _response_coro(**kwargs)
return created["coro"]
with warnings.catch_warnings(record=True) as caught:
warnings.simplefilter("always")
with patch(
"agent.async_utils.asyncio.run_coroutine_threadsafe",
side_effect=RuntimeError("scheduler down"),
):
cb = make_approval_callback(_request_permission, loop, session_id="s1", timeout=0.01)
result = cb("rm -rf /", "dangerous")
gc.collect()
assert result == "deny"
assert created["coro"] is not None
assert created["coro"].cr_frame is None
runtime_warnings = [
w for w in caught
if issubclass(w.category, RuntimeWarning)
and "was never awaited" in str(w.message)
and "_response_coro" in str(w.message)
]
assert runtime_warnings == []