Teknium
409 lines
17 KiB
Python
409 lines
17 KiB
Python
"""Windows subprocess compatibility helpers.
|
|
|
|
* ``["npm", ...]`` — on Windows ``npm`` is ``npm.cmd``, a batch shim; ``Popen`` fails with
|
|
WinError 193 because CreateProcessW can't run a ``.cmd`` without ``shell=True``/PATHEXT.
|
|
* ``start_new_session=True`` — POSIX ``os.setsid()`` detach; silently ignored on Windows, whose
|
|
equivalent is the ``CREATE_NEW_PROCESS_GROUP | CREATE_NO_WINDOW`` creationflags bundle.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import os
|
|
import re
|
|
import shutil
|
|
import subprocess
|
|
import sys
|
|
from typing import Mapping, Sequence
|
|
|
|
__all__ = [
|
|
"IS_WINDOWS",
|
|
"resolve_node_command",
|
|
"split_command_line",
|
|
"suppress_platform_ver_console",
|
|
"windows_detach_flags",
|
|
"windows_detach_flags_without_breakaway",
|
|
"windows_hide_flags",
|
|
"windows_detach_popen_kwargs",
|
|
"bounded_git_probe",
|
|
"bounded_probe_run",
|
|
"noninteractive_git_env",
|
|
"NO_DRIVER_DIFF_FLAGS",
|
|
"pid_is_hermes",
|
|
]
|
|
|
|
# Flags that neutralize *attribute-scoped* diff drivers on any diff-rendering git command. A
|
|
# malicious repo can name a driver in ``.gitattributes`` (``* diff=evil``) and point it at an
|
|
# arbitrary program via ``[diff "evil"] command=/textconv=`` in ``.git/config``; because the
|
|
# attacker chooses the name, ``GIT_CONFIG_KEY`` overrides in ``noninteractive_git_env`` cannot
|
|
# enumerate it — only these flags do. ``--no-ext-diff`` kills ``command=``; ``--no-textconv`` kills
|
|
# ``textconv=``; each alone leaves the other live. Smudge/clean filters are neutralized by the env
|
|
# layer's ``core.hooksPath`` + running against the index without checkout.
|
|
NO_DRIVER_DIFF_FLAGS = ("--no-ext-diff", "--no-textconv")
|
|
|
|
# Only these subcommands accept ``NO_DRIVER_DIFF_FLAGS`` — ``status`` and friends reject them
|
|
# (``unknown option``), so the helper gates on this set rather than blanket-prepending.
|
|
_DIFF_RENDERING_SUBCOMMANDS = frozenset({"diff", "show", "log", "blame"})
|
|
|
|
# Options that consume the FOLLOWING token, so that value is never mistaken for the subcommand
|
|
# (``-C diff`` is a path; ``-c diff=x`` is a config pair).
|
|
_GIT_VALUE_OPTS = {"-C", "-c", "--git-dir", "--work-tree", "--namespace", "--exec-path"}
|
|
|
|
|
|
def harden_git_argv(args: Sequence[str]) -> list[str]:
|
|
"""Copy of subcommand-first git *args* (no leading ``"git"``) with :data:`NO_DRIVER_DIFF_FLAGS`
|
|
inserted right after a diff-rendering subcommand; other subcommands are returned unchanged.
|
|
|
|
Pair with :func:`noninteractive_git_env`: the env layer disables fsmonitor/hooks/pager/editor/
|
|
credential sinks, this closes the one class (attacker-named attribute drivers) env cannot reach.
|
|
"""
|
|
out = list(args)
|
|
i = 0
|
|
while i < len(out):
|
|
tok = out[i]
|
|
if tok in _GIT_VALUE_OPTS:
|
|
i += 2
|
|
continue
|
|
if tok.startswith("-"):
|
|
i += 1
|
|
continue
|
|
if tok in _DIFF_RENDERING_SUBCOMMANDS:
|
|
return out[: i + 1] + list(NO_DRIVER_DIFF_FLAGS) + out[i + 1 :]
|
|
return out # first non-option token is a non-diff subcommand
|
|
return out
|
|
|
|
|
|
IS_WINDOWS = sys.platform == "win32"
|
|
|
|
# Private launcher-to-child metadata. This is diagnostic state, not user config.
|
|
_WINDOWS_GATEWAY_BREAKAWAY_ENV = "_HERMES_GATEWAY_BREAKAWAY"
|
|
|
|
|
|
def split_command_line(line: str) -> list[str]:
|
|
"""Split a user-supplied command line into tokens, Windows-safely.
|
|
|
|
``shlex.split`` (posix=True) treats every backslash as an escape, mangling Windows paths. On
|
|
Windows use ``posix=False`` and strip one layer of matching quotes per token; on POSIX this is
|
|
exactly ``shlex.split``. Raises ValueError on unbalanced quotes.
|
|
"""
|
|
import shlex
|
|
|
|
if not IS_WINDOWS:
|
|
return shlex.split(line)
|
|
out: list[str] = []
|
|
for tok in shlex.split(line, posix=False):
|
|
if len(tok) >= 2 and tok[0] == tok[-1] and tok[0] in ("'", '"'):
|
|
tok = tok[1:-1]
|
|
out.append(tok)
|
|
return out
|
|
|
|
|
|
def resolve_node_command(name: str, argv: Sequence[str]) -> list[str]:
|
|
"""Resolve a Node-ecosystem command name (``npm``, ``npx``, ``yarn``…) to an absolute-path argv.
|
|
|
|
On Windows these ship as ``.cmd`` batch shims that CreateProcessW won't execute by bare name;
|
|
``shutil.which`` resolves via PATHEXT to a fully-qualified path whose extension routes it
|
|
through ``cmd.exe /c``.
|
|
"""
|
|
resolved = shutil.which(name)
|
|
return [resolved or name, *argv]
|
|
|
|
|
|
# Win32 CreationFlags — defined here because CREATE_NO_WINDOW / DETACHED_PROCESS aren't guaranteed
|
|
# to exist on stdlib subprocess for older Pythons or non-Windows builds.
|
|
_CREATE_NEW_PROCESS_GROUP = 0x00000200
|
|
# DETACHED_PROCESS (0x00000008) is intentionally NOT part of any flag bundle — do not re-add it
|
|
# (the recurring console-flash bug): (1) MSDN: CREATE_NO_WINDOW "is ignored if used with either
|
|
# CREATE_NEW_CONSOLE or DETACHED_PROCESS"; (2) a DETACHED_PROCESS child has NO console, so every
|
|
# console-subsystem descendant (git, gh, cmd, node, powershell, …) allocates its own — a visible
|
|
# flash per spawn, including inside third-party libraries no per-site sweep can reach. A
|
|
# CREATE_NO_WINDOW child instead OWNS a hidden console all descendants inherit (A/B verified on
|
|
# Windows 11 by the desktop backend fix).
|
|
_CREATE_NO_WINDOW = 0x08000000
|
|
# Escape any Win32 job object the parent belongs to. Without this a detached child inherits the
|
|
# parent's job, and when that parent (Electron, Tauri, Windows Terminal, the Desktop bootstrap
|
|
# installer) dies the OS tears down the whole job — taking the "detached" child with it. Critical
|
|
# for the post-update gateway watcher spawned from inside Electron's job.
|
|
_CREATE_BREAKAWAY_FROM_JOB = 0x01000000
|
|
|
|
|
|
def windows_detach_flags() -> int:
|
|
"""Win32 creationflags detaching a child from the parent console/group; 0 elsewhere.
|
|
|
|
Pair with the default ``start_new_session=False`` (POSIX uses ``start_new_session=True``).
|
|
CREATE_NEW_PROCESS_GROUP stops Ctrl+C propagating; CREATE_NO_WINDOW gives the child a hidden
|
|
console descendants inherit; CREATE_BREAKAWAY_FROM_JOB escapes Electron/Tauri job objects. A
|
|
job that forbids breakaway yields PermissionError from Popen — callers catch OSError and fall
|
|
back to :func:`windows_detach_flags_without_breakaway`.
|
|
"""
|
|
if not IS_WINDOWS:
|
|
return 0
|
|
return _CREATE_NEW_PROCESS_GROUP | _CREATE_NO_WINDOW | _CREATE_BREAKAWAY_FROM_JOB
|
|
|
|
|
|
def windows_detach_flags_without_breakaway() -> int:
|
|
""":func:`windows_detach_flags` minus ``CREATE_BREAKAWAY_FROM_JOB``; 0 on non-Windows."""
|
|
if not IS_WINDOWS:
|
|
return 0
|
|
return _CREATE_NEW_PROCESS_GROUP | _CREATE_NO_WINDOW
|
|
|
|
|
|
def windows_hide_flags() -> int:
|
|
"""Win32 creationflags hiding the child's console without detaching it; 0 elsewhere.
|
|
|
|
For short-lived synchronous helpers (``taskkill``, ``where``, version probes): no flash, but the
|
|
child stays in the parent's process group and job so Ctrl+C and job teardown still propagate.
|
|
Stdio is inherited, so ``capture_output=True`` works.
|
|
"""
|
|
return _CREATE_NO_WINDOW if IS_WINDOWS else 0
|
|
|
|
|
|
def suppress_platform_ver_console() -> None:
|
|
"""Stub ``platform._syscmd_ver`` on Windows so it never flashes a console. No-op elsewhere.
|
|
|
|
``platform.win32_ver()`` shells out ``cmd /c ver`` without CREATE_NO_WINDOW, so a windowless
|
|
parent (pythonw gateway, kanban workers) flashes a cmd window whenever a dependency touches
|
|
``platform.uname()`` at import. With the stub, ``win32_ver()`` takes its documented fallback to
|
|
``sys.getwindowsversion()`` — same data, in-process. Call before heavy imports.
|
|
"""
|
|
if not IS_WINDOWS:
|
|
return
|
|
try:
|
|
import platform
|
|
|
|
if hasattr(platform, "_syscmd_ver"):
|
|
def _quiet_syscmd_ver(system="", release="", version="",
|
|
supported_platforms=("win32", "win16", "dos")):
|
|
return system, release, version
|
|
|
|
platform._syscmd_ver = _quiet_syscmd_ver
|
|
except Exception:
|
|
pass # Purely cosmetic hardening — never let it break startup.
|
|
|
|
|
|
def windows_detach_popen_kwargs() -> dict:
|
|
"""Popen kwargs detaching a child on Windows, or ``start_new_session=True`` on POSIX.
|
|
|
|
Bare ``start_new_session=True`` is accepted but has no effect on Windows: the child stays
|
|
attached to the parent console and dies when it closes.
|
|
"""
|
|
if IS_WINDOWS:
|
|
return {"creationflags": windows_detach_flags()}
|
|
return {"start_new_session": True}
|
|
|
|
|
|
# GIT_CONFIG_KEY_n/VALUE_n overrides for internal git children: no credential/askpass prompts, no
|
|
# repo-configured fsmonitor/hooks/pager/editor/external-diff programs.
|
|
_GIT_CONFIG_INJECT_PREFIXES = ("GIT_CONFIG_KEY_", "GIT_CONFIG_VALUE_")
|
|
_GIT_CONFIG_OVERRIDES = {
|
|
"credential.helper": "",
|
|
"core.askPass": "",
|
|
"core.fsmonitor": "false",
|
|
"core.untrackedCache": "false",
|
|
"core.hooksPath": os.devnull,
|
|
"core.pager": "cat",
|
|
"core.editor": "true",
|
|
"sequence.editor": "true",
|
|
"diff.external": "",
|
|
}
|
|
|
|
|
|
def noninteractive_git_env(base: "Mapping[str, str] | None" = None) -> dict[str, str]:
|
|
"""Environment for *internal* git invocations that must never prompt.
|
|
|
|
Copy of ``base`` (default ``os.environ``) with ``GIT_TERMINAL_PROMPT=0`` (fail instead of
|
|
prompting), ``GCM_INTERACTIVE=Never`` (no Git Credential Manager dialog), and isolated git
|
|
config: inherited ``GIT_CONFIG_*`` injection, global/system config, pagers, editors, fsmonitor,
|
|
external diff and hooks are all disabled so a user's repo/global config cannot hang or mutate
|
|
Hermes's plumbing calls. ``GIT_ASKPASS``/``SSH_ASKPASS`` are deliberately left alone: a
|
|
*working* askpass helper or ssh-agent should still succeed non-interactively. Pair with
|
|
``stdin=subprocess.DEVNULL``. Internal plumbing only — the agent-facing terminal tool has its
|
|
own policy layer and visible PTY.
|
|
"""
|
|
env = dict(base if base is not None else os.environ)
|
|
env["GIT_TERMINAL_PROMPT"] = "0"
|
|
env["GCM_INTERACTIVE"] = "Never"
|
|
# Drop caller-supplied config injection; the GIT_CONFIG_COUNT block is rebuilt below so
|
|
# ambient -c values cannot re-enable pagers, hooks, fsmonitor, editors or credential prompts.
|
|
for key in list(env):
|
|
if key == "GIT_CONFIG_PARAMETERS" or key.startswith(_GIT_CONFIG_INJECT_PREFIXES):
|
|
env.pop(key, None)
|
|
env.pop("GIT_CONFIG_COUNT", None)
|
|
env["GIT_CONFIG_GLOBAL"] = os.devnull
|
|
env["GIT_CONFIG_SYSTEM"] = os.devnull
|
|
env["GIT_CONFIG_NOSYSTEM"] = "1"
|
|
env["GIT_PAGER"] = "cat"
|
|
env["PAGER"] = "cat"
|
|
env["GIT_EDITOR"] = "true"
|
|
env["GIT_CONFIG_COUNT"] = str(len(_GIT_CONFIG_OVERRIDES))
|
|
for idx, (key, value) in enumerate(_GIT_CONFIG_OVERRIDES.items()):
|
|
env[f"GIT_CONFIG_KEY_{idx}"] = key
|
|
env[f"GIT_CONFIG_VALUE_{idx}"] = value
|
|
return env
|
|
|
|
|
|
def _process_start_time(pid: int) -> int | None:
|
|
"""The repository's stable process-start fingerprint, if available."""
|
|
try:
|
|
from gateway.status import get_process_start_time
|
|
|
|
return get_process_start_time(pid)
|
|
except Exception:
|
|
return None
|
|
|
|
|
|
def _text_names_hermes(text: str) -> bool:
|
|
r"""True when *text* names Hermes at a path-segment / token boundary.
|
|
|
|
A bare ``"hermes" in text`` substring test would also match unrelated processes whose paths
|
|
merely contain the letters (``...\shermesa\...``) — the false-positive class this prevents.
|
|
"""
|
|
return any(token.startswith(("hermes", ".hermes"))
|
|
for token in re.split(r"[\\/\s=,;\"']+", text.lower()))
|
|
|
|
|
|
def _process_command_is_hermes(pid: int) -> bool:
|
|
"""Best-effort check that *pid* currently runs Hermes code."""
|
|
try:
|
|
import psutil
|
|
|
|
process = psutil.Process(pid)
|
|
command = " ".join(process.cmdline() or [])
|
|
executable = process.exe() or ""
|
|
return _text_names_hermes(f"{command} {executable}")
|
|
except Exception:
|
|
return False
|
|
|
|
|
|
def pid_is_hermes(pid: int, *, expected_start_time: int | None = None) -> bool:
|
|
"""Whether it is safe to use ``taskkill`` for *pid*.
|
|
|
|
The PID must be valid, currently exist, and identify a Hermes process. When the caller captured
|
|
a start-time fingerprint before the destructive action, the live process must still have the
|
|
same ``(pid, start_time)`` identity. Any ambiguity fails closed.
|
|
"""
|
|
if not isinstance(pid, int) or isinstance(pid, bool) or pid <= 0:
|
|
return False
|
|
if not IS_WINDOWS:
|
|
if expected_start_time is None:
|
|
return True
|
|
try:
|
|
return _process_start_time(pid) == expected_start_time
|
|
except Exception:
|
|
return False
|
|
try:
|
|
current_start_time = _process_start_time(pid)
|
|
except Exception:
|
|
return False
|
|
if current_start_time is None:
|
|
return False
|
|
if expected_start_time is not None and current_start_time != expected_start_time:
|
|
return False
|
|
try:
|
|
return _process_command_is_hermes(pid)
|
|
except Exception:
|
|
return False
|
|
|
|
|
|
def kill_process_tree(proc: "subprocess.Popen") -> None:
|
|
"""Best-effort terminate *proc* and its descendants on both platforms; never raises.
|
|
|
|
``proc.kill()`` alone only terminates the direct child. This is cleanup on an already-failing
|
|
path whose contract is to fail open, so every failure (access denied, already reaped) is
|
|
swallowed rather than escaping the caller's ``except``.
|
|
"""
|
|
try:
|
|
from agent.deadline import kill_process_tree as _deadline_kill_tree
|
|
|
|
_deadline_kill_tree(proc.pid)
|
|
except Exception:
|
|
_legacy_kill_process_tree(proc)
|
|
return
|
|
# Ensure Popen's own bookkeeping sees the exit so communicate()/wait() cannot hang.
|
|
try:
|
|
proc.kill()
|
|
except OSError:
|
|
pass
|
|
|
|
|
|
def _legacy_kill_process_tree(proc: "subprocess.Popen") -> None:
|
|
"""Local tree-kill fallback when agent.deadline is unavailable (partial install, cycle)."""
|
|
if not IS_WINDOWS:
|
|
# Verify the child leads its own process group before signalling, never a shared group.
|
|
try:
|
|
import signal as _signal
|
|
|
|
pgid = os.getpgid(proc.pid)
|
|
if pgid == proc.pid:
|
|
os.killpg(pgid, _signal.SIGKILL) # windows-footgun: ok — inside `if not IS_WINDOWS` gate
|
|
except Exception:
|
|
pass
|
|
try:
|
|
proc.kill()
|
|
except OSError:
|
|
pass
|
|
if IS_WINDOWS:
|
|
# No identity guard on purpose: *proc* is our own retained Popen handle, so the PID cannot
|
|
# be recycled while we hold it. The fail-closed ``pid_is_hermes`` guard is for BARE pids.
|
|
try:
|
|
subprocess.run(["taskkill", "/T", "/F", "/PID", str(proc.pid)],
|
|
stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL,
|
|
stdin=subprocess.DEVNULL, timeout=2, check=False,
|
|
creationflags=windows_hide_flags())
|
|
except Exception:
|
|
pass
|
|
|
|
|
|
def bounded_probe_run(
|
|
argv: Sequence[str], *, timeout: float, errors: str = "replace",
|
|
env: "Mapping[str, str] | None" = None,
|
|
) -> "subprocess.CompletedProcess[str] | None":
|
|
"""Deadlock-safe ``subprocess.run(argv, capture_output=True, timeout=…)`` for fail-open probes.
|
|
|
|
Returns a ``CompletedProcess`` when the child finished within *timeout* (any exit code), or
|
|
``None`` on spawn failure or timeout.
|
|
"""
|
|
_popen_kwargs: dict = {"creationflags": windows_hide_flags()} if IS_WINDOWS else {"process_group": 0}
|
|
try:
|
|
proc = subprocess.Popen(
|
|
list(argv), stdout=subprocess.PIPE, stderr=subprocess.PIPE, stdin=subprocess.DEVNULL,
|
|
text=True, encoding="utf-8", errors=errors,
|
|
env=dict(env) if env is not None else None, **_popen_kwargs)
|
|
except Exception:
|
|
return None
|
|
try:
|
|
stdout, stderr = proc.communicate(timeout=timeout)
|
|
except Exception:
|
|
# Timeout OR any other communicate() failure (torn-down pipe, decode error): tree-kill and
|
|
# drain bounded — leaving it running would leak the suspended-descendant class this guards.
|
|
kill_process_tree(proc)
|
|
try:
|
|
proc.communicate(timeout=1)
|
|
except Exception:
|
|
pass
|
|
return None
|
|
return subprocess.CompletedProcess(list(argv), proc.returncode, stdout, stderr)
|
|
|
|
|
|
def bounded_git_probe(argv: Sequence[str], *, timeout: float) -> str:
|
|
"""Run a short ``git`` probe and return stripped stdout, or ``""`` on ANY failure.
|
|
|
|
On Windows ``run()``'s post-timeout cleanup calls an unbounded ``communicate()``; a suspended
|
|
descendant git.exe holding the pipe handles then blocks forever. Here: bounded ``communicate``,
|
|
tree-kill plus a 1s drain, then abandon the pipes; on POSIX the probe gets its own process
|
|
group so cleanup also takes down credential/remote helpers.
|
|
|
|
Security (GHSA-7x36-8jrh-v4pw): these probes run automatically against whatever directory the
|
|
session sits in, before any tool call or trust prompt, and an index refresh executes the
|
|
repo-configured ``core.fsmonitor`` program. Every probe therefore runs under
|
|
:func:`noninteractive_git_env`; diff-rendering callers additionally pass
|
|
:data:`NO_DRIVER_DIFF_FLAGS` (attribute-scoped drivers can't be disabled via env).
|
|
"""
|
|
result = bounded_probe_run(argv, timeout=timeout, env=noninteractive_git_env())
|
|
if result is None or result.returncode != 0:
|
|
return ""
|
|
return (result.stdout or "").strip()
|
|
|
|
|
|
# Backward-compat alias — existing call sites/tests import the historical name.
|
|
_kill_git_process_tree = kill_process_tree
|