perf(delivery): ship verified Python bytecode caches

2026-10-28 16:40:52 +08:00
parent 81b5258808
commit 70007a36b9
8 changed files with 303 additions and 7 deletions
@@ -0,0 +1,111 @@
+"""Build checked-hash caches from trusted source; never execute cache contents."""
+
+import importlib.util
+import os
+import re
+import stat
+import sys
+from pathlib import Path
+
+from .policy import BYTECODE_ROOTS, DeliveryError, bytecode_source, relative_name
+
+MAX_SOURCE_BYTES = 4 * 1024**2
+
+
+def _stamp(info: os.stat_result) -> tuple[int, ...]:
+    return (
+        info.st_dev,
+        info.st_ino,
+        info.st_mode,
+        info.st_nlink,
+        info.st_uid,
+        info.st_gid,
+        info.st_size,
+        info.st_mtime_ns,
+        info.st_ctime_ns,
+    )
+
+
+def source_bytes(path: Path) -> bytes:
+    descriptor = os.open(path, os.O_RDONLY | os.O_NOFOLLOW | os.O_CLOEXEC | os.O_NONBLOCK)
+    with os.fdopen(descriptor, "rb") as stream:
+        before = os.fstat(stream.fileno())
+        if (
+            not stat.S_ISREG(before.st_mode)
+            or before.st_nlink != 1
+            or stat.S_IMODE(before.st_mode) not in (0o444, 0o555)
+            or before.st_size > MAX_SOURCE_BYTES
+        ):
+            raise DeliveryError("bytecode source requires a bounded read-only ordinary file")
+        data = stream.read(MAX_SOURCE_BYTES + 1)
+        after = os.fstat(stream.fileno())
+    if (
+        len(data) != before.st_size
+        or _stamp(before) != _stamp(after)
+        or _stamp(after) != _stamp(path.lstat())
+    ):
+        raise DeliveryError("bytecode source changed during verification")
+    return data
+
+
+def verify_header(header: bytes, source: Path) -> None:
+    # Bundle SHA-256 verification authenticates every cache byte. This extra
+    # pairing binds the CPython header to the actual, separately hashed source.
+    # Never deserialize or execute an incoming marshal payload to inspect it.
+    if sys.version_info[:3] != (3, 13, 16):
+        raise DeliveryError("bytecode verification requires the pinned delivery interpreter")
+    expected = importlib.util.MAGIC_NUMBER + b"\x03\x00\x00\x00"
+    if len(header) != 16 or header[:8] != expected:
+        raise DeliveryError("only pinned checked-hash Python bytecode is supported")
+    if header[8:] != importlib.util.source_hash(source_bytes(source)):
+        raise DeliveryError("bytecode source hash differs")
+
+
+def compile_payload(payload: Path, release_id: str) -> int:
+    import py_compile
+
+    if (
+        sys.version_info[:3] != (3, 13, 16)
+        or sys.implementation.cache_tag != "cpython-313"
+        or not re.fullmatch(r"sd-[0-9a-f]{12}", release_id)
+    ):
+        raise DeliveryError("bytecode build requires the pinned delivery interpreter")
+    sources = []
+    for directory, directories, files in os.walk(payload, followlinks=False):
+        directories.sort()
+        if "__pycache__" in directories:
+            raise DeliveryError("inherited bytecode cache refused")
+        if any((Path(directory) / name).is_symlink() for name in directories):
+            raise DeliveryError("bytecode build directory link refused")
+        for name in sorted(files):
+            path = Path(directory) / name
+            relative = path.relative_to(payload).as_posix()
+            if relative.startswith(BYTECODE_ROOTS) and relative.endswith(".py"):
+                relative_name(relative)
+                sources.append((path, relative))
+                if len(sources) > 10000:
+                    raise DeliveryError("bytecode source count exceeds build bound")
+    for source, relative in sources:
+        before = _stamp(source.lstat())
+        source_bytes(source)
+        # Explicit cfile prevents an inherited PYTHONPYCACHEPREFIX from sending
+        # output outside this new bundle. Source filenames never expose Host paths.
+        cache = source.parent / "__pycache__" / (source.stem + ".cpython-313.pyc")
+        if bytecode_source(cache.relative_to(payload).as_posix()) != relative:
+            raise DeliveryError("bytecode source pairing differs")
+        if cache.exists() or cache.is_symlink():
+            raise DeliveryError("existing bytecode cache refused")
+        py_compile.compile(
+            str(source),
+            cfile=str(cache),
+            dfile=f"/opt/kf-onlinejudge/releases/{release_id}/payload/{relative}",
+            doraise=True,
+            optimize=0,
+            invalidation_mode=py_compile.PycInvalidationMode.CHECKED_HASH,
+        )
+        cache.chmod(0o444)
+        with cache.open("rb") as stream:
+            verify_header(stream.read(16), source)
+        if _stamp(source.lstat()) != before:
+            raise DeliveryError("source changed during bytecode compilation")
+    return len(sources)