PEP 446 – Make newly created file descriptors non-inheritable
- Author:
- Victor Stinner <vstinner at python.org>
- Status:
- Final
- Type:
- Standards Track
- Created:
- 05-Aug-2013
- Python-Version:
- 3.4
- Replaces:
- 433
Table of Contents
- Abstract
- Rationale
- Inheritance of File Descriptors
- Inheritance of File Descriptors on Windows
- Only Inherit Some Handles on Windows
- Inheritance of File Descriptors on UNIX
- Issues with Inheritable File Descriptors
- Security Vulnerability
- Issues fixed in the subprocess module
- Atomic Creation of non-inheritable File Descriptors
- Status of Python 3.3
- Closing All Open File Descriptors
- Proposal
- Backward Compatibility
- Related Work
- Rejected Alternatives
- Python Issues
- Copyright
Abstract
Leaking file descriptors in child processes causes various annoying
issues and is a known major security vulnerability. Using the
subprocess
module with the close_fds parameter set to True
is
not possible in all cases.
This PEP proposes to make all file descriptors created by Python non-inheritable by default to reduce the risk of these issues. This PEP fixes also a race condition in multi-threaded applications on operating systems supporting atomic flags to create non-inheritable file descriptors.
We are aware of the code breakage this is likely to cause, and doing it anyway for the good of mankind. (Details in the section “Backward Compatibility” below.)
Rationale
Inheritance of File Descriptors
Each operating system handles the inheritance of file descriptors differently. Windows creates non-inheritable handles by default, whereas UNIX and the POSIX API on Windows create inheritable file descriptors by default. Python prefers the POSIX API over the native Windows API, to have a single code base and to use the same type for file descriptors, and so it creates inheritable file descriptors.
There is one exception: os.pipe()
creates non-inheritable pipes on
Windows, whereas it creates inheritable pipes on UNIX. The reason is an
implementation artifact: os.pipe()
calls CreatePipe()
on Windows
(native API), whereas it calls pipe()
on UNIX (POSIX API). The call
to CreatePipe()
was added in Python in 1994, before the introduction
of pipe()
in the POSIX API in Windows 98. The issue #4708 proposes to change os.pipe()
on
Windows to create inheritable pipes.
Inheritance of File Descriptors on Windows
On Windows, the native type of file objects is handles (C type
HANDLE
). These handles have a HANDLE_FLAG_INHERIT
flag which
defines if a handle can be inherited in a child process or not. For the
POSIX API, the C runtime (CRT) also provides file descriptors (C type
int
). The handle of a file descriptor can be retrieve using the
function _get_osfhandle(fd)
. A file descriptor can be created from a
handle using the function _open_osfhandle(handle)
.
Using CreateProcess(),
handles are only inherited if their inheritable flag
(HANDLE_FLAG_INHERIT
) is set and the bInheritHandles
parameter of CreateProcess()
is TRUE
; all file descriptors
except standard streams (0, 1, 2) are closed in the child process, even
if bInheritHandles
is TRUE
. Using the spawnv()
function, all
inheritable handles and all inheritable file descriptors are inherited
in the child process. This function uses the undocumented fields
cbReserved2 and lpReserved2 of the STARTUPINFO
structure to pass an array of file descriptors.
To replace standard streams (stdin, stdout, stderr) using
CreateProcess()
, the STARTF_USESTDHANDLES
flag must be set in
the dwFlags field of the STARTUPINFO
structure and the
bInheritHandles parameter of CreateProcess()
must be set to
TRUE
. So when at least one standard stream is replaced, all
inheritable handles are inherited by the child process.
The default value of the close_fds parameter of subprocess
process
is True
(bInheritHandles=FALSE
) if stdin, stdout and
stderr parameters are None
, False
(bInheritHandles=TRUE
)
otherwise.
See also:
Only Inherit Some Handles on Windows
Since Windows Vista, CreateProcess()
supports an extension of the
STARTUPINFO structure: the STARTUPINFOEX structure.
Using this new structure, it is possible to specify a list of handles to
inherit: PROC_THREAD_ATTRIBUTE_HANDLE_LIST
. Read Programmatically
controlling which handles are inherited by new processes in Win32
(Raymond Chen, Dec 2011) for more information.
Before Windows Vista, it is possible to make handles inheritable and
call CreateProcess()
with bInheritHandles=TRUE
. This option
works if all other handles are non-inheritable. There is a race
condition: if another thread calls CreateProcess()
with
bInheritHandles=TRUE
, handles will also be inherited in the second
process.
Microsoft suggests to use a lock to avoid the race condition: read Q315939: PRB: Child Inherits Unintended Handles During CreateProcess Call (last review: November 2006). The Python issue #16500 “Add an atfork module” proposes to add such lock, it can be used to make handles non-inheritable without the race condition. Such lock only protects against a race condition between Python threads; C threads are not protected.
Another option is to duplicate handles that must be inherited, passing the
values of the duplicated handles to the child process, so the child
process can steal duplicated handles using DuplicateHandle()
with DUPLICATE_CLOSE_SOURCE
. Handle values change between the
parent and the child process because the handles are duplicated (twice);
the parent and/or the child process must be adapted to handle this
change. If the child program cannot be modified, an intermediate program
can be used to steal handles from the parent process before spawning the
final child program. The intermediate program has to pass the handle from the
child process to the parent process. The parent may have to close
duplicated handles if all handles were not stolen, for example if the
intermediate process fails. If the command line is used to pass the
handle values, the command line must be modified when handles are
duplicated, because their values are modified.
This PEP does not include a solution to this problem because there is no perfect solution working on all Windows versions. This point is deferred until use cases relying on handle or file descriptor inheritance on Windows are well known, so we can choose the best solution and carefully test its implementation.
Inheritance of File Descriptors on UNIX
POSIX provides a close-on-exec flag on file descriptors to automatically
close a file descriptor when the C function execv()
is
called. File descriptors with the close-on-exec flag cleared are
inherited in the child process, file descriptors with the flag set are
closed in the child process.
The flag can be set in two syscalls (one to get current flags, a second
to set new flags) using fcntl()
:
int flags, res;
flags = fcntl(fd, F_GETFD);
if (flags == -1) { /* handle the error */ }
flags |= FD_CLOEXEC;
/* or "flags &= ~FD_CLOEXEC;" to clear the flag */
res = fcntl(fd, F_SETFD, flags);
if (res == -1) { /* handle the error */ }
FreeBSD, Linux, Mac OS X, NetBSD, OpenBSD and QNX also support setting the flag in a single syscall using ioctl():
int res;
res = ioctl(fd, FIOCLEX, 0);
if (!res) { /* handle the error */ }
NOTE: The close-on-exec flag has no effect on fork()
: all file
descriptors are inherited by the child process. The Python issue #16500
“Add an atfork module” proposes to
add a new atfork
module to execute code at fork, which may be used to
automatically close file descriptors.
Issues with Inheritable File Descriptors
Most of the time, inheritable file descriptors “leaked” to child processes are not noticed, because they don’t cause major bugs. It does not mean that these bugs must not be fixed.
Two common issues with inherited file descriptors:
- On Windows, a directory cannot be removed before all file handles open
in the directory are closed. The same issue can be seen with files,
except if the file was created with the
FILE_SHARE_DELETE
flag (O_TEMPORARY
mode foropen()
). - If a listening socket is leaked to a child process, the socket address cannot be reused before the parent and child processes terminated. For example, if a web server spawns a new program to handle a process, and the server restarts while the program is not done, the server cannot start because the TCP port is still in use.
Example of issues in open source projects:
- Mozilla (Firefox): open since 2002-05
- dbus library: fixed in 2008-05 (dbus commit), close file descriptors in the child process
- autofs: fixed in 2009-02, set the CLOEXEC flag
- qemu: fixed in 2009-12 (qemu commit), set CLOEXEC flag
- Tor: fixed in 2010-12, set CLOEXEC flag
- OCaml: open since 2011-04, “PR#5256: Processes opened using Unix.open_process* inherit all opened file descriptors (including sockets)”
- ØMQ: open since 2012-08
- Squid: open since 2012-07
See also: Excuse me son, but your code is leaking !!! (Dan Walsh, March 2012) for SELinux issues with leaked file descriptors.
Security Vulnerability
Leaking sensitive file handles and file descriptors can lead to security vulnerabilities. An untrusted child process might read sensitive data like passwords or take control of the parent process though a leaked file descriptor. With a leaked listening socket, a child process can accept new connections to read sensitive data.
Example of vulnerabilities:
- Hijacking Apache https by mod_php (2003)
- Apache: Apr should set FD_CLOEXEC if APR_FOPEN_NOCLEANUP is not set: fixed in 2009
- PHP: system() (and similar) don’t cleanup opened handles of Apache: open since 2006
- CWE-403: Exposure of File Descriptor to Unintended Control Sphere (2008)
- OpenSSH Security Advisory: portable-keysign-rand-helper.adv (2011)
Read also the CERT Secure Coding Standards: FIO42-C. Ensure files are properly closed when they are no longer needed.
Issues fixed in the subprocess module
Inherited file descriptors caused 4 issues in the subprocess
module:
- Issue #2320: Race condition in subprocess using stdin (opened in 2008)
- Issue #3006: subprocess.Popen causes socket to remain open after close (opened in 2008)
- Issue #7213: subprocess leaks open file descriptors between Popen instances causing hangs (opened in 2009)
- Issue #12786: subprocess wait() hangs when stdin is closed (opened in 2011)
These issues were fixed in Python 3.2 by 4 different changes in the
subprocess
module:
- Pipes are now non-inheritable;
- The default value of the close_fds parameter is now
True
, with one exception on Windows: the default value isFalse
if at least one standard stream is replaced; - A new pass_fds parameter has been added;
- Creation of a
_posixsubprocess
module implemented in C.
Atomic Creation of non-inheritable File Descriptors
In a multi-threaded application, an inheritable file descriptor may be created just before a new program is spawned, before the file descriptor is made non-inheritable. In this case, the file descriptor is leaked to the child process. This race condition could be avoided if the file descriptor is created directly non-inheritable.
FreeBSD, Linux, Mac OS X, Windows and many other operating systems support creating non-inheritable file descriptors with the inheritable flag cleared atomically at the creation of the file descriptor.
A new WSA_FLAG_NO_HANDLE_INHERIT
flag for WSASocket()
was added
in Windows 7 SP1 and Windows Server 2008 R2 SP1 to create
non-inheritable sockets. If this flag is used on an older Windows
version (ex: Windows XP SP3), WSASocket()
fails with
WSAEPROTOTYPE
.
On UNIX, new flags were added for files and sockets:
O_CLOEXEC
: available on Linux (2.6.23), FreeBSD (8.3), Mac OS 10.8, OpenBSD 5.0, Solaris 11, QNX, BeOS, next NetBSD release (6.1?). This flag is part of POSIX.1-2008.SOCK_CLOEXEC
flag forsocket()
andsocketpair()
, available on Linux 2.6.27, OpenBSD 5.2, NetBSD 6.0.fcntl()
:F_DUPFD_CLOEXEC
flag, available on Linux 2.6.24, OpenBSD 5.0, FreeBSD 9.1, NetBSD 6.0, Solaris 11. This flag is part of POSIX.1-2008.fcntl()
:F_DUP2FD_CLOEXEC
flag, available on FreeBSD 9.1 and Solaris 11.recvmsg()
:MSG_CMSG_CLOEXEC
, available on Linux 2.6.23, NetBSD 6.0.
On Linux older than 2.6.23, O_CLOEXEC
flag is simply ignored. So
fcntl()
must be called to check if the file descriptor is
non-inheritable: O_CLOEXEC
is not supported if the FD_CLOEXEC
flag is missing. On Linux older than 2.6.27, socket()
or
socketpair()
fail with errno
set to EINVAL
if the
SOCK_CLOEXEC
flag is set in the socket type.
New functions:
dup3()
: available on Linux 2.6.27 (and glibc 2.9)pipe2()
: available on Linux 2.6.27 (and glibc 2.9)accept4()
: available on Linux 2.6.28 (and glibc 2.10)
On Linux older than 2.6.28, accept4()
fails with errno
set to
ENOSYS
.
Summary:
Operating System | Atomic File | Atomic Socket |
---|---|---|
FreeBSD | 8.3 (2012) | X |
Linux | 2.6.23 (2007) | 2.6.27 (2008) |
Mac OS X | 10.8 (2012) | X |
NetBSD | 6.1 (?) | 6.0 (2012) |
OpenBSD | 5.0 (2011) | 5.2 (2012) |
Solaris | 11 (2011) | X |
Windows | XP (2001) | Seven SP1 (2011), 2008 R2 SP1 (2011) |
Legend:
- “Atomic File”: first version of the operating system supporting
creating atomically a non-inheritable file descriptor using
open()
- “Atomic Socket”: first version of the operating system supporting creating atomically a non-inheritable socket
- “X”: not supported yet
See also:
- Secure File Descriptor Handling (Ulrich Drepper, 2008)
- Ghosts of Unix past, part 2: Conflated designs (Neil Brown, 2010) explains the
history of
O_CLOEXEC
andO_NONBLOCK
flags - File descriptor handling changes in 2.6.27
- FreeBSD: atomic close on exec
Status of Python 3.3
Python 3.3 creates inheritable file descriptors on all platforms, except
os.pipe()
which creates non-inheritable file descriptors on Windows.
New constants and functions related to the atomic creation of
non-inheritable file descriptors were added to Python 3.3:
os.O_CLOEXEC
, os.pipe2()
and socket.SOCK_CLOEXEC
.
On UNIX, the subprocess
module closes all file descriptors in the
child process by default, except standard streams (0, 1, 2) and file
descriptors of the pass_fds parameter. If the close_fds parameter is
set to False
, all inheritable file descriptors are inherited in the
child process.
On Windows, the subprocess
closes all handles and file descriptors
in the child process by default. If at least one standard stream (stdin,
stdout or stderr) is replaced (ex: redirected into a pipe), all
inheritable handles and file descriptors 0, 1 and 2 are inherited in the
child process.
Using the functions of the os.execv*()
and os.spawn*()
families,
all inheritable handles and all inheritable file descriptors are
inherited by the child process.
On UNIX, the multiprocessing
module uses os.fork()
and so all
file descriptors are inherited by child processes.
On Windows, all inheritable handles and file descriptors 0, 1 and 2 are
inherited by the child process using the multiprocessing
module, all
file descriptors except standard streams are closed.
Summary:
Module | FD on UNIX | Handles on Windows | FD on Windows |
---|---|---|---|
subprocess, default | STD, pass_fds | none | STD |
subprocess, replace stdout | STD, pass_fds | all | STD |
subprocess, close_fds=False | all | all | STD |
multiprocessing | not applicable | all | STD |
os.execv(), os.spawn() | all | all | all |
Legend:
- “all”: all inheritable file descriptors or handles are inherited in the child process
- “none”: all handles are closed in the child process
- “STD”: only file descriptors 0 (stdin), 1 (stdout) and 2 (stderr) are inherited in the child process
- “pass_fds”: file descriptors of the pass_fds parameter of the subprocess are inherited
- “not applicable”: on UNIX, the multiprocessing uses
fork()
, so this case is not affected by this PEP.
Closing All Open File Descriptors
On UNIX, the subprocess
module closes almost all file descriptors in
the child process. This operation requires MAXFD system calls, where
MAXFD is the maximum number of file descriptors, even if there are only
few open file descriptors. This maximum can be read using:
os.sysconf("SC_OPEN_MAX")
.
The operation can be slow if MAXFD is large. For example, on a FreeBSD
buildbot with MAXFD=655,000
, the operation took 300 ms: see
issue #11284: slow close file descriptors.
On Linux, Python 3.3 gets the list of all open file descriptors from
/proc/<PID>/fd/
, and so performances depends on the number of open
file descriptors, not on MAXFD.
See also:
- Python issue #1663329:
subprocess close_fds perform poor if
SC_OPEN_MAX
is high - Squid Bug #837033: Squid should set CLOEXEC on opened FDs. “32k+ close() calls in each child process take a long time ([12-56] seconds) in Xen PV guests.”
Proposal
Non-inheritable File Descriptors
The following functions are modified to make newly created file descriptors non-inheritable by default:
asyncore.dispatcher.create_socket()
io.FileIO
io.open()
open()
os.dup()
os.fdopen()
os.open()
os.openpty()
os.pipe()
select.devpoll()
select.epoll()
select.kqueue()
socket.socket()
socket.socket.accept()
socket.socket.dup()
socket.socket.fromfd()
socket.socketpair()
os.dup2()
still creates inheritable by default, see below.
When available, atomic flags are used to make file descriptors non-inheritable. The atomicity is not guaranteed because a fallback is required when atomic flags are not available.
New Functions And Methods
New functions available on all platforms:
os.get_inheritable(fd: int)
: returnTrue
if the file descriptor can be inherited by child processes,False
otherwise.os.set_inheritable(fd: int, inheritable: bool)
: set the inheritable flag of the specified file descriptor.
New functions only available on Windows:
os.get_handle_inheritable(handle: int)
: returnTrue
if the handle can be inherited by child processes,False
otherwise.os.set_handle_inheritable(handle: int, inheritable: bool)
: set the inheritable flag of the specified handle.
New methods:
socket.socket.get_inheritable()
: returnTrue
if the socket can be inherited by child processes,False
otherwise.socket.socket.set_inheritable(inheritable: bool)
: set the inheritable flag of the specified socket.
Other Changes
On UNIX, subprocess makes file descriptors of the pass_fds parameter
inheritable. The file descriptor is made inheritable in the child
process after the fork()
and before execv()
, so the inheritable
flag of file descriptors is unchanged in the parent process.
os.dup2()
has a new optional inheritable parameter: os.dup2(fd,
fd2, inheritable=True)
. fd2 is created inheritable by default, but
non-inheritable if inheritable is False
.
os.dup2()
behaves differently than os.dup()
because the most
common use case of os.dup2()
is to replace the file descriptors of
the standard streams: stdin
(0
), stdout
(1
) and
stderr
(2
). Standard streams are expected to be inherited by
child processes.
Backward Compatibility
This PEP break applications relying on inheritance of file descriptors.
Developers are encouraged to reuse the high-level Python module
subprocess
which handles the inheritance of file descriptors in a
portable way.
Applications using the subprocess
module with the pass_fds
parameter or using only os.dup2()
to redirect standard streams should
not be affected.
Python no longer conform to POSIX, since file descriptors are now made non-inheritable by default. Python was not designed to conform to POSIX, but was designed to develop portable applications.
Rejected Alternatives
Add a new open_noinherit() function
In June 2007, Henning von Bargen proposed on the python-dev mailing list
to add a new open_noinherit() function to fix issues of inherited file
descriptors in child processes. At this time, the default value of the
close_fds parameter of the subprocess module was False
.
Read the mail thread: [Python-Dev] Proposal for a new function “open_noinherit” to avoid problems with subprocesses and security risks.
PEP 433
PEP 433, “Easier suppression of file descriptor inheritance”, was a previous attempt proposing various other alternatives, but no consensus could be reached.
Python Issues
- #10115: Support accept4() for atomic setting of flags at socket creation
- #12105: open() does not able to set flags, such as O_CLOEXEC
- #12107: TCP listening sockets created without FD_CLOEXEC flag
- #16850: Add “e” mode to open(): close-and-exec (O_CLOEXEC) / O_NOINHERIT
- #16860: Use O_CLOEXEC in the tempfile module
- #16946: subprocess: _close_open_fd_range_safe() does not set close-on-exec flag on Linux < 2.6.23 if O_CLOEXEC is defined
- #17070: Use the new cloexec to improve security and avoid bugs
- #18571: Implementation of the PEP 446: non-inheritable file descriptors
Copyright
This document has been placed into the public domain.
Source: https://github.com/python/peps/blob/main/peps/pep-0446.rst
Last modified: 2023-09-09 17:39:29 GMT