OpenE2K/rust
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

std: Clean up process spawn impl on unix

Browse Source 
* De-indent quite a bit by removing usage of FnOnce closures
* Clearly separate code for the parent/child after the fork
* Use `fs2::{File, OpenOptions}` instead of calling `open` manually
* Use RAII to close I/O objects wherever possible
* Remove loop for closing all file descriptors, all our own ones are now
  `CLOEXEC` by default so they cannot be inherited
This commit is contained in:
Alex Crichton 2015-04-03 15:34:15 -07:00
parent d6c72306c8
commit 33a2191d0b
6 changed files with 214 additions and 250 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/libstd/process.rs
View File

@ -340,7 +340,7 @@ fn setup_io(io: &StdioImp, fd: libc::c_int, readable: bool)
(Some(AnonPipe::from_fd(fd)), None)
}
Piped => {
let (reader, writer) = try!(unsafe { pipe2::anon_pipe() });
let (reader, writer) = try!(pipe2::anon_pipe());
if readable {
(Some(reader), Some(writer))
} else {

2
src/libstd/sys/unix/c.rs
View File

@ -159,6 +159,8 @@ extern {
pub fn utimes(filename: *const libc::c_char,
times: *const libc::timeval) -> libc::c_int;
pub fn gai_strerror(errcode: libc::c_int) -> *const libc::c_char;
pub fn setgroups(ngroups: libc::c_int,
ptr: *const libc::c_void) -> libc::c_int;
}
#[cfg(any(target_os = "macos", target_os = "ios"))]

8
src/libstd/sys/unix/fs2.rs
View File

@ -205,13 +205,17 @@ impl OpenOptions {
impl File {
pub fn open(path: &Path, opts: &OpenOptions) -> io::Result<File> {
let path = try!(cstr(path));
File::open_c(&path, opts)
}
pub fn open_c(path: &CStr, opts: &OpenOptions) -> io::Result<File> {
let flags = opts.flags | match (opts.read, opts.write) {
(true, true) => libc::O_RDWR,
(false, true) => libc::O_WRONLY,
(true, false) |
(false, false) => libc::O_RDONLY,
};
let path = try!(cstr(path));
let fd = try!(cvt_r(|| unsafe {
libc::open(path.as_ptr(), flags, opts.mode)
}));
@ -220,6 +224,8 @@ impl File {
Ok(File(fd))
}
pub fn into_fd(self) -> FileDesc { self.0 }
pub fn file_attr(&self) -> io::Result<FileAttr> {
let mut stat: libc::stat = unsafe { mem::zeroed() };
try!(cvt(unsafe { libc::fstat(self.0.raw(), &mut stat) }));

11
src/libstd/sys/unix/pipe2.rs
View File

@ -20,11 +20,10 @@ use libc;
pub struct AnonPipe(FileDesc);
pub unsafe fn anon_pipe() -> io::Result<(AnonPipe, AnonPipe)> {
pub fn anon_pipe() -> io::Result<(AnonPipe, AnonPipe)> {
let mut fds = [0; 2];
if libc::pipe(fds.as_mut_ptr()) == 0 {
Ok((AnonPipe::from_fd(fds[0]),
AnonPipe::from_fd(fds[1])))
if unsafe { libc::pipe(fds.as_mut_ptr()) == 0 } {
Ok((AnonPipe::from_fd(fds[0]), AnonPipe::from_fd(fds[1])))
} else {
Err(io::Error::last_os_error())
}
@ -45,7 +44,7 @@ impl AnonPipe {
self.0.write(buf)
}
pub fn raw(&self) -> libc::c_int {
self.0.raw()
pub fn into_fd(self) -> FileDesc {
self.0
}
}

423
src/libstd/sys/unix/process2.rs
View File

@ -13,14 +13,14 @@ use os::unix::prelude::*;
use collections::HashMap;
use env;
use ffi::{OsString, OsStr, CString};
use ffi::{OsString, OsStr, CString, CStr};
use fmt;
use io::{self, Error, ErrorKind};
use libc::{self, pid_t, c_void, c_int, gid_t, uid_t};
use mem;
use ptr;
use sys::pipe2::AnonPipe;
use sys::{self, retry, c, cvt};
use sys::fs2::{File, OpenOptions};
////////////////////////////////////////////////////////////////////////////////
// Command
@ -128,221 +128,178 @@ impl Process {
}
pub fn spawn(cfg: &Command,
in_fd: Option<AnonPipe>, out_fd: Option<AnonPipe>, err_fd: Option<AnonPipe>)
-> io::Result<Process>
{
use libc::funcs::posix88::unistd::{fork, dup2, close, chdir, execvp};
in_fd: Option<AnonPipe>,
out_fd: Option<AnonPipe>,
err_fd: Option<AnonPipe>) -> io::Result<Process> {
let dirp = cfg.cwd.as_ref().map(|c| c.as_ptr()).unwrap_or(ptr::null());
mod rustrt {
extern {
pub fn rust_unset_sigprocmask();
let (envp, _a, _b) = make_envp(cfg.env.as_ref());
let (argv, _a) = make_argv(&cfg.program, &cfg.args);
let (input, output) = try!(sys::pipe2::anon_pipe());
let pid = unsafe {
match libc::fork() {
0 => {
drop(input);
Process::child_after_fork(cfg, output, argv, envp, dirp,
in_fd, out_fd, err_fd)
}
n if n < 0 => return Err(Error::last_os_error()),
n => n,
}
};
let p = Process{ pid: pid };
drop(output);
let mut bytes = [0; 8];
// loop to handle EINTR
loop {
match input.read(&mut bytes) {
Ok(0) => return Ok(p),
Ok(8) => {
assert!(combine(CLOEXEC_MSG_FOOTER) == combine(&bytes[4.. 8]),
"Validation on the CLOEXEC pipe failed: {:?}", bytes);
let errno = combine(&bytes[0.. 4]);
assert!(p.wait().is_ok(),
"wait() should either return Ok or panic");
return Err(Error::from_raw_os_error(errno))
}
Err(ref e) if e.kind() == ErrorKind::Interrupted => {}
Err(e) => {
assert!(p.wait().is_ok(),
"wait() should either return Ok or panic");
panic!("the CLOEXEC pipe failed: {:?}", e)
},
Ok(..) => { // pipe I/O up to PIPE_BUF bytes should be atomic
assert!(p.wait().is_ok(),
"wait() should either return Ok or panic");
panic!("short read on the CLOEXEC pipe")
}
}
}
unsafe fn set_cloexec(fd: c_int) {
let ret = c::ioctl(fd, c::FIOCLEX);
assert_eq!(ret, 0);
fn combine(arr: &[u8]) -> i32 {
let a = arr[0] as u32;
let b = arr[1] as u32;
let c = arr[2] as u32;
let d = arr[3] as u32;
((a << 24) | (b << 16) | (c << 8) | (d << 0)) as i32
}
}
// And at this point we've reached a special time in the life of the
// child. The child must now be considered hamstrung and unable to
// do anything other than syscalls really. Consider the following
// scenario:
//
// 1. Thread A of process 1 grabs the malloc() mutex
// 2. Thread B of process 1 forks(), creating thread C
// 3. Thread C of process 2 then attempts to malloc()
// 4. The memory of process 2 is the same as the memory of
// process 1, so the mutex is locked.
//
// This situation looks a lot like deadlock, right? It turns out
// that this is what pthread_atfork() takes care of, which is
// presumably implemented across platforms. The first thing that
// threads to *before* forking is to do things like grab the malloc
// mutex, and then after the fork they unlock it.
//
// Despite this information, libnative's spawn has been witnessed to
// deadlock on both OSX and FreeBSD. I'm not entirely sure why, but
// all collected backtraces point at malloc/free traffic in the
// child spawned process.
//
// For this reason, the block of code below should contain 0
// invocations of either malloc of free (or their related friends).
//
// As an example of not having malloc/free traffic, we don't close
// this file descriptor by dropping the FileDesc (which contains an
// allocation). Instead we just close it manually. This will never
// have the drop glue anyway because this code never returns (the
// child will either exec() or invoke libc::exit)
unsafe fn child_after_fork(cfg: &Command,
mut output: AnonPipe,
argv: *const *const libc::c_char,
envp: *const libc::c_void,
dirp: *const libc::c_char,
in_fd: Option<AnonPipe>,
out_fd: Option<AnonPipe>,
err_fd: Option<AnonPipe>) -> ! {
fn fail(output: &mut AnonPipe) -> ! {
let errno = sys::os::errno() as u32;
let bytes = [
(errno >> 24) as u8,
(errno >> 16) as u8,
(errno >> 8) as u8,
(errno >> 0) as u8,
CLOEXEC_MSG_FOOTER[0], CLOEXEC_MSG_FOOTER[1],
CLOEXEC_MSG_FOOTER[2], CLOEXEC_MSG_FOOTER[3]
];
// pipe I/O up to PIPE_BUF bytes should be atomic, and then we want
// to be sure we *don't* run at_exit destructors as we're being torn
// down regardless
assert!(output.write(&bytes).is_ok());
unsafe { libc::_exit(1) }
}
#[cfg(all(target_os = "android", target_arch = "aarch64"))]
unsafe fn getdtablesize() -> c_int {
libc::sysconf(libc::consts::os::sysconf::_SC_OPEN_MAX) as c_int
}
// If a stdio file descriptor is set to be ignored, we don't
// actually close it, but rather open up /dev/null into that
// file descriptor. Otherwise, the first file descriptor opened
// up in the child would be numbered as one of the stdio file
// descriptors, which is likely to wreak havoc.
let setup = |src: Option<AnonPipe>, dst: c_int| {
src.map(|p| p.into_fd()).or_else(|| {
let mut opts = OpenOptions::new();
opts.read(dst == libc::STDIN_FILENO);
opts.write(dst != libc::STDIN_FILENO);
let devnull = CStr::from_ptr(b"/dev/null\0".as_ptr()
as *const _);
File::open_c(devnull, &opts).ok().map(|f| f.into_fd())
}).map(|fd| {
fd.unset_cloexec();
retry(|| libc::dup2(fd.raw(), dst)) != -1
}).unwrap_or(false)
};
#[cfg(not(all(target_os = "android", target_arch = "aarch64")))]
unsafe fn getdtablesize() -> c_int {
libc::funcs::bsd44::getdtablesize()
}
if !setup(in_fd, libc::STDIN_FILENO) { fail(&mut output) }
if !setup(out_fd, libc::STDOUT_FILENO) { fail(&mut output) }
if !setup(err_fd, libc::STDERR_FILENO) { fail(&mut output) }
let dirp = cfg.cwd.as_ref().map(|c| c.as_ptr()).unwrap_or(ptr::null());
with_envp(cfg.env.as_ref(), |envp: *const c_void| {
with_argv(&cfg.program, &cfg.args, |argv: *const *const libc::c_char| unsafe {
let (input, mut output) = try!(sys::pipe2::anon_pipe());
// We may use this in the child, so perform allocations before the
// fork
let devnull = b"/dev/null\0";
set_cloexec(output.raw());
let pid = fork();
if pid < 0 {
return Err(Error::last_os_error())
} else if pid > 0 {
#[inline]
fn combine(arr: &[u8]) -> i32 {
let a = arr[0] as u32;
let b = arr[1] as u32;
let c = arr[2] as u32;
let d = arr[3] as u32;
((a << 24) | (b << 16) | (c << 8) | (d << 0)) as i32
}
let p = Process{ pid: pid };
drop(output);
let mut bytes = [0; 8];
// loop to handle EINTER
loop {
match input.read(&mut bytes) {
Ok(8) => {
assert!(combine(CLOEXEC_MSG_FOOTER) == combine(&bytes[4.. 8]),
"Validation on the CLOEXEC pipe failed: {:?}", bytes);
let errno = combine(&bytes[0.. 4]);
assert!(p.wait().is_ok(),
"wait() should either return Ok or panic");
return Err(Error::from_raw_os_error(errno))
}
Ok(0) => return Ok(p),
Err(ref e) if e.kind() == ErrorKind::Interrupted => {}
Err(e) => {
assert!(p.wait().is_ok(),
"wait() should either return Ok or panic");
panic!("the CLOEXEC pipe failed: {:?}", e)
},
Ok(..) => { // pipe I/O up to PIPE_BUF bytes should be atomic
assert!(p.wait().is_ok(),
"wait() should either return Ok or panic");
panic!("short read on the CLOEXEC pipe")
}
}
}
}
// And at this point we've reached a special time in the life of the
// child. The child must now be considered hamstrung and unable to
// do anything other than syscalls really. Consider the following
// scenario:
//
// 1. Thread A of process 1 grabs the malloc() mutex
// 2. Thread B of process 1 forks(), creating thread C
// 3. Thread C of process 2 then attempts to malloc()
// 4. The memory of process 2 is the same as the memory of
// process 1, so the mutex is locked.
//
// This situation looks a lot like deadlock, right? It turns out
// that this is what pthread_atfork() takes care of, which is
// presumably implemented across platforms. The first thing that
// threads to *before* forking is to do things like grab the malloc
// mutex, and then after the fork they unlock it.
//
// Despite this information, libnative's spawn has been witnessed to
// deadlock on both OSX and FreeBSD. I'm not entirely sure why, but
// all collected backtraces point at malloc/free traffic in the
// child spawned process.
//
// For this reason, the block of code below should contain 0
// invocations of either malloc of free (or their related friends).
//
// As an example of not having malloc/free traffic, we don't close
// this file descriptor by dropping the FileDesc (which contains an
// allocation). Instead we just close it manually. This will never
// have the drop glue anyway because this code never returns (the
// child will either exec() or invoke libc::exit)
let _ = libc::close(input.raw());
fn fail(output: &mut AnonPipe) -> ! {
let errno = sys::os::errno() as u32;
let bytes = [
(errno >> 24) as u8,
(errno >> 16) as u8,
(errno >> 8) as u8,
(errno >> 0) as u8,
CLOEXEC_MSG_FOOTER[0], CLOEXEC_MSG_FOOTER[1],
CLOEXEC_MSG_FOOTER[2], CLOEXEC_MSG_FOOTER[3]
];
// pipe I/O up to PIPE_BUF bytes should be atomic
assert!(output.write(&bytes).is_ok());
unsafe { libc::_exit(1) }
}
rustrt::rust_unset_sigprocmask();
// If a stdio file descriptor is set to be ignored, we don't
// actually close it, but rather open up /dev/null into that
// file descriptor. Otherwise, the first file descriptor opened
// up in the child would be numbered as one of the stdio file
// descriptors, which is likely to wreak havoc.
let setup = |src: Option<AnonPipe>, dst: c_int| {
let src = match src {
None => {
let flags = if dst == libc::STDIN_FILENO {
libc::O_RDONLY
} else {
libc::O_RDWR
};
libc::open(devnull.as_ptr() as *const _, flags, 0)
}
Some(obj) => {
let fd = obj.raw();
// Leak the memory and the file descriptor. We're in the
// child now an all our resources are going to be
// cleaned up very soon
mem::forget(obj);
fd
}
};
src != -1 && retry(|| dup2(src, dst)) != -1
};
if !setup(in_fd, libc::STDIN_FILENO) { fail(&mut output) }
if !setup(out_fd, libc::STDOUT_FILENO) { fail(&mut output) }
if !setup(err_fd, libc::STDERR_FILENO) { fail(&mut output) }
// close all other fds
for fd in (3..getdtablesize()).rev() {
if fd != output.raw() {
let _ = close(fd as c_int);
}
}
match cfg.gid {
Some(u) => {
if libc::setgid(u as libc::gid_t) != 0 {
fail(&mut output);
}
}
None => {}
}
match cfg.uid {
Some(u) => {
// When dropping privileges from root, the `setgroups` call
// will remove any extraneous groups. If we don't call this,
// then even though our uid has dropped, we may still have
// groups that enable us to do super-user things. This will
// fail if we aren't root, so don't bother checking the
// return value, this is just done as an optimistic
// privilege dropping function.
extern {
fn setgroups(ngroups: libc::c_int,
ptr: *const libc::c_void) -> libc::c_int;
}
let _ = setgroups(0, ptr::null());
if libc::setuid(u as libc::uid_t) != 0 {
fail(&mut output);
}
}
None => {}
}
if cfg.detach {
// Don't check the error of setsid because it fails if we're the
// process leader already. We just forked so it shouldn't return
// error, but ignore it anyway.
let _ = libc::setsid();
}
if !dirp.is_null() && chdir(dirp) == -1 {
fail(&mut output);
}
if !envp.is_null() {
*sys::os::environ() = envp as *const _;
}
let _ = execvp(*argv, argv as *mut _);
if let Some(u) = cfg.gid {
if libc::setgid(u as libc::gid_t) != 0 {
fail(&mut output);
})
})
}
}
if let Some(u) = cfg.uid {
// When dropping privileges from root, the `setgroups` call
// will remove any extraneous groups. If we don't call this,
// then even though our uid has dropped, we may still have
// groups that enable us to do super-user things. This will
// fail if we aren't root, so don't bother checking the
// return value, this is just done as an optimistic
// privilege dropping function.
let _ = c::setgroups(0, ptr::null());
if libc::setuid(u as libc::uid_t) != 0 {
fail(&mut output);
}
}
if cfg.detach {
// Don't check the error of setsid because it fails if we're the
// process leader already. We just forked so it shouldn't return
// error, but ignore it anyway.
let _ = libc::setsid();
}
if !dirp.is_null() && libc::chdir(dirp) == -1 {
fail(&mut output);
}
if !envp.is_null() {
*sys::os::environ() = envp as *const _;
}
let _ = libc::execvp(*argv, argv as *mut _);
fail(&mut output)
}
pub fn wait(&self) -> io::Result<ExitStatus> {
@ -364,8 +321,8 @@ impl Process {
}
}
fn with_argv<T,F>(prog: &CString, args: &[CString], cb: F) -> T
where F : FnOnce(*const *const libc::c_char) -> T
fn make_argv(prog: &CString, args: &[CString])
-> (*const *const libc::c_char, Vec<*const libc::c_char>)
{
let mut ptrs: Vec<*const libc::c_char> = Vec::with_capacity(args.len()+1);
@ -380,40 +337,38 @@ fn with_argv<T,F>(prog: &CString, args: &[CString], cb: F) -> T
// Add a terminating null pointer (required by libc).
ptrs.push(ptr::null());
cb(ptrs.as_ptr())
(ptrs.as_ptr(), ptrs)
}
fn with_envp<T, F>(env: Option<&HashMap<OsString, OsString>>, cb: F) -> T
where F : FnOnce(*const c_void) -> T
fn make_envp(env: Option<&HashMap<OsString, OsString>>)
-> (*const c_void, Vec<Vec<u8>>, Vec<*const libc::c_char>)
{
// On posixy systems we can pass a char** for envp, which is a
// null-terminated array of "k=v\0" strings. Since we must create
// these strings locally, yet expose a raw pointer to them, we
// create a temporary vector to own the CStrings that outlives the
// call to cb.
match env {
Some(env) => {
let mut tmps = Vec::with_capacity(env.len());
if let Some(env) = env {
let mut tmps = Vec::with_capacity(env.len());
for pair in env {
let mut kv = Vec::new();
kv.push_all(pair.0.as_bytes());
kv.push('=' as u8);
kv.push_all(pair.1.as_bytes());
kv.push(0); // terminating null
tmps.push(kv);
}
// As with `with_argv`, this is unsafe, since cb could leak the pointers.
let mut ptrs: Vec<*const libc::c_char> =
tmps.iter()
.map(|tmp| tmp.as_ptr() as *const libc::c_char)
.collect();
ptrs.push(ptr::null());
cb(ptrs.as_ptr() as *const c_void)
for pair in env {
let mut kv = Vec::new();
kv.push_all(pair.0.as_bytes());
kv.push('=' as u8);
kv.push_all(pair.1.as_bytes());
kv.push(0); // terminating null
tmps.push(kv);
}
_ => cb(ptr::null())
let mut ptrs: Vec<*const libc::c_char> =
tmps.iter()
.map(|tmp| tmp.as_ptr() as *const libc::c_char)
.collect();
ptrs.push(ptr::null());
(ptrs.as_ptr() as *const _, tmps, ptrs)
} else {
(0 as *const _, Vec::new(), Vec::new())
}
}

18
src/libstd/sys/windows/pipe2.rs
View File

@ -22,22 +22,24 @@ pub struct AnonPipe {
fd: c_int
}
pub unsafe fn anon_pipe() -> io::Result<(AnonPipe, AnonPipe)> {
pub fn anon_pipe() -> io::Result<(AnonPipe, AnonPipe)> {
// Windows pipes work subtly differently than unix pipes, and their
// inheritance has to be handled in a different way that I do not
// fully understand. Here we explicitly make the pipe non-inheritable,
// which means to pass it to a subprocess they need to be duplicated
// first, as in std::run.
let mut fds = [0; 2];
match libc::pipe(fds.as_mut_ptr(), 1024 as ::libc::c_uint,
(libc::O_BINARY | libc::O_NOINHERIT) as c_int) {
0 => {
assert!(fds[0] != -1 && fds[0] != 0);
assert!(fds[1] != -1 && fds[1] != 0);
unsafe {
match libc::pipe(fds.as_mut_ptr(), 1024 as ::libc::c_uint,
(libc::O_BINARY | libc::O_NOINHERIT) as c_int) {
0 => {
assert!(fds[0] != -1 && fds[0] != 0);
assert!(fds[1] != -1 && fds[1] != 0);
Ok((AnonPipe::from_fd(fds[0]), AnonPipe::from_fd(fds[1])))
Ok((AnonPipe::from_fd(fds[0]), AnonPipe::from_fd(fds[1])))
}
_ => Err(io::Error::last_os_error()),
}
_ => Err(io::Error::last_os_error()),
}
}