.\" Copyright, the authors of the Linux man-pages project
.\"
.\" SPDX-License-Identifier: Linux-man-pages-copyleft
.\"
.TH pidfd_open 2 (date) "Linux man-pages (unreleased)"
.SH NAME
pidfd_open \- obtain a file descriptor that refers to a task
.SH LIBRARY
Standard C library
.RI ( libc ,\~ \-lc )
.SH SYNOPSIS
.nf
.BR "#include <sys/syscall.h>" "      /* Definition of " SYS_* " constants */"
.B #include <unistd.h>
.P
.BI "int syscall(SYS_pidfd_open, pid_t " pid ", unsigned int " flags );
.fi
.P
.IR Note :
glibc provides no wrapper for
.BR pidfd_open (),
necessitating the use of
.BR syscall (2).
.SH DESCRIPTION
The
.BR pidfd_open ()
system call creates a file descriptor that refers to
the task referenced by
.IR pid .
The file descriptor is returned as the function result;
the close-on-exec flag is set on the file descriptor.
.P
The
.I flags
argument either has the value 0, or contains the following flags:
.TP
.BR PIDFD_NONBLOCK " (since Linux 5.10)"
.\" commit 4da9af0014b51c8b015ed8c622440ef28912efe6
Return a nonblocking file descriptor.
If the task referred to by the file descriptor has not yet terminated,
then an attempt to wait on the file descriptor using
.BR waitid (2)
will immediately return the error
.B EAGAIN
rather than blocking.
.TP
.BR PIDFD_THREAD " (since Linux 6.9)"
.\" commit 64bef697d33b75fc06c5789b3f8108680271529f
Create a file descriptor that refers to a specific thread,
rather than a process (thread-group leader).
If this flag is not set,
.I pid
must refer to a process.
.SH RETURN VALUE
On success,
.BR pidfd_open ()
returns a file descriptor (a nonnegative integer).
On error, \-1 is returned and
.I errno
is set to indicate the error.
.SH ERRORS
.TP
.B EINVAL
.I flags
is not valid.
.TP
.B EINVAL
.I pid
is not valid.
.TP
.B EMFILE
The per-process limit on the number of open file descriptors has been reached
(see the description of
.B RLIMIT_NOFILE
in
.BR getrlimit (2)).
.TP
.B ENFILE
The system-wide limit on the total number of open files has been reached.
.TP
.B ENODEV
The anonymous inode filesystem is not available in this kernel.
.TP
.B ENOMEM
Insufficient kernel memory was available.
.TP
.B ESRCH
The process specified by
.I pid
does not exist.
.SH STANDARDS
Linux.
.SH HISTORY
Linux 5.3.
.SH NOTES
The following code sequence can be used to obtain a file descriptor
for the child of
.BR fork (2):
.P
.in +4n
.EX
pid = fork();
if (pid > 0) {     /* If parent */
    pidfd = pidfd_open(pid, 0);
    ...
}
.EE
.in
.P
Even if the child has already terminated by the time of the
.BR pidfd_open ()
call, its PID will not have been recycled and the returned
file descriptor will refer to the resulting zombie process.
Note, however, that this is guaranteed only if the following
conditions hold true:
.IP \[bu] 3
the disposition of
.B SIGCHLD
has not been explicitly set to
.B SIG_IGN
(see
.BR sigaction (2));
.IP \[bu]
the
.B SA_NOCLDWAIT
flag was not specified while establishing a handler for
.B SIGCHLD
or while setting the disposition of that signal to
.B SIG_DFL
(see
.BR sigaction (2));
and
.IP \[bu]
the zombie process was not reaped elsewhere in the program
(e.g., either by an asynchronously executed signal handler or by
.BR wait (2)
or similar in another thread).
.P
If any of these conditions does not hold,
then the child process (along with a PID file descriptor that refers to it)
should instead be created using
.BR clone (2)
with the
.B CLONE_PIDFD
flag.
.\"
.SS Use cases for PID file descriptors
A PID file descriptor returned by
.BR pidfd_open ()
(or by
.BR clone (2)
with the
.B CLONE_PID
flag) can be used for the following purposes:
.IP \[bu] 3
The
.BR pidfd_send_signal (2)
system call can be used to send a signal to the process referred to by
a PID file descriptor.
.IP \[bu]
A PID file descriptor can be monitored using
.BR poll (2),
.BR select (2),
and
.BR epoll (7).
.IP
When the task that it refers to terminates and becomes a zombie,
these interfaces indicate the file descriptor as readable
.RB ( EPOLLIN ).
When the task is reaped, these interfaces produce a hangup event
.\" commit 43f0df54c96f
.RB ( EPOLLHUP ).
.IP
Note, however, that in the current implementation,
nothing can be read from the file descriptor
.RB ( read (2)
on the file descriptor fails with the error
.BR EINVAL ).
The polling behavior depends on whether
.B PIDFD_THREAD
flag was used when obtaining the file descriptor:
.RS
.IP \[bu] 3
With
.BR PIDFD_THREAD ,
the file descriptor becomes readable
when the task exits and becomes a zombie,
even if the thread-group is not empty.
.IP \[bu]
Without
.BR PIDFD_THREAD ,
the file descriptor becomes readable
only when the last thread in the thread group exits.
.RE
.IP \[bu]
If the PID file descriptor refers to a child of the calling process,
then it can be waited on using
.BR waitid (2).
.IP \[bu]
The
.BR pidfd_getfd (2)
system call can be used to obtain a duplicate of a file descriptor
of another process referred to by a PID file descriptor.
.IP \[bu]
A PID file descriptor can be used as the argument of
.BR setns (2)
in order to move into one or more of the same namespaces as the process
referred to by the file descriptor.
.IP \[bu]
A PID file descriptor can be used as the argument of
.BR process_madvise (2)
in order to provide advice on the memory usage patterns of the process
referred to by the file descriptor.
.P
The
.BR pidfd_open ()
system call is the preferred way of obtaining a PID file descriptor
for an already existing process.
The alternative is to obtain a file descriptor by opening a
.IR /proc/ pid
directory.
However, the latter technique is possible only if the
.BR proc (5)
filesystem is mounted;
furthermore, the file descriptor obtained in this way is
.I not
pollable and can't be waited on with
.BR waitid (2).
.SH EXAMPLES
The program below opens a PID file descriptor for the
process whose PID is specified as its command-line argument.
It then uses
.BR poll (2)
to monitor the file descriptor for process exit, as indicated by an
.B EPOLLIN
event.
.\"
.SS Program source
\&
.\" SRC BEGIN (pidfd_open.c)
.EX
#define _GNU_SOURCE
#include <poll.h>
#include <stdio.h>
#include <stdlib.h>
#include <sys/syscall.h>
#include <sys/types.h>
#include <unistd.h>
\&
static int
pidfd_open(pid_t pid, unsigned int flags)
{
    return syscall(SYS_pidfd_open, pid, flags);
}
\&
int
main(int argc, char *argv[])
{
    int            pidfd, ready;
    struct pollfd  pollfd;
\&
    if (argc != 2) {
        fprintf(stderr, "Usage: %s <pid>\[rs]n", argv[0]);
        exit(EXIT_SUCCESS);
    }
\&
    pidfd = pidfd_open(atoi(argv[1]), 0);
    if (pidfd == \-1) {
        perror("pidfd_open");
        exit(EXIT_FAILURE);
    }
\&
    pollfd.fd = pidfd;
    pollfd.events = POLLIN;
\&
    ready = poll(&pollfd, 1, \-1);
    if (ready == \-1) {
        perror("poll");
        exit(EXIT_FAILURE);
    }
\&
    printf("Events (%#x): POLLIN is %sset\[rs]n", pollfd.revents,
           (pollfd.revents & POLLIN) ? "" : "not ");
\&
    close(pidfd);
    exit(EXIT_SUCCESS);
}
.EE
.\" SRC END
.SH SEE ALSO
.BR clone (2),
.BR kill (2),
.BR pidfd_getfd (2),
.BR pidfd_send_signal (2),
.BR poll (2),
.BR process_madvise (2),
.BR select (2),
.BR setns (2),
.BR waitid (2),
.BR epoll (7)