path: root/man2/semop.2

.\" Copyright 1993 Giorgio Ciucci (giorgio@crcc.it)
.\"
.\" SPDX-License-Identifier: Linux-man-pages-copyleft
.\"
.\" Modified 1996-10-22, Eric S. Raymond <esr@thyrsus.com>
.\" Modified 2002-01-08, Michael Kerrisk <mtk.manpages@gmail.com>
.\" Modified 2003-04-28, Ernie Petrides <petrides@redhat.com>
.\" Modified 2004-05-27, Michael Kerrisk <mtk.manpages@gmail.com>
.\" Modified, 11 Nov 2004, Michael Kerrisk <mtk.manpages@gmail.com>
.\"	Language and formatting clean-ups
.\"	Added notes on /proc files
.\" 2005-04-08, mtk, Noted kernel version numbers for semtimedop()
.\" 2007-07-09, mtk, Added an EXAMPLE code segment.
.\"
.TH semop 2 2023-10-31 "Linux man-pages 6.7"
.SH NAME
semop, semtimedop \- System V semaphore operations
.SH LIBRARY
Standard C library
.RI ( libc ", " \-lc )
.SH SYNOPSIS
.nf
.B #include <sys/sem.h>
.P
.BI "int semop(int " semid ", struct sembuf *" sops ", size_t " nsops );
.BI "int semtimedop(int " semid ", struct sembuf *" sops ", size_t " nsops ,
.BI "               const struct timespec *_Nullable " timeout );
.fi
.P
.RS -4
Feature Test Macro Requirements for glibc (see
.BR feature_test_macros (7)):
.RE
.P
.BR semtimedop ():
.nf
    _GNU_SOURCE
.fi
.SH DESCRIPTION
Each semaphore in a System\ V semaphore set
has the following associated values:
.P
.in +4n
.EX
unsigned short  semval;   /* semaphore value */
unsigned short  semzcnt;  /* # waiting for zero */
unsigned short  semncnt;  /* # waiting for increase */
pid_t           sempid;   /* PID of process that last
                             modified the semaphore value */
.EE
.in
.P
.BR semop ()
performs operations on selected semaphores in the set indicated by
.IR semid .
Each of the
.I nsops
elements in the array pointed to by
.I sops
is a structure that
specifies an operation to be performed on a single semaphore.
The elements of this structure are of type
.IR "struct sembuf" ,
containing the following members:
.P
.in +4n
.EX
unsigned short sem_num;  /* semaphore number */
short          sem_op;   /* semaphore operation */
short          sem_flg;  /* operation flags */
.EE
.in
.P
Flags recognized in
.I sem_flg
are
.B IPC_NOWAIT
and
.BR SEM_UNDO .
If an operation specifies
.BR SEM_UNDO ,
it will be automatically undone when the process terminates.
.P
The set of operations contained in
.I sops
is performed in
.IR "array order" ,
and
.IR atomically ,
that is, the operations are performed either as a complete unit,
or not at all.
The behavior of the system call if not all operations can be
performed immediately depends on the presence of the
.B IPC_NOWAIT
flag in the individual
.I sem_flg
fields, as noted below.
.P
Each operation is performed on the
.IR sem_num \-th
semaphore of the semaphore set, where the first semaphore of the set
is numbered 0.
There are three types of operation, distinguished by the value of
.IR sem_op .
.P
If
.I sem_op
is a positive integer, the operation adds this value to
the semaphore value
.RI  ( semval ).
Furthermore, if
.B SEM_UNDO
is specified for this operation, the system subtracts the value
.I sem_op
from the semaphore adjustment
.RI ( semadj )
value for this semaphore.
This operation can always proceed\[em]it never forces a thread to wait.
The calling process must have alter permission on the semaphore set.
.P
If
.I sem_op
is zero, the process must have read permission on the semaphore
set.
This is a "wait-for-zero" operation: if
.I semval
is zero, the operation can immediately proceed.
Otherwise, if
.B IPC_NOWAIT
is specified in
.IR sem_flg ,
.BR semop ()
fails with
.I errno
set to
.B EAGAIN
(and none of the operations in
.I sops
is performed).
Otherwise,
.I semzcnt
(the count of threads waiting until this semaphore's value becomes zero)
is incremented by one and the thread sleeps until
one of the following occurs:
.IP \[bu] 3
.I semval
becomes 0, at which time the value of
.I semzcnt
is decremented.
.IP \[bu]
The semaphore set
is removed:
.BR semop ()
fails, with
.I errno
set to
.BR EIDRM .
.IP \[bu]
The calling thread catches a signal:
the value of
.I semzcnt
is decremented and
.BR semop ()
fails, with
.I errno
set to
.BR EINTR .
.P
If
.I sem_op
is less than zero, the process must have alter permission on the
semaphore set.
If
.I semval
is greater than or equal to the absolute value of
.IR sem_op ,
the operation can proceed immediately:
the absolute value of
.I sem_op
is subtracted from
.IR semval ,
and, if
.B SEM_UNDO
is specified for this operation, the system adds the absolute value of
.I sem_op
to the semaphore adjustment
.RI ( semadj )
value for this semaphore.
If the absolute value of
.I sem_op
is greater than
.IR semval ,
and
.B IPC_NOWAIT
is specified in
.IR sem_flg ,
.BR semop ()
fails, with
.I errno
set to
.B EAGAIN
(and none of the operations in
.I sops
is performed).
Otherwise,
.I semncnt
(the counter of threads waiting for this semaphore's value to increase)
is incremented by one and the thread sleeps until
one of the following occurs:
.IP \[bu] 3
.I semval
becomes greater than or equal to the absolute value of
.IR sem_op :
the operation now proceeds, as described above.
.IP \[bu]
The semaphore set is removed from the system:
.BR semop ()
fails, with
.I errno
set to
.BR EIDRM .
.IP \[bu]
The calling thread catches a signal:
the value of
.I semncnt
is decremented and
.BR semop ()
fails, with
.I errno
set to
.BR EINTR .
.P
On successful completion, the
.I sempid
value for each semaphore specified in the array pointed to by
.I sops
is set to the caller's process ID.
In addition, the
.I sem_otime
.\" and
.\" .I sem_ctime
is set to the current time.
.SS semtimedop()
.BR semtimedop ()
behaves identically to
.BR semop ()
except that in those cases where the calling thread would sleep,
the duration of that sleep is limited by the amount of elapsed
time specified by the
.I timespec
structure whose address is passed in the
.I timeout
argument.
(This sleep interval will be rounded up to the system clock granularity,
and kernel scheduling delays mean that the interval
may overrun by a small amount.)
If the specified time limit has been reached,
.BR semtimedop ()
fails with
.I errno
set to
.B EAGAIN
(and none of the operations in
.I sops
is performed).
If the
.I timeout
argument is NULL,
then
.BR semtimedop ()
behaves exactly like
.BR semop ().
.P
Note that if
.BR semtimedop ()
is interrupted by a signal, causing the call to fail with the error
.BR EINTR ,
the contents of
.I timeout
are left unchanged.
.SH RETURN VALUE
On success,
.BR semop ()
and
.BR semtimedop ()
return 0.
On failure, they return \-1, and set
.I errno
to indicate the error.
.SH ERRORS
.TP
.B E2BIG
The argument
.I nsops
is greater than
.BR SEMOPM ,
the maximum number of operations allowed per system
call.
.TP
.B EACCES
The calling process does not have the permissions required
to perform the specified semaphore operations,
and does not have the
.B CAP_IPC_OWNER
capability in the user namespace that governs its IPC namespace.
.TP
.B EAGAIN
An operation could not proceed immediately and either
.B IPC_NOWAIT
was specified in
.I sem_flg
or the time limit specified in
.I timeout
expired.
.TP
.B EFAULT
An address specified in either the
.I sops
or the
.I timeout
argument isn't accessible.
.TP
.B EFBIG
For some operation the value of
.I sem_num
is less than 0 or greater than or equal to the number
of semaphores in the set.
.TP
.B EIDRM
The semaphore set was removed.
.TP
.B EINTR
While blocked in this system call, the thread caught a signal; see
.BR signal (7).
.TP
.B EINVAL
The semaphore set doesn't exist, or
.I semid
is less than zero, or
.I nsops
has a nonpositive value.
.TP
.B ENOMEM
The
.I sem_flg
of some operation specified
.B SEM_UNDO
and the system does not have enough memory to allocate the undo
structure.
.TP
.B ERANGE
For some operation
.I sem_op+semval
is greater than
.BR SEMVMX ,
the implementation dependent maximum value for
.IR semval .
.SH STANDARDS
POSIX.1-2008.
.SH VERSIONS
Linux 2.5.52 (backported into Linux 2.4.22),
glibc 2.3.3.
POSIX.1-2001, SVr4.
.\" SVr4 documents additional error conditions EINVAL, EFBIG, ENOSPC.
.SH NOTES
The
.I sem_undo
structures of a process aren't inherited by the child produced by
.BR fork (2),
but they are inherited across an
.BR execve (2)
system call.
.P
.BR semop ()
is never automatically restarted after being interrupted by a signal handler,
regardless of the setting of the
.B SA_RESTART
flag when establishing a signal handler.
.P
A semaphore adjustment
.RI ( semadj )
value is a per-process, per-semaphore integer that is the negated sum
of all operations performed on a semaphore specifying the
.B SEM_UNDO
flag.
Each process has a list of
.I semadj
values\[em]one value for each semaphore on which it has operated using
.BR SEM_UNDO .
When a process terminates, each of its per-semaphore
.I semadj
values is added to the corresponding semaphore,
thus undoing the effect of that process's operations on the semaphore
(but see BUGS below).
When a semaphore's value is directly set using the
.B SETVAL
or
.B SETALL
request to
.BR semctl (2),
the corresponding
.I semadj
values in all processes are cleared.
The
.BR clone (2)
.B CLONE_SYSVSEM
flag allows more than one process to share a
.I semadj
list; see
.BR clone (2)
for details.
.P
The \fIsemval\fP, \fIsempid\fP, \fIsemzcnt\fP, and \fIsemnct\fP values
for a semaphore can all be retrieved using appropriate
.BR semctl (2)
calls.
.SS Semaphore limits
The following limits on semaphore set resources affect the
.BR semop ()
call:
.TP
.B SEMOPM
Maximum number of operations allowed for one
.BR semop ()
call.
Before Linux 3.19,
.\" commit e843e7d2c88b7db107a86bd2c7145dc715c058f4
the default value for this limit was 32.
Since Linux 3.19, the default value is 500.
On Linux, this limit can be read and modified via the third field of
.IR /proc/sys/kernel/sem .
.\" This /proc file is not available in Linux 2.2 and earlier -- MTK
.IR Note :
this limit should not be raised above 1000,
.\" See comment in Linux 3.19 source file include/uapi/linux/sem.h
because of the risk of that
.BR semop ()
fails due to kernel memory fragmentation when allocating memory to copy the
.I sops
array.
.TP
.B SEMVMX
Maximum allowable value for
.IR semval :
implementation dependent (32767).
.P
The implementation has no intrinsic limits for
the adjust on exit maximum value
.RB ( SEMAEM ),
the system wide maximum number of undo structures
.RB ( SEMMNU )
and the per-process maximum number of undo entries system parameters.
.SH BUGS
When a process terminates, its set of associated
.I semadj
structures is used to undo the effect of all of the
semaphore operations it performed with the
.B SEM_UNDO
flag.
This raises a difficulty: if one (or more) of these semaphore adjustments
would result in an attempt to decrease a semaphore's value below zero,
what should an implementation do?
One possible approach would be to block until all the semaphore
adjustments could be performed.
This is however undesirable since it could force process termination to
block for arbitrarily long periods.
Another possibility is that such semaphore adjustments could be ignored
altogether (somewhat analogously to failing when
.B IPC_NOWAIT
is specified for a semaphore operation).
Linux adopts a third approach: decreasing the semaphore value
as far as possible (i.e., to zero) and allowing process
termination to proceed immediately.
.P
In Linux 2.6.x, x <= 10, there is a bug that in some circumstances
prevents a thread that is waiting for a semaphore value to become
zero from being woken up when the value does actually become zero.
This bug is fixed in Linux 2.6.11.
.\" The bug report:
.\" http://marc.theaimsgroup.com/?l=linux-kernel&m=110260821123863&w=2
.\" the fix:
.\" http://marc.theaimsgroup.com/?l=linux-kernel&m=110261701025794&w=2
.SH EXAMPLES
The following code segment uses
.BR semop ()
to atomically wait for the value of semaphore 0 to become zero,
and then increment the semaphore value by one.
.P
.in +4n
.EX
struct sembuf sops[2];
int semid;
\&
/* Code to set \fIsemid\fP omitted */
\&
sops[0].sem_num = 0;        /* Operate on semaphore 0 */
sops[0].sem_op = 0;         /* Wait for value to equal 0 */
sops[0].sem_flg = 0;
\&
sops[1].sem_num = 0;        /* Operate on semaphore 0 */
sops[1].sem_op = 1;         /* Increment value by one */
sops[1].sem_flg = 0;
\&
if (semop(semid, sops, 2) == \-1) {
    perror("semop");
    exit(EXIT_FAILURE);
}
.EE
.in
.P
A further example of the use of
.BR semop ()
can be found in
.BR shmop (2).
.SH SEE ALSO
.BR clone (2),
.BR semctl (2),
.BR semget (2),
.BR sigaction (2),
.BR capabilities (7),
.BR sem_overview (7),
.BR sysvipc (7),
.BR time (7)