55 строки
1.5 KiB
Plaintext
55 строки
1.5 KiB
Plaintext
|
.\" -*- nroff -*-
|
||
|
|
.\" Copyright (c) 2015 University of Houston. All rights reserved.
|
||
|
|
.\" Copyright (c) 2015 Mellanox Technologies, Inc.
|
||
|
|
.\" $COPYRIGHT$
|
||
|
|
.de Vb
|
||
|
|
.ft CW
|
||
|
|
.nf
|
||
|
|
..
|
||
|
|
.de Ve
|
||
|
|
.ft R
|
||
|
|
|
||
|
|
.fi
|
||
|
|
..
|
||
|
|
.TH "SHMEM\\_FENCE" "3" "#OMPI_DATE#" "#PACKAGE_VERSION#" "#PACKAGE_NAME#"
|
||
|
|
.SH NAME
|
||
|
|
|
||
|
|
shmem_fence \- Provides a separate ordering on the sequence of puts issued by this PE to each destination
|
||
|
|
PE.
|
||
|
|
.SH SYNOPSIS
|
||
|
|
|
||
|
|
C or C++:
|
||
|
|
.Vb
|
||
|
|
#include <mpp/shmem.h>
|
||
|
|
|
||
|
|
void shmem_fence(void);
|
||
|
|
.Ve
|
||
|
|
Fortran:
|
||
|
|
.Vb
|
||
|
|
INCLUDE "mpp/shmem.fh"
|
||
|
|
|
||
|
|
CALL SHMEM_FENCE
|
||
|
|
.Ve
|
||
|
|
.SH DESCRIPTION
|
||
|
|
|
||
|
|
The \fBshmem_fence()\fP
|
||
|
|
routine provides an ordering on the put operations issued by the calling
|
||
|
|
PE prior to the call to \fBshmem_fence()\fP
|
||
|
|
relative to the put operations issued by the
|
||
|
|
calling PE following the call to \fBshmem_fence()\fP\&.
|
||
|
|
It guarantees that all such prior put operations
|
||
|
|
issued to a particular destination PE are fully written to the symmetric memory of
|
||
|
|
that destination PE, before any such following put operations to that same destination PE
|
||
|
|
are written to the symmetric memory of that destination PE.
|
||
|
|
Note that the ordering is provided separately on the sequences of puts from the calling PE to
|
||
|
|
each distinct destination PE. The \fBshmem_quiet()\fP
|
||
|
|
routine should be used instead if ordering
|
||
|
|
of puts is required when multiple destination PEs are involved.
|
||
|
|
.SH NOTES
|
||
|
|
|
||
|
|
The shmem_quiet function should be called if ordering of puts is desired when multiple remote
|
||
|
|
PEs are involved.
|
||
|
|
.SH SEE ALSO
|
||
|
|
|
||
|
|
\fIintro_shmem\fP(3)
|