shmem_barrier_all

Registers the arrival of a PE at a barrier and blocks the PE until all other
PEs arrive at the barrier and all local updates and remote memory updates on
the default context are completed.

Definitions

C/C++ Synopsis

void shmem_barrier_all(void);

Arguments

None

Description

The shmem_barrier_all routine is a mechanism for synchronizing all PEs in
the world team at once.This routineblocks the PE until all PEs have called
shmem_barrier_all. In a multithreaded OpenSHMEM program, only the calling
thread is blocked.

Prior to synchronizing with other PEs, shmem_barrier_all ensures completion
of all previously issued memory stores and remote memory updates issued on
the default context via OpenSHMEM AMOs and RMA routine calls such
as shmem_int_add, shmem_put32, shmem_put_nbi, and shmem_get_nbi.

Return Values

None.

Notes

The shmem_barrier_all routine is equivalent to calling shmem_ctx_quiet on
the default context followed by calling shmem_team_sync on the world team.

The shmem_barrier_all routine can be used to portably ensure that memory
access operations observe remote updates in the order enforced by
initiator PEs.

Calls to shmem_ctx_quiet can be performed prior to calling the barrier
routine to ensure completion of operations issued on additional contexts.

Examples

C/C++ Example

The following shmem_barrier_all example is for C11 programs:

#include <stdio.h>
#include <shmem.h>

int main(void)
{
   static int x = 1010;

   shmem_init();
   int me = shmem_my_pe();
   int npes = shmem_n_pes();

   /* put to next  PE in a circular fashion */
   shmem_p(&x, 4, (me + 1) % npes);

   /* synchronize all PEs */
   shmem_barrier_all();
   printf("%d: x = %d\n", me, x);
   shmem_finalize();
   return 0;
}