MPI_Barrier not working inside a loop

问题

I have running some tests on the MPI functions to understand how it works and have got a weird result with the MPI_Barrier: it does what everyone would expect if I use it in code like

int main(int argc, char *argv[])
{
  <some code>
  MPI_Barrier(MPI_COMM_WORLD);
  <more code>
  MPI_Barrier(MPI_COMM_WORLD);
  <...>
}

but when I call it from inside a loop i get random results. To be specific, I have the following test code:

#include "mpi.h" 
#include <stdio.h>

int main(int argc, char *argv[]) 
{ 
  int i, rb, rank, nprocs;

  MPI_Init(&argc,&argv); 
  MPI_Comm_size(MPI_COMM_WORLD,&nprocs); 
  MPI_Comm_rank(MPI_COMM_WORLD,&rank);

  i=0;
  while(i<5)
  {
    rb=MPI_Barrier(MPI_COMM_WORLD);
    printf("Itartion %d.  I am %d of %d. MPIBarrierRes: %d\n", i, rank, nprocs, rb);
    i++;
  } 
  MPI_Finalize(); 
  return 0; 
}

When i run it with 3 tasks i randomly get:

Itartion 0.  I am 0 of 3. MPIBarrierRes: 0
Itartion 0.  I am 2 of 3. MPIBarrierRes: 0
Itartion 0.  I am 1 of 3. MPIBarrierRes: 0
Itartion 1.  I am 0 of 3. MPIBarrierRes: 0
Itartion 1.  I am 1 of 3. MPIBarrierRes: 0
Itartion 1.  I am 2 of 3. MPIBarrierRes: 0
Itartion 2.  I am 0 of 3. MPIBarrierRes: 0
Itartion 2.  I am 1 of 3. MPIBarrierRes: 0
Itartion 2.  I am 2 of 3. MPIBarrierRes: 0
Itartion 3.  I am 0 of 3. MPIBarrierRes: 0
Itartion 3.  I am 1 of 3. MPIBarrierRes: 0
Itartion 3.  I am 2 of 3. MPIBarrierRes: 0
Itartion 4.  I am 0 of 3. MPIBarrierRes: 0
Itartion 4.  I am 1 of 3. MPIBarrierRes: 0
Itartion 4.  I am 2 of 3. MPIBarrierRes: 0

which is waht i shall expect, or just the oposite:

Itartion 0.  I am 2 of 3. MPIBarrierRes: 0
Itartion 1.  I am 2 of 3. MPIBarrierRes: 0
Itartion 2.  I am 2 of 3. MPIBarrierRes: 0
Itartion 3.  I am 2 of 3. MPIBarrierRes: 0
Itartion 4.  I am 2 of 3. MPIBarrierRes: 0
Itartion 0.  I am 0 of 3. MPIBarrierRes: 0
Itartion 1.  I am 0 of 3. MPIBarrierRes: 0
Itartion 2.  I am 0 of 3. MPIBarrierRes: 0
Itartion 3.  I am 0 of 3. MPIBarrierRes: 0
Itartion 4.  I am 0 of 3. MPIBarrierRes: 0
Itartion 0.  I am 1 of 3. MPIBarrierRes: 0
Itartion 1.  I am 1 of 3. MPIBarrierRes: 0
Itartion 2.  I am 1 of 3. MPIBarrierRes: 0
Itartion 3.  I am 1 of 3. MPIBarrierRes: 0
Itartion 4.  I am 1 of 3. MPIBarrierRes: 0

or something in between, like

Itartion 0.  I am 1 of 3. MPIBarrierRes: 0
Itartion 0.  I am 0 of 3. MPIBarrierRes: 0
Itartion 1.  I am 0 of 3. MPIBarrierRes: 0
Itartion 0.  I am 2 of 3. MPIBarrierRes: 0
Itartion 1.  I am 1 of 3. MPIBarrierRes: 0
Itartion 2.  I am 0 of 3. MPIBarrierRes: 0
Itartion 1.  I am 2 of 3. MPIBarrierRes: 0
Itartion 2.  I am 1 of 3. MPIBarrierRes: 0
Itartion 3.  I am 0 of 3. MPIBarrierRes: 0
Itartion 2.  I am 2 of 3. MPIBarrierRes: 0
Itartion 3.  I am 1 of 3. MPIBarrierRes: 0
Itartion 4.  I am 0 of 3. MPIBarrierRes: 0
Itartion 3.  I am 2 of 3. MPIBarrierRes: 0
Itartion 4.  I am 1 of 3. MPIBarrierRes: 0
Itartion 4.  I am 2 of 3. MPIBarrierRes: 0

Can anyone tell me if there is some conflict between MPI_Barrier and loops? (I have only found warnings to avoid deadlocks using loops of different sizes in different tasks.) If there is one, what can i do to force tasks to wait for each other before starting a new iteration of the loop? If there isn't, what is wrong with this code?

Thanks!

回答1:

It has nothing to do with your loops. There's no sequentialization of IO implied by MPI. If you need prints to come out in order, you'll have to explicitly send them to one rank, and print them all there.

来源：https://stackoverflow.com/questions/6840625/mpi-barrier-not-working-inside-a-loop

标签

loops

mpi

barrier