How actually MPI_Waitall works - c++

Question

Does this example contradict the manual? The manual states that both the array of requests and the array of statuses must be of the same size. To be more precise, both arrays should be at least as long as it indicated by the count argument. Yet in the below example status array size is 2, not 4. Also, the example contradicts this statement from the manual
The error-free execution of MPI_Waitall(count, array_of_requests,
array_of_statuses) has the same effect as the execution of
MPI_Wait(&array_of_request[i], &array_of_statuses[i]), for
i=0,...,count-1, in some arbitrary order.
#include "mpi.h"
#include <stdio.h>
int main(argc,argv)
int argc;
char *argv[]; {
int numtasks, rank, next, prev, buf[2], tag1=1, tag2=2;
MPI_Request reqs[4];
MPI_Status stats[2];
MPI_Init(&argc,&argv);
MPI_Comm_size(MPI_COMM_WORLD, &numtasks);
MPI_Comm_rank(MPI_COMM_WORLD, &rank);
prev = rank-1;
next = rank+1;
if (rank == 0) prev = numtasks - 1;
if (rank == (numtasks - 1)) next = 0;
MPI_Irecv(&buf[0], 1, MPI_INT, prev, tag1, MPI_COMM_WORLD, &reqs[0]);
MPI_Irecv(&buf[1], 1, MPI_INT, next, tag2, MPI_COMM_WORLD, &reqs[1]);
MPI_Isend(&rank, 1, MPI_INT, prev, tag2, MPI_COMM_WORLD, &reqs[2]);
MPI_Isend(&rank, 1, MPI_INT, next, tag1, MPI_COMM_WORLD, &reqs[3]);
{ do some work }
MPI_Waitall(4, reqs, stats);
MPI_Finalize();
}
P.S. Definition of main looks strange. The return value is missing. Is it prehistoric C or typo?

Answer 1

Yes, this example contradicts the manual. If you compare the example with the Fortran version, you'll see that the Fortran version is correct in that the status array is large enough (strangely enough, it's a 2D array but thanks to implicit interfaces and storage association it can be seen as a 1D array of size MPI_STATUS_SIZE * 2 which is larger than 4 provided MPI_STATUS_SIZE is larger than 1 (on my system it's 5).
And yes, the missing return statement is an error; however some compilers resort to just emitting a warning for omitting the return statement in main(). Also, the prehistoricity of the code can be seen in the K&R style declaration of the arguments.