Hi,
if you run this under a debugger and look at how MPI_Scatterv is
invoked, you will find that
- sendcounts = {1, 1, 1}
- resizedtype has size 32
- recvcount*sizeof(MPI_INTEGER) = 32 on task 0, but 16 on task 1 and 2
=> too much data is sent to tasks 1 and 2, hence the error.
in this case
Dear Jeff Hammond
Thanks a lot for the reply. I have tried with mpiexec, I am getting the
same error. But according to this link:
http://stackoverflow.com/questions/7549316/mpi-partition-matrix-into-blocks
it is possible. Any suggestions/ advice?
_
*SAVE WATER ** ~ **SAVE ENERGY**~ **~ **SAVE
Based upon the symbols in the backtrace, you are using Intel MPI, not
Open-MPI. If there is a bug in the MPI library, it is likely also in
MPICH, so you might try to reproduce this in MPICH. You can also try to
run with Open-MPI. If you see a problem in both Intel MPI/MPICH and
Open-MPI, it is a
I want to scatter matrix from root to other processors using scatterv. I am
creating a communicator topology using *mpi_cart_create*. As an example I
have the below code in fortran:
PROGRAM SendRecv
USE mpi
IMPLICIT none
integer, PARAMETER :: m = 4, n = 4
integer, DIMENSION(m,n) :: a, b,h
integer
