Parallel with communication of Matrix(SMatrix(ComplexF64)) using MPI

New to Julia
1

I am new to julia.
I have a question about MPI parallel computing.
I would like to communicate the matrices in each process as follows.
The code I have created is shown below.
I would like this code to send the matrix in the i-th process to the i-1st and i+1st processes, but so far it is not working.
Thank you in advance for your help.

using StaticArrays
using LinearAlgebra
using MPI
L=500
k_mesh=range(-π, length=L, step=2π / L)

function Hamiltonian(kx,ky)
    return @SMatrix[cos(kx) im*sin(ky);-im*sin(ky) cos(kx)]
end

function main()
    comm = MPI.COMM_WORLD
    myrank = MPI.Comm_rank(comm)
    size = MPI.Comm_size(comm)
    function start_and_end(N, comm)
        nprocs = MPI.Comm_size(comm)
        myrank = MPI.Comm_rank(comm)
        if N % nprocs != 0
            println("エラー！ N % プロセス数 は0である必要があります。")
            MPI.Abort(comm, 1)  # MPIを正しく終了させます
        end
        nbun = div(N, nprocs)
        ista = myrank * nbun + 1
        iend = ista + nbun - 1
        return ista, iend, nbun
    end
    dst = mod(myrank+1, size)
    src = mod(myrank-1, size)
    ista, iend , nbun = start_and_end(L,comm)
    k_mesh_i=deepcopy(k_mesh[ista:iend])
    Ham_i = Hamiltonian.(k_mesh,k_mesh_i')
    Ham_upper=deepcopy(Ham_i[1,:])
    Ham_downer=deepcopy(Ham_i[nbun,:])
    rreq = MPI.Irecv!(Ham_upper, comm; source=src, tag=src+32)
    sreq = MPI.Isend(Ham_downer, comm; dest=dst, tag=myrank+32)

    stats = MPI.Waitall([rreq, sreq])
    print("$rank: Sending   $rank -> $dst = $sreq\n")
    print("$rank: Received $src -> $rank = $rreq\n")
end

MPI.Init()
main()
MPI.Finalize()