Use lon long indices on large loops and vector access

  double *_yv, const int nyv, double *_zv, const int nzv, double *vec_vkzm, const int jlmf, const int jlml,

  const int nkv, const int nlmmt, double delks, double frsh

) {

  int nvtot = nxv * nyv * nzv;

  int nkvs = nkv * nkv;

  int nkvmo = nkv - 1; 

  int nkvvmo = nkvmo * nkv;

  int nvxy = nxv * nyv;

  long long nvtot = nxv * nyv * nzv;

  long long nkvs = nkv * nkv;

  long long nkvmo = nkv - 1; 

  long long nkvvmo = nkvmo * nkv;

  long long nvxy = nxv * nyv;

  const dcomplex uim = 0.0 + I * 1.0;

  dcomplex cc0 = 0.0 + I * 0.0;

 // Inizializza global_vec_w e vec_wsum sulla CPU in parallelo

#pragma omp parallel for collapse(2)

  for (int j80 = jlmf - 1; j80 < jlml; j80++) {

    for (int jxy50 = 0; jxy50 < nkvs; jxy50++) {

    for (long long jxy50 = 0; jxy50 < nkvs; jxy50++) {

      int j80_index = j80 - jlmf + 1;

      dcomplex *vec_w = global_vec_w + nkvs * j80_index;

      int wk_index = nlmmt * jxy50;

      long long wk_index = nlmmt * jxy50;

      dcomplex wk_value = vec_tt1_wk[wk_index + j80_index];

      int jy50 = jxy50 / nkv;

      int jx50 = jxy50 % nkv;

      long long jy50 = jxy50 / nkv;

      long long jx50 = jxy50 % nkv;

      vec_w[(nkv * jx50) + jy50] = wk_value;

    } // jxy50 loop

  }

  map(to: _xv[0:nxv], _yv[0:nyv], _zv[0:nzv]) \

  map(to: vkv[0:nkv], vec_vkzm[0:nkvs])

  {

    // Esegue il calcolo principale in un unico kernel GPU

    // con parallelizzazione sui cicli più grandi (nvtot e nkvs)

    // Run the main calculation in a single GPU kernel

    // with parallelization on the largest loops (nvtot & nkvs)

#pragma omp target teams distribute parallel for collapse(2)

    for (int ixyz = 0; ixyz < nvtot; ixyz++) {

      for (int jy60x55 = 0; jy60x55 < nkvs ; jy60x55++) {

    for (long long ixyz = 0; ixyz < nvtot; ixyz++) {

      for (long long jy60x55 = 0; jy60x55 < nkvs ; jy60x55++) {

        int iz75 = ixyz / nvxy;

        int iy70 = (ixyz % nvxy) / nxv;

        int ix65 = ixyz % nxv;

        long long iz75 = ixyz / nvxy;

        long long iy70 = (ixyz % nvxy) / nxv;

        long long ix65 = ixyz % nxv;

        double z = _zv[iz75] + frsh;

        double y = _yv[iy70];

        double x = _xv[ix65];

        int jy60 = jy60x55 / nkv;

        int jx55 = jy60x55 % nkv;

        int w_index = (jx55 * nkv) + jy60;

        long long jy60 = jy60x55 / nkv;

        long long jx55 = jy60x55 % nkv;

        long long w_index = (jx55 * nkv) + jy60;

        double vky = vkv[jy60];

        double factor = (jy60 == 0 || jy60 == nkvmo) ? 0.5 : 1.0;

        double vkx, vkz;

          term = phas * factor;

        }

        // L'ultima loop è ora seriale, garantendo la correttezza dei risultati

	// The last loop can now be serialized, granting for correctness

	// of the results

        for (int j80 = jlmf - 1; j80 < jlml; j80++) {

          int j80_index = j80 - jlmf + 1;

          dcomplex *vec_w = global_vec_w + nkvs * j80_index;

          long wsum_index = (j80_index * nvtot) + ixyz;

          long long wsum_index = (j80_index * nvtot) + ixyz;

          vec_wsum[wsum_index] += delks * vec_w[w_index] * term;

        }

      } // jy60x55 loop