Apply ffmt: remove blanks after #ifdef, fix Fypp continuation alignment

sbryngelson · sbryngelson · commit 01fd1b6ea720 · 2026-03-21T14:52:17.000-04:00
diff --git a/src/common/m_checker_common.fpp b/src/common/m_checker_common.fpp
@@ -32,7 +32,6 @@ contains
     end subroutine s_check_inputs_common
 
 #ifndef MFC_SIMULATION
-
     !> @brief Verifies that the total number of grid cells meets the minimum required by the number of dimensions and MPI ranks.
     impure subroutine s_check_total_cells
 
diff --git a/src/common/m_chemistry.fpp b/src/common/m_chemistry.fpp
@@ -195,11 +195,10 @@ contains
             if (chem_params%transport_model == 1) then
                 ! Note: Added 'i' and 'eqn' to private list.
                 $:GPU_PARALLEL_LOOP(collapse=3,  private='[x, y, z, i, eqn, Ys_L, Ys_R, Ys_cell, Xs_L, Xs_R, &
-                                    & mass_diffusivities_mixavg1, mass_diffusivities_mixavg2, mass_diffusivities_mixavg_Cell, h_l, &
-                                    & h_r, Xs_cell, h_k, dXk_dxi, Mass_Diffu_Flux, Mass_Diffu_Energy, MW_L, MW_R, MW_cell, Rgas_L, &
-                                    & Rgas_R, T_L, T_R, P_L, P_R, rho_L, rho_R, rho_cell, rho_Vic, lambda_L, lambda_R, &
-                                        & lambda_Cell, &
-                                    & dT_dxi, grid_spacing]', copyin='[offsets]')
+                                    & mass_diffusivities_mixavg1, mass_diffusivities_mixavg2, mass_diffusivities_mixavg_Cell, &
+                                    & h_l, h_r, Xs_cell, h_k, dXk_dxi, Mass_Diffu_Flux, Mass_Diffu_Energy, MW_L, MW_R, MW_cell, &
+                                    & Rgas_L, Rgas_R, T_L, T_R, P_L, P_R, rho_L, rho_R, rho_cell, rho_Vic, lambda_L, lambda_R, &
+                                    & lambda_Cell, dT_dxi, grid_spacing]', copyin='[offsets]')
                 do z = isc3%beg, isc3%end
                     do y = isc2%beg, isc2%end
                         do x = isc1%beg, isc1%end
diff --git a/src/common/m_model.fpp b/src/common/m_model.fpp
@@ -1004,7 +1004,6 @@ contains
     end subroutine s_distance_normals_2D
 
 #ifdef MFC_SIMULATION
-
     subroutine s_instantiate_STL_models()
 
         ! Variables for IBM+STL
diff --git a/src/common/m_mpi_common.fpp b/src/common/m_mpi_common.fpp
@@ -119,7 +119,6 @@ contains
         integer, dimension(1)                               :: airfoil_glb, airfoil_loc, airfoil_start
 
 #ifdef MFC_MPI
-
         ! Generic loop iterator
         integer :: i, j
         integer :: ierr !< Generic flag used to identify and report MPI errors
@@ -202,7 +201,6 @@ contains
         integer, dimension(3)                               :: sf_start_idx
 
 #ifdef MFC_MPI
-
         ! Generic loop iterator
         integer :: i, j, q, k, l, m_ds, n_ds, p_ds, ierr
 
@@ -989,7 +987,6 @@ contains
     subroutine s_mpi_decompose_computational_domain
 
 #ifdef MFC_MPI
-
         integer :: num_procs_x, num_procs_y, num_procs_z !< Optimal number of processors in the x-, y- and z-directions
         !> Non-optimal number of processors in the x-, y- and z-directions
         real(wp) :: tmp_num_procs_x, tmp_num_procs_y, tmp_num_procs_z
diff --git a/src/common/m_nvtx.f90 b/src/common/m_nvtx.f90
@@ -28,7 +28,6 @@ module m_nvtx
     end type nvtxEventAttributes
 
 #if defined(MFC_GPU) && defined(__PGI)
-
     interface nvtxRangePush
         ! push range with custom label and standard color
         subroutine nvtxRangePushA(name) bind(C, name='nvtxRangePushA')
diff --git a/src/common/m_phase_change.fpp b/src/common/m_phase_change.fpp
@@ -9,7 +9,6 @@
 module m_phase_change
 
 #ifndef MFC_POST_PROCESS
-
     use m_derived_types        !< Definitions of the derived types
     use m_global_parameters    !< Definitions of the global parameters
     use m_mpi_proxy            !< Message passing interface (MPI) module proxy
@@ -100,8 +99,8 @@ contains
 
         ! starting equilibrium solver
 
-        $:GPU_PARALLEL_LOOP(collapse=3, private='[i, j, k, l, p_infOV, p_infpT, p_infSL, sk, hk, gk, ek, rhok, pS, pSOV, pSSL, TS, &
-                            & TSOV, TSatOV, TSatSL, TSSL, rhoe, dynE, rhos, rho, rM, m1, m2, MCT, TvF]')
+        $:GPU_PARALLEL_LOOP(collapse=3, private='[i, j, k, l, p_infOV, p_infpT, p_infSL, sk, hk, gk, ek, rhok, pS, pSOV, pSSL, &
+                            & TS, TSOV, TSatOV, TSatSL, TSSL, rhoe, dynE, rhos, rho, rM, m1, m2, MCT, TvF]')
         do j = 0, m
             do k = 0, n
                 do l = 0, p
diff --git a/src/common/m_variables_conversion.fpp b/src/common/m_variables_conversion.fpp
@@ -384,7 +384,6 @@ contains
         $:GPU_UPDATE(device='[gammas, gs_min, pi_infs, ps_inf, cvs, qvs, qvps, Gs_vc]')
 
 #ifdef MFC_SIMULATION
-
         if (viscous) then
             @:ALLOCATE(Res_vc(1:2, 1:Re_size_max))
             do i = 1, 2
diff --git a/src/post_process/m_data_input.f90 b/src/post_process/m_data_input.f90
@@ -288,7 +288,6 @@ impure subroutine s_read_parallel_data_files(t_step)
         integer, intent(in) :: t_step
 
 #ifdef MFC_MPI
-
         real(wp), allocatable, dimension(:)  :: x_cb_glb, y_cb_glb, z_cb_glb
         integer                              :: ifile, ierr, data_size, filetype, stride
         integer, dimension(MPI_STATUS_SIZE)  :: status
diff --git a/src/post_process/m_global_parameters.fpp b/src/post_process/m_global_parameters.fpp
@@ -161,7 +161,6 @@ module m_global_parameters
     integer, allocatable, dimension(:) :: start_idx        !< Starting cell-center index of local processor in global grid
     integer                            :: num_ibs          !< Number of immersed boundaries
 #ifdef MFC_MPI
-
     type(mpi_io_var), public                      :: MPI_IO_DATA
     type(mpi_io_ib_var), public                   :: MPI_IO_IB_DATA
     type(mpi_io_levelset_var), public             :: MPI_IO_levelset_DATA
@@ -934,7 +933,6 @@ contains
         if (parallel_io .neqv. .true.) return
 
 #ifdef MFC_MPI
-
         ! Option for Lustre file system (Darter/Comet/Stampede)
         write (mpiiofs, '(A)') '/lustre_'
         mpiiofs = trim(mpiiofs)
@@ -975,7 +973,6 @@ contains
         deallocate (adv)
 
 #ifdef MFC_MPI
-
         if (parallel_io) then
             deallocate (start_idx)
             do i = 1, sys_size
diff --git a/src/post_process/m_mpi_proxy.fpp b/src/post_process/m_mpi_proxy.fpp
@@ -29,7 +29,6 @@ contains
     impure subroutine s_initialize_mpi_proxy_module
 
 #ifdef MFC_MPI
-
         integer :: i    !< Generic loop iterator
         integer :: ierr !< Generic flag used to identify and report MPI errors
         ! Allocating and configuring the receive counts and the displacement vector variables used in variable-gather communication
@@ -315,7 +314,6 @@ contains
     impure subroutine s_finalize_mpi_proxy_module
 
 #ifdef MFC_MPI
-
         ! Deallocating the receive counts and the displacement vector variables used in variable-gather communication procedures
         if ((format == 1 .and. n > 0) .or. n == 0) then
             deallocate (recvcounts)
diff --git a/src/pre_process/m_data_output.fpp b/src/pre_process/m_data_output.fpp
@@ -409,7 +409,6 @@ contains
         type(integer_field), dimension(1:num_dims, -1:1), intent(in) :: bc_type
 
 #ifdef MFC_MPI
-
         integer                              :: ifile, ierr, data_size
         integer, dimension(MPI_STATUS_SIZE)  :: status
         integer(KIND=MPI_OFFSET_KIND)        :: disp
diff --git a/src/pre_process/m_global_parameters.fpp b/src/pre_process/m_global_parameters.fpp
@@ -143,7 +143,6 @@ module m_global_parameters
     integer, allocatable, dimension(:)  :: proc_coords        !< Processor coordinates in MPI_CART_COMM
     integer, allocatable, dimension(:)  :: start_idx          !< Starting cell-center index of local processor in global grid
 #ifdef MFC_MPI
-
     type(mpi_io_var), public :: MPI_IO_DATA
     character(LEN=name_len)  :: mpiiofs
     integer                  :: mpi_info_int !< MPI info for parallel IO with Lustre file systems
@@ -846,7 +845,6 @@ contains
                                            & bubbles_lagrange, m, n, p, num_dims, igr, ib)
 
 #ifdef MFC_MPI
-
         if (qbmm .and. .not. polytropic) then
             allocate (MPI_IO_DATA%view(1:sys_size + 2*nb*nnode))
             allocate (MPI_IO_DATA%var(1:sys_size + 2*nb*nnode))
@@ -913,7 +911,6 @@ contains
         if (parallel_io .neqv. .true.) return
 
 #ifdef MFC_MPI
-
         ! Option for Lustre file system (Darter/Comet/Stampede)
         write (mpiiofs, '(A)') '/lustre_'
         mpiiofs = trim(mpiiofs)
@@ -947,7 +944,6 @@ contains
         deallocate (proc_coords)
 
 #ifdef MFC_MPI
-
         if (parallel_io) then
             deallocate (start_idx)
             do i = 1, sys_size
diff --git a/src/pre_process/m_grid.f90 b/src/pre_process/m_grid.f90
@@ -158,7 +158,6 @@ end subroutine s_generate_serial_grid
     impure subroutine s_generate_parallel_grid
 
 #ifdef MFC_MPI
-
         real(wp) :: length !< domain lengths
         ! Locations of cell boundaries
         real(wp), allocatable, dimension(:) :: x_cb_glb, y_cb_glb, z_cb_glb !< Locations of cell boundaries
diff --git a/src/pre_process/m_mpi_proxy.fpp b/src/pre_process/m_mpi_proxy.fpp
@@ -22,7 +22,6 @@ contains
     impure subroutine s_mpi_bcast_user_inputs
 
 #ifdef MFC_MPI
-
         ! Generic loop iterator
         integer :: i, j
         ! Generic flag used to identify and report MPI errors
diff --git a/src/pre_process/m_start_up.fpp b/src/pre_process/m_start_up.fpp
@@ -396,7 +396,6 @@ contains
     impure subroutine s_read_parallel_grid_data_files
 
 #ifdef MFC_MPI
-
         real(wp), allocatable, dimension(:)  :: x_cb_glb, y_cb_glb, z_cb_glb
         integer                              :: ifile, ierr, data_size
         integer, dimension(MPI_STATUS_SIZE)  :: status
@@ -496,7 +495,6 @@ contains
         type(scalar_field), dimension(sys_size), intent(inout) :: q_cons_vf_in
 
 #ifdef MFC_MPI
-
         integer                              :: ifile, ierr, data_size
         integer, dimension(MPI_STATUS_SIZE)  :: status
         integer(KIND=MPI_OFFSET_KIND)        :: disp
diff --git a/src/simulation/m_bubbles_EL.fpp b/src/simulation/m_bubbles_EL.fpp
@@ -198,8 +198,7 @@ contains
 
         $:GPU_UPDATE(device='[lag_id, bub_R0, Rmax_stats, Rmin_stats, gas_mg, gas_betaT, gas_betaC, bub_dphidt, gas_p, gas_mv, &
                      & intfc_rad, intfc_vel, mtn_pos, mtn_posPrev, mtn_vel, mtn_s, intfc_draddt, intfc_dveldt, gas_dpdt, &
-                         & gas_dmvdt, &
-                     & mtn_dposdt, mtn_dveldt, nBubs]')
+                     & gas_dmvdt, mtn_dposdt, mtn_dveldt, nBubs]')
 
         Rmax_glb = min(dflt_real, -dflt_real)
         Rmin_glb = max(dflt_real, -dflt_real)
diff --git a/src/simulation/m_cbc.fpp b/src/simulation/m_cbc.fpp
@@ -641,10 +641,9 @@ contains
                 ! FD2 or FD4 of RHS at j = 0
                 $:GPU_PARALLEL_LOOP(collapse=2, private='[r, k, alpha_rho, vel, adv_local, mf, dvel_ds, dadv_ds, Re_cbc, &
                                     & dalpha_rho_ds, dpres_ds, dvel_dt, dadv_dt, dalpha_rho_dt, L, lambda, Ys, dYs_dt, dYs_ds, &
-                                        & h_k, &
-                                    & Cp_i, Gamma_i, Xs, drho_dt, dpres_dt, dpi_inf_dt, dqv_dt, dgamma_dt, rho, pres, E, H, gamma, &
-                                    & pi_inf, qv, c, Ma, T, sum_Enthalpies, Cv, Cp, e_mix, Mw, R_gas, vel_K_sum, vel_dv_dt_sum, i, &
-                                    & j]', copyin='[dir_idx]')
+                                    & h_k, Cp_i, Gamma_i, Xs, drho_dt, dpres_dt, dpi_inf_dt, dqv_dt, dgamma_dt, rho, pres, E, H, &
+                                    & gamma, pi_inf, qv, c, Ma, T, sum_Enthalpies, Cv, Cp, e_mix, Mw, R_gas, vel_K_sum, &
+                                    & vel_dv_dt_sum, i, j]', copyin='[dir_idx]')
                 do r = is3%beg, is3%end
                     do k = is2%beg, is2%end
                         ! Transferring the Primitive Variables
diff --git a/src/simulation/m_data_output.fpp b/src/simulation/m_data_output.fpp
@@ -705,7 +705,6 @@ contains
         type(integer_field), dimension(1:num_dims, -1:1), intent(in) :: bc_type
 
 #ifdef MFC_MPI
-
         integer                              :: ifile, ierr, data_size
         integer, dimension(MPI_STATUS_SIZE)  :: status
         integer(kind=MPI_OFFSET_kind)        :: disp
@@ -948,7 +947,6 @@ contains
         integer, intent(in) :: time_step
 
 #ifdef MFC_MPI
-
         character(LEN=path_len + 2*name_len) :: file_loc
         integer(kind=MPI_OFFSET_kind)        :: disp
         integer(kind=MPI_OFFSET_kind)        :: m_MOK, n_MOK, p_MOK
diff --git a/src/simulation/m_fftw.fpp b/src/simulation/m_fftw.fpp
@@ -129,7 +129,6 @@ contains
 
         if (bc_y%beg >= 0) return
 #if defined(MFC_GPU)
-
         $:GPU_PARALLEL_LOOP(collapse=3)
         do k = 1, sys_size
             do j = 0, m
@@ -300,7 +299,6 @@ contains
 
         @:DEALLOCATE(data_real_gpu, data_fltr_cmplx_gpu, data_cmplx_gpu)
 #if defined(__PGI)
-
         ierr = cufftDestroy(fwd_plan_gpu)
         ierr = cufftDestroy(bwd_plan_gpu)
 #else
diff --git a/src/simulation/m_global_parameters.fpp b/src/simulation/m_global_parameters.fpp
@@ -1255,8 +1255,7 @@ contains
         $:GPU_UPDATE(device='[alt_soundspeed, acoustic_source, num_source]')
         $:GPU_UPDATE(device='[dt, sys_size, buff_size, pref, rhoref, gamma_idx, pi_inf_idx, E_idx, alf_idx, stress_idx, mpp_lim, &
                      & bubbles_euler, hypoelasticity, alt_soundspeed, avg_state, model_eqns, mixture_err, grid_geometry, &
-                         & cyl_coord, &
-                     & mp_weno, weno_eps, teno_CT, hyperelasticity, hyper_model, elasticity, xi_idx, B_idx, low_Mach]')
+                     & cyl_coord, mp_weno, weno_eps, teno_CT, hyperelasticity, hyper_model, elasticity, xi_idx, B_idx, low_Mach]')
 
         $:GPU_UPDATE(device='[Bx0]')
 
@@ -1328,7 +1327,6 @@ contains
         if (parallel_io .neqv. .true.) return
 
 #ifdef MFC_MPI
-
         ! Option for Lustre file system (Darter/Comet/Stampede)
         write (mpiiofs, '(A)') '/lustre_'
         mpiiofs = trim(mpiiofs)
diff --git a/src/simulation/m_ib_patches.fpp b/src/simulation/m_ib_patches.fpp
@@ -426,8 +426,8 @@ contains
         call get_bounding_indices(center(2) - ca_in, center(2) + ca_in, y_cc, jl, jr)
         call get_bounding_indices(center(3) - ca_in, center(3) + ca_in, z_cc, ll, lr)
 
-        $:GPU_PARALLEL_LOOP(private='[i, j, l, xyz_local, k, f]', copyin='[encoded_patch_id, center, inverse_rotation, offset, ma, &
-                            & ca_in, airfoil_grid_u, airfoil_grid_l, z_min, z_max]', collapse=3)
+        $:GPU_PARALLEL_LOOP(private='[i, j, l, xyz_local, k, f]', copyin='[encoded_patch_id, center, inverse_rotation, offset, &
+                            & ma, ca_in, airfoil_grid_u, airfoil_grid_l, z_min, z_max]', collapse=3)
         do l = ll, lr
             do j = jl, jr
                 do i = il, ir
@@ -525,8 +525,8 @@ contains
         ! Checking whether the rectangle covers a particular cell in the domain and verifying whether the current patch has the
         ! permission to write to that cell. If both queries check out, the primitive variables of the current patch are assigned to
         ! this cell.
-        $:GPU_PARALLEL_LOOP(private='[i, j, xy_local]', &
-                            & copyin='[encoded_patch_id, center, length, inverse_rotation, x_cc, y_cc]', collapse=2)
+        $:GPU_PARALLEL_LOOP(private='[i, j, xy_local]', copyin='[encoded_patch_id, center, length, inverse_rotation, x_cc, &
+                            & y_cc]', collapse=2)
         do j = jl, jr
             do i = il, ir
                 ! get the x and y coordinates in the local IB frame
@@ -793,8 +793,8 @@ contains
         call get_bounding_indices(center(2) - maxval(ellipse_coeffs)*2._wp, center(2) + maxval(ellipse_coeffs)*2._wp, y_cc, jl, jr)
 
         ! Checking whether the ellipse covers a particular cell in the domain
-        $:GPU_PARALLEL_LOOP(private='[i, j, xy_local]', copyin='[encoded_patch_id, center, ellipse_coeffs, inverse_rotation, x_cc, &
-                            & y_cc]', collapse=2)
+        $:GPU_PARALLEL_LOOP(private='[i, j, xy_local]', copyin='[encoded_patch_id, center, ellipse_coeffs, inverse_rotation, &
+                            & x_cc, y_cc]', collapse=2)
         do j = jl, jr
             do i = il, ir
                 ! get the x and y coordinates in the local IB frame
@@ -871,9 +871,8 @@ contains
         call get_bounding_indices(bbox_min(1), bbox_max(1), x_cc, il, ir)
         call get_bounding_indices(bbox_min(2), bbox_max(2), y_cc, jl, jr)
 
-        $:GPU_PARALLEL_LOOP(private='[i, j, xy_local, eta]', &
-                            & copyin='[patch_id, encoded_patch_id, center, inverse_rotation, offset, &
-                            & spc, threshold]', collapse=2)
+        $:GPU_PARALLEL_LOOP(private='[i, j, xy_local, eta]', copyin='[patch_id, encoded_patch_id, center, inverse_rotation, &
+                            & offset, spc, threshold]', collapse=2)
         do i = il, ir
             do j = jl, jr
                 xy_local = [x_cc(i) - center(1), y_cc(j) - center(2), 0._wp]
diff --git a/src/simulation/m_ibm.fpp b/src/simulation/m_ibm.fpp
@@ -393,7 +393,8 @@ contains
 
         bounds_error = .false.
 
-        $:GPU_PARALLEL_LOOP(private='[q, gp, i, j, k, physical_loc, patch_id, dist, norm, dim, bound, dir, index, temp_loc, s_cc]', copy='[bounds_error]')
+        $:GPU_PARALLEL_LOOP(private='[q, gp, i, j, k, physical_loc, patch_id, dist, norm, dim, bound, dir, index, temp_loc, &
+                            & s_cc]', copy='[bounds_error]')
         do q = 1, num_gps
             gp = ghost_points_in(q)
             i = gp%loc(1)
diff --git a/src/simulation/m_igr.fpp b/src/simulation/m_igr.fpp
@@ -419,8 +419,7 @@ contains
                 #:if not MFC_CASE_OPTIMIZATION or num_dims > 1
                     $:GPU_PARALLEL_LOOP(collapse=3, private='[j, k, l, rho_L, rho_R, gamma_L, gamma_R, pi_inf_L, pi_inf_R, mu_L, &
                                         & mu_R, vel_L, vel_R, pres_L, pres_R, alpha_L, alpha_R, alpha_rho_L, alpha_rho_R, F_L, &
-                                            & F_R, &
-                                        & E_L, E_R, cfl, dvel, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
+                                        & F_R, E_L, E_R, cfl, dvel, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
                     do l = 0, p
                         do k = 0, n
                             do j = -1, m
@@ -818,8 +817,7 @@ contains
                 #:if not MFC_CASE_OPTIMIZATION or num_dims > 2
                     $:GPU_PARALLEL_LOOP(collapse=3, private='[j, k, l, rho_L, rho_R, gamma_L, gamma_R, pi_inf_L, pi_inf_R, mu_L, &
                                         & mu_R, vel_L, vel_R, pres_L, pres_R, alpha_L, alpha_R, alpha_rho_L, alpha_rho_R, F_L, &
-                                            & F_R, &
-                                        & E_L, E_R, cfl, dvel, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
+                                        & F_R, E_L, E_R, cfl, dvel, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
                     do l = 0, p
                         do k = 0, n
                             do j = -1, m
@@ -1308,8 +1306,7 @@ contains
                 #:if not MFC_CASE_OPTIMIZATION or num_dims > 1
                     $:GPU_PARALLEL_LOOP(collapse=3, private='[j, k, l, rho_L, rho_R, gamma_L, gamma_R, pi_inf_L, pi_inf_R, mu_L, &
                                         & mu_R, vel_L, vel_R, pres_L, pres_R, alpha_L, alpha_R, alpha_rho_L, alpha_rho_R, F_L, &
-                                            & F_R, &
-                                        & E_L, E_R, cfl, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
+                                        & F_R, E_L, E_R, cfl, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
                     do l = 0, p
                         do k = -1, n
                             do j = 0, m
@@ -1688,8 +1685,7 @@ contains
                 #:if not MFC_CASE_OPTIMIZATION or num_dims > 2
                     $:GPU_PARALLEL_LOOP(collapse=3, private='[j, k, l, rho_L, rho_R, gamma_L, gamma_R, pi_inf_L, pi_inf_R, mu_L, &
                                         & mu_R, vel_L, vel_R, pres_L, pres_R, alpha_L, alpha_R, alpha_rho_L, alpha_rho_R, F_L, &
-                                            & F_R, &
-                                        & E_L, E_R, cfl, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
+                                        & F_R, E_L, E_R, cfl, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
                     do l = 0, p
                         do k = -1, n
                             do j = 0, m
@@ -2152,10 +2148,9 @@ contains
             end if
         else if (idir == 3) then
             #:if not MFC_CASE_OPTIMIZATION or num_dims > 2
-                $:GPU_PARALLEL_LOOP(collapse=3, private='[j, k, l, rho_L, rho_R, gamma_L, gamma_R, pi_inf_L, pi_inf_R, mu_L, mu_R, &
-                                    & vel_L, vel_R, pres_L, pres_R, alpha_L, alpha_R, alpha_rho_L, alpha_rho_R, F_L, F_R, E_L, &
-                                        & E_R, &
-                                    & cfl, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
+                $:GPU_PARALLEL_LOOP(collapse=3, private='[j, k, l, rho_L, rho_R, gamma_L, gamma_R, pi_inf_L, pi_inf_R, mu_L, &
+                                    & mu_R, vel_L, vel_R, pres_L, pres_R, alpha_L, alpha_R, alpha_rho_L, alpha_rho_R, F_L, F_R, &
+                                    & E_L, E_R, cfl, dvel_small, rho_sf_small, vflux_L_arr, vflux_R_arr]')
                 do l = -1, p
                     do k = 0, n
                         do j = 0, m
diff --git a/src/simulation/m_mpi_proxy.fpp b/src/simulation/m_mpi_proxy.fpp
diff --git a/src/simulation/m_qbmm.fpp b/src/simulation/m_qbmm.fpp
diff --git a/src/simulation/m_riemann_solvers.fpp b/src/simulation/m_riemann_solvers.fpp
diff --git a/src/simulation/m_start_up.fpp b/src/simulation/m_start_up.fpp
