Finishing touches

added ci added ci things2 things added single precision to test suite and generated golden files added working precision to case_dicts, testing, and documentation fixed code not passing using double working precision
MFlowCode · May 16, 2023 · 843a50b · 843a50b
1 parent ebadc94
commit 843a50b
Show file tree

Hide file tree

Showing 32 changed files with 191 additions and 184 deletions.
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -86,17 +86,29 @@ jobs:
           ${{ matrix.intel-command }}
           /bin/bash mfc.sh build -j $(nproc) ${{ matrix.debug }}
 
-      - name: Test Suite (Debug)
+      - name: Test Suite (Debug, Double Precision)
         if: matrix.debug == '--debug'
         run: |
           ${{ matrix.intel-command }}
           /bin/bash mfc.sh test  -j $(nproc) --debug
 
-      - name: Test Suite (No Debug)
+      - name: Test Suite (Debug, Single Precision)
+        if: matrix.debug == '--debug'
+        run: |
+          ${{ matrix.intel-command }}
+          /bin/bash mfc.sh test  -j $(nproc) --debug --single
+
+      - name: Test Suite (No Debug, Double Precision)
         if: matrix.debug == '--no-debug'
         run: |
           ${{ matrix.intel-command }}
           /bin/bash mfc.sh test -j $(nproc) -a
+
+      - name: Test Suite (No Debug, Single Precision)
+        if: matrix.debug == '--no-debug'
+        run: |
+          ${{ matrix.intel-command }}
+          /bin/bash mfc.sh test -j $(nproc) -a --single  
           
 
   self-cpu-release:

diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -13,6 +13,7 @@ option(MFC_SIMULATION    "Build simulation"    OFF)
 option(MFC_POST_PROCESS  "Build post_process"  OFF)
 option(MFC_DOCUMENTATION "Build documentation" OFF)
 option(MFC_ALL           "Build everything"    OFF)
+option(MFC_SINGLE_PRECISION "Build double precision" OFF)
 
 if (MFC_BUILD_ALL)
     set(MFC_PRE_PROCESS   ON FORCE)
@@ -156,6 +157,12 @@ if (CMAKE_BUILD_TYPE STREQUAL "Debug")
     add_compile_definitions(MFC_DEBUG)
 endif()
 
+if (MFC_SINGLE_PRECISION)
+    add_compile_definitions(MFC_SINGLE_PRECISION)
+else()
+    add_compile_definitions(MFC_DOUBLE_PRECISION)
+endif()
+
 ## === HANDLE_SOURCES
 # Gather F90 source files for a given target, including common/, and preprocessing .fpp -> .f90.
 # Outputs:

diff --git a/docs/documentation/case.md b/docs/documentation/case.md
@@ -76,8 +76,11 @@ Definition of the parameters is described in the following subsections.
 | Parameter        | Type           | Description                      |
 | ---:             |    :----:      |          :---                    |
 | `run_time_info`  | Logical        | Output run-time information      |
+| `working_precision` | Integer     | Simulation working precision     |
 
 - `run_time_info` generates a text file that includes run-time information including the CFL number(s) at each time-step.
+- `working_precision` sets the working precision to either (1) - single precision, or
+(2) - double precision.
 
 ### 2. Computational Domain
 

diff --git a/misc/run-phoenix-release-cpu.sh b/misc/run-phoenix-release-cpu.sh
@@ -11,4 +11,5 @@
 cd $SLURM_SUBMIT_DIR                            # Change to working directory
 echo $(pwd)
 . ./mfc.sh load -c p -m g
-./mfc.sh test -j 12 -b mpirun -a
+./mfc.sh test -j 12 -b mpirun -a --no-single
+./mfc.sh test -j 12 -b mpirun -a --single
diff --git a/misc/run-phoenix-release-gpu.sh b/misc/run-phoenix-release-gpu.sh
@@ -11,4 +11,5 @@
 cd $SLURM_SUBMIT_DIR                            # Change to working directory
 echo $(pwd)
 . ./mfc.sh load -c p -m g
-./mfc.sh test -j 1 -b mpirun -a --gpu
+./mfc.sh test -j 1 -b mpirun -a --gpu --no-single
+./mfc.sh test -j 1 -b mpirun -a --gpu --single
diff --git a/src/common/include/inline_conversions.fpp b/src/common/include/inline_conversions.fpp
@@ -32,11 +32,11 @@
 
             if (mpp_lim .and. (num_fluids > 1)) then
                 c = (1._wp/gamma + 1._wp)* &
-                      (pres + pi_inf)/rho
+                      (pres + pi_inf/(gamma+1._wp))/rho
             else
                 c = &
                     (1._wp/gamma + 1._wp)* &    
-                    (pres + pi_inf)/ &
+                    (pres + pi_inf/(gamma + 1._wp))/ &
                     (rho*(1._wp - adv(num_fluids)))
             end if
         else 

diff --git a/src/common/m_constants.fpp b/src/common/m_constants.fpp
@@ -4,7 +4,20 @@
 
 module m_constants
 
-    use m_precision_select
+    use mpi
+
+    integer, parameter :: sp = selected_real_kind(6, 37)
+    integer, parameter :: dp = selected_real_kind(15, 307)
+
+#ifdef MFC_DOUBLE_PRECISION
+    integer, parameter :: wp = dp
+    integer, parameter :: mpi_p = MPI_DOUBLE_PRECISION
+#endif
+
+#ifdef MFC_SINGLE_PRECISION
+    integer, parameter :: wp = sp
+    integer, parameter :: mpi_p = MPI_REAL
+#endif
 
     character, parameter :: dflt_char = ' ' !< Default string value
 

diff --git a/src/common/m_derived_types.f90 b/src/common/m_derived_types.f90
@@ -7,7 +7,7 @@
 module m_derived_types
 
     use m_constants !< Constants
-    use m_precision_select
+
 
     implicit none
 

diff --git a/src/common/m_eigen_solver.f90 b/src/common/m_eigen_solver.f90
@@ -8,7 +8,7 @@
 !!              modifications for compatibility.
 module m_eigen_solver
 
-    use m_precision_select
+    use m_constants
 
     implicit none
 

diff --git a/src/common/m_helper.f90 b/src/common/m_helper.f90
@@ -92,11 +92,11 @@ end subroutine s_compute_finite_difference_coefficients ! --------------
         !! @param ntmp is the output number bubble density
     subroutine s_comp_n_from_prim(vftmp, Rtmp, ntmp, weights)
         !$acc routine seq
-        real(kind(0._wp)), intent(IN) :: vftmp
-        real(kind(0._wp)), dimension(nb), intent(IN) :: Rtmp
-        real(kind(0._wp)), intent(OUT) :: ntmp
-        real(kind(0._wp)) :: R3
-        real(kind(0._wp)), dimension(nb) :: weights
+        real(wp), intent(IN) :: vftmp
+        real(wp), dimension(nb), intent(IN) :: Rtmp
+        real(wp), intent(OUT) :: ntmp
+        real(wp) :: R3
+        real(wp), dimension(nb) :: weights
 
         R3 = dot_product(weights, Rtmp**3._wp)
         ntmp = (3._wp/(4._wp*pi))*vftmp/R3
@@ -105,11 +105,11 @@ end subroutine s_comp_n_from_prim
 
     subroutine s_comp_n_from_cons(vftmp, nRtmp, ntmp, weights)
         !$acc routine seq
-        real(kind(0._wp)), intent(IN) :: vftmp
-        real(kind(0._wp)), dimension(nb), intent(IN) :: nRtmp
-        real(kind(0._wp)), intent(OUT) :: ntmp  
-        real(kind(0._wp)) :: nR3
-        real(kind(0._wp)), dimension(nb) :: weights
+        real(wp), intent(IN) :: vftmp
+        real(wp), dimension(nb), intent(IN) :: nRtmp
+        real(wp), intent(OUT) :: ntmp  
+        real(wp) :: nR3
+        real(wp), dimension(nb) :: weights
 
         nR3 = dot_product(weights, nRtmp**3._wp)
         ntmp = sqrt((4._wp*pi/3._wp)*nR3/vftmp)

diff --git a/src/common/m_precision_select.f90 b/src/common/m_precision_select.f90
diff --git a/src/common/m_variables_conversion.fpp b/src/common/m_variables_conversion.fpp
@@ -158,7 +158,7 @@ contains
 
             pres = ( &
                    energy - &
-                   0.5_wp*(mom**2._wp)/rho - &
+                   (5._wp * (10._wp ** (-1)))*(mom**2._wp)/rho - &
                    pi_inf - E_e &
                    )/gamma
         end if
@@ -235,7 +235,6 @@ contains
 
         real(wp), optional, dimension(2), intent(OUT) :: Re_K
         real(wp), dimension(num_fluids) :: alpha_rho_K, alpha_K
-        real(wp), optional, dimension(2), intent(OUT) :: Re_K
 
         real(wp), optional, intent(OUT) :: G_K
         real(wp), optional, dimension(num_fluids), intent(IN) :: G
@@ -510,7 +509,7 @@ contains
         real(wp), dimension(num_fluids), intent(IN) :: alpha_rho_K, alpha_K !<
             !! Partial densities and volume fractions
 
-        real(kind(0d0)), dimension(2), intent(OUT) :: Re_K
+        real(wp), dimension(2), intent(OUT) :: Re_K
 
         integer, intent(IN) :: k, l, r
         integer :: i, j !< Generic loop iterators
@@ -731,7 +730,7 @@ contains
 
         integer :: i, j, k, l !< Generic loop iterators
 
-        real(kind(0._wp)) :: ntmp
+        real(wp) :: ntmp
 
         if (bubbles) then
             allocate(nRtmp(nb))

diff --git a/src/post_process/m_checker.f90 b/src/post_process/m_checker.f90
@@ -65,9 +65,6 @@ subroutine s_check_inputs()
             ! Constraints on model equations and number of fluids in the flow
         elseif (all(model_eqns /= (/1, 2, 3, 4/))) then
             call s_mpi_abort('Unsupported value of model_eqns. Exiting ...')
-        elseif (wp == single_precision .and. precision == 2) then
-            call s_mpi_abort('Unsupported combination of working precision'// &
-                            'and silo precision')
         elseif (num_fluids /= dflt_int &
                 .and. &
                 (num_fluids < 1 .or. num_fluids > num_fluids)) then

diff --git a/src/post_process/m_data_output.fpp b/src/post_process/m_data_output.fpp
@@ -140,7 +140,7 @@ contains
 
         if (n == 0) then
             allocate (q_root_sf(0:m_root, 0:0, 0:0))
-            if (precision == 1 .and. wp == double_precision) then
+            if (precision == 1) then
                 allocate (q_root_sf_s(0:m_root, 0:0, 0:0))
             end if
         end if
@@ -769,10 +769,10 @@ contains
             ! and write it to the formatted database master file.
             if (n == 0) then
 
-                if (precision == 1 .and. wp == double_precision) then
+                if (precision == 1 .and. wp ==  dp) then
                     x_cc_s = real(x_cc, sp)
                     q_sf_s = real(q_sf, sp)
-                elseif (precision == 1 .and. wp == single_precision) then
+                elseif (precision == 1 .and. wp == sp) then
                     x_cc_s = x_cc
                     q_sf_s = q_sf
                 end if
@@ -866,7 +866,7 @@ contains
                 ! Finally, each of the local processor(s) proceeds to write
                 ! the flow variable data that it is responsible for to the
                 ! formatted database slave file.
-                if (wp == double_precision) then
+                if (wp == dp) then
                     if (precision == 1) then
                         do i = -offset_x%beg, m + offset_x%end
                             do j = -offset_y%beg, n + offset_y%end
@@ -895,7 +895,7 @@ contains
                             end do
                         end if
                     end if 
-                elseif (wp == single_precision) then
+                elseif (wp == dp) then
                     do i = -offset_x%beg, m + offset_x%end
                         do j = -offset_y%beg, n + offset_y%end
                             do k = -offset_z%beg, p + offset_z%end

diff --git a/src/post_process/m_global_parameters.f90 b/src/post_process/m_global_parameters.f90
@@ -629,28 +629,28 @@ end subroutine s_initialize_global_parameters_module ! --------------------
     subroutine s_initialize_nonpoly
 
         integer :: ir
-        real(kind(0._wp)) :: rhol0
-        real(kind(0._wp)) :: pl0
-        real(kind(0._wp)) :: uu
-        real(kind(0._wp)) :: D_m
-        real(kind(0._wp)) :: temp
-        real(kind(0._wp)) :: omega_ref
-        real(kind(0._wp)), dimension(Nb) :: chi_vw0
-        real(kind(0._wp)), dimension(Nb) :: cp_m0
-        real(kind(0._wp)), dimension(Nb) :: k_m0
-        real(kind(0._wp)), dimension(Nb) :: rho_m0
-        real(kind(0._wp)), dimension(Nb) :: x_vw
+        real(wp) :: rhol0
+        real(wp) :: pl0
+        real(wp) :: uu
+        real(wp) :: D_m
+        real(wp) :: temp
+        real(wp) :: omega_ref
+        real(wp), dimension(Nb) :: chi_vw0
+        real(wp), dimension(Nb) :: cp_m0
+        real(wp), dimension(Nb) :: k_m0
+        real(wp), dimension(Nb) :: rho_m0
+        real(wp), dimension(Nb) :: x_vw
 
         ! liquid physical properties
-        real(kind(0._wp)) :: mul0, ss, pv, gamma_v, M_v, mu_v
+        real(wp) :: mul0, ss, pv, gamma_v, M_v, mu_v
 
         ! gas physical properties
-        real(kind(0._wp)) :: gamma_m, gamma_n, M_n, mu_n
+        real(wp) :: gamma_m, gamma_n, M_n, mu_n
 
         ! polytropic index used to compute isothermal natural frequency
-        real(kind(0._wp)), parameter :: k_poly = 1._wp
+        real(wp), parameter :: k_poly = 1._wp
         ! universal gas constant
-        real(kind(0._wp)), parameter :: Ru = 8314._wp
+        real(wp), parameter :: Ru = 8314._wp
 
         rhol0 = rhoref
         pl0 = pref
@@ -749,13 +749,13 @@ end subroutine s_initialize_nonpoly
     !> Subroutine to compute the transfer coefficient for non-polytropic gas modeling
     subroutine s_transcoeff(omega, peclet, Re_trans, Im_trans)
 
-        real(kind(0._wp)), intent(IN) :: omega
-        real(kind(0._wp)), intent(IN) :: peclet
-        real(kind(0._wp)), intent(OUT) :: Re_trans
-        real(kind(0._wp)), intent(OUT) :: Im_trans
+        real(wp), intent(IN) :: omega
+        real(wp), intent(IN) :: peclet
+        real(wp), intent(OUT) :: Re_trans
+        real(wp), intent(OUT) :: Im_trans
         complex :: trans, c1, c2, c3
         complex :: imag = (0., 1.)
-        real(kind(0._wp)) :: f_transcoeff
+        real(wp) :: f_transcoeff
 
         c1 = imag*omega*peclet
         c2 = CSQRT(c1)
@@ -844,12 +844,12 @@ subroutine s_simpson(Npt)
 
         integer, intent(IN) :: Npt
         integer :: ir
-        real(kind(0._wp)) :: R0mn
-        real(kind(0._wp)) :: R0mx
-        real(kind(0._wp)) :: dphi
-        real(kind(0._wp)) :: tmp
-        real(kind(0._wp)) :: sd
-        real(kind(0._wp)), dimension(Npt) :: phi
+        real(wp) :: R0mn
+        real(wp) :: R0mx
+        real(wp) :: dphi
+        real(wp) :: tmp
+        real(wp) :: sd
+        real(wp), dimension(Npt) :: phi
 
         ! nondiml. min. & max. initial radii for numerical quadrature
         !sd   = 0.05D0

diff --git a/src/pre_process/m_assign_variables.f90 b/src/pre_process/m_assign_variables.f90
@@ -238,15 +238,12 @@ subroutine s_assign_patch_species_primitive_variables_bubbles(patch_id, j, k, l,
         real(wp) :: orig_pi_inf
         real(wp) :: muR, muV
 
-        real(wp), dimension(sys_size) :: orig_prim_vf !<
+        real(wp), dimension(sys_size) :: orig_prim_vf !< Vector to hold original values of cell for smoothing purposes
         real(wp), dimension(int(E_idx - mom_idx%beg)) :: vel    !< velocity
         real(wp) :: pres   !< pressure
         real(wp) :: x_centroid, y_centroid
         real(wp) :: epsilon, beta
 
-        real(kind(0d0)), dimension(sys_size) :: orig_prim_vf !<
-            !! Vector to hold original values of cell for smoothing purposes
-
         integer :: i  !< Generic loop iterator
         integer :: smooth_patch_id
 
@@ -648,15 +645,12 @@ subroutine s_assign_patch_species_primitive_variables(patch_id, j, k, l, &
             !! function, respectively, obtained from the combination of primitive
             !! variables of the current and smoothing patches
 
-        real(wp), dimension(sys_size) :: orig_prim_vf !<
+        real(wp), dimension(sys_size) :: orig_prim_vf !< Vector to hold original values of cell for smoothing purposes
         real(wp), dimension(int(E_idx - mom_idx%beg)) :: vel    !< velocity
         real(wp) :: pres   !< pressure
         real(wp) :: x_centroid, y_centroid
         real(wp) :: epsilon, beta
 
-        real(kind(0d0)), dimension(sys_size) :: orig_prim_vf !<
-        ! Vector to hold original values of cell for smoothing purposes
-
         integer :: smooth_patch_id
 
         integer :: i !< generic loop iterator