taskset setting updated

Add _fh variants of fderivs, fdderivs, kodis, and lopsided that accept
a caller-provided fh work array instead of allocating one internally.
Declare two shared work arrays in compute_rhs_bssn (fh_work2 for
symmetry_bd(2) callers, fh_work3 for symmetry_bd(3) callers) and pass
them to all ~84 subroutine calls, eliminating ~77 redundant automatic
array allocations (~591 MB churn per RHS call, ~2.3 GB per timestep).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

AMSS_NCKU_Input.py

@@ -16,7 +16,7 @@ import numpy
 File_directory   = "GW150914"                    ## output file directory
 Output_directory = "binary_output"               ## binary data file directory
                                                  ## The file directory name should not be too long
-MPI_processes    = 1                              ## number of processes (MPI removed, single-process mode)
+MPI_processes    = 64                             ## number of mpi processes used in the simulation
 
 GPU_Calculation  = "no"                          ## Use GPU or not 
                                                  ## (prefer "no" in the current version, because the GPU part may have bugs when integrated in this Python interface)

AMSS_NCKU_source/ABE.C

@@ -20,11 +20,7 @@ using namespace std;
 #include <map.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "misc.h"
 #include "macrodef.h"

AMSS_NCKU_source/Ansorg.h

@@ -19,11 +19,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #define PI M_PI
 

AMSS_NCKU_source/Block.h

@@ -2,11 +2,7 @@
 #ifndef BLOCK_H
 #define BLOCK_H
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 #include "macrodef.h" //need dim here; Vertex or Cell
 #include "var.h"
 #include "MyList.h"

AMSS_NCKU_source/IntPnts0.C

@@ -4,11 +4,7 @@
 #include <stdarg.h>
 #include <string.h>
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "myglobal.h"
 

AMSS_NCKU_source/MPatch.C

@@ -341,9 +341,8 @@ void Patch::Interp_Points(MyList<var> *VarList,
                           double *Shellf, int Symmetry)
 {
   // NOTE: we do not Synchnize variables here, make sure of that before calling this routine
-  int myrank, nprocs;
+  int myrank;
   MPI_Comm_rank(MPI_COMM_WORLD, &myrank);
-  MPI_Comm_size(MPI_COMM_WORLD, &nprocs);
 
   int ordn = 2 * ghost_width;
   MyList<var> *varl;
@@ -355,18 +354,24 @@ void Patch::Interp_Points(MyList<var> *VarList,
     varl = varl->next;
   }
 
-  memset(Shellf, 0, sizeof(double) * NN * num_var);
+  double *shellf;
+  shellf = new double[NN * num_var];
+  memset(shellf, 0, sizeof(double) * NN * num_var);
 
-  // owner_rank[j] records which MPI rank owns point j
-  // All ranks traverse the same block list so they all agree on ownership
-  int *owner_rank;
-  owner_rank = new int[NN];
-  for (int j = 0; j < NN; j++)
-    owner_rank[j] = -1;
+  // we use weight to monitor code, later some day we can move it for optimization
+  int *weight;
+  weight = new int[NN];
+  memset(weight, 0, sizeof(int) * NN);
+
+  double *DH, *llb, *uub;
+  DH = new double[dim];
 
-  double DH[dim], llb[dim], uub[dim];
   for (int i = 0; i < dim; i++)
+  {
     DH[i] = getdX(i);
+  }
+  llb = new double[dim];
+  uub = new double[dim];
 
   for (int j = 0; j < NN; j++) // run along points
   {
@@ -398,6 +403,12 @@ void Patch::Interp_Points(MyList<var> *VarList,
       bool flag = true;
       for (int i = 0; i < dim; i++)
       {
+// NOTE: our dividing structure is (exclude ghost)
+// -1 0
+//       1  2
+// so (0,1) does not belong to any part for vertex structure
+// here we put (0,0.5) to left part and (0.5,1) to right part
+// BUT for cell structure the bbox is (-1.5,0.5) and (0.5,2.5), there is no missing region at all
 #ifdef Vertex
 #ifdef Cell
 #error Both Cell and Vertex are defined
@@ -422,7 +433,6 @@ void Patch::Interp_Points(MyList<var> *VarList,
       if (flag)
       {
         notfind = false;
-        owner_rank[j] = BP->rank;
         if (myrank == BP->rank)
         {
           //---> interpolation
@@ -430,11 +440,14 @@ void Patch::Interp_Points(MyList<var> *VarList,
           int k = 0;
           while (varl) // run along variables
           {
-            f_global_interp(BP->shape, BP->X[0], BP->X[1], BP->X[2], BP->fgfs[varl->data->sgfn], Shellf[j * num_var + k],
+            //              shellf[j*num_var+k] = Parallel::global_interp(dim,BP->shape,BP->X,BP->fgfs[varl->data->sgfn],
+            //	  		                                    pox,ordn,varl->data->SoA,Symmetry);
+            f_global_interp(BP->shape, BP->X[0], BP->X[1], BP->X[2], BP->fgfs[varl->data->sgfn], shellf[j * num_var + k],
                             pox[0], pox[1], pox[2], ordn, varl->data->SoA, Symmetry);
             varl = varl->next;
             k++;
           }
+          weight[j] = 1;
         }
       }
       if (Bp == ble)
@@ -443,327 +456,103 @@ void Patch::Interp_Points(MyList<var> *VarList,
     }
   }
 
-  // Replace MPI_Allreduce with per-owner MPI_Bcast:
-  // Group consecutive points by owner rank and broadcast each group.
-  // Since each point's data is non-zero only on the owner rank,
-  // Bcast from owner is equivalent to Allreduce(MPI_SUM) but much cheaper.
+  MPI_Allreduce(shellf, Shellf, NN * num_var, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  int *Weight;
+  Weight = new int[NN];
+  MPI_Allreduce(weight, Weight, NN, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+
+  //  misc::tillherecheck("print me");
+
+  for (int i = 0; i < NN; i++)
   {
-    int j = 0;
-    while (j < NN)
+    if (Weight[i] > 1)
     {
-      int cur_owner = owner_rank[j];
-      if (cur_owner < 0)
-      {
-        if (myrank == 0)
-        {
-          cout << "ERROR: Patch::Interp_Points fails to find point (";
-          for (int d = 0; d < dim; d++)
-          {
-            cout << XX[d][j];
-            if (d < dim - 1)
-              cout << ",";
-            else
-              cout << ")";
-          }
-          cout << " on Patch (";
-          for (int d = 0; d < dim; d++)
-          {
-            cout << bbox[d] << "+" << lli[d] * DH[d];
-            if (d < dim - 1)
-              cout << ",";
-            else
-              cout << ")--";
-          }
-          cout << "(";
-          for (int d = 0; d < dim; d++)
-          {
-            cout << bbox[dim + d] << "-" << uui[d] * DH[d];
-            if (d < dim - 1)
-              cout << ",";
-            else
-              cout << ")" << endl;
-          }
-          MPI_Abort(MPI_COMM_WORLD, 1);
-        }
-        j++;
-        continue;
-      }
-      // Find contiguous run of points with the same owner
-      int jstart = j;
-      while (j < NN && owner_rank[j] == cur_owner)
-        j++;
-      int count = (j - jstart) * num_var;
-      MPI_Bcast(Shellf + jstart * num_var, count, MPI_DOUBLE, cur_owner, MPI_COMM_WORLD);
+      if (myrank == 0)
+        cout << "WARNING: Patch::Interp_Points meets multiple weight" << endl;
+      for (int j = 0; j < num_var; j++)
+        Shellf[j + i * num_var] = Shellf[j + i * num_var] / Weight[i];
     }
-  }
-
-  delete[] owner_rank;
-}
-void Patch::Interp_Points(MyList<var> *VarList,
-                          int NN, double **XX,
-                          double *Shellf, int Symmetry,
-                          int Nmin_consumer, int Nmax_consumer)
-{
-  // Targeted point-to-point overload: each owner sends each point only to
-  // the one rank that needs it for integration (consumer), reducing
-  // communication volume by ~nprocs times compared to the Bcast version.
-  int myrank, nprocs;
-  MPI_Comm_rank(MPI_COMM_WORLD, &myrank);
-  MPI_Comm_size(MPI_COMM_WORLD, &nprocs);
-
-  int ordn = 2 * ghost_width;
-  MyList<var> *varl;
-  int num_var = 0;
-  varl = VarList;
-  while (varl)
-  {
-    num_var++;
-    varl = varl->next;
-  }
-
-  memset(Shellf, 0, sizeof(double) * NN * num_var);
-
-  // owner_rank[j] records which MPI rank owns point j
-  int *owner_rank;
-  owner_rank = new int[NN];
-  for (int j = 0; j < NN; j++)
-    owner_rank[j] = -1;
-
-  double DH[dim], llb[dim], uub[dim];
-  for (int i = 0; i < dim; i++)
-    DH[i] = getdX(i);
-
-  // --- Interpolation phase (identical to original) ---
-  for (int j = 0; j < NN; j++)
-  {
-    double pox[dim];
-    for (int i = 0; i < dim; i++)
-    {
-      pox[i] = XX[i][j];
-      if (myrank == 0 && (XX[i][j] < bbox[i] + lli[i] * DH[i] || XX[i][j] > bbox[dim + i] - uui[i] * DH[i]))
-      {
-        cout << "Patch::Interp_Points: point (";
-        for (int k = 0; k < dim; k++)
-        {
-          cout << XX[k][j];
-          if (k < dim - 1)
-            cout << ",";
-          else
-            cout << ") is out of current Patch." << endl;
-        }
-        MPI_Abort(MPI_COMM_WORLD, 1);
-      }
-    }
-
-    MyList<Block> *Bp = blb;
-    bool notfind = true;
-    while (notfind && Bp)
-    {
-      Block *BP = Bp->data;
-
-      bool flag = true;
-      for (int i = 0; i < dim; i++)
-      {
-#ifdef Vertex
-#ifdef Cell
-#error Both Cell and Vertex are defined
-#endif
-        llb[i] = (feq(BP->bbox[i], bbox[i], DH[i] / 2)) ? BP->bbox[i] + lli[i] * DH[i] : BP->bbox[i] + (ghost_width - 0.5) * DH[i];
-        uub[i] = (feq(BP->bbox[dim + i], bbox[dim + i], DH[i] / 2)) ? BP->bbox[dim + i] - uui[i] * DH[i] : BP->bbox[dim + i] - (ghost_width - 0.5) * DH[i];
-#else
-#ifdef Cell
-        llb[i] = (feq(BP->bbox[i], bbox[i], DH[i] / 2)) ? BP->bbox[i] + lli[i] * DH[i] : BP->bbox[i] + ghost_width * DH[i];
-        uub[i] = (feq(BP->bbox[dim + i], bbox[dim + i], DH[i] / 2)) ? BP->bbox[dim + i] - uui[i] * DH[i] : BP->bbox[dim + i] - ghost_width * DH[i];
-#else
-#error Not define Vertex nor Cell
-#endif
-#endif
-        if (XX[i][j] - llb[i] < -DH[i] / 2 || XX[i][j] - uub[i] > DH[i] / 2)
-        {
-          flag = false;
-          break;
-        }
-      }
-
-      if (flag)
-      {
-        notfind = false;
-        owner_rank[j] = BP->rank;
-        if (myrank == BP->rank)
-        {
-          varl = VarList;
-          int k = 0;
-          while (varl)
-          {
-            f_global_interp(BP->shape, BP->X[0], BP->X[1], BP->X[2], BP->fgfs[varl->data->sgfn], Shellf[j * num_var + k],
-                            pox[0], pox[1], pox[2], ordn, varl->data->SoA, Symmetry);
-            varl = varl->next;
-            k++;
-          }
-        }
-      }
-      if (Bp == ble)
-        break;
-      Bp = Bp->next;
-    }
-  }
-
-  // --- Error check for unfound points ---
-  for (int j = 0; j < NN; j++)
-  {
-    if (owner_rank[j] < 0 && myrank == 0)
+    else if (Weight[i] == 0 && myrank == 0)
     {
       cout << "ERROR: Patch::Interp_Points fails to find point (";
-      for (int d = 0; d < dim; d++)
+      for (int j = 0; j < dim; j++)
       {
-        cout << XX[d][j];
-        if (d < dim - 1)
+        cout << XX[j][i];
+        if (j < dim - 1)
           cout << ",";
         else
           cout << ")";
       }
       cout << " on Patch (";
-      for (int d = 0; d < dim; d++)
+      for (int j = 0; j < dim; j++)
       {
-        cout << bbox[d] << "+" << lli[d] * DH[d];
-        if (d < dim - 1)
+        cout << bbox[j] << "+" << lli[j] * getdX(j);
+        if (j < dim - 1)
           cout << ",";
         else
           cout << ")--";
       }
       cout << "(";
-      for (int d = 0; d < dim; d++)
+      for (int j = 0; j < dim; j++)
       {
-        cout << bbox[dim + d] << "-" << uui[d] * DH[d];
-        if (d < dim - 1)
+        cout << bbox[dim + j] << "-" << uui[j] * getdX(j);
+        if (j < dim - 1)
           cout << ",";
         else
           cout << ")" << endl;
       }
+#if 0
+       checkBlock();
+#else
+      cout << "splited domains:" << endl;
+      {
+        MyList<Block> *Bp = blb;
+        while (Bp)
+        {
+          Block *BP = Bp->data;
+
+          for (int i = 0; i < dim; i++)
+          {
+#ifdef Vertex
+#ifdef Cell
+#error Both Cell and Vertex are defined
+#endif
+            llb[i] = (feq(BP->bbox[i], bbox[i], DH[i] / 2)) ? BP->bbox[i] + lli[i] * DH[i] : BP->bbox[i] + (ghost_width - 0.5) * DH[i];
+            uub[i] = (feq(BP->bbox[dim + i], bbox[dim + i], DH[i] / 2)) ? BP->bbox[dim + i] - uui[i] * DH[i] : BP->bbox[dim + i] - (ghost_width - 0.5) * DH[i];
+#else
+#ifdef Cell
+            llb[i] = (feq(BP->bbox[i], bbox[i], DH[i] / 2)) ? BP->bbox[i] + lli[i] * DH[i] : BP->bbox[i] + ghost_width * DH[i];
+            uub[i] = (feq(BP->bbox[dim + i], bbox[dim + i], DH[i] / 2)) ? BP->bbox[dim + i] - uui[i] * DH[i] : BP->bbox[dim + i] - ghost_width * DH[i];
+#else
+#error Not define Vertex nor Cell
+#endif
+#endif
+          }
+          cout << "(";
+          for (int j = 0; j < dim; j++)
+          {
+            cout << llb[j] << ":" << uub[j];
+            if (j < dim - 1)
+              cout << ",";
+            else
+              cout << ")" << endl;
+          }
+          if (Bp == ble)
+            break;
+          Bp = Bp->next;
+        }
+      }
+#endif
       MPI_Abort(MPI_COMM_WORLD, 1);
     }
   }
 
-  // --- Targeted point-to-point communication phase ---
-  // Compute consumer_rank[j] using the same deterministic formula as surface_integral
-  int *consumer_rank = new int[NN];
-  {
-    int mp = NN / nprocs;
-    int Lp = NN - nprocs * mp;
-    for (int j = 0; j < NN; j++)
-    {
-      if (j < Lp * (mp + 1))
-        consumer_rank[j] = j / (mp + 1);
-      else
-        consumer_rank[j] = Lp + (j - Lp * (mp + 1)) / mp;
-    }
-  }
-
-  // Count sends and recvs per rank
-  int *send_count = new int[nprocs];
-  int *recv_count = new int[nprocs];
-  memset(send_count, 0, sizeof(int) * nprocs);
-  memset(recv_count, 0, sizeof(int) * nprocs);
-
-  for (int j = 0; j < NN; j++)
-  {
-    int own = owner_rank[j];
-    int con = consumer_rank[j];
-    if (own == con)
-      continue; // local — no communication needed
-    if (own == myrank)
-      send_count[con]++;
-    if (con == myrank)
-      recv_count[own]++;
-  }
-
-  // Build send buffers: for each destination rank, pack (index, data) pairs
-  // Each entry: 1 int (point index j) + num_var doubles
-  int total_send = 0, total_recv = 0;
-  int *send_offset = new int[nprocs];
-  int *recv_offset = new int[nprocs];
-  for (int r = 0; r < nprocs; r++)
-  {
-    send_offset[r] = total_send;
-    total_send += send_count[r];
-    recv_offset[r] = total_recv;
-    total_recv += recv_count[r];
-  }
-
-  // Pack send buffers: each message contains (j, data[0..num_var-1]) per point
-  int stride = 1 + num_var; // 1 double for index + num_var doubles for data
-  double *sendbuf = new double[total_send * stride];
-  double *recvbuf = new double[total_recv * stride];
-
-  // Temporary counters for packing
-  int *pack_pos = new int[nprocs];
-  memset(pack_pos, 0, sizeof(int) * nprocs);
-
-  for (int j = 0; j < NN; j++)
-  {
-    int own = owner_rank[j];
-    int con = consumer_rank[j];
-    if (own != myrank || con == myrank)
-      continue;
-    int pos = (send_offset[con] + pack_pos[con]) * stride;
-    sendbuf[pos] = (double)j; // point index
-    for (int v = 0; v < num_var; v++)
-      sendbuf[pos + 1 + v] = Shellf[j * num_var + v];
-    pack_pos[con]++;
-  }
-
-  // Post non-blocking recvs and sends
-  int n_req = 0;
-  for (int r = 0; r < nprocs; r++)
-  {
-    if (recv_count[r] > 0) n_req++;
-    if (send_count[r] > 0) n_req++;
-  }
-
-  MPI_Request *reqs = new MPI_Request[n_req];
-  int req_idx = 0;
-
-  for (int r = 0; r < nprocs; r++)
-  {
-    if (recv_count[r] > 0)
-    {
-      MPI_Irecv(recvbuf + recv_offset[r] * stride,
-                recv_count[r] * stride, MPI_DOUBLE,
-                r, 0, MPI_COMM_WORLD, &reqs[req_idx++]);
-    }
-  }
-  for (int r = 0; r < nprocs; r++)
-  {
-    if (send_count[r] > 0)
-    {
-      MPI_Isend(sendbuf + send_offset[r] * stride,
-                send_count[r] * stride, MPI_DOUBLE,
-                r, 0, MPI_COMM_WORLD, &reqs[req_idx++]);
-    }
-  }
-
-  if (n_req > 0)
-    MPI_Waitall(n_req, reqs, MPI_STATUSES_IGNORE);
-
-  // Unpack recv buffers into Shellf
-  for (int i = 0; i < total_recv; i++)
-  {
-    int pos = i * stride;
-    int j = (int)recvbuf[pos];
-    for (int v = 0; v < num_var; v++)
-      Shellf[j * num_var + v] = recvbuf[pos + 1 + v];
-  }
-
-  delete[] reqs;
-  delete[] sendbuf;
-  delete[] recvbuf;
-  delete[] pack_pos;
-  delete[] send_offset;
-  delete[] recv_offset;
-  delete[] send_count;
-  delete[] recv_count;
-  delete[] consumer_rank;
-  delete[] owner_rank;
+  delete[] shellf;
+  delete[] weight;
+  delete[] Weight;
+  delete[] DH;
+  delete[] llb;
+  delete[] uub;
 }
 void Patch::Interp_Points(MyList<var> *VarList,
                           int NN, double **XX,
@@ -784,22 +573,24 @@ void Patch::Interp_Points(MyList<var> *VarList,
     varl = varl->next;
   }
 
-  memset(Shellf, 0, sizeof(double) * NN * num_var);
+  double *shellf;
+  shellf = new double[NN * num_var];
+  memset(shellf, 0, sizeof(double) * NN * num_var);
 
-  // owner_rank[j] stores the global rank that owns point j
-  int *owner_rank;
-  owner_rank = new int[NN];
-  for (int j = 0; j < NN; j++)
-    owner_rank[j] = -1;
+  // we use weight to monitor code, later some day we can move it for optimization
+  int *weight;
+  weight = new int[NN];
+  memset(weight, 0, sizeof(int) * NN);
 
-  // Build global-to-local rank translation for Comm_here
-  MPI_Group world_group, local_group;
-  MPI_Comm_group(MPI_COMM_WORLD, &world_group);
-  MPI_Comm_group(Comm_here, &local_group);
+  double *DH, *llb, *uub;
+  DH = new double[dim];
 
-  double DH[dim], llb[dim], uub[dim];
   for (int i = 0; i < dim; i++)
+  {
     DH[i] = getdX(i);
+  }
+  llb = new double[dim];
+  uub = new double[dim];
 
   for (int j = 0; j < NN; j++) // run along points
   {
@@ -831,6 +622,12 @@ void Patch::Interp_Points(MyList<var> *VarList,
       bool flag = true;
       for (int i = 0; i < dim; i++)
       {
+// NOTE: our dividing structure is (exclude ghost)
+// -1 0
+//       1  2
+// so (0,1) does not belong to any part for vertex structure
+// here we put (0,0.5) to left part and (0.5,1) to right part
+// BUT for cell structure the bbox is (-1.5,0.5) and (0.5,2.5), there is no missing region at all
 #ifdef Vertex
 #ifdef Cell
 #error Both Cell and Vertex are defined
@@ -855,7 +652,6 @@ void Patch::Interp_Points(MyList<var> *VarList,
       if (flag)
       {
         notfind = false;
-        owner_rank[j] = BP->rank;
         if (myrank == BP->rank)
         {
           //---> interpolation
@@ -863,11 +659,14 @@ void Patch::Interp_Points(MyList<var> *VarList,
           int k = 0;
           while (varl) // run along variables
           {
-            f_global_interp(BP->shape, BP->X[0], BP->X[1], BP->X[2], BP->fgfs[varl->data->sgfn], Shellf[j * num_var + k],
+            //              shellf[j*num_var+k] = Parallel::global_interp(dim,BP->shape,BP->X,BP->fgfs[varl->data->sgfn],
+            //	  		                                    pox,ordn,varl->data->SoA,Symmetry);
+            f_global_interp(BP->shape, BP->X[0], BP->X[1], BP->X[2], BP->fgfs[varl->data->sgfn], shellf[j * num_var + k],
                             pox[0], pox[1], pox[2], ordn, varl->data->SoA, Symmetry);
             varl = varl->next;
             k++;
           }
+          weight[j] = 1;
         }
       }
       if (Bp == ble)
@@ -876,35 +675,97 @@ void Patch::Interp_Points(MyList<var> *VarList,
     }
   }
 
-  // Collect unique global owner ranks and translate to local ranks in Comm_here
-  // Then broadcast each owner's points via MPI_Bcast on Comm_here
-  {
-    int j = 0;
-    while (j < NN)
-    {
-      int cur_owner_global = owner_rank[j];
-      if (cur_owner_global < 0)
-      {
-        // Point not found — skip (error check disabled for sub-communicator levels)
-        j++;
-        continue;
-      }
-      // Translate global rank to local rank in Comm_here
-      int cur_owner_local;
-      MPI_Group_translate_ranks(world_group, 1, &cur_owner_global, local_group, &cur_owner_local);
+  MPI_Allreduce(shellf, Shellf, NN * num_var, MPI_DOUBLE, MPI_SUM, Comm_here);
+  int *Weight;
+  Weight = new int[NN];
+  MPI_Allreduce(weight, Weight, NN, MPI_INT, MPI_SUM, Comm_here);
 
-      // Find contiguous run of points with the same owner
-      int jstart = j;
-      while (j < NN && owner_rank[j] == cur_owner_global)
-        j++;
-      int count = (j - jstart) * num_var;
-      MPI_Bcast(Shellf + jstart * num_var, count, MPI_DOUBLE, cur_owner_local, Comm_here);
+  //  misc::tillherecheck("print me");
+  //  if(lmyrank == 0) cout<<"myrank = "<<myrank<<"print me"<<endl;
+
+  for (int i = 0; i < NN; i++)
+  {
+    if (Weight[i] > 1)
+    {
+      if (lmyrank == 0)
+        cout << "WARNING: Patch::Interp_Points meets multiple weight" << endl;
+      for (int j = 0; j < num_var; j++)
+        Shellf[j + i * num_var] = Shellf[j + i * num_var] / Weight[i];
     }
+#if 0 // for not involved levels, this may fail     
+     else if(Weight[i] == 0 && lmyrank == 0)
+     {
+       cout<<"ERROR: Patch::Interp_Points fails to find point (";
+       for(int j=0;j<dim;j++)
+       {
+	  cout<<XX[j][i];
+	  if(j<dim-1) cout<<",";
+	  else        cout<<")";
+       }
+       cout<<" on Patch (";
+       for(int j=0;j<dim;j++)
+       {
+	  cout<<bbox[j]<<"+"<<lli[j]*getdX(j);
+	  if(j<dim-1) cout<<",";
+	  else        cout<<")--";
+       }
+       cout<<"(";
+       for(int j=0;j<dim;j++)
+       {
+	  cout<<bbox[dim+j]<<"-"<<uui[j]*getdX(j);
+	  if(j<dim-1) cout<<",";
+	  else        cout<<")"<<endl;
+       }
+#if 0
+       checkBlock();
+#else
+  cout<<"splited domains:"<<endl;
+  {
+     MyList<Block> *Bp=blb;
+     while(Bp)
+     {
+	Block *BP=Bp->data;
+
+	for(int i=0;i<dim;i++)
+	{
+#ifdef Vertex
+#ifdef Cell
+#error Both Cell and Vertex are defined
+#endif    
+          llb[i] = (feq(BP->bbox[i]    ,bbox[i]    ,DH[i]/2)) ? BP->bbox[i]+lli[i]*DH[i]     : BP->bbox[i]    +(ghost_width-0.5)*DH[i];
+          uub[i] = (feq(BP->bbox[dim+i],bbox[dim+i],DH[i]/2)) ? BP->bbox[dim+i]-uui[i]*DH[i] : BP->bbox[dim+i]-(ghost_width-0.5)*DH[i];
+#else
+#ifdef Cell
+          llb[i] = (feq(BP->bbox[i]    ,bbox[i]    ,DH[i]/2)) ? BP->bbox[i]+lli[i]*DH[i]     : BP->bbox[i]    +ghost_width*DH[i];
+          uub[i] = (feq(BP->bbox[dim+i],bbox[dim+i],DH[i]/2)) ? BP->bbox[dim+i]-uui[i]*DH[i] : BP->bbox[dim+i]-ghost_width*DH[i];
+#else
+#error Not define Vertex nor Cell
+#endif
+#endif 
+	}       
+       cout<<"(";
+       for(int j=0;j<dim;j++)
+       {
+	  cout<<llb[j]<<":"<<uub[j];
+	  if(j<dim-1) cout<<",";
+	  else        cout<<")"<<endl;
+       }
+	if(Bp == ble) break;
+	Bp=Bp->next;
+     }
+  }
+#endif       
+       MPI_Abort(MPI_COMM_WORLD,1);
+     }
+#endif
   }
 
-  MPI_Group_free(&world_group);
-  MPI_Group_free(&local_group);
-  delete[] owner_rank;
+  delete[] shellf;
+  delete[] weight;
+  delete[] Weight;
+  delete[] DH;
+  delete[] llb;
+  delete[] uub;
 }
 void Patch::checkBlock()
 {

AMSS_NCKU_source/MPatch.h

@@ -2,11 +2,7 @@
 #ifndef PATCH_H
 #define PATCH_H
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 #include "MyList.h"
 #include "Block.h"
 #include "var.h"
@@ -43,10 +39,6 @@ public:
 
    bool Find_Point(double *XX);
 
-   void Interp_Points(MyList<var> *VarList,
-                      int NN, double **XX,
-                      double *Shellf, int Symmetry,
-                      int Nmin_consumer, int Nmax_consumer);
    void Interp_Points(MyList<var> *VarList,
                       int NN, double **XX,
                       double *Shellf, int Symmetry, MPI_Comm Comm_here);

AMSS_NCKU_source/Newton.C

@@ -8,11 +8,7 @@
 #include <limits.h>
 #include <float.h>
 #include <math.h>
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "util_Table.h"
 #include "cctk.h"

AMSS_NCKU_source/NullShellPatch.h

@@ -23,11 +23,7 @@ using namespace std;
 #include <complex.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 #include "MyList.h"
 #include "Block.h"
 #include "Parallel.h"

AMSS_NCKU_source/NullShellPatch2.h

@@ -23,11 +23,7 @@ using namespace std;
 #include <complex.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 #include "MyList.h"
 #include "Block.h"
 #include "Parallel.h"

AMSS_NCKU_source/Parallel.C

@@ -3756,484 +3756,6 @@ void Parallel::Sync(MyList<Patch> *PatL, MyList<var> *VarList, int Symmetry)
   delete[] transfer_src;
   delete[] transfer_dst;
 }
-// Merged Sync: collect all intra-patch and inter-patch grid segment lists,
-// then issue a single transfer() call instead of N+1 separate ones.
-void Parallel::Sync_merged(MyList<Patch> *PatL, MyList<var> *VarList, int Symmetry)
-{
-  int cpusize;
-  MPI_Comm_size(MPI_COMM_WORLD, &cpusize);
-
-  MyList<Parallel::gridseg> **combined_src = new MyList<Parallel::gridseg> *[cpusize];
-  MyList<Parallel::gridseg> **combined_dst = new MyList<Parallel::gridseg> *[cpusize];
-  for (int node = 0; node < cpusize; node++)
-    combined_src[node] = combined_dst[node] = 0;
-
-  // Phase A: Intra-patch ghost exchange segments
-  MyList<Patch> *Pp = PatL;
-  while (Pp)
-  {
-    Patch *Pat = Pp->data;
-    MyList<Parallel::gridseg> *dst_ghost = build_ghost_gsl(Pat);
-
-    for (int node = 0; node < cpusize; node++)
-    {
-      MyList<Parallel::gridseg> *src_owned = build_owned_gsl0(Pat, node);
-      MyList<Parallel::gridseg> *tsrc = 0, *tdst = 0;
-      build_gstl(src_owned, dst_ghost, &tsrc, &tdst);
-
-      if (tsrc)
-      {
-        if (combined_src[node])
-          combined_src[node]->catList(tsrc);
-        else
-          combined_src[node] = tsrc;
-      }
-      if (tdst)
-      {
-        if (combined_dst[node])
-          combined_dst[node]->catList(tdst);
-        else
-          combined_dst[node] = tdst;
-      }
-
-      if (src_owned)
-        src_owned->destroyList();
-    }
-
-    if (dst_ghost)
-      dst_ghost->destroyList();
-
-    Pp = Pp->next;
-  }
-
-  // Phase B: Inter-patch buffer exchange segments
-  MyList<Parallel::gridseg> *dst_buffer = build_buffer_gsl(PatL);
-  for (int node = 0; node < cpusize; node++)
-  {
-    MyList<Parallel::gridseg> *src_owned = build_owned_gsl(PatL, node, 5, Symmetry);
-    MyList<Parallel::gridseg> *tsrc = 0, *tdst = 0;
-    build_gstl(src_owned, dst_buffer, &tsrc, &tdst);
-
-    if (tsrc)
-    {
-      if (combined_src[node])
-        combined_src[node]->catList(tsrc);
-      else
-        combined_src[node] = tsrc;
-    }
-    if (tdst)
-    {
-      if (combined_dst[node])
-        combined_dst[node]->catList(tdst);
-      else
-        combined_dst[node] = tdst;
-    }
-
-    if (src_owned)
-      src_owned->destroyList();
-  }
-  if (dst_buffer)
-    dst_buffer->destroyList();
-
-  // Phase C: Single transfer
-  transfer(combined_src, combined_dst, VarList, VarList, Symmetry);
-
-  // Phase D: Cleanup
-  for (int node = 0; node < cpusize; node++)
-  {
-    if (combined_src[node])
-      combined_src[node]->destroyList();
-    if (combined_dst[node])
-      combined_dst[node]->destroyList();
-  }
-  delete[] combined_src;
-  delete[] combined_dst;
-}
-// SyncCache constructor
-Parallel::SyncCache::SyncCache()
-    : valid(false), cpusize(0), combined_src(0), combined_dst(0),
-      send_lengths(0), recv_lengths(0), send_bufs(0), recv_bufs(0),
-      send_buf_caps(0), recv_buf_caps(0), reqs(0), stats(0), max_reqs(0)
-{
-}
-// SyncCache invalidate: free grid segment lists but keep buffers
-void Parallel::SyncCache::invalidate()
-{
-  if (!valid)
-    return;
-  for (int i = 0; i < cpusize; i++)
-  {
-    if (combined_src[i])
-      combined_src[i]->destroyList();
-    if (combined_dst[i])
-      combined_dst[i]->destroyList();
-    combined_src[i] = combined_dst[i] = 0;
-    send_lengths[i] = recv_lengths[i] = 0;
-  }
-  valid = false;
-}
-// SyncCache destroy: free everything
-void Parallel::SyncCache::destroy()
-{
-  invalidate();
-  if (combined_src) delete[] combined_src;
-  if (combined_dst) delete[] combined_dst;
-  if (send_lengths) delete[] send_lengths;
-  if (recv_lengths) delete[] recv_lengths;
-  if (send_buf_caps) delete[] send_buf_caps;
-  if (recv_buf_caps) delete[] recv_buf_caps;
-  for (int i = 0; i < cpusize; i++)
-  {
-    if (send_bufs && send_bufs[i]) delete[] send_bufs[i];
-    if (recv_bufs && recv_bufs[i]) delete[] recv_bufs[i];
-  }
-  if (send_bufs) delete[] send_bufs;
-  if (recv_bufs) delete[] recv_bufs;
-  if (reqs) delete[] reqs;
-  if (stats) delete[] stats;
-  combined_src = combined_dst = 0;
-  send_lengths = recv_lengths = 0;
-  send_buf_caps = recv_buf_caps = 0;
-  send_bufs = recv_bufs = 0;
-  reqs = 0; stats = 0;
-  cpusize = 0; max_reqs = 0;
-}
-// transfer_cached: reuse pre-allocated buffers from SyncCache
-void Parallel::transfer_cached(MyList<Parallel::gridseg> **src, MyList<Parallel::gridseg> **dst,
-                               MyList<var> *VarList1, MyList<var> *VarList2,
-                               int Symmetry, SyncCache &cache)
-{
-  int myrank;
-  MPI_Comm_size(MPI_COMM_WORLD, &cache.cpusize);
-  MPI_Comm_rank(MPI_COMM_WORLD, &myrank);
-  int cpusize = cache.cpusize;
-
-  int req_no = 0;
-  int node;
-
-  for (node = 0; node < cpusize; node++)
-  {
-    if (node == myrank)
-    {
-      int length = data_packer(0, src[myrank], dst[myrank], node, PACK, VarList1, VarList2, Symmetry);
-      cache.recv_lengths[node] = length;
-      if (length > 0)
-      {
-        if (length > cache.recv_buf_caps[node])
-        {
-          if (cache.recv_bufs[node]) delete[] cache.recv_bufs[node];
-          cache.recv_bufs[node] = new double[length];
-          cache.recv_buf_caps[node] = length;
-        }
-        data_packer(cache.recv_bufs[node], src[myrank], dst[myrank], node, PACK, VarList1, VarList2, Symmetry);
-      }
-    }
-    else
-    {
-      // send
-      int slength = data_packer(0, src[myrank], dst[myrank], node, PACK, VarList1, VarList2, Symmetry);
-      cache.send_lengths[node] = slength;
-      if (slength > 0)
-      {
-        if (slength > cache.send_buf_caps[node])
-        {
-          if (cache.send_bufs[node]) delete[] cache.send_bufs[node];
-          cache.send_bufs[node] = new double[slength];
-          cache.send_buf_caps[node] = slength;
-        }
-        data_packer(cache.send_bufs[node], src[myrank], dst[myrank], node, PACK, VarList1, VarList2, Symmetry);
-        MPI_Isend((void *)cache.send_bufs[node], slength, MPI_DOUBLE, node, 1, MPI_COMM_WORLD, cache.reqs + req_no++);
-      }
-      // recv
-      int rlength = data_packer(0, src[node], dst[node], node, UNPACK, VarList1, VarList2, Symmetry);
-      cache.recv_lengths[node] = rlength;
-      if (rlength > 0)
-      {
-        if (rlength > cache.recv_buf_caps[node])
-        {
-          if (cache.recv_bufs[node]) delete[] cache.recv_bufs[node];
-          cache.recv_bufs[node] = new double[rlength];
-          cache.recv_buf_caps[node] = rlength;
-        }
-        MPI_Irecv((void *)cache.recv_bufs[node], rlength, MPI_DOUBLE, node, 1, MPI_COMM_WORLD, cache.reqs + req_no++);
-      }
-    }
-  }
-
-  MPI_Waitall(req_no, cache.reqs, cache.stats);
-
-  for (node = 0; node < cpusize; node++)
-    if (cache.recv_bufs[node] && cache.recv_lengths[node] > 0)
-      data_packer(cache.recv_bufs[node], src[node], dst[node], node, UNPACK, VarList1, VarList2, Symmetry);
-}
-// Sync_cached: build grid segment lists on first call, reuse on subsequent calls
-void Parallel::Sync_cached(MyList<Patch> *PatL, MyList<var> *VarList, int Symmetry, SyncCache &cache)
-{
-  if (!cache.valid)
-  {
-    int cpusize;
-    MPI_Comm_size(MPI_COMM_WORLD, &cpusize);
-    cache.cpusize = cpusize;
-
-    // Allocate cache arrays if needed
-    if (!cache.combined_src)
-    {
-      cache.combined_src = new MyList<Parallel::gridseg> *[cpusize];
-      cache.combined_dst = new MyList<Parallel::gridseg> *[cpusize];
-      cache.send_lengths = new int[cpusize];
-      cache.recv_lengths = new int[cpusize];
-      cache.send_bufs = new double *[cpusize];
-      cache.recv_bufs = new double *[cpusize];
-      cache.send_buf_caps = new int[cpusize];
-      cache.recv_buf_caps = new int[cpusize];
-      for (int i = 0; i < cpusize; i++)
-      {
-        cache.send_bufs[i] = cache.recv_bufs[i] = 0;
-        cache.send_buf_caps[i] = cache.recv_buf_caps[i] = 0;
-      }
-      cache.max_reqs = 2 * cpusize;
-      cache.reqs = new MPI_Request[cache.max_reqs];
-      cache.stats = new MPI_Status[cache.max_reqs];
-    }
-
-    for (int node = 0; node < cpusize; node++)
-    {
-      cache.combined_src[node] = cache.combined_dst[node] = 0;
-      cache.send_lengths[node] = cache.recv_lengths[node] = 0;
-    }
-
-    // Build intra-patch segments (same as Sync_merged Phase A)
-    MyList<Patch> *Pp = PatL;
-    while (Pp)
-    {
-      Patch *Pat = Pp->data;
-      MyList<Parallel::gridseg> *dst_ghost = build_ghost_gsl(Pat);
-      for (int node = 0; node < cpusize; node++)
-      {
-        MyList<Parallel::gridseg> *src_owned = build_owned_gsl0(Pat, node);
-        MyList<Parallel::gridseg> *tsrc = 0, *tdst = 0;
-        build_gstl(src_owned, dst_ghost, &tsrc, &tdst);
-        if (tsrc)
-        {
-          if (cache.combined_src[node])
-            cache.combined_src[node]->catList(tsrc);
-          else
-            cache.combined_src[node] = tsrc;
-        }
-        if (tdst)
-        {
-          if (cache.combined_dst[node])
-            cache.combined_dst[node]->catList(tdst);
-          else
-            cache.combined_dst[node] = tdst;
-        }
-        if (src_owned) src_owned->destroyList();
-      }
-      if (dst_ghost) dst_ghost->destroyList();
-      Pp = Pp->next;
-    }
-
-    // Build inter-patch segments (same as Sync_merged Phase B)
-    MyList<Parallel::gridseg> *dst_buffer = build_buffer_gsl(PatL);
-    for (int node = 0; node < cpusize; node++)
-    {
-      MyList<Parallel::gridseg> *src_owned = build_owned_gsl(PatL, node, 5, Symmetry);
-      MyList<Parallel::gridseg> *tsrc = 0, *tdst = 0;
-      build_gstl(src_owned, dst_buffer, &tsrc, &tdst);
-      if (tsrc)
-      {
-        if (cache.combined_src[node])
-          cache.combined_src[node]->catList(tsrc);
-        else
-          cache.combined_src[node] = tsrc;
-      }
-      if (tdst)
-      {
-        if (cache.combined_dst[node])
-          cache.combined_dst[node]->catList(tdst);
-        else
-          cache.combined_dst[node] = tdst;
-      }
-      if (src_owned) src_owned->destroyList();
-    }
-    if (dst_buffer) dst_buffer->destroyList();
-
-    cache.valid = true;
-  }
-
-  // Use cached lists with buffer-reusing transfer
-  transfer_cached(cache.combined_src, cache.combined_dst, VarList, VarList, Symmetry, cache);
-}
-// Sync_start: pack and post MPI_Isend/Irecv, return immediately
-void Parallel::Sync_start(MyList<Patch> *PatL, MyList<var> *VarList, int Symmetry,
-                          SyncCache &cache, AsyncSyncState &state)
-{
-  // Ensure cache is built
-  if (!cache.valid)
-  {
-    // Build cache (same logic as Sync_cached)
-    int cpusize;
-    MPI_Comm_size(MPI_COMM_WORLD, &cpusize);
-    cache.cpusize = cpusize;
-
-    if (!cache.combined_src)
-    {
-      cache.combined_src = new MyList<Parallel::gridseg> *[cpusize];
-      cache.combined_dst = new MyList<Parallel::gridseg> *[cpusize];
-      cache.send_lengths = new int[cpusize];
-      cache.recv_lengths = new int[cpusize];
-      cache.send_bufs = new double *[cpusize];
-      cache.recv_bufs = new double *[cpusize];
-      cache.send_buf_caps = new int[cpusize];
-      cache.recv_buf_caps = new int[cpusize];
-      for (int i = 0; i < cpusize; i++)
-      {
-        cache.send_bufs[i] = cache.recv_bufs[i] = 0;
-        cache.send_buf_caps[i] = cache.recv_buf_caps[i] = 0;
-      }
-      cache.max_reqs = 2 * cpusize;
-      cache.reqs = new MPI_Request[cache.max_reqs];
-      cache.stats = new MPI_Status[cache.max_reqs];
-    }
-
-    for (int node = 0; node < cpusize; node++)
-    {
-      cache.combined_src[node] = cache.combined_dst[node] = 0;
-      cache.send_lengths[node] = cache.recv_lengths[node] = 0;
-    }
-
-    MyList<Patch> *Pp = PatL;
-    while (Pp)
-    {
-      Patch *Pat = Pp->data;
-      MyList<Parallel::gridseg> *dst_ghost = build_ghost_gsl(Pat);
-      for (int node = 0; node < cpusize; node++)
-      {
-        MyList<Parallel::gridseg> *src_owned = build_owned_gsl0(Pat, node);
-        MyList<Parallel::gridseg> *tsrc = 0, *tdst = 0;
-        build_gstl(src_owned, dst_ghost, &tsrc, &tdst);
-        if (tsrc)
-        {
-          if (cache.combined_src[node])
-            cache.combined_src[node]->catList(tsrc);
-          else
-            cache.combined_src[node] = tsrc;
-        }
-        if (tdst)
-        {
-          if (cache.combined_dst[node])
-            cache.combined_dst[node]->catList(tdst);
-          else
-            cache.combined_dst[node] = tdst;
-        }
-        if (src_owned) src_owned->destroyList();
-      }
-      if (dst_ghost) dst_ghost->destroyList();
-      Pp = Pp->next;
-    }
-
-    MyList<Parallel::gridseg> *dst_buffer = build_buffer_gsl(PatL);
-    for (int node = 0; node < cpusize; node++)
-    {
-      MyList<Parallel::gridseg> *src_owned = build_owned_gsl(PatL, node, 5, Symmetry);
-      MyList<Parallel::gridseg> *tsrc = 0, *tdst = 0;
-      build_gstl(src_owned, dst_buffer, &tsrc, &tdst);
-      if (tsrc)
-      {
-        if (cache.combined_src[node])
-          cache.combined_src[node]->catList(tsrc);
-        else
-          cache.combined_src[node] = tsrc;
-      }
-      if (tdst)
-      {
-        if (cache.combined_dst[node])
-          cache.combined_dst[node]->catList(tdst);
-        else
-          cache.combined_dst[node] = tdst;
-      }
-      if (src_owned) src_owned->destroyList();
-    }
-    if (dst_buffer) dst_buffer->destroyList();
-    cache.valid = true;
-  }
-
-  // Now pack and post async MPI operations
-  int myrank;
-  MPI_Comm_rank(MPI_COMM_WORLD, &myrank);
-  int cpusize = cache.cpusize;
-  state.req_no = 0;
-  state.active = true;
-
-  MyList<Parallel::gridseg> **src = cache.combined_src;
-  MyList<Parallel::gridseg> **dst = cache.combined_dst;
-
-  for (int node = 0; node < cpusize; node++)
-  {
-    if (node == myrank)
-    {
-      int length = data_packer(0, src[myrank], dst[myrank], node, PACK, VarList, VarList, Symmetry);
-      cache.recv_lengths[node] = length;
-      if (length > 0)
-      {
-        if (length > cache.recv_buf_caps[node])
-        {
-          if (cache.recv_bufs[node]) delete[] cache.recv_bufs[node];
-          cache.recv_bufs[node] = new double[length];
-          cache.recv_buf_caps[node] = length;
-        }
-        data_packer(cache.recv_bufs[node], src[myrank], dst[myrank], node, PACK, VarList, VarList, Symmetry);
-      }
-    }
-    else
-    {
-      int slength = data_packer(0, src[myrank], dst[myrank], node, PACK, VarList, VarList, Symmetry);
-      cache.send_lengths[node] = slength;
-      if (slength > 0)
-      {
-        if (slength > cache.send_buf_caps[node])
-        {
-          if (cache.send_bufs[node]) delete[] cache.send_bufs[node];
-          cache.send_bufs[node] = new double[slength];
-          cache.send_buf_caps[node] = slength;
-        }
-        data_packer(cache.send_bufs[node], src[myrank], dst[myrank], node, PACK, VarList, VarList, Symmetry);
-        MPI_Isend((void *)cache.send_bufs[node], slength, MPI_DOUBLE, node, 2, MPI_COMM_WORLD, cache.reqs + state.req_no++);
-      }
-      int rlength = data_packer(0, src[node], dst[node], node, UNPACK, VarList, VarList, Symmetry);
-      cache.recv_lengths[node] = rlength;
-      if (rlength > 0)
-      {
-        if (rlength > cache.recv_buf_caps[node])
-        {
-          if (cache.recv_bufs[node]) delete[] cache.recv_bufs[node];
-          cache.recv_bufs[node] = new double[rlength];
-          cache.recv_buf_caps[node] = rlength;
-        }
-        MPI_Irecv((void *)cache.recv_bufs[node], rlength, MPI_DOUBLE, node, 2, MPI_COMM_WORLD, cache.reqs + state.req_no++);
-      }
-    }
-  }
-}
-// Sync_finish: wait for async MPI operations and unpack
-void Parallel::Sync_finish(SyncCache &cache, AsyncSyncState &state,
-                           MyList<var> *VarList, int Symmetry)
-{
-  if (!state.active)
-    return;
-
-  MPI_Waitall(state.req_no, cache.reqs, cache.stats);
-
-  int cpusize = cache.cpusize;
-  MyList<Parallel::gridseg> **src = cache.combined_src;
-  MyList<Parallel::gridseg> **dst = cache.combined_dst;
-
-  for (int node = 0; node < cpusize; node++)
-    if (cache.recv_bufs[node] && cache.recv_lengths[node] > 0)
-      data_packer(cache.recv_bufs[node], src[node], dst[node], node, UNPACK, VarList, VarList, Symmetry);
-
-  state.active = false;
-}
 // collect buffer grid segments or blocks for the periodic boundary condition of given patch
 // ---------------------------------------------------
 // |con |                                       |con |

AMSS_NCKU_source/Parallel.h

@@ -81,42 +81,6 @@ namespace Parallel
                    int Symmetry);
   void Sync(Patch *Pat, MyList<var> *VarList, int Symmetry);
   void Sync(MyList<Patch> *PatL, MyList<var> *VarList, int Symmetry);
-  void Sync_merged(MyList<Patch> *PatL, MyList<var> *VarList, int Symmetry);
-
-  struct SyncCache {
-    bool valid;
-    int cpusize;
-    MyList<gridseg> **combined_src;
-    MyList<gridseg> **combined_dst;
-    int *send_lengths;
-    int *recv_lengths;
-    double **send_bufs;
-    double **recv_bufs;
-    int *send_buf_caps;
-    int *recv_buf_caps;
-    MPI_Request *reqs;
-    MPI_Status *stats;
-    int max_reqs;
-    SyncCache();
-    void invalidate();
-    void destroy();
-  };
-
-  void Sync_cached(MyList<Patch> *PatL, MyList<var> *VarList, int Symmetry, SyncCache &cache);
-  void transfer_cached(MyList<gridseg> **src, MyList<gridseg> **dst,
-                       MyList<var> *VarList1, MyList<var> *VarList2,
-                       int Symmetry, SyncCache &cache);
-
-  struct AsyncSyncState {
-    int req_no;
-    bool active;
-    AsyncSyncState() : req_no(0), active(false) {}
-  };
-
-  void Sync_start(MyList<Patch> *PatL, MyList<var> *VarList, int Symmetry,
-                  SyncCache &cache, AsyncSyncState &state);
-  void Sync_finish(SyncCache &cache, AsyncSyncState &state,
-                   MyList<var> *VarList, int Symmetry);
   void OutBdLow2Hi(Patch *Patc, Patch *Patf,
                    MyList<var> *VarList1 /* source */, MyList<var> *VarList2 /* target */,
                    int Symmetry);

AMSS_NCKU_source/ShellPatch.h

@@ -2,11 +2,7 @@
 #ifndef SHELLPATCH_H
 #define SHELLPATCH_H
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 #include "MyList.h"
 #include "Block.h"
 #include "Parallel.h"

AMSS_NCKU_source/TwoPunctures.h

@@ -1,8 +1,7 @@
+
 #ifndef TWO_PUNCTURES_H
 #define TWO_PUNCTURES_H
 
-#include <omp.h>
-
 #define StencilSize 19
 #define N_PlaneRelax 1
 #define NRELAX 200
@@ -33,7 +32,7 @@ private:
        int npoints_A, npoints_B, npoints_phi;
 
        double target_M_plus, target_M_minus;
-
+       
        double admMass;
 
        double adm_tol;
@@ -43,17 +42,32 @@ private:
 
        int ntotal;
 
-       // ===== Precomputed spectral derivative matrices =====
-       double *D1_A, *D2_A;
-       double *D1_B, *D2_B;
-       double *DF1_phi, *DF2_phi;
-
-       // ===== Pre-allocated workspace for LineRelax (per-thread) =====
-       int max_threads;
-       double **ws_diag_be, **ws_e_be, **ws_f_be, **ws_b_be, **ws_x_be;
-       double **ws_l_be, **ws_u_be, **ws_d_be, **ws_y_be;
-       double **ws_diag_al, **ws_e_al, **ws_f_al, **ws_b_al, **ws_x_al;
-       double **ws_l_al, **ws_u_al, **ws_d_al, **ws_y_al;
+       // Pre-allocated workspace buffers for hot-path allocation elimination
+       // LineRelax_be workspace (sized for n2)
+       double *ws_diag_be, *ws_e_be, *ws_f_be, *ws_b_be, *ws_x_be;
+       // LineRelax_al workspace (sized for n1)
+       double *ws_diag_al, *ws_e_al, *ws_f_al, *ws_b_al, *ws_x_al;
+       // ThomasAlgorithm workspace (sized for max(n1,n2))
+       double *ws_thomas_y;
+       // JFD_times_dv workspace (sized for nvar)
+       double *ws_jfd_values;
+       derivs ws_jfd_dU, ws_jfd_U;
+       // chebft_Zeros workspace (sized for max(n1,n2,n3)+1)
+       double *ws_cheb_c;
+       // fourft workspace (sized for max(n1,n2,n3)/2+1 each)
+       double *ws_four_a, *ws_four_b;
+       // Derivatives_AB3 workspace
+       double *ws_deriv_p, *ws_deriv_dp, *ws_deriv_d2p;
+       double *ws_deriv_q, *ws_deriv_dq;
+       double *ws_deriv_r, *ws_deriv_dr;
+       int *ws_deriv_indx;
+       // F_of_v workspace
+       double *ws_fov_sources;
+       double *ws_fov_values;
+       derivs ws_fov_U;
+       // J_times_dv workspace
+       double *ws_jtdv_values;
+       derivs ws_jtdv_dU, ws_jtdv_U;
 
        struct parameters
        {
@@ -71,28 +85,6 @@ public:
                     int Newtonmaxit);
        ~TwoPunctures();
 
-       // 02/07: New/modified methods
-       void allocate_workspace();
-       void free_workspace();
-       void precompute_derivative_matrices();
-       void build_cheb_deriv_matrices(int n, double *D1, double *D2);
-       void build_fourier_deriv_matrices(int N, double *DF1, double *DF2);
-       void Derivatives_AB3_MatMul(int nvar, int n1, int n2, int n3, derivs v);
-       void ThomasAlgorithm_ws(int N, double *b, double *a, double *c, double *x, double *q,
-                                double *l, double *u_ws, double *d, double *y);
-       void LineRelax_be_omp(double *dv,
-                         int const i, int const k, int const nvar,
-                         int const n1, int const n2, int const n3,
-                         double const *rhs, int const *ncols, int **cols,
-                         double **JFD, int tid);
-       void LineRelax_al_omp(double *dv,
-                         int const j, int const k, int const nvar,
-                         int const n1, int const n2, int const n3,
-                         double const *rhs, int const *ncols,
-                         int **cols, double **JFD, int tid);
-       void relax_omp(double *dv, int const nvar, int const n1, int const n2, int const n3,
-                  double const *rhs, int const *ncols, int **cols, double **JFD);
-
        void Solve();
        void set_initial_guess(derivs v);
        int index(int i, int j, int k, int l, int a, int b, int c, int d);
@@ -151,11 +143,23 @@ public:
        double BY_KKofxyz(double x, double y, double z);
        void SetMatrix_JFD(int nvar, int n1, int n2, int n3, derivs u, int *ncols, int **cols, double **Matrix);
        void J_times_dv(int nvar, int n1, int n2, int n3, derivs dv, double *Jdv, derivs u);
+       void relax(double *dv, int const nvar, int const n1, int const n2, int const n3,
+                  double const *rhs, int const *ncols, int **cols, double **JFD);
+       void LineRelax_be(double *dv,
+                         int const i, int const k, int const nvar,
+                         int const n1, int const n2, int const n3,
+                         double const *rhs, int const *ncols, int **cols,
+                         double **JFD);
        void JFD_times_dv(int i, int j, int k, int nvar, int n1, int n2,
                          int n3, derivs dv, derivs u, double *values);
        void LinEquations(double A, double B, double X, double R,
                          double x, double r, double phi,
                          double y, double z, derivs dU, derivs U, double *values);
+       void LineRelax_al(double *dv,
+                         int const j, int const k, int const nvar,
+                         int const n1, int const n2, int const n3,
+                         double const *rhs, int const *ncols,
+                         int **cols, double **JFD);
        void ThomasAlgorithm(int N, double *b, double *a, double *c, double *x, double *q);
        void Save(char *fname);
        // provided by Vasileios Paschalidis (vpaschal@illinois.edu)
@@ -164,4 +168,4 @@ public:
        void SpecCoef(parameters par, int ivar, double *v, double *cf);
 };
 
-#endif /* TWO_PUNCTURES_H */
+#endif /* TWO_PUNCTURES_H */

AMSS_NCKU_source/Z4c_class.h

@@ -19,11 +19,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "cgh.h"
 #include "ShellPatch.h"

AMSS_NCKU_source/bssnEM_class.h

@@ -19,11 +19,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "cgh.h"
 #include "ShellPatch.h"

AMSS_NCKU_source/bssnEScalar_class.h

@@ -19,11 +19,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "cgh.h"
 #include "ShellPatch.h"

AMSS_NCKU_source/bssn_class.C

@@ -730,12 +730,6 @@ void bssn_class::Initialize()
     PhysTime = StartTime;
     Setup_Black_Hole_position();
   }
-
-  // Initialize sync caches (per-level, for predictor and corrector)
-  sync_cache_pre = new Parallel::SyncCache[GH->levels];
-  sync_cache_cor = new Parallel::SyncCache[GH->levels];
-  sync_cache_rp_coarse = new Parallel::SyncCache[GH->levels];
-  sync_cache_rp_fine = new Parallel::SyncCache[GH->levels];
 }
 
 //================================================================================================
@@ -987,32 +981,6 @@ bssn_class::~bssn_class()
   delete Azzz;
 #endif
 
-  // Destroy sync caches before GH
-  if (sync_cache_pre)
-  {
-    for (int i = 0; i < GH->levels; i++)
-      sync_cache_pre[i].destroy();
-    delete[] sync_cache_pre;
-  }
-  if (sync_cache_cor)
-  {
-    for (int i = 0; i < GH->levels; i++)
-      sync_cache_cor[i].destroy();
-    delete[] sync_cache_cor;
-  }
-  if (sync_cache_rp_coarse)
-  {
-    for (int i = 0; i < GH->levels; i++)
-      sync_cache_rp_coarse[i].destroy();
-    delete[] sync_cache_rp_coarse;
-  }
-  if (sync_cache_rp_fine)
-  {
-    for (int i = 0; i < GH->levels; i++)
-      sync_cache_rp_fine[i].destroy();
-    delete[] sync_cache_rp_fine;
-  }
-
   delete GH;
 #ifdef WithShell
   delete SH;
@@ -2213,7 +2181,6 @@ void bssn_class::Evolve(int Steps)
     GH->Regrid(Symmetry, BH_num, Porgbr, Porg0,
                SynchList_cor, OldStateList, StateList, SynchList_pre,
                fgt(PhysTime - dT_mon, StartTime, dT_mon / 2), ErrorMonitor);
-    for (int il = 0; il < GH->levels; il++) { sync_cache_pre[il].invalidate(); sync_cache_cor[il].invalidate(); sync_cache_rp_coarse[il].invalidate(); sync_cache_rp_fine[il].invalidate(); }
 #endif
 
 #if (REGLEV == 0 && (PSTR == 1 || PSTR == 2))
@@ -2429,7 +2396,6 @@ void bssn_class::RecursiveStep(int lev)
   GH->Regrid_Onelevel(lev, Symmetry, BH_num, Porgbr, Porg0,
                       SynchList_cor, OldStateList, StateList, SynchList_pre,
                       fgt(PhysTime - dT_lev, StartTime, dT_lev / 2), ErrorMonitor);
-  for (int il = 0; il < GH->levels; il++) { sync_cache_pre[il].invalidate(); sync_cache_cor[il].invalidate(); sync_cache_rp_coarse[il].invalidate(); sync_cache_rp_fine[il].invalidate(); }
 #endif
 }
 
@@ -2608,7 +2574,6 @@ void bssn_class::ParallelStep()
   GH->Regrid_Onelevel(GH->mylev, Symmetry, BH_num, Porgbr, Porg0,
                       SynchList_cor, OldStateList, StateList, SynchList_pre,
                       fgt(PhysTime - dT_lev, StartTime, dT_lev / 2), ErrorMonitor);
-  for (int il = 0; il < GH->levels; il++) { sync_cache_pre[il].invalidate(); sync_cache_cor[il].invalidate(); sync_cache_rp_coarse[il].invalidate(); sync_cache_rp_fine[il].invalidate(); }
 #endif
 }
 
@@ -2775,7 +2740,6 @@ void bssn_class::ParallelStep()
         GH->Regrid_Onelevel(lev + 1, Symmetry, BH_num, Porgbr, Porg0,
                             SynchList_cor, OldStateList, StateList, SynchList_pre,
                             fgt(PhysTime - dT_levp1, StartTime, dT_levp1 / 2), ErrorMonitor);
-        for (int il = 0; il < GH->levels; il++) { sync_cache_pre[il].invalidate(); sync_cache_cor[il].invalidate(); sync_cache_rp_coarse[il].invalidate(); sync_cache_rp_fine[il].invalidate(); }
 
         //               a_stream.clear();
         //               a_stream.str("");
@@ -2790,7 +2754,6 @@ void bssn_class::ParallelStep()
       GH->Regrid_Onelevel(lev, Symmetry, BH_num, Porgbr, Porg0,
                           SynchList_cor, OldStateList, StateList, SynchList_pre,
                           fgt(PhysTime - dT_lev, StartTime, dT_lev / 2), ErrorMonitor);
-      for (int il = 0; il < GH->levels; il++) { sync_cache_pre[il].invalidate(); sync_cache_cor[il].invalidate(); sync_cache_rp_coarse[il].invalidate(); sync_cache_rp_fine[il].invalidate(); }
 
       //               a_stream.clear();
       //               a_stream.str("");
@@ -2809,7 +2772,6 @@ void bssn_class::ParallelStep()
           GH->Regrid_Onelevel(lev - 1, Symmetry, BH_num, Porgbr, Porg0,
                               SynchList_cor, OldStateList, StateList, SynchList_pre,
                               fgt(PhysTime - dT_lev, StartTime, dT_levm1 / 2), ErrorMonitor);
-          for (int il = 0; il < GH->levels; il++) { sync_cache_pre[il].invalidate(); sync_cache_cor[il].invalidate(); sync_cache_rp_coarse[il].invalidate(); sync_cache_rp_fine[il].invalidate(); }
 
           //               a_stream.clear();
           //               a_stream.str("");
@@ -2825,7 +2787,6 @@ void bssn_class::ParallelStep()
           GH->Regrid_Onelevel(lev - 1, Symmetry, BH_num, Porgbr, Porg0,
                               SynchList_cor, OldStateList, StateList, SynchList_pre,
                               fgt(PhysTime - dT_lev, StartTime, dT_levm1 / 2), ErrorMonitor);
-          for (int il = 0; il < GH->levels; il++) { sync_cache_pre[il].invalidate(); sync_cache_cor[il].invalidate(); sync_cache_rp_coarse[il].invalidate(); sync_cache_rp_fine[il].invalidate(); }
 
           //               a_stream.clear();
           //               a_stream.str("");
@@ -3197,7 +3158,21 @@ void bssn_class::Step(int lev, int YN)
     }
     Pp = Pp->next;
   }
-  // NOTE: error check deferred to after Shell Patch computation to reduce MPI_Allreduce calls
+  // check error information
+  {
+    int erh = ERROR;
+    MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+  }
+  if (ERROR)
+  {
+    Parallel::Dump_Data(GH->PatL[lev], StateList, 0, PhysTime, dT_lev);
+    if (myrank == 0)
+    {
+      if (ErrorMonitor->outfile)
+        ErrorMonitor->outfile << "find NaN in state variables at t = " << PhysTime << ", lev = " << lev << endl;
+      MPI_Abort(MPI_COMM_WORLD, 1);
+    }
+  }
 
 #ifdef WithShell
   // evolve Shell Patches
@@ -3215,9 +3190,9 @@ void bssn_class::Step(int lev, int YN)
         {
 #if (AGM == 0)
           f_enforce_ga(cg->shape,
-                       cg->fgfs[gxx0->sgfn], cg->fgfs[gxy0->sgfn], cg->fgfs[gxz0->sgfn],
+                       cg->fgfs[gxx0->sgfn], cg->fgfs[gxy0->sgfn], cg->fgfs[gxz0->sgfn], 
                        cg->fgfs[gyy0->sgfn], cg->fgfs[gyz0->sgfn], cg->fgfs[gzz0->sgfn],
-                       cg->fgfs[Axx0->sgfn], cg->fgfs[Axy0->sgfn], cg->fgfs[Axz0->sgfn],
+                       cg->fgfs[Axx0->sgfn], cg->fgfs[Axy0->sgfn], cg->fgfs[Axz0->sgfn], 
                        cg->fgfs[Ayy0->sgfn], cg->fgfs[Ayz0->sgfn], cg->fgfs[Azz0->sgfn]);
 #endif
 
@@ -3341,16 +3316,25 @@ void bssn_class::Step(int lev, int YN)
 #endif
   }
 
-  // Non-blocking error reduction overlapped with Sync to hide Allreduce latency
-  MPI_Request err_req;
+  // check error information
   {
     int erh = ERROR;
-    MPI_Iallreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD, &err_req);
+    MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+  }
+
+  if (ERROR)
+  {
+    SH->Dump_Data(StateList, 0, PhysTime, dT_lev);
+    if (myrank == 0)
+    {
+      if (ErrorMonitor->outfile)
+        ErrorMonitor->outfile << "find NaN in state variables on Shell Patches at t = " << PhysTime << endl;
+      MPI_Abort(MPI_COMM_WORLD, 1);
+    }
   }
 #endif
 
-  Parallel::AsyncSyncState async_pre;
-  Parallel::Sync_start(GH->PatL[lev], SynchList_pre, Symmetry, sync_cache_pre[lev], async_pre);
+  Parallel::Sync(GH->PatL[lev], SynchList_pre, Symmetry);
 
 #ifdef WithShell
   if (lev == 0)
@@ -3363,29 +3347,12 @@ void bssn_class::Step(int lev, int YN)
     {
       prev_clock = curr_clock;
       curr_clock = clock();
-      cout << " Shell stuff synchronization used "
-           << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC)
+      cout << " Shell stuff synchronization used " 
+           << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC) 
            << " seconds! " << endl;
     }
   }
 #endif
-  Parallel::Sync_finish(sync_cache_pre[lev], async_pre, SynchList_pre, Symmetry);
-
-#ifdef WithShell
-  // Complete non-blocking error reduction and check
-  MPI_Wait(&err_req, MPI_STATUS_IGNORE);
-  if (ERROR)
-  {
-    Parallel::Dump_Data(GH->PatL[lev], StateList, 0, PhysTime, dT_lev);
-    SH->Dump_Data(StateList, 0, PhysTime, dT_lev);
-    if (myrank == 0)
-    {
-      if (ErrorMonitor->outfile)
-        ErrorMonitor->outfile << "find NaN in state variables at t = " << PhysTime << ", lev = " << lev << endl;
-      MPI_Abort(MPI_COMM_WORLD, 1);
-    }
-  }
-#endif
 
 #if (MAPBH == 0)
   // for black hole position
@@ -3561,7 +3528,24 @@ void bssn_class::Step(int lev, int YN)
       Pp = Pp->next;
     }
 
-    // NOTE: error check deferred to after Shell Patch computation to reduce MPI_Allreduce calls
+    // check error information
+    {
+      int erh = ERROR;
+      MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+    }
+
+    if (ERROR)
+    {
+      Parallel::Dump_Data(GH->PatL[lev], SynchList_pre, 0, PhysTime, dT_lev);
+      if (myrank == 0)
+      {
+        if (ErrorMonitor->outfile)
+          ErrorMonitor->outfile << "find NaN in RK4 substep#" << iter_count 
+                                << " variables at t = " << PhysTime 
+                                << ", lev = " << lev << endl;
+        MPI_Abort(MPI_COMM_WORLD, 1);
+      }
+    }
 
 #ifdef WithShell
     // evolve Shell Patches
@@ -3579,9 +3563,9 @@ void bssn_class::Step(int lev, int YN)
           {
 #if (AGM == 0)
             f_enforce_ga(cg->shape,
-                         cg->fgfs[gxx->sgfn], cg->fgfs[gxy->sgfn], cg->fgfs[gxz->sgfn],
+                         cg->fgfs[gxx->sgfn], cg->fgfs[gxy->sgfn], cg->fgfs[gxz->sgfn], 
                          cg->fgfs[gyy->sgfn], cg->fgfs[gyz->sgfn], cg->fgfs[gzz->sgfn],
-                         cg->fgfs[Axx->sgfn], cg->fgfs[Axy->sgfn], cg->fgfs[Axz->sgfn],
+                         cg->fgfs[Axx->sgfn], cg->fgfs[Axy->sgfn], cg->fgfs[Axz->sgfn], 
                          cg->fgfs[Ayy->sgfn], cg->fgfs[Ayz->sgfn], cg->fgfs[Azz->sgfn]);
 #elif (AGM == 1)
             if (iter_count == 3)
@@ -3701,16 +3685,26 @@ void bssn_class::Step(int lev, int YN)
         sPp = sPp->next;
       }
     }
-    // Non-blocking error reduction overlapped with Sync to hide Allreduce latency
-    MPI_Request err_req_cor;
+    // check error information
     {
       int erh = ERROR;
-      MPI_Iallreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD, &err_req_cor);
+      MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+    }
+    if (ERROR)
+    {
+      SH->Dump_Data(SynchList_pre, 0, PhysTime, dT_lev);
+      if (myrank == 0)
+      {
+        if (ErrorMonitor->outfile)
+          ErrorMonitor->outfile << "find NaN on Shell Patches in RK4 substep#" 
+                                << iter_count << " variables at t = " 
+                                << PhysTime << endl;
+        MPI_Abort(MPI_COMM_WORLD, 1);
+      }
     }
 #endif
 
-    Parallel::AsyncSyncState async_cor;
-    Parallel::Sync_start(GH->PatL[lev], SynchList_cor, Symmetry, sync_cache_cor[lev], async_cor);
+    Parallel::Sync(GH->PatL[lev], SynchList_cor, Symmetry);
 
 #ifdef WithShell
     if (lev == 0)
@@ -3723,31 +3717,12 @@ void bssn_class::Step(int lev, int YN)
       {
         prev_clock = curr_clock;
         curr_clock = clock();
-        cout << " Shell stuff synchronization used "
-             << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC)
+        cout << " Shell stuff synchronization used " 
+             << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC) 
              << " seconds! " << endl;
       }
     }
 #endif
-    Parallel::Sync_finish(sync_cache_cor[lev], async_cor, SynchList_cor, Symmetry);
-
-#ifdef WithShell
-    // Complete non-blocking error reduction and check
-    MPI_Wait(&err_req_cor, MPI_STATUS_IGNORE);
-    if (ERROR)
-    {
-      Parallel::Dump_Data(GH->PatL[lev], SynchList_pre, 0, PhysTime, dT_lev);
-      SH->Dump_Data(SynchList_pre, 0, PhysTime, dT_lev);
-      if (myrank == 0)
-      {
-        if (ErrorMonitor->outfile)
-          ErrorMonitor->outfile << "find NaN in RK4 substep#" << iter_count
-                                << " variables at t = " << PhysTime
-                                << ", lev = " << lev << endl;
-        MPI_Abort(MPI_COMM_WORLD, 1);
-      }
-    }
-#endif
 
 #if (MAPBH == 0)
     // for black hole position
@@ -4059,7 +4034,22 @@ void bssn_class::Step(int lev, int YN)
     }
     Pp = Pp->next;
   }
-  // NOTE: error check deferred to after Shell Patch computation to reduce MPI_Allreduce calls
+  // check error information
+  {
+    int erh = ERROR;
+    MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+  }
+  if (ERROR)
+  {
+    Parallel::Dump_Data(GH->PatL[lev], StateList, 0, PhysTime, dT_lev);
+    if (myrank == 0)
+    {
+      if (ErrorMonitor->outfile)
+        ErrorMonitor->outfile << "find NaN in state variables at t = " << PhysTime 
+                              << ", lev = " << lev << endl;
+      MPI_Abort(MPI_COMM_WORLD, 1);
+    }
+  }
 
 #ifdef WithShell
   // evolve Shell Patches
@@ -4077,15 +4067,15 @@ void bssn_class::Step(int lev, int YN)
         {
 #if (AGM == 0)
           f_enforce_ga(cg->shape,
-                       cg->fgfs[gxx0->sgfn], cg->fgfs[gxy0->sgfn], cg->fgfs[gxz0->sgfn],
+                       cg->fgfs[gxx0->sgfn], cg->fgfs[gxy0->sgfn], cg->fgfs[gxz0->sgfn], 
                        cg->fgfs[gyy0->sgfn], cg->fgfs[gyz0->sgfn], cg->fgfs[gzz0->sgfn],
-                       cg->fgfs[Axx0->sgfn], cg->fgfs[Axy0->sgfn], cg->fgfs[Axz0->sgfn],
+                       cg->fgfs[Axx0->sgfn], cg->fgfs[Axy0->sgfn], cg->fgfs[Axz0->sgfn], 
                        cg->fgfs[Ayy0->sgfn], cg->fgfs[Ayz0->sgfn], cg->fgfs[Azz0->sgfn]);
 #endif
 
           if (f_compute_rhs_bssn_ss(cg->shape, TRK4, cg->X[0], cg->X[1], cg->X[2],
-                                    cg->fgfs[fngfs + ShellPatch::gx],
-                                    cg->fgfs[fngfs + ShellPatch::gy],
+                                    cg->fgfs[fngfs + ShellPatch::gx], 
+                                    cg->fgfs[fngfs + ShellPatch::gy], 
                                     cg->fgfs[fngfs + ShellPatch::gz],
                                     cg->fgfs[fngfs + ShellPatch::drhodx], 
                                     cg->fgfs[fngfs + ShellPatch::drhody], 
@@ -4200,16 +4190,25 @@ void bssn_class::Step(int lev, int YN)
   }
 #endif
   }
-  // Non-blocking error reduction overlapped with Sync to hide Allreduce latency
-  MPI_Request err_req;
+  // check error information
   {
     int erh = ERROR;
-    MPI_Iallreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD, &err_req);
+    MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+  }
+  if (ERROR)
+  {
+    SH->Dump_Data(StateList, 0, PhysTime, dT_lev);
+    if (myrank == 0)
+    {
+      if (ErrorMonitor->outfile)
+        ErrorMonitor->outfile << "find NaN in state variables on Shell Patches at t = " 
+                              << PhysTime << endl;
+      MPI_Abort(MPI_COMM_WORLD, 1);
+    }
   }
 #endif
 
-  Parallel::AsyncSyncState async_pre;
-  Parallel::Sync_start(GH->PatL[lev], SynchList_pre, Symmetry, sync_cache_pre[lev], async_pre);
+  Parallel::Sync(GH->PatL[lev], SynchList_pre, Symmetry);
 
 #ifdef WithShell
   if (lev == 0)
@@ -4222,27 +4221,9 @@ void bssn_class::Step(int lev, int YN)
     {
       prev_clock = curr_clock;
       curr_clock = clock();
-      cout << " Shell stuff synchronization used "
-           << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC)
-           << " seconds! " << endl;
-    }
-  }
-#endif
-  Parallel::Sync_finish(sync_cache_pre[lev], async_pre, SynchList_pre, Symmetry);
-
-#ifdef WithShell
-  // Complete non-blocking error reduction and check
-  MPI_Wait(&err_req, MPI_STATUS_IGNORE);
-  if (ERROR)
-  {
-    Parallel::Dump_Data(GH->PatL[lev], StateList, 0, PhysTime, dT_lev);
-    SH->Dump_Data(StateList, 0, PhysTime, dT_lev);
-    if (myrank == 0)
-    {
-      if (ErrorMonitor->outfile)
-        ErrorMonitor->outfile << "find NaN in state variables at t = " << PhysTime
-                              << ", lev = " << lev << endl;
-      MPI_Abort(MPI_COMM_WORLD, 1);
+      cout << " Shell stuff synchronization used " 
+      << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC) 
+      << " seconds! " << endl;
     }
   }
 #endif
@@ -4405,7 +4386,23 @@ void bssn_class::Step(int lev, int YN)
       Pp = Pp->next;
     }
 
-    // NOTE: error check deferred to after Shell Patch computation to reduce MPI_Allreduce calls
+    // check error information
+    {
+      int erh = ERROR;
+      MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+    }
+    if (ERROR)
+    {
+      Parallel::Dump_Data(GH->PatL[lev], SynchList_pre, 0, PhysTime, dT_lev);
+      if (myrank == 0)
+      {
+        if (ErrorMonitor->outfile)
+          ErrorMonitor->outfile << "find NaN in RK4 substep#" << iter_count 
+                                << " variables at t = " << PhysTime 
+                                << ", lev = " << lev << endl;
+        MPI_Abort(MPI_COMM_WORLD, 1);
+      }
+    }
 
 #ifdef WithShell
     // evolve Shell Patches
@@ -4423,9 +4420,9 @@ void bssn_class::Step(int lev, int YN)
           {
 #if (AGM == 0)
             f_enforce_ga(cg->shape,
-                         cg->fgfs[gxx->sgfn], cg->fgfs[gxy->sgfn], cg->fgfs[gxz->sgfn],
+                         cg->fgfs[gxx->sgfn], cg->fgfs[gxy->sgfn], cg->fgfs[gxz->sgfn], 
                          cg->fgfs[gyy->sgfn], cg->fgfs[gyz->sgfn], cg->fgfs[gzz->sgfn],
-                         cg->fgfs[Axx->sgfn], cg->fgfs[Axy->sgfn], cg->fgfs[Axz->sgfn],
+                         cg->fgfs[Axx->sgfn], cg->fgfs[Axy->sgfn], cg->fgfs[Axz->sgfn], 
                          cg->fgfs[Ayy->sgfn], cg->fgfs[Ayz->sgfn], cg->fgfs[Azz->sgfn]);
 #elif (AGM == 1)
             if (iter_count == 3)
@@ -4545,16 +4542,25 @@ void bssn_class::Step(int lev, int YN)
         sPp = sPp->next;
       }
     }
-    // Non-blocking error reduction overlapped with Sync to hide Allreduce latency
-    MPI_Request err_req_cor;
+    // check error information
     {
       int erh = ERROR;
-      MPI_Iallreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD, &err_req_cor);
+      MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+    }
+    if (ERROR)
+    {
+      SH->Dump_Data(SynchList_pre, 0, PhysTime, dT_lev);
+      if (myrank == 0)
+      {
+        if (ErrorMonitor->outfile)
+          ErrorMonitor->outfile << "find NaN on Shell Patches in RK4 substep#" << iter_count 
+                                << " variables at t = " << PhysTime << endl;
+        MPI_Abort(MPI_COMM_WORLD, 1);
+      }
     }
 #endif
 
-    Parallel::AsyncSyncState async_cor;
-    Parallel::Sync_start(GH->PatL[lev], SynchList_cor, Symmetry, sync_cache_cor[lev], async_cor);
+    Parallel::Sync(GH->PatL[lev], SynchList_cor, Symmetry);
 
 #ifdef WithShell
     if (lev == 0)
@@ -4567,30 +4573,11 @@ void bssn_class::Step(int lev, int YN)
       {
         prev_clock = curr_clock;
         curr_clock = clock();
-        cout << " Shell stuff synchronization used "
-             << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC)
+        cout << " Shell stuff synchronization used " 
+             << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC) 
              << " seconds! " << endl;
       }
     }
-#endif
-    Parallel::Sync_finish(sync_cache_cor[lev], async_cor, SynchList_cor, Symmetry);
-
-#ifdef WithShell
-    // Complete non-blocking error reduction and check
-    MPI_Wait(&err_req_cor, MPI_STATUS_IGNORE);
-    if (ERROR)
-    {
-      Parallel::Dump_Data(GH->PatL[lev], SynchList_pre, 0, PhysTime, dT_lev);
-      SH->Dump_Data(SynchList_pre, 0, PhysTime, dT_lev);
-      if (myrank == 0)
-      {
-        if (ErrorMonitor->outfile)
-          ErrorMonitor->outfile << "find NaN in RK4 substep#" << iter_count
-                                << " variables at t = " << PhysTime
-                                << ", lev = " << lev << endl;
-        MPI_Abort(MPI_COMM_WORLD, 1);
-      }
-    }
 #endif
     // for black hole position
     if (BH_num > 0 && lev == GH->levels - 1)
@@ -4956,19 +4943,11 @@ void bssn_class::Step(int lev, int YN)
 
   //   misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"after Predictor rhs calculation");
 
-  // Non-blocking error reduction overlapped with Sync to hide Allreduce latency
-  MPI_Request err_req;
+  // check error information
   {
     int erh = ERROR;
-    MPI_Iallreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, GH->Commlev[lev], &err_req);
+    MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, GH->Commlev[lev]);
   }
-
-  //   misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"before Predictor sync");
-
-  Parallel::Sync_cached(GH->PatL[lev], SynchList_pre, Symmetry, sync_cache_pre[lev]);
-
-  // Complete non-blocking error reduction and check
-  MPI_Wait(&err_req, MPI_STATUS_IGNORE);
   if (ERROR)
   {
     Parallel::Dump_Data(GH->PatL[lev], StateList, 0, PhysTime, dT_lev);
@@ -4980,6 +4959,10 @@ void bssn_class::Step(int lev, int YN)
     }
   }
 
+  //   misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"before Predictor sync");
+
+  Parallel::Sync(GH->PatL[lev], SynchList_pre, Symmetry);
+
 #if (MAPBH == 0)
   // for black hole position
   if (BH_num > 0 && lev == GH->levels - 1)
@@ -5157,34 +5140,30 @@ void bssn_class::Step(int lev, int YN)
 
     //   misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"before Corrector error check");
 
-    // Non-blocking error reduction overlapped with Sync to hide Allreduce latency
-    MPI_Request err_req_cor;
+    // check error information
     {
       int erh = ERROR;
-      MPI_Iallreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, GH->Commlev[lev], &err_req_cor);
+      MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, GH->Commlev[lev]);
     }
-
-    //    misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"before Corrector sync");
-
-    Parallel::Sync_cached(GH->PatL[lev], SynchList_cor, Symmetry, sync_cache_cor[lev]);
-
-    //    misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"after Corrector sync");
-
-    // Complete non-blocking error reduction and check
-    MPI_Wait(&err_req_cor, MPI_STATUS_IGNORE);
     if (ERROR)
     {
       Parallel::Dump_Data(GH->PatL[lev], SynchList_pre, 0, PhysTime, dT_lev);
       if (myrank == 0)
       {
         if (ErrorMonitor->outfile)
-          ErrorMonitor->outfile << "find NaN in RK4 substep#" << iter_count
-                                << " variables at t = " << PhysTime
+          ErrorMonitor->outfile << "find NaN in RK4 substep#" << iter_count 
+                                << " variables at t = " << PhysTime 
                                 << ", lev = " << lev << endl;
         MPI_Abort(MPI_COMM_WORLD, 1);
       }
     }
 
+    //    misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"before Corrector sync");
+
+    Parallel::Sync(GH->PatL[lev], SynchList_cor, Symmetry);
+
+    //    misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"after Corrector sync");
+
 #if (MAPBH == 0)
     // for black hole position
     if (BH_num > 0 && lev == GH->levels - 1)
@@ -5468,11 +5447,21 @@ void bssn_class::SHStep()
 #if (PSTR == 1 || PSTR == 2)
 //   misc::tillherecheck(GH->Commlev[lev],GH->start_rank[lev],"before Predictor's error check");
 #endif
-  // Non-blocking error reduction overlapped with Synch to hide Allreduce latency
-  MPI_Request err_req;
+  // check error information
   {
     int erh = ERROR;
-    MPI_Iallreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD, &err_req);
+    MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+  }
+
+  if (ERROR)
+  {
+    SH->Dump_Data(StateList, 0, PhysTime, dT_lev);
+    if (myrank == 0)
+    {
+      if (ErrorMonitor->outfile)
+        ErrorMonitor->outfile << "find NaN in state variables on Shell Patches at t = " << PhysTime << endl;
+      MPI_Abort(MPI_COMM_WORLD, 1);
+    }
   }
 
   {
@@ -5484,25 +5473,12 @@ void bssn_class::SHStep()
     {
       prev_clock = curr_clock;
       curr_clock = clock();
-      cout << " Shell stuff synchronization used "
-           << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC)
+      cout << " Shell stuff synchronization used " 
+           << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC) 
            << " seconds! " << endl;
     }
   }
 
-  // Complete non-blocking error reduction and check
-  MPI_Wait(&err_req, MPI_STATUS_IGNORE);
-  if (ERROR)
-  {
-    SH->Dump_Data(StateList, 0, PhysTime, dT_lev);
-    if (myrank == 0)
-    {
-      if (ErrorMonitor->outfile)
-        ErrorMonitor->outfile << "find NaN in state variables on Shell Patches at t = " << PhysTime << endl;
-      MPI_Abort(MPI_COMM_WORLD, 1);
-    }
-  }
-
   // corrector
   for (iter_count = 1; iter_count < 4; iter_count++)
   {
@@ -5645,11 +5621,21 @@ void bssn_class::SHStep()
         sPp = sPp->next;
       }
     }
-    // Non-blocking error reduction overlapped with Synch to hide Allreduce latency
-    MPI_Request err_req_cor;
+    // check error information
     {
       int erh = ERROR;
-      MPI_Iallreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD, &err_req_cor);
+      MPI_Allreduce(&erh, &ERROR, 1, MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+    }
+    if (ERROR)
+    {
+      SH->Dump_Data(SynchList_pre, 0, PhysTime, dT_lev);
+      if (myrank == 0)
+      {
+        if (ErrorMonitor->outfile)
+          ErrorMonitor->outfile << "find NaN on Shell Patches in RK4 substep#" << iter_count 
+                                << " variables at t = " << PhysTime << endl;
+        MPI_Abort(MPI_COMM_WORLD, 1);
+      }
     }
 
     {
@@ -5661,26 +5647,12 @@ void bssn_class::SHStep()
       {
         prev_clock = curr_clock;
         curr_clock = clock();
-        cout << " Shell stuff synchronization used "
-             << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC)
+        cout << " Shell stuff synchronization used " 
+             << (double)(curr_clock - prev_clock) / ((double)CLOCKS_PER_SEC) 
              << " seconds! " << endl;
       }
     }
 
-    // Complete non-blocking error reduction and check
-    MPI_Wait(&err_req_cor, MPI_STATUS_IGNORE);
-    if (ERROR)
-    {
-      SH->Dump_Data(SynchList_pre, 0, PhysTime, dT_lev);
-      if (myrank == 0)
-      {
-        if (ErrorMonitor->outfile)
-          ErrorMonitor->outfile << "find NaN on Shell Patches in RK4 substep#" << iter_count
-                                << " variables at t = " << PhysTime << endl;
-        MPI_Abort(MPI_COMM_WORLD, 1);
-      }
-    }
-
     sPp = SH->PatL;
     while (sPp)
     {
@@ -5809,7 +5781,7 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB,
 //       misc::tillherecheck(GH->Commlev[GH->mylev],GH->start_rank[GH->mylev],a_stream.str());
 #endif
 
-      Parallel::Sync_cached(GH->PatL[lev - 1], SynchList_pre, Symmetry, sync_cache_rp_coarse[lev]);
+      Parallel::Sync(GH->PatL[lev - 1], SynchList_pre, Symmetry);
 
 #if (PSTR == 1 || PSTR == 2)
 //       a_stream.clear();
@@ -5870,7 +5842,7 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB,
 //       misc::tillherecheck(GH->Commlev[GH->mylev],GH->start_rank[GH->mylev],a_stream.str());
 #endif
 
-      Parallel::Sync_cached(GH->PatL[lev - 1], SL, Symmetry, sync_cache_rp_coarse[lev]);
+      Parallel::Sync(GH->PatL[lev - 1], SL, Symmetry);
 
 #if (PSTR == 1 || PSTR == 2)
 //       a_stream.clear();
@@ -5908,7 +5880,7 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB,
 #endif
     }
 
-    Parallel::Sync_cached(GH->PatL[lev], SL, Symmetry, sync_cache_rp_fine[lev]);
+    Parallel::Sync(GH->PatL[lev], SL, Symmetry);
 
 #if (PSTR == 1 || PSTR == 2)
 //    a_stream.clear();
@@ -5966,7 +5938,7 @@ void bssn_class::RestrictProlong_aux(int lev, int YN, bool BB,
       Parallel::Restrict_bam(GH->PatL[lev - 1], GH->PatL[lev], SL, SynchList_pre, GH->rsul[lev], Symmetry);
 #endif
 
-      Parallel::Sync_cached(GH->PatL[lev - 1], SynchList_pre, Symmetry, sync_cache_rp_coarse[lev]);
+      Parallel::Sync(GH->PatL[lev - 1], SynchList_pre, Symmetry);
 
 #if (RPB == 0)
       Ppc = GH->PatL[lev - 1];
@@ -5998,7 +5970,7 @@ void bssn_class::RestrictProlong_aux(int lev, int YN, bool BB,
       Parallel::Restrict_bam(GH->PatL[lev - 1], GH->PatL[lev], SL, SL, GH->rsul[lev], Symmetry);
 #endif
 
-      Parallel::Sync_cached(GH->PatL[lev - 1], SL, Symmetry, sync_cache_rp_coarse[lev]);
+      Parallel::Sync(GH->PatL[lev - 1], SL, Symmetry);
 
 #if (RPB == 0)
       Ppc = GH->PatL[lev - 1];
@@ -6022,7 +5994,7 @@ void bssn_class::RestrictProlong_aux(int lev, int YN, bool BB,
 #endif
     }
 
-    Parallel::Sync_cached(GH->PatL[lev], SL, Symmetry, sync_cache_rp_fine[lev]);
+    Parallel::Sync(GH->PatL[lev], SL, Symmetry);
   }
 }
 
@@ -6073,7 +6045,7 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB)
       Parallel::Restrict_bam(GH->PatL[lev - 1], GH->PatL[lev], SynchList_cor, SynchList_pre, GH->rsul[lev], Symmetry);
 #endif
 
-      Parallel::Sync_cached(GH->PatL[lev - 1], SynchList_pre, Symmetry, sync_cache_rp_coarse[lev]);
+      Parallel::Sync(GH->PatL[lev - 1], SynchList_pre, Symmetry);
 
 #if (RPB == 0)
       Ppc = GH->PatL[lev - 1];
@@ -6107,7 +6079,7 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB)
       Parallel::Restrict_bam(GH->PatL[lev - 1], GH->PatL[lev], SynchList_cor, StateList, GH->rsul[lev], Symmetry);
 #endif
 
-      Parallel::Sync_cached(GH->PatL[lev - 1], StateList, Symmetry, sync_cache_rp_coarse[lev]);
+      Parallel::Sync(GH->PatL[lev - 1], StateList, Symmetry);
 
 #if (RPB == 0)
       Ppc = GH->PatL[lev - 1];
@@ -6131,7 +6103,7 @@ void bssn_class::RestrictProlong(int lev, int YN, bool BB)
 #endif
     }
 
-    Parallel::Sync_cached(GH->PatL[lev], SynchList_cor, Symmetry, sync_cache_rp_fine[lev]);
+    Parallel::Sync(GH->PatL[lev], SynchList_cor, Symmetry);
   }
 }
 
@@ -6214,10 +6186,10 @@ void bssn_class::ProlongRestrict(int lev, int YN, bool BB)
 #else
       Parallel::Restrict_after(GH->PatL[lev - 1], GH->PatL[lev], SynchList_cor, StateList, Symmetry);
 #endif
-      Parallel::Sync_cached(GH->PatL[lev - 1], StateList, Symmetry, sync_cache_rp_coarse[lev]);
+      Parallel::Sync(GH->PatL[lev - 1], StateList, Symmetry);
     }
 
-    Parallel::Sync_cached(GH->PatL[lev], SynchList_cor, Symmetry, sync_cache_rp_fine[lev]);
+    Parallel::Sync(GH->PatL[lev], SynchList_cor, Symmetry);
   }
 }
 #undef MIXOUTB

AMSS_NCKU_source/bssn_class.h

@@ -19,11 +19,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "macrodef.h"
 #include "cgh.h"
@@ -130,11 +126,6 @@ public:
        MyList<var> *OldStateList, *DumpList;
        MyList<var> *ConstraintList;
 
-       Parallel::SyncCache *sync_cache_pre;  // per-level cache for predictor sync
-       Parallel::SyncCache *sync_cache_cor;  // per-level cache for corrector sync
-       Parallel::SyncCache *sync_cache_rp_coarse;  // RestrictProlong sync on PatL[lev-1]
-       Parallel::SyncCache *sync_cache_rp_fine;    // RestrictProlong sync on PatL[lev]
-
        monitor *ErrorMonitor, *Psi4Monitor, *BHMonitor, *MAPMonitor;
        monitor *ConVMonitor;
        surface_integral *Waveshell;

AMSS_NCKU_source/bssn_gpu_class.h

@@ -19,11 +19,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "macrodef.h"
 #include "cgh.h"

AMSS_NCKU_source/bssn_rhs.f90

@@ -83,6 +83,10 @@
   real*8, dimension(ex(1),ex(2),ex(3)) :: gupxx,gupxy,gupxz
   real*8, dimension(ex(1),ex(2),ex(3)) :: gupyy,gupyz,gupzz
 
+! shared work arrays (memory pool) to avoid repeated allocation in subroutines
+  real*8, dimension(-1:ex(1),-1:ex(2),-1:ex(3)) :: fh_work2  ! for fderivs_fh/fdderivs_fh (ghost_width=2)
+  real*8, dimension(-2:ex(1),-2:ex(2),-2:ex(3)) :: fh_work3  ! for kodis_fh/lopsided_fh (ghost_width=3)
+
   real*8,dimension(3) ::SSS,AAS,ASA,SAA,ASS,SAS,SSA
   real*8            :: dX, dY, dZ, PI
   real*8, parameter :: ZEO = 0.d0,ONE = 1.D0, TWO = 2.D0, FOUR = 4.D0
@@ -151,22 +155,22 @@
   gyy = dyy + ONE
   gzz = dzz + ONE
 
-  call fderivs(ex,betax,betaxx,betaxy,betaxz,X,Y,Z,ANTI, SYM, SYM,Symmetry,Lev)
-  call fderivs(ex,betay,betayx,betayy,betayz,X,Y,Z, SYM,ANTI, SYM,Symmetry,Lev)
-  call fderivs(ex,betaz,betazx,betazy,betazz,X,Y,Z, SYM, SYM,ANTI,Symmetry,Lev)
+  call fderivs_fh(ex,betax,betaxx,betaxy,betaxz,X,Y,Z,ANTI, SYM, SYM,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,betay,betayx,betayy,betayz,X,Y,Z, SYM,ANTI, SYM,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,betaz,betazx,betazy,betazz,X,Y,Z, SYM, SYM,ANTI,Symmetry,Lev,fh_work2)
   
   div_beta = betaxx + betayy + betazz
  
-  call fderivs(ex,chi,chix,chiy,chiz,X,Y,Z,SYM,SYM,SYM,symmetry,Lev)
+  call fderivs_fh(ex,chi,chix,chiy,chiz,X,Y,Z,SYM,SYM,SYM,symmetry,Lev,fh_work2)
 
   chi_rhs = F2o3 *chin1*( alpn1 * trK - div_beta ) !rhs for chi
 
-  call fderivs(ex,dxx,gxxx,gxxy,gxxz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,Lev)
-  call fderivs(ex,gxy,gxyx,gxyy,gxyz,X,Y,Z,ANTI,ANTI,SYM ,Symmetry,Lev)
-  call fderivs(ex,gxz,gxzx,gxzy,gxzz,X,Y,Z,ANTI,SYM ,ANTI,Symmetry,Lev)
-  call fderivs(ex,dyy,gyyx,gyyy,gyyz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,Lev)
-  call fderivs(ex,gyz,gyzx,gyzy,gyzz,X,Y,Z,SYM ,ANTI,ANTI,Symmetry,Lev)
-  call fderivs(ex,dzz,gzzx,gzzy,gzzz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,Lev)
+  call fderivs_fh(ex,dxx,gxxx,gxxy,gxxz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,gxy,gxyx,gxyy,gxyz,X,Y,Z,ANTI,ANTI,SYM ,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,gxz,gxzx,gxzy,gxzz,X,Y,Z,ANTI,SYM ,ANTI,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,dyy,gyyx,gyyy,gyyz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,gyz,gyzx,gyzy,gyzz,X,Y,Z,SYM ,ANTI,ANTI,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,dzz,gzzx,gzzy,gzzz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,Lev,fh_work2)
 
   gxx_rhs = - TWO * alpn1 * Axx    -  F2o3 * gxx * div_beta          + &
               TWO *(  gxx * betaxx +   gxy * betayx +   gxz * betazx)
@@ -284,8 +288,8 @@
           (gupyy * gupzz       + gupyz * gupyz)* Ayz
 
 ! Right hand side for Gam^i without shift terms...
-  call fderivs(ex,Lap,Lapx,Lapy,Lapz,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev)
-  call fderivs(ex,trK,Kx,Ky,Kz,X,Y,Z,SYM,SYM,SYM,symmetry,Lev)
+  call fderivs_fh(ex,Lap,Lapx,Lapy,Lapz,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,trK,Kx,Ky,Kz,X,Y,Z,SYM,SYM,SYM,symmetry,Lev,fh_work2)
 
    Gamx_rhs = - TWO * (   Lapx * Rxx +   Lapy * Rxy +   Lapz * Rxz ) + &
         TWO * alpn1 * (                                                &
@@ -314,12 +318,12 @@
                         Gamzxx * Rxx + Gamzyy * Ryy + Gamzzz * Rzz   + &
                 TWO * ( Gamzxy * Rxy + Gamzxz * Rxz + Gamzyz * Ryz ) )
 
-  call fdderivs(ex,betax,gxxx,gxyx,gxzx,gyyx,gyzx,gzzx,&
-                X,Y,Z,ANTI,SYM, SYM ,Symmetry,Lev)
-  call fdderivs(ex,betay,gxxy,gxyy,gxzy,gyyy,gyzy,gzzy,&
-                X,Y,Z,SYM ,ANTI,SYM ,Symmetry,Lev)
-  call fdderivs(ex,betaz,gxxz,gxyz,gxzz,gyyz,gyzz,gzzz,&
-                X,Y,Z,SYM ,SYM, ANTI,Symmetry,Lev)
+  call fdderivs_fh(ex,betax,gxxx,gxyx,gxzx,gyyx,gyzx,gzzx,&
+                X,Y,Z,ANTI,SYM, SYM ,Symmetry,Lev,fh_work2)
+  call fdderivs_fh(ex,betay,gxxy,gxyy,gxzy,gyyy,gyzy,gzzy,&
+                X,Y,Z,SYM ,ANTI,SYM ,Symmetry,Lev,fh_work2)
+  call fdderivs_fh(ex,betaz,gxxz,gxyz,gxzz,gyyz,gyzz,gzzz,&
+                X,Y,Z,SYM ,SYM, ANTI,Symmetry,Lev,fh_work2)
 
   fxx = gxxx + gxyy + gxzz
   fxy = gxyx + gyyy + gyzz
@@ -332,9 +336,9 @@
   Gamza =       gupxx * Gamzxx + gupyy * Gamzyy + gupzz * Gamzzz + &
           TWO*( gupxy * Gamzxy + gupxz * Gamzxz + gupyz * Gamzyz )
 
-  call fderivs(ex,Gamx,Gamxx,Gamxy,Gamxz,X,Y,Z,ANTI,SYM ,SYM ,Symmetry,Lev)
-  call fderivs(ex,Gamy,Gamyx,Gamyy,Gamyz,X,Y,Z,SYM ,ANTI,SYM ,Symmetry,Lev)
-  call fderivs(ex,Gamz,Gamzx,Gamzy,Gamzz,X,Y,Z,SYM ,SYM ,ANTI,Symmetry,Lev)
+  call fderivs_fh(ex,Gamx,Gamxx,Gamxy,Gamxz,X,Y,Z,ANTI,SYM ,SYM ,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,Gamy,Gamyx,Gamyy,Gamyz,X,Y,Z,SYM ,ANTI,SYM ,Symmetry,Lev,fh_work2)
+  call fderivs_fh(ex,Gamz,Gamzx,Gamzy,Gamzz,X,Y,Z,SYM ,SYM ,ANTI,Symmetry,Lev,fh_work2)
 
   Gamx_rhs =               Gamx_rhs +  F2o3 *  Gamxa * div_beta        - &
                      Gamxa * betaxx - Gamya * betaxy - Gamza * betaxz  + &
@@ -377,27 +381,27 @@
   gzzz = gxz * Gamxzz + gyz * Gamyzz + gzz * Gamzzz
 
 !compute Ricci tensor for tilted metric
-   call fdderivs(ex,dxx,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM ,SYM ,SYM ,symmetry,Lev)
+   call fdderivs_fh(ex,dxx,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM ,SYM ,SYM ,symmetry,Lev,fh_work2)
    Rxx =   gupxx * fxx + gupyy * fyy + gupzz * fzz + &
          ( gupxy * fxy + gupxz * fxz + gupyz * fyz ) * TWO
 
-   call fdderivs(ex,dyy,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM ,SYM ,SYM ,symmetry,Lev)
+   call fdderivs_fh(ex,dyy,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM ,SYM ,SYM ,symmetry,Lev,fh_work2)
    Ryy =   gupxx * fxx + gupyy * fyy + gupzz * fzz + &
          ( gupxy * fxy + gupxz * fxz + gupyz * fyz ) * TWO
 
-   call fdderivs(ex,dzz,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM ,SYM ,SYM ,symmetry,Lev)
+   call fdderivs_fh(ex,dzz,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM ,SYM ,SYM ,symmetry,Lev,fh_work2)
    Rzz =   gupxx * fxx + gupyy * fyy + gupzz * fzz + &
          ( gupxy * fxy + gupxz * fxz + gupyz * fyz ) * TWO
 
-   call fdderivs(ex,gxy,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,ANTI, ANTI,SYM ,symmetry,Lev)
+   call fdderivs_fh(ex,gxy,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,ANTI, ANTI,SYM ,symmetry,Lev,fh_work2)
    Rxy =   gupxx * fxx + gupyy * fyy + gupzz * fzz + &
          ( gupxy * fxy + gupxz * fxz + gupyz * fyz ) * TWO
 
-   call fdderivs(ex,gxz,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,ANTI ,SYM ,ANTI,symmetry,Lev)
+   call fdderivs_fh(ex,gxz,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,ANTI ,SYM ,ANTI,symmetry,Lev,fh_work2)
    Rxz =   gupxx * fxx + gupyy * fyy + gupzz * fzz + &
          ( gupxy * fxy + gupxz * fxz + gupyz * fyz ) * TWO
 
-   call fdderivs(ex,gyz,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM ,ANTI ,ANTI,symmetry,Lev)
+   call fdderivs_fh(ex,gyz,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM ,ANTI ,ANTI,symmetry,Lev,fh_work2)
    Ryz =   gupxx * fxx + gupyy * fyy + gupzz * fzz + &
          ( gupxy * fxy + gupxz * fxz + gupyz * fyz ) * TWO
 
@@ -602,7 +606,7 @@
             Gamxzz * gxzy + Gamyzz * gyzy + Gamzzz * gzzy  + &
             Gamxyz * gzzx + Gamyyz * gzzy + Gamzyz * gzzz )
 !covariant second derivative of chi respect to tilted metric
-  call fdderivs(ex,chi,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev)
+  call fdderivs_fh(ex,chi,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev,fh_work2)
 
   fxx = fxx - Gamxxx * chix - Gamyxx * chiy - Gamzxx * chiz
   fxy = fxy - Gamxxy * chix - Gamyxy * chiy - Gamzxy * chiz
@@ -628,8 +632,8 @@
   Ryz = Ryz + (fyz - chiy*chiz/chin1/TWO + gyz * f)/chin1/TWO
 
 ! covariant second derivatives of the lapse respect to physical metric
-  call fdderivs(ex,Lap,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z, &
-                SYM,SYM,SYM,symmetry,Lev)
+  call fdderivs_fh(ex,Lap,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z, &
+                SYM,SYM,SYM,symmetry,Lev,fh_work2)
 
   gxxx = (gupxx * chix + gupxy * chiy + gupxz * chiz)/chin1
   gxxy = (gupxy * chix + gupyy * chiy + gupyz * chiz)/chin1
@@ -812,7 +816,7 @@
   betay_rhs = FF*dtSfy
   betaz_rhs = FF*dtSfz
 
-  call fderivs(ex,chi,dtSfx_rhs,dtSfy_rhs,dtSfz_rhs,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev)
+  call fderivs_fh(ex,chi,dtSfx_rhs,dtSfy_rhs,dtSfz_rhs,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev,fh_work2)
   reta = gupxx * dtSfx_rhs * dtSfx_rhs + gupyy * dtSfy_rhs * dtSfy_rhs + gupzz * dtSfz_rhs * dtSfz_rhs + &
        TWO * (gupxy * dtSfx_rhs * dtSfy_rhs + gupxz * dtSfx_rhs * dtSfz_rhs + gupyz * dtSfy_rhs * dtSfz_rhs)
   reta = 1.31d0/2*dsqrt(reta/chin1)/(1-dsqrt(chin1))**2
@@ -824,7 +828,7 @@
   betay_rhs = FF*dtSfy
   betaz_rhs = FF*dtSfz
 
-  call fderivs(ex,chi,dtSfx_rhs,dtSfy_rhs,dtSfz_rhs,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev)
+  call fderivs_fh(ex,chi,dtSfx_rhs,dtSfy_rhs,dtSfz_rhs,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev,fh_work2)
   reta = gupxx * dtSfx_rhs * dtSfx_rhs + gupyy * dtSfy_rhs * dtSfy_rhs + gupzz * dtSfz_rhs * dtSfz_rhs + &
        TWO * (gupxy * dtSfx_rhs * dtSfy_rhs + gupxz * dtSfx_rhs * dtSfz_rhs + gupyz * dtSfy_rhs * dtSfz_rhs)
   reta = 1.31d0/2*dsqrt(reta/chin1)/(1-chin1)**2
@@ -832,7 +836,7 @@
   dtSfy_rhs = Gamy_rhs - reta*dtSfy
   dtSfz_rhs = Gamz_rhs - reta*dtSfz
 #elif (GAUGE == 4)
-  call fderivs(ex,chi,dtSfx_rhs,dtSfy_rhs,dtSfz_rhs,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev)
+  call fderivs_fh(ex,chi,dtSfx_rhs,dtSfy_rhs,dtSfz_rhs,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev,fh_work2)
   reta = gupxx * dtSfx_rhs * dtSfx_rhs + gupyy * dtSfy_rhs * dtSfy_rhs + gupzz * dtSfz_rhs * dtSfz_rhs + &
        TWO * (gupxy * dtSfx_rhs * dtSfy_rhs + gupxz * dtSfx_rhs * dtSfz_rhs + gupyz * dtSfy_rhs * dtSfz_rhs)
   reta = 1.31d0/2*dsqrt(reta/chin1)/(1-dsqrt(chin1))**2
@@ -844,7 +848,7 @@
   dtSfy_rhs = ZEO
   dtSfz_rhs = ZEO
 #elif (GAUGE == 5)
-  call fderivs(ex,chi,dtSfx_rhs,dtSfy_rhs,dtSfz_rhs,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev)
+  call fderivs_fh(ex,chi,dtSfx_rhs,dtSfy_rhs,dtSfz_rhs,X,Y,Z,SYM,SYM,SYM,Symmetry,Lev,fh_work2)
   reta = gupxx * dtSfx_rhs * dtSfx_rhs + gupyy * dtSfy_rhs * dtSfy_rhs + gupzz * dtSfz_rhs * dtSfz_rhs + &
        TWO * (gupxy * dtSfx_rhs * dtSfy_rhs + gupxz * dtSfx_rhs * dtSfz_rhs + gupyz * dtSfy_rhs * dtSfz_rhs)
   reta = 1.31d0/2*dsqrt(reta/chin1)/(1-chin1)**2
@@ -947,51 +951,51 @@
 
 !!!!!!!!!advection term part
 
-  call lopsided(ex,X,Y,Z,gxx,gxx_rhs,betax,betay,betaz,Symmetry,SSS)
-  call lopsided(ex,X,Y,Z,gxy,gxy_rhs,betax,betay,betaz,Symmetry,AAS)
-  call lopsided(ex,X,Y,Z,gxz,gxz_rhs,betax,betay,betaz,Symmetry,ASA)
-  call lopsided(ex,X,Y,Z,gyy,gyy_rhs,betax,betay,betaz,Symmetry,SSS)
-  call lopsided(ex,X,Y,Z,gyz,gyz_rhs,betax,betay,betaz,Symmetry,SAA)
-  call lopsided(ex,X,Y,Z,gzz,gzz_rhs,betax,betay,betaz,Symmetry,SSS)
+  call lopsided_fh(ex,X,Y,Z,gxx,gxx_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,gxy,gxy_rhs,betax,betay,betaz,Symmetry,AAS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,gxz,gxz_rhs,betax,betay,betaz,Symmetry,ASA,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,gyy,gyy_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,gyz,gyz_rhs,betax,betay,betaz,Symmetry,SAA,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,gzz,gzz_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
 
-  call lopsided(ex,X,Y,Z,Axx,Axx_rhs,betax,betay,betaz,Symmetry,SSS)
-  call lopsided(ex,X,Y,Z,Axy,Axy_rhs,betax,betay,betaz,Symmetry,AAS)
-  call lopsided(ex,X,Y,Z,Axz,Axz_rhs,betax,betay,betaz,Symmetry,ASA)
-  call lopsided(ex,X,Y,Z,Ayy,Ayy_rhs,betax,betay,betaz,Symmetry,SSS)
-  call lopsided(ex,X,Y,Z,Ayz,Ayz_rhs,betax,betay,betaz,Symmetry,SAA)
-  call lopsided(ex,X,Y,Z,Azz,Azz_rhs,betax,betay,betaz,Symmetry,SSS)
+  call lopsided_fh(ex,X,Y,Z,Axx,Axx_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,Axy,Axy_rhs,betax,betay,betaz,Symmetry,AAS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,Axz,Axz_rhs,betax,betay,betaz,Symmetry,ASA,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,Ayy,Ayy_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,Ayz,Ayz_rhs,betax,betay,betaz,Symmetry,SAA,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,Azz,Azz_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
 
-  call lopsided(ex,X,Y,Z,chi,chi_rhs,betax,betay,betaz,Symmetry,SSS)
-  call lopsided(ex,X,Y,Z,trK,trK_rhs,betax,betay,betaz,Symmetry,SSS)
+  call lopsided_fh(ex,X,Y,Z,chi,chi_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,trK,trK_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
 
-  call lopsided(ex,X,Y,Z,Gamx,Gamx_rhs,betax,betay,betaz,Symmetry,ASS)
-  call lopsided(ex,X,Y,Z,Gamy,Gamy_rhs,betax,betay,betaz,Symmetry,SAS)
-  call lopsided(ex,X,Y,Z,Gamz,Gamz_rhs,betax,betay,betaz,Symmetry,SSA)
+  call lopsided_fh(ex,X,Y,Z,Gamx,Gamx_rhs,betax,betay,betaz,Symmetry,ASS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,Gamy,Gamy_rhs,betax,betay,betaz,Symmetry,SAS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,Gamz,Gamz_rhs,betax,betay,betaz,Symmetry,SSA,fh_work3)
 !!
-  call lopsided(ex,X,Y,Z,Lap,Lap_rhs,betax,betay,betaz,Symmetry,SSS)
+  call lopsided_fh(ex,X,Y,Z,Lap,Lap_rhs,betax,betay,betaz,Symmetry,SSS,fh_work3)
 
 #if (GAUGE == 0 || GAUGE == 1 || GAUGE == 2 || GAUGE == 3 || GAUGE == 4 || GAUGE == 5 || GAUGE == 6 || GAUGE == 7)
-  call lopsided(ex,X,Y,Z,betax,betax_rhs,betax,betay,betaz,Symmetry,ASS)
-  call lopsided(ex,X,Y,Z,betay,betay_rhs,betax,betay,betaz,Symmetry,SAS)
-  call lopsided(ex,X,Y,Z,betaz,betaz_rhs,betax,betay,betaz,Symmetry,SSA)
+  call lopsided_fh(ex,X,Y,Z,betax,betax_rhs,betax,betay,betaz,Symmetry,ASS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,betay,betay_rhs,betax,betay,betaz,Symmetry,SAS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,betaz,betaz_rhs,betax,betay,betaz,Symmetry,SSA,fh_work3)
 #endif
 
 #if (GAUGE == 0 || GAUGE == 2 || GAUGE == 3 || GAUGE == 6 || GAUGE == 7)
-  call lopsided(ex,X,Y,Z,dtSfx,dtSfx_rhs,betax,betay,betaz,Symmetry,ASS)
-  call lopsided(ex,X,Y,Z,dtSfy,dtSfy_rhs,betax,betay,betaz,Symmetry,SAS)
-  call lopsided(ex,X,Y,Z,dtSfz,dtSfz_rhs,betax,betay,betaz,Symmetry,SSA)
+  call lopsided_fh(ex,X,Y,Z,dtSfx,dtSfx_rhs,betax,betay,betaz,Symmetry,ASS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,dtSfy,dtSfy_rhs,betax,betay,betaz,Symmetry,SAS,fh_work3)
+  call lopsided_fh(ex,X,Y,Z,dtSfz,dtSfz_rhs,betax,betay,betaz,Symmetry,SSA,fh_work3)
 #endif
 
   if(eps>0)then 
 ! usual Kreiss-Oliger dissipation      
-  call kodis(ex,X,Y,Z,chi,chi_rhs,SSS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,trK,trK_rhs,SSS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,dxx,gxx_rhs,SSS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,gxy,gxy_rhs,AAS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,gxz,gxz_rhs,ASA,Symmetry,eps)
-  call kodis(ex,X,Y,Z,dyy,gyy_rhs,SSS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,gyz,gyz_rhs,SAA,Symmetry,eps)
-  call kodis(ex,X,Y,Z,dzz,gzz_rhs,SSS,Symmetry,eps)
+  call kodis_fh(ex,X,Y,Z,chi,chi_rhs,SSS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,trK,trK_rhs,SSS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,dxx,gxx_rhs,SSS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,gxy,gxy_rhs,AAS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,gxz,gxz_rhs,ASA,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,dyy,gyy_rhs,SSS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,gyz,gyz_rhs,SAA,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,dzz,gzz_rhs,SSS,Symmetry,eps,fh_work3)
 #if 0
 #define i 42
 #define j 40
@@ -1005,7 +1009,7 @@ endif
 #undef k
 !!stop
 #endif
-  call kodis(ex,X,Y,Z,Axx,Axx_rhs,SSS,Symmetry,eps)
+  call kodis_fh(ex,X,Y,Z,Axx,Axx_rhs,SSS,Symmetry,eps,fh_work3)
 #if 0
 #define i 42
 #define j 40
@@ -1019,25 +1023,25 @@ endif
 #undef k
 !!stop
 #endif
-  call kodis(ex,X,Y,Z,Axy,Axy_rhs,AAS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,Axz,Axz_rhs,ASA,Symmetry,eps)
-  call kodis(ex,X,Y,Z,Ayy,Ayy_rhs,SSS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,Ayz,Ayz_rhs,SAA,Symmetry,eps)
-  call kodis(ex,X,Y,Z,Azz,Azz_rhs,SSS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,Gamx,Gamx_rhs,ASS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,Gamy,Gamy_rhs,SAS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,Gamz,Gamz_rhs,SSA,Symmetry,eps)
+  call kodis_fh(ex,X,Y,Z,Axy,Axy_rhs,AAS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,Axz,Axz_rhs,ASA,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,Ayy,Ayy_rhs,SSS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,Ayz,Ayz_rhs,SAA,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,Azz,Azz_rhs,SSS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,Gamx,Gamx_rhs,ASS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,Gamy,Gamy_rhs,SAS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,Gamz,Gamz_rhs,SSA,Symmetry,eps,fh_work3)
 
 #if 1 
 !! bam does not apply dissipation on gauge variables
-  call kodis(ex,X,Y,Z,Lap,Lap_rhs,SSS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,betax,betax_rhs,ASS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,betay,betay_rhs,SAS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,betaz,betaz_rhs,SSA,Symmetry,eps)
+  call kodis_fh(ex,X,Y,Z,Lap,Lap_rhs,SSS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,betax,betax_rhs,ASS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,betay,betay_rhs,SAS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,betaz,betaz_rhs,SSA,Symmetry,eps,fh_work3)
 #if (GAUGE == 0 || GAUGE == 2 || GAUGE == 3 || GAUGE == 6 || GAUGE == 7)
-  call kodis(ex,X,Y,Z,dtSfx,dtSfx_rhs,ASS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,dtSfy,dtSfy_rhs,SAS,Symmetry,eps)
-  call kodis(ex,X,Y,Z,dtSfz,dtSfz_rhs,SSA,Symmetry,eps)
+  call kodis_fh(ex,X,Y,Z,dtSfx,dtSfx_rhs,ASS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,dtSfy,dtSfy_rhs,SAS,Symmetry,eps,fh_work3)
+  call kodis_fh(ex,X,Y,Z,dtSfz,dtSfz_rhs,SSA,Symmetry,eps,fh_work3)
 #endif
 #endif
 
@@ -1078,12 +1082,12 @@ endif
 
 ! mov_Res_j = gupkj*(-1/chi d_k chi*A_ij + D_k A_ij) - 2/3 d_j trK - 8 PI s_j where D respect to physical metric
 ! store D_i A_jk - 1/chi d_i chi*A_jk in gjk_i
-  call fderivs(ex,Axx,gxxx,gxxy,gxxz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,0)
-  call fderivs(ex,Axy,gxyx,gxyy,gxyz,X,Y,Z,ANTI,ANTI,SYM ,Symmetry,0)
-  call fderivs(ex,Axz,gxzx,gxzy,gxzz,X,Y,Z,ANTI,SYM ,ANTI,Symmetry,0)
-  call fderivs(ex,Ayy,gyyx,gyyy,gyyz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,0)
-  call fderivs(ex,Ayz,gyzx,gyzy,gyzz,X,Y,Z,SYM ,ANTI,ANTI,Symmetry,0)
-  call fderivs(ex,Azz,gzzx,gzzy,gzzz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,0)
+  call fderivs_fh(ex,Axx,gxxx,gxxy,gxxz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,0,fh_work2)
+  call fderivs_fh(ex,Axy,gxyx,gxyy,gxyz,X,Y,Z,ANTI,ANTI,SYM ,Symmetry,0,fh_work2)
+  call fderivs_fh(ex,Axz,gxzx,gxzy,gxzz,X,Y,Z,ANTI,SYM ,ANTI,Symmetry,0,fh_work2)
+  call fderivs_fh(ex,Ayy,gyyx,gyyy,gyyz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,0,fh_work2)
+  call fderivs_fh(ex,Ayz,gyzx,gyzy,gyzz,X,Y,Z,SYM ,ANTI,ANTI,Symmetry,0,fh_work2)
+  call fderivs_fh(ex,Azz,gzzx,gzzy,gzzz,X,Y,Z,SYM ,SYM ,SYM ,Symmetry,0,fh_work2)
 
   gxxx = gxxx - (  Gamxxx * Axx + Gamyxx * Axy + Gamzxx * Axz &
                  + Gamxxx * Axx + Gamyxx * Axy + Gamzxx * Axz) - chix*Axx/chin1

AMSS_NCKU_source/cgh.C

@@ -20,11 +20,7 @@ using namespace std;
 #include <map.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "macrodef.h"
 #include "misc.h"

AMSS_NCKU_source/cgh.h

@@ -2,11 +2,7 @@
 #ifndef CGH_H
 #define CGH_H
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 #include "MyList.h"
 #include "MPatch.h"
 #include "macrodef.h"

AMSS_NCKU_source/checkpoint.h

@@ -19,11 +19,7 @@ using namespace std;
 #include <time.h>
 #include <stdlib.h>
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "var.h"
 #include "MyList.h"

AMSS_NCKU_source/diff_new.f90

@@ -69,7 +69,6 @@
   fy = ZEO
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -152,7 +151,6 @@
 
   fx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -220,7 +218,6 @@
 
   fy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -285,7 +282,6 @@
 
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -375,7 +371,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -474,7 +469,6 @@
 
   fxx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -537,7 +531,6 @@
 
   fyy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -601,7 +594,6 @@
 
   fzz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -665,7 +657,6 @@
 
   fxy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -728,7 +719,6 @@
 
   fxz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -790,7 +780,6 @@
 
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -877,7 +866,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1009,7 +997,6 @@
   fy = ZEO
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1116,6 +1103,103 @@
 
   end subroutine fderivs
 !-----------------------------------------------------------------------------
+! fderivs variant: reuses caller-provided fh work array (memory pool)
+!-----------------------------------------------------------------------------
+  subroutine fderivs_fh(ex,f,fx,fy,fz,X,Y,Z,SYM1,SYM2,SYM3, &
+                         symmetry,onoff,fh)
+  implicit none
+
+  integer,                               intent(in ):: ex(1:3),symmetry,onoff
+  real*8,  dimension(ex(1),ex(2),ex(3)), intent(in ):: f
+  real*8,  dimension(ex(1),ex(2),ex(3)), intent(out):: fx,fy,fz
+  real*8,                                intent(in) :: X(ex(1)),Y(ex(2)),Z(ex(3))
+  real*8,                                intent(in ):: SYM1,SYM2,SYM3
+  real*8,dimension(-1:ex(1),-1:ex(2),-1:ex(3)),intent(inout):: fh
+
+  real*8 :: dX,dY,dZ
+  real*8, dimension(3) :: SoA
+  integer :: imin,jmin,kmin,imax,jmax,kmax,i,j,k
+  real*8 :: d12dx,d12dy,d12dz,d2dx,d2dy,d2dz
+  integer, parameter :: NO_SYMM = 0, EQ_SYMM = 1, OCTANT = 2
+  real*8,  parameter :: ZEO=0.d0,ONE=1.d0
+  real*8,  parameter :: TWO=2.d0,EIT=8.d0
+  real*8,  parameter :: F12=1.2d1
+
+  dX = X(2)-X(1)
+  dY = Y(2)-Y(1)
+  dZ = Z(2)-Z(1)
+
+  imax = ex(1)
+  jmax = ex(2)
+  kmax = ex(3)
+
+  imin = 1
+  jmin = 1
+  kmin = 1
+  if(Symmetry > NO_SYMM .and. dabs(Z(1)) < dZ) kmin = -1
+  if(Symmetry > EQ_SYMM .and. dabs(X(1)) < dX) imin = -1
+  if(Symmetry > EQ_SYMM .and. dabs(Y(1)) < dY) jmin = -1
+
+  SoA(1) = SYM1
+  SoA(2) = SYM2
+  SoA(3) = SYM3
+
+  call symmetry_bd(2,ex,f,fh,SoA)
+
+  d12dx = ONE/F12/dX
+  d12dy = ONE/F12/dY
+  d12dz = ONE/F12/dZ
+
+  d2dx = ONE/TWO/dX
+  d2dy = ONE/TWO/dY
+  d2dz = ONE/TWO/dZ
+
+  fx = ZEO
+  fy = ZEO
+  fz = ZEO
+
+  do k=1,ex(3)-1
+  do j=1,ex(2)-1
+  do i=1,ex(1)-1
+#if 0
+   if(i+2 <= imax .and. i-2 >= imin)then
+      fx(i,j,k)=d12dx*(fh(i-2,j,k)-EIT*fh(i-1,j,k)+EIT*fh(i+1,j,k)-fh(i+2,j,k))
+    elseif(i+1 <= imax .and. i-1 >= imin)then
+      fx(i,j,k)=d2dx*(-fh(i-1,j,k)+fh(i+1,j,k))
+    endif
+        if(j+2 <= jmax .and. j-2 >= jmin)then
+      fy(i,j,k)=d12dy*(fh(i,j-2,k)-EIT*fh(i,j-1,k)+EIT*fh(i,j+1,k)-fh(i,j+2,k))
+    elseif(j+1 <= jmax .and. j-1 >= jmin)then
+     fy(i,j,k)=d2dy*(-fh(i,j-1,k)+fh(i,j+1,k))
+    endif
+        if(k+2 <= kmax .and. k-2 >= kmin)then
+      fz(i,j,k)=d12dz*(fh(i,j,k-2)-EIT*fh(i,j,k-1)+EIT*fh(i,j,k+1)-fh(i,j,k+2))
+    elseif(k+1 <= kmax .and. k-1 >= kmin)then
+      fz(i,j,k)=d2dz*(-fh(i,j,k-1)+fh(i,j,k+1))
+    endif
+#else
+   if(i+2 <= imax .and. i-2 >= imin .and. &
+      j+2 <= jmax .and. j-2 >= jmin .and. &
+      k+2 <= kmax .and. k-2 >= kmin) then
+      fx(i,j,k)=d12dx*(fh(i-2,j,k)-EIT*fh(i-1,j,k)+EIT*fh(i+1,j,k)-fh(i+2,j,k))
+      fy(i,j,k)=d12dy*(fh(i,j-2,k)-EIT*fh(i,j-1,k)+EIT*fh(i,j+1,k)-fh(i,j+2,k))
+      fz(i,j,k)=d12dz*(fh(i,j,k-2)-EIT*fh(i,j,k-1)+EIT*fh(i,j,k+1)-fh(i,j,k+2))
+   elseif(i+1 <= imax .and. i-1 >= imin .and. &
+          j+1 <= jmax .and. j-1 >= jmin .and. &
+          k+1 <= kmax .and. k-1 >= kmin) then
+      fx(i,j,k)=d2dx*(-fh(i-1,j,k)+fh(i+1,j,k))
+      fy(i,j,k)=d2dy*(-fh(i,j-1,k)+fh(i,j+1,k))
+      fz(i,j,k)=d2dz*(-fh(i,j,k-1)+fh(i,j,k+1))
+   endif
+#endif
+  enddo
+  enddo
+  enddo
+
+  return
+
+  end subroutine fderivs_fh
+!-----------------------------------------------------------------------------
 !
 ! single derivatives dx
 !
@@ -1164,7 +1248,6 @@
 
   fx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1241,7 +1324,6 @@
 
   fy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1312,7 +1394,6 @@
 
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1417,7 +1498,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1593,7 +1673,6 @@
 
   fxx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1661,7 +1740,6 @@
 
   fyy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1731,7 +1809,6 @@
 
   fzz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1801,7 +1878,6 @@
 
   fxy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1872,7 +1948,6 @@
 
   fxz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1941,7 +2016,6 @@
 
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -1963,6 +2037,162 @@
 
   end subroutine fddyz
 
+!-----------------------------------------------------------------------------
+! fdderivs variant: reuses caller-provided fh work array (memory pool)
+!-----------------------------------------------------------------------------
+  subroutine fdderivs_fh(ex,f,fxx,fxy,fxz,fyy,fyz,fzz,X,Y,Z, &
+                          SYM1,SYM2,SYM3,symmetry,onoff,fh)
+  implicit none
+
+  integer,                             intent(in ):: ex(1:3),symmetry,onoff
+  real*8, dimension(ex(1),ex(2),ex(3)),intent(in ):: f
+  real*8, dimension(ex(1),ex(2),ex(3)),intent(out):: fxx,fxy,fxz,fyy,fyz,fzz
+  real*8,                              intent(in ):: X(ex(1)),Y(ex(2)),Z(ex(3))
+  real*8,                              intent(in ):: SYM1,SYM2,SYM3
+  real*8,dimension(-1:ex(1),-1:ex(2),-1:ex(3)),intent(inout):: fh
+
+  real*8 :: dX,dY,dZ
+  real*8, dimension(3) :: SoA
+  integer :: imin,jmin,kmin,imax,jmax,kmax,i,j,k
+  real*8  :: Sdxdx,Sdydy,Sdzdz,Fdxdx,Fdydy,Fdzdz
+  real*8  :: Sdxdy,Sdxdz,Sdydz,Fdxdy,Fdxdz,Fdydz
+  integer, parameter :: NO_SYMM = 0, EQ_SYMM = 1, OCTANT = 2
+  real*8, parameter :: ZEO=0.d0, ONE=1.d0, TWO=2.d0, F1o4=2.5d-1
+  real*8, parameter :: F8=8.d0, F16=1.6d1, F30=3.d1
+  real*8, parameter :: F1o12=ONE/1.2d1, F1o144=ONE/1.44d2
+
+  dX = X(2)-X(1)
+  dY = Y(2)-Y(1)
+  dZ = Z(2)-Z(1)
+
+  imax = ex(1)
+  jmax = ex(2)
+  kmax = ex(3)
+
+  imin = 1
+  jmin = 1
+  kmin = 1
+  if(Symmetry > NO_SYMM .and. dabs(Z(1)) < dZ) kmin = -1
+  if(Symmetry > EQ_SYMM .and. dabs(X(1)) < dX) imin = -1
+  if(Symmetry > EQ_SYMM .and. dabs(Y(1)) < dY) jmin = -1
+
+  SoA(1) = SYM1
+  SoA(2) = SYM2
+  SoA(3) = SYM3
+
+  call symmetry_bd(2,ex,f,fh,SoA)
+
+  Sdxdx =  ONE /( dX * dX )
+  Sdydy =  ONE /( dY * dY )
+  Sdzdz =  ONE /( dZ * dZ )
+
+  Fdxdx = F1o12 /( dX * dX )
+  Fdydy = F1o12 /( dY * dY )
+  Fdzdz = F1o12 /( dZ * dZ )
+
+  Sdxdy = F1o4 /( dX * dY )
+  Sdxdz = F1o4 /( dX * dZ )
+  Sdydz = F1o4 /( dY * dZ )
+
+  Fdxdy = F1o144 /( dX * dY )
+  Fdxdz = F1o144 /( dX * dZ )
+  Fdydz = F1o144 /( dY * dZ )
+
+  fxx = ZEO
+  fyy = ZEO
+  fzz = ZEO
+  fxy = ZEO
+  fxz = ZEO
+  fyz = ZEO
+
+  do k=1,ex(3)-1
+  do j=1,ex(2)-1
+  do i=1,ex(1)-1
+#if 0
+   if(i+2 <= imax .and. i-2 >= imin)then
+   fxx(i,j,k) = Fdxdx*(-fh(i-2,j,k)+F16*fh(i-1,j,k)-F30*fh(i,j,k) &
+                       -fh(i+2,j,k)+F16*fh(i+1,j,k)              )
+   elseif(i+1 <= imax .and. i-1 >= imin)then
+   fxx(i,j,k) = Sdxdx*(fh(i-1,j,k)-TWO*fh(i,j,k)+fh(i+1,j,k))
+   endif
+        if(j+2 <= jmax .and. j-2 >= jmin)then
+   fyy(i,j,k) = Fdydy*(-fh(i,j-2,k)+F16*fh(i,j-1,k)-F30*fh(i,j,k) &
+                       -fh(i,j+2,k)+F16*fh(i,j+1,k)              )
+   elseif(j+1 <= jmax .and. j-1 >= jmin)then
+   fyy(i,j,k) = Sdydy*(fh(i,j-1,k)-TWO*fh(i,j,k)+fh(i,j+1,k))
+   endif
+        if(k+2 <= kmax .and. k-2 >= kmin)then
+   fzz(i,j,k) = Fdzdz*(-fh(i,j,k-2)+F16*fh(i,j,k-1)-F30*fh(i,j,k) &
+                       -fh(i,j,k+2)+F16*fh(i,j,k+1)              )
+   elseif(k+1 <= kmax .and. k-1 >= kmin)then
+   fzz(i,j,k) = Sdzdz*(fh(i,j,k-1)-TWO*fh(i,j,k)+fh(i,j,k+1))
+   endif
+       if(i+2 <= imax .and. i-2 >= imin .and. j+2 <= jmax .and. j-2 >= jmin)then
+   fxy(i,j,k) = Fdxdy*(     (fh(i-2,j-2,k)-F8*fh(i-1,j-2,k)+F8*fh(i+1,j-2,k)-fh(i+2,j-2,k))  &
+                       -F8 *(fh(i-2,j-1,k)-F8*fh(i-1,j-1,k)+F8*fh(i+1,j-1,k)-fh(i+2,j-1,k))  &
+                       +F8 *(fh(i-2,j+1,k)-F8*fh(i-1,j+1,k)+F8*fh(i+1,j+1,k)-fh(i+2,j+1,k))  &
+                       -    (fh(i-2,j+2,k)-F8*fh(i-1,j+2,k)+F8*fh(i+1,j+2,k)-fh(i+2,j+2,k)))
+   elseif(i+1 <= imax .and. i-1 >= imin .and. j+1 <= jmax .and. j-1 >= jmin)then
+   fxy(i,j,k) = Sdxdy*(fh(i-1,j-1,k)-fh(i+1,j-1,k)-fh(i-1,j+1,k)+fh(i+1,j+1,k))
+   endif
+       if(i+2 <= imax .and. i-2 >= imin .and. k+2 <= kmax .and. k-2 >= kmin)then
+   fxz(i,j,k) = Fdxdz*(     (fh(i-2,j,k-2)-F8*fh(i-1,j,k-2)+F8*fh(i+1,j,k-2)-fh(i+2,j,k-2))  &
+                       -F8 *(fh(i-2,j,k-1)-F8*fh(i-1,j,k-1)+F8*fh(i+1,j,k-1)-fh(i+2,j,k-1))  &
+                       +F8 *(fh(i-2,j,k+1)-F8*fh(i-1,j,k+1)+F8*fh(i+1,j,k+1)-fh(i+2,j,k+1))  &
+                       -    (fh(i-2,j,k+2)-F8*fh(i-1,j,k+2)+F8*fh(i+1,j,k+2)-fh(i+2,j,k+2)))
+   elseif(i+1 <= imax .and. i-1 >= imin .and. k+1 <= kmax .and. k-1 >= kmin)then
+   fxz(i,j,k) = Sdxdz*(fh(i-1,j,k-1)-fh(i+1,j,k-1)-fh(i-1,j,k+1)+fh(i+1,j,k+1))
+   endif
+       if(j+2 <= jmax .and. j-2 >= jmin .and. k+2 <= kmax .and. k-2 >= kmin)then
+   fyz(i,j,k) = Fdydz*(     (fh(i,j-2,k-2)-F8*fh(i,j-1,k-2)+F8*fh(i,j+1,k-2)-fh(i,j+2,k-2))  &
+                       -F8 *(fh(i,j-2,k-1)-F8*fh(i,j-1,k-1)+F8*fh(i,j+1,k-1)-fh(i,j+2,k-1))  &
+                       +F8 *(fh(i,j-2,k+1)-F8*fh(i,j-1,k+1)+F8*fh(i,j+1,k+1)-fh(i,j+2,k+1))  &
+                       -    (fh(i,j-2,k+2)-F8*fh(i,j-1,k+2)+F8*fh(i,j+1,k+2)-fh(i,j+2,k+2)))
+   elseif(j+1 <= jmax .and. j-1 >= jmin .and. k+1 <= kmax .and. k-1 >= kmin)then
+   fyz(i,j,k) = Sdydz*(fh(i,j-1,k-1)-fh(i,j+1,k-1)-fh(i,j-1,k+1)+fh(i,j+1,k+1))
+   endif
+#else
+! for bam comparison
+   if(i+2 <= imax .and. i-2 >= imin .and. &
+      j+2 <= jmax .and. j-2 >= jmin .and. &
+      k+2 <= kmax .and. k-2 >= kmin) then
+   fxx(i,j,k) = Fdxdx*(-fh(i-2,j,k)+F16*fh(i-1,j,k)-F30*fh(i,j,k) &
+                       -fh(i+2,j,k)+F16*fh(i+1,j,k)              )
+   fyy(i,j,k) = Fdydy*(-fh(i,j-2,k)+F16*fh(i,j-1,k)-F30*fh(i,j,k) &
+                       -fh(i,j+2,k)+F16*fh(i,j+1,k)              )
+   fzz(i,j,k) = Fdzdz*(-fh(i,j,k-2)+F16*fh(i,j,k-1)-F30*fh(i,j,k) &
+                       -fh(i,j,k+2)+F16*fh(i,j,k+1)              )
+   fxy(i,j,k) = Fdxdy*(     (fh(i-2,j-2,k)-F8*fh(i-1,j-2,k)+F8*fh(i+1,j-2,k)-fh(i+2,j-2,k))  &
+                       -F8 *(fh(i-2,j-1,k)-F8*fh(i-1,j-1,k)+F8*fh(i+1,j-1,k)-fh(i+2,j-1,k))  &
+                       +F8 *(fh(i-2,j+1,k)-F8*fh(i-1,j+1,k)+F8*fh(i+1,j+1,k)-fh(i+2,j+1,k))  &
+                       -    (fh(i-2,j+2,k)-F8*fh(i-1,j+2,k)+F8*fh(i+1,j+2,k)-fh(i+2,j+2,k)))
+   fxz(i,j,k) = Fdxdz*(     (fh(i-2,j,k-2)-F8*fh(i-1,j,k-2)+F8*fh(i+1,j,k-2)-fh(i+2,j,k-2))  &
+                       -F8 *(fh(i-2,j,k-1)-F8*fh(i-1,j,k-1)+F8*fh(i+1,j,k-1)-fh(i+2,j,k-1))  &
+                       +F8 *(fh(i-2,j,k+1)-F8*fh(i-1,j,k+1)+F8*fh(i+1,j,k+1)-fh(i+2,j,k+1))  &
+                       -    (fh(i-2,j,k+2)-F8*fh(i-1,j,k+2)+F8*fh(i+1,j,k+2)-fh(i+2,j,k+2)))
+   fyz(i,j,k) = Fdydz*(     (fh(i,j-2,k-2)-F8*fh(i,j-1,k-2)+F8*fh(i,j+1,k-2)-fh(i,j+2,k-2))  &
+                       -F8 *(fh(i,j-2,k-1)-F8*fh(i,j-1,k-1)+F8*fh(i,j+1,k-1)-fh(i,j+2,k-1))  &
+                       +F8 *(fh(i,j-2,k+1)-F8*fh(i,j-1,k+1)+F8*fh(i,j+1,k+1)-fh(i,j+2,k+1))  &
+                       -    (fh(i,j-2,k+2)-F8*fh(i,j-1,k+2)+F8*fh(i,j+1,k+2)-fh(i,j+2,k+2)))
+   elseif(i+1 <= imax .and. i-1 >= imin .and. &
+          j+1 <= jmax .and. j-1 >= jmin .and. &
+          k+1 <= kmax .and. k-1 >= kmin) then
+   fxx(i,j,k) = Sdxdx*(fh(i-1,j,k)-TWO*fh(i,j,k)+fh(i+1,j,k))
+   fyy(i,j,k) = Sdydy*(fh(i,j-1,k)-TWO*fh(i,j,k)+fh(i,j+1,k))
+   fzz(i,j,k) = Sdzdz*(fh(i,j,k-1)-TWO*fh(i,j,k)+fh(i,j,k+1))
+   fxy(i,j,k) = Sdxdy*(fh(i-1,j-1,k)-fh(i+1,j-1,k)-fh(i-1,j+1,k)+fh(i+1,j+1,k))
+   fxz(i,j,k) = Sdxdz*(fh(i-1,j,k-1)-fh(i+1,j,k-1)-fh(i-1,j,k+1)+fh(i+1,j,k+1))
+   fyz(i,j,k) = Sdydz*(fh(i,j-1,k-1)-fh(i,j+1,k-1)-fh(i,j-1,k+1)+fh(i,j+1,k+1))
+   endif
+#endif
+   enddo
+   enddo
+   enddo
+
+  return
+
+  end subroutine fdderivs_fh
+
 #elif (ghost_width == 4)
 ! sixth order code
 
@@ -2034,7 +2264,6 @@
   fy = ZEO
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2151,7 +2380,6 @@
 
   fx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2237,7 +2465,6 @@
 
   fy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2314,7 +2541,6 @@
 
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2433,7 +2659,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2621,7 +2846,6 @@
 
   fxx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2694,7 +2918,6 @@
 
   fyy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2770,7 +2993,6 @@
 
   fzz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2846,7 +3068,6 @@
 
   fxy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -2927,7 +3148,6 @@
 
   fxz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -3006,7 +3226,6 @@
 
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -3114,7 +3333,6 @@
   fy = ZEO
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -3251,7 +3469,6 @@
 
   fx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -3347,7 +3564,6 @@
 
   fy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -3432,7 +3648,6 @@
 
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -3568,7 +3783,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -3841,7 +4055,6 @@
 
   fxx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -3923,7 +4136,6 @@
 
   fyy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -4008,7 +4220,6 @@
   
   fzz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -4093,7 +4304,6 @@
 
   fxy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -4196,7 +4406,6 @@
 
   fxz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
@@ -4297,7 +4506,6 @@
 
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1

AMSS_NCKU_source/diff_new_sh.f90

@@ -81,7 +81,6 @@
   fy = ZEO
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -180,7 +179,6 @@
 
   fx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -264,7 +262,6 @@
 
   fy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -345,7 +342,6 @@
 
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -447,7 +443,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -558,7 +553,6 @@
 
   fxx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -633,7 +627,6 @@
 
   fyy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -709,7 +702,6 @@
 
   fzz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -785,7 +777,6 @@
 
   fxy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -860,7 +851,6 @@
 
   fxz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -934,7 +924,6 @@
 
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1030,7 +1019,6 @@
   fy = ZEO
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1146,7 +1134,6 @@
 
   fx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1240,7 +1227,6 @@
 
   fy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1328,7 +1314,6 @@
 
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1445,7 +1430,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1596,7 +1580,6 @@
 
   fxx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1676,7 +1659,6 @@
 
   fyy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1758,7 +1740,6 @@
 
   fzz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1840,7 +1821,6 @@
 
   fxy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -1923,7 +1903,6 @@
 
   fxz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2004,7 +1983,6 @@
 
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2109,7 +2087,6 @@
   fy = ZEO
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2242,7 +2219,6 @@
 
   fx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2345,7 +2321,6 @@
 
   fy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2439,7 +2414,6 @@
 
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2570,7 +2544,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2770,7 +2743,6 @@
 
   fxx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2855,7 +2827,6 @@
 
   fyy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -2943,7 +2914,6 @@
 
   fzz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -3031,7 +3001,6 @@
 
   fxy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -3124,7 +3093,6 @@
 
   fxz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -3215,7 +3183,6 @@
 
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -3335,7 +3302,6 @@
   fy = ZEO
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -3488,7 +3454,6 @@
 
   fx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -3601,7 +3566,6 @@
 
   fy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -3703,7 +3667,6 @@
 
   fz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -3851,7 +3814,6 @@
   fxz = ZEO
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -4136,7 +4098,6 @@
 
   fxx = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -4230,7 +4191,6 @@
 
   fyy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -4327,7 +4287,6 @@
   
   fzz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -4424,7 +4383,6 @@
 
   fxy = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -4539,7 +4497,6 @@
 
   fxz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -4652,7 +4609,6 @@
 
   fyz = ZEO
 
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -4723,7 +4679,6 @@ subroutine fderivs_shc(ex,f,fx,fy,fz,crho,sigma,R,SYM1,SYM2,SYM3,Symmetry,Lev,ss
 #if 0  
   integer :: i,j,k
   
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -4774,7 +4729,6 @@ subroutine fdderivs_shc(ex,f,fxx,fxy,fxz,fyy,fyz,fzz,crho,sigma,R,SYM1,SYM2,SYM3
 #if 0  
   integer :: i,j,k
   
-  !$omp parallel do collapse(2) private(i,j,k)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)

AMSS_NCKU_source/enforce_algebra.f90

@@ -27,7 +27,6 @@
 
 !~~~~~~>
 
-  !$omp parallel do collapse(2) private(i,j,k,lgxx,lgyy,lgzz,ldetg,lgupxx,lgupxy,lgupxz,lgupyy,lgupyz,lgupzz,ltrA,lscale)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
@@ -105,7 +104,6 @@
 
 !~~~~~~>
 
-  !$omp parallel do collapse(2) private(i,j,k,lgxx,lgyy,lgzz,lscale,lgxy,lgxz,lgyz,lgupxx,lgupxy,lgupxz,lgupyy,lgupyz,lgupzz,ltrA)
   do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)

AMSS_NCKU_source/expansion.C

@@ -6,11 +6,7 @@
 #include <stdio.h>
 #include <assert.h>
 #include <math.h>
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "util_Table.h"
 #include "cctk.h"

AMSS_NCKU_source/find_horizons.C

@@ -6,11 +6,7 @@
 #include <stdio.h>
 #include <assert.h>
 #include <math.h>
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "cctk.h"
 

AMSS_NCKU_source/kodiss.f90

@@ -65,8 +65,7 @@ real*8,intent(in) :: eps
 !                       dx^4
 
 !  note the sign (-1)^r-1, now r=2
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -160,8 +159,7 @@ integer, parameter :: NO_SYMM=0, OCTANT=2
 
   call symmetry_bd(3,ex,f,fh,SoA)
 
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -217,6 +215,99 @@ integer, parameter :: NO_SYMM=0, OCTANT=2
 
   end subroutine kodis
 
+!-----------------------------------------------------------------------------
+! kodis variant: reuses caller-provided fh work array (memory pool)
+!-----------------------------------------------------------------------------
+subroutine kodis_fh(ex,X,Y,Z,f,f_rhs,SoA,Symmetry,eps,fh)
+
+implicit none
+! argument variables
+integer,intent(in) :: Symmetry
+integer,dimension(3),intent(in)::ex
+real*8, dimension(1:3), intent(in) :: SoA
+double precision,intent(in),dimension(ex(1))::X
+double precision,intent(in),dimension(ex(2))::Y
+double precision,intent(in),dimension(ex(3))::Z
+double precision,intent(in),dimension(ex(1),ex(2),ex(3))::f
+double precision,intent(inout),dimension(ex(1),ex(2),ex(3))::f_rhs
+real*8,intent(in) :: eps
+real*8,dimension(-2:ex(1),-2:ex(2),-2:ex(3)),intent(inout):: fh
+! local variables
+integer :: imin,jmin,kmin,imax,jmax,kmax
+integer :: i,j,k
+real*8  :: dX,dY,dZ
+real*8, parameter :: ONE=1.d0,SIX=6.d0,FIT=1.5d1,TWT=2.d1
+real*8,parameter::cof=6.4d1   ! 2^6
+integer, parameter :: NO_SYMM=0, OCTANT=2
+
+  dX = X(2)-X(1)
+  dY = Y(2)-Y(1)
+  dZ = Z(2)-Z(1)
+
+  imax = ex(1)
+  jmax = ex(2)
+  kmax = ex(3)
+
+  imin = 1
+  jmin = 1
+  kmin = 1
+
+  if(Symmetry > NO_SYMM .and. dabs(Z(1)) < dZ) kmin = -2
+  if(Symmetry == OCTANT .and. dabs(X(1)) < dX) imin = -2
+  if(Symmetry == OCTANT .and. dabs(Y(1)) < dY) jmin = -2
+
+  call symmetry_bd(3,ex,f,fh,SoA)
+
+  do k=1,ex(3)
+  do j=1,ex(2)
+  do i=1,ex(1)
+
+  if(i-3 >= imin .and. i+3 <= imax .and. &
+     j-3 >= jmin .and. j+3 <= jmax .and. &
+     k-3 >= kmin .and. k+3 <= kmax) then
+#if 0
+   f_rhs(i,j,k)       = f_rhs(i,j,k) + eps/dX/cof * (     &
+                              (fh(i-3,j,k)+fh(i+3,j,k)) - &
+                          SIX*(fh(i-2,j,k)+fh(i+2,j,k)) + &
+                          FIT*(fh(i-1,j,k)+fh(i+1,j,k)) - &
+                          TWT* fh(i,j,k)            )
+   f_rhs(i,j,k)       = f_rhs(i,j,k) + eps/dY/cof * (     &
+                              (fh(i,j-3,k)+fh(i,j+3,k)) - &
+                          SIX*(fh(i,j-2,k)+fh(i,j+2,k)) + &
+                          FIT*(fh(i,j-1,k)+fh(i,j+1,k)) - &
+                          TWT* fh(i,j,k)            )
+   f_rhs(i,j,k)       = f_rhs(i,j,k) + eps/dZ/cof * (     &
+                              (fh(i,j,k-3)+fh(i,j,k+3)) - &
+                          SIX*(fh(i,j,k-2)+fh(i,j,k+2)) + &
+                          FIT*(fh(i,j,k-1)+fh(i,j,k+1)) - &
+                          TWT* fh(i,j,k)            )
+#else
+   f_rhs(i,j,k)       = f_rhs(i,j,k) + eps/cof *( (     &
+                              (fh(i-3,j,k)+fh(i+3,j,k)) - &
+                          SIX*(fh(i-2,j,k)+fh(i+2,j,k)) + &
+                          FIT*(fh(i-1,j,k)+fh(i+1,j,k)) - &
+                          TWT* fh(i,j,k)            )/dX + &
+                                                  (     &
+                              (fh(i,j-3,k)+fh(i,j+3,k)) - &
+                          SIX*(fh(i,j-2,k)+fh(i,j+2,k)) + &
+                          FIT*(fh(i,j-1,k)+fh(i,j+1,k)) - &
+                          TWT* fh(i,j,k)            )/dY + &
+                                                  (     &
+                              (fh(i,j,k-3)+fh(i,j,k+3)) - &
+                          SIX*(fh(i,j,k-2)+fh(i,j,k+2)) + &
+                          FIT*(fh(i,j,k-1)+fh(i,j,k+1)) - &
+                          TWT* fh(i,j,k)            )/dZ )
+#endif
+  endif
+
+  enddo
+  enddo
+  enddo
+
+  return
+
+  end subroutine kodis_fh
+
 #elif (ghost_width == 4)
 ! sixth order code
 !------------------------------------------------------------------------------------------------------------------------------
@@ -275,8 +366,7 @@ real*8,intent(in) :: eps
 !                                              dx^8
 
 !  note the sign (-1)^r-1, now r=4
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -388,8 +478,7 @@ real*8,intent(in) :: eps
 !                                                              dx^10
 
 !  note the sign (-1)^r-1, now r=5
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 

AMSS_NCKU_source/kodiss_sh.f90

@@ -80,8 +80,7 @@ real*8,intent(in) :: eps
 !                       dx^4
 
 !  note the sign (-1)^r-1, now r=2
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -179,8 +178,7 @@ real*8,intent(in) :: eps
 !                       dx^4
 
 !  note the sign (-1)^r-1, now r=2
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -275,8 +273,7 @@ integer, parameter :: NO_SYMM=0, EQ_SYMM=1, OCTANT=2
 
   call symmetry_stbd(2,ex,f,fh,SoA)
 
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -372,8 +369,7 @@ integer, parameter :: NO_SYMM=0, EQ_SYMM=1, OCTANT=2
 
   call symmetry_stbd(3,ex,f,fh,SoA)
 
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -514,8 +510,7 @@ integer, parameter :: NO_SYMM=0, EQ_SYMM=1, OCTANT=2
 
   call symmetry_stbd(3,ex,f,fh,SoA)
 
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -603,8 +598,7 @@ integer, parameter :: NO_SYMM=0, EQ_SYMM=1, OCTANT=2
 
   call symmetry_stbd(3,ex,f,fh,SoA)
 
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -700,8 +694,7 @@ real*8,intent(in) :: eps
 !                                              dx^8
 
 !  note the sign (-1)^r-1, now r=4
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -801,8 +794,7 @@ integer, parameter :: NO_SYMM=0, EQ_SYMM=1, OCTANT=2
 
   call symmetry_stbd(4,ex,f,fh,SoA)
 
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -911,8 +903,7 @@ real*8,intent(in) :: eps
 !                                                              dx^10
 
 !  note the sign (-1)^r-1, now r=5
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 
@@ -1015,8 +1006,7 @@ integer, parameter :: NO_SYMM=0, EQ_SYMM=1, OCTANT=2
 
   call symmetry_stbd(5,ex,f,fh,SoA)
 
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)
+  do k=1,ex(3)
   do j=1,ex(2)
   do i=1,ex(1)
 

AMSS_NCKU_source/lopsidediff.f90

@@ -68,8 +68,7 @@ subroutine lopsided(ex,X,Y,Z,f,f_rhs,Sfx,Sfy,Sfz,Symmetry,SoA)
 
 ! upper bound set ex-1 only for efficiency, 
 ! the loop body will set ex 0 also
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)-1
+  do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
 ! x direction   
@@ -234,8 +233,7 @@ subroutine lopsided(ex,X,Y,Z,f,f_rhs,Sfx,Sfy,Sfz,Symmetry,SoA)
 
 ! upper bound set ex-1 only for efficiency, 
 ! the loop body will set ex 0 also
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)-1
+  do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
 #if 0  
@@ -489,6 +487,160 @@ subroutine lopsided(ex,X,Y,Z,f,f_rhs,Sfx,Sfy,Sfz,Symmetry,SoA)
 
   end subroutine lopsided
 
+!-----------------------------------------------------------------------------
+! lopsided variant: reuses caller-provided fh work array (memory pool)
+!-----------------------------------------------------------------------------
+subroutine lopsided_fh(ex,X,Y,Z,f,f_rhs,Sfx,Sfy,Sfz,Symmetry,SoA,fh)
+  implicit none
+
+!~~~~~~> Input parameters:
+
+  integer, intent(in)  :: ex(1:3),Symmetry
+  real*8,  intent(in)  :: X(1:ex(1)),Y(1:ex(2)),Z(1:ex(3))
+  real*8,dimension(ex(1),ex(2),ex(3)),intent(in)   :: f,Sfx,Sfy,Sfz
+
+  real*8,dimension(ex(1),ex(2),ex(3)),intent(inout):: f_rhs
+  real*8,dimension(3),intent(in) ::SoA
+  real*8,dimension(-2:ex(1),-2:ex(2),-2:ex(3)),intent(inout):: fh
+
+!~~~~~~> local variables:
+  integer :: imin,jmin,kmin,imax,jmax,kmax,i,j,k
+  real*8 :: dX,dY,dZ
+  real*8 :: d12dx,d12dy,d12dz,d2dx,d2dy,d2dz
+  real*8,  parameter :: ZEO=0.d0,ONE=1.d0, F3=3.d0
+  real*8,  parameter :: TWO=2.d0,F6=6.0d0,F18=1.8d1
+  real*8,  parameter :: F12=1.2d1, F10=1.d1,EIT=8.d0
+  integer, parameter :: NO_SYMM = 0, EQ_SYMM = 1, OCTANT = 2
+
+  dX = X(2)-X(1)
+  dY = Y(2)-Y(1)
+  dZ = Z(2)-Z(1)
+
+  d12dx = ONE/F12/dX
+  d12dy = ONE/F12/dY
+  d12dz = ONE/F12/dZ
+
+  d2dx = ONE/TWO/dX
+  d2dy = ONE/TWO/dY
+  d2dz = ONE/TWO/dZ
+
+  imax = ex(1)
+  jmax = ex(2)
+  kmax = ex(3)
+
+  imin = 1
+  jmin = 1
+  kmin = 1
+  if(Symmetry > NO_SYMM .and. dabs(Z(1)) < dZ) kmin = -2
+  if(Symmetry > EQ_SYMM .and. dabs(X(1)) < dX) imin = -2
+  if(Symmetry > EQ_SYMM .and. dabs(Y(1)) < dY) jmin = -2
+
+  call symmetry_bd(3,ex,f,fh,SoA)
+
+! upper bound set ex-1 only for efficiency,
+! the loop body will set ex 0 also
+  do k=1,ex(3)-1
+  do j=1,ex(2)-1
+  do i=1,ex(1)-1
+#if 0
+!! old code - same as original lopsided
+#else
+!! new code, 2012dec27, based on bam
+! x direction
+    if(Sfx(i,j,k) > ZEO)then
+      if(i+3 <= imax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                   &
+                  Sfx(i,j,k)*d12dx*(-F3*fh(i-1,j,k)-F10*fh(i,j,k)+F18*fh(i+1,j,k) &
+                                    -F6*fh(i+2,j,k)+    fh(i+3,j,k))
+     elseif(i+2 <= imax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                           &
+                  Sfx(i,j,k)*d12dx*(fh(i-2,j,k)-EIT*fh(i-1,j,k)+EIT*fh(i+1,j,k)-fh(i+2,j,k))
+     elseif(i+1 <= imax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)-                                                   &
+                  Sfx(i,j,k)*d12dx*(-F3*fh(i+1,j,k)-F10*fh(i,j,k)+F18*fh(i-1,j,k) &
+                                    -F6*fh(i-2,j,k)+    fh(i-3,j,k))
+     endif
+   elseif(Sfx(i,j,k) < ZEO)then
+      if(i-3 >= imin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)-                                                   &
+                  Sfx(i,j,k)*d12dx*(-F3*fh(i+1,j,k)-F10*fh(i,j,k)+F18*fh(i-1,j,k) &
+                                    -F6*fh(i-2,j,k)+    fh(i-3,j,k))
+     elseif(i-2 >= imin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                           &
+                  Sfx(i,j,k)*d12dx*(fh(i-2,j,k)-EIT*fh(i-1,j,k)+EIT*fh(i+1,j,k)-fh(i+2,j,k))
+     elseif(i-1 >= imin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                   &
+                  Sfx(i,j,k)*d12dx*(-F3*fh(i-1,j,k)-F10*fh(i,j,k)+F18*fh(i+1,j,k) &
+                                    -F6*fh(i+2,j,k)+    fh(i+3,j,k))
+     endif
+   endif
+
+! y direction
+    if(Sfy(i,j,k) > ZEO)then
+      if(j+3 <= jmax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                   &
+                  Sfy(i,j,k)*d12dy*(-F3*fh(i,j-1,k)-F10*fh(i,j,k)+F18*fh(i,j+1,k) &
+                                    -F6*fh(i,j+2,k)+    fh(i,j+3,k))
+     elseif(j+2 <= jmax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                           &
+                  Sfy(i,j,k)*d12dy*(fh(i,j-2,k)-EIT*fh(i,j-1,k)+EIT*fh(i,j+1,k)-fh(i,j+2,k))
+     elseif(j+1 <= jmax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)-                                                   &
+                  Sfy(i,j,k)*d12dy*(-F3*fh(i,j+1,k)-F10*fh(i,j,k)+F18*fh(i,j-1,k) &
+                                    -F6*fh(i,j-2,k)+    fh(i,j-3,k))
+     endif
+   elseif(Sfy(i,j,k) < ZEO)then
+      if(j-3 >= jmin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)-                                                   &
+                  Sfy(i,j,k)*d12dy*(-F3*fh(i,j+1,k)-F10*fh(i,j,k)+F18*fh(i,j-1,k) &
+                                    -F6*fh(i,j-2,k)+    fh(i,j-3,k))
+     elseif(j-2 >= jmin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                           &
+                  Sfy(i,j,k)*d12dy*(fh(i,j-2,k)-EIT*fh(i,j-1,k)+EIT*fh(i,j+1,k)-fh(i,j+2,k))
+     elseif(j-1 >= jmin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                   &
+                  Sfy(i,j,k)*d12dy*(-F3*fh(i,j-1,k)-F10*fh(i,j,k)+F18*fh(i,j+1,k) &
+                                    -F6*fh(i,j+2,k)+    fh(i,j+3,k))
+     endif
+   endif
+
+! z direction
+    if(Sfz(i,j,k) > ZEO)then
+      if(k+3 <= kmax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                   &
+                  Sfz(i,j,k)*d12dz*(-F3*fh(i,j,k-1)-F10*fh(i,j,k)+F18*fh(i,j,k+1) &
+                                    -F6*fh(i,j,k+2)+    fh(i,j,k+3))
+     elseif(k+2 <= kmax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                           &
+                  Sfz(i,j,k)*d12dz*(fh(i,j,k-2)-EIT*fh(i,j,k-1)+EIT*fh(i,j,k+1)-fh(i,j,k+2))
+     elseif(k+1 <= kmax)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)-                                                   &
+                  Sfz(i,j,k)*d12dz*(-F3*fh(i,j,k+1)-F10*fh(i,j,k)+F18*fh(i,j,k-1) &
+                                    -F6*fh(i,j,k-2)+    fh(i,j,k-3))
+     endif
+   elseif(Sfz(i,j,k) < ZEO)then
+      if(k-3 >= kmin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)-                                                   &
+                  Sfz(i,j,k)*d12dz*(-F3*fh(i,j,k+1)-F10*fh(i,j,k)+F18*fh(i,j,k-1) &
+                                    -F6*fh(i,j,k-2)+    fh(i,j,k-3))
+     elseif(k-2 >= kmin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                           &
+                  Sfz(i,j,k)*d12dz*(fh(i,j,k-2)-EIT*fh(i,j,k-1)+EIT*fh(i,j,k+1)-fh(i,j,k+2))
+     elseif(k-1 >= kmin)then
+     f_rhs(i,j,k)=f_rhs(i,j,k)+                                                   &
+                  Sfz(i,j,k)*d12dz*(-F3*fh(i,j,k-1)-F10*fh(i,j,k)+F18*fh(i,j,k+1) &
+                                    -F6*fh(i,j,k+2)+    fh(i,j,k+3))
+     endif
+   endif
+#endif
+  enddo
+  enddo
+  enddo
+
+  return
+
+  end subroutine lopsided_fh
+
 #elif (ghost_width == 4)
 ! sixth order code
 ! Compute advection terms in right hand sides of field equations
@@ -560,8 +712,7 @@ subroutine lopsided(ex,X,Y,Z,f,f_rhs,Sfx,Sfy,Sfz,Symmetry,SoA)
 
 ! upper bound set ex-1 only for efficiency, 
 ! the loop body will set ex 0 also
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)-1
+  do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
 ! x direction   
@@ -777,8 +928,7 @@ subroutine lopsided(ex,X,Y,Z,f,f_rhs,Sfx,Sfy,Sfz,Symmetry,SoA)
 
 ! upper bound set ex-1 only for efficiency, 
 ! the loop body will set ex 0 also
-  !$omp parallel do collapse(2) private(i,j,k)
-  do k=1,ex(3)-1
+  do k=1,ex(3)-1
   do j=1,ex(2)-1
   do i=1,ex(1)-1
 ! x direction   

AMSS_NCKU_source/makefile

@@ -16,12 +16,6 @@ include makefile.inc
 .cu.o:
 	$(Cu) $(CUDA_APP_FLAGS) -c $< -o $@ $(CUDA_LIB_PATH)
 
-TwoPunctures.o: TwoPunctures.C
-	${CXX} $(CXXAPPFLAGS) -qopenmp -c $< -o $@
-
-TwoPunctureABE.o: TwoPunctureABE.C
-	${CXX} $(CXXAPPFLAGS) -qopenmp -c $< -o $@
-
 # Input files
 C++FILES = ABE.o Ansorg.o Block.o misc.o monitor.o Parallel.o MPatch.o var.o\
            cgh.o bssn_class.o surface_integral.o ShellPatch.o\
@@ -95,14 +89,14 @@ $(CUDAFILES): bssn_gpu.h gpu_mem.h gpu_rhsSS_mem.h
 misc.o : zbesh.o
 
 # projects
-ABE: $(C++FILES) $(F90FILES) $(F77FILES) $(AHFDOBJS)
-	$(CLINKER) $(CXXAPPFLAGS) -qopenmp -o $@ $(C++FILES) $(F90FILES) $(F77FILES) $(AHFDOBJS) $(LDLIBS)
+ABE: $(C++FILES) $(F90FILES) $(F77FILES) $(AHFDOBJS) 
+	$(CLINKER) $(CXXAPPFLAGS) -o $@ $(C++FILES) $(F90FILES) $(F77FILES) $(AHFDOBJS) $(LDLIBS)
 	
 ABEGPU: $(C++FILES_GPU) $(F90FILES) $(F77FILES) $(AHFDOBJS) $(CUDAFILES)
 	$(CLINKER) $(CXXAPPFLAGS) -o $@ $(C++FILES_GPU) $(F90FILES) $(F77FILES) $(AHFDOBJS) $(CUDAFILES) $(LDLIBS)
 
 TwoPunctureABE: $(TwoPunctureFILES)
-	$(CLINKER) $(CXXAPPFLAGS) -qopenmp -o $@ $(TwoPunctureFILES) $(LDLIBS)
+	$(CLINKER) $(CXXAPPFLAGS) -o $@ $(TwoPunctureFILES) $(LDLIBS)
 
 clean:
 	rm *.o ABE ABEGPU TwoPunctureABE make.log -f

AMSS_NCKU_source/makefile.inc

@@ -6,23 +6,25 @@
 ## Intel oneAPI version with oneMKL (Optimized for performance)
 filein  = -I/usr/include/ -I${MKLROOT}/include
 
-## Using OpenMP-threaded MKL for parallel performance
+## Using sequential MKL (OpenMP disabled for better single-threaded performance)
 ## Added -lifcore for Intel Fortran runtime and -limf for Intel math library
-LDLIBS  = -L${MKLROOT}/lib -lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core -liomp5 -lifcore -limf -lpthread -lm -ldl
+LDLIBS  = -L${MKLROOT}/lib -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lifcore -limf -lpthread -lm -ldl
 
-## Aggressive optimization flags + PGO Phase 2 (profile-guided optimization)
-## -fprofile-instr-use: use collected profile data to guide optimization decisions
-##   (branch prediction, basic block layout, inlining, loop unrolling)
-PROFDATA     = /home/amss/AMSS-NCKU/pgo_profile/default.profdata
-CXXAPPFLAGS  = -O3 -march=native -fp-model fast=2 -fma -ipo -qopenmp \
-               -DMPI_STUB -Dfortran3 -Dnewc -I${MKLROOT}/include
-f90appflags  = -O3 -march=native -fp-model fast=2 -fma -ipo -qopenmp \
+## Aggressive optimization flags:
+## -O3: Maximum optimization
+## -xHost: Optimize for the host CPU architecture (Intel/AMD compatible)
+## -fp-model fast=2: Aggressive floating-point optimizations
+## -fma: Enable fused multiply-add instructions
+## Note: OpenMP has been disabled (-qopenmp removed) due to performance issues
+CXXAPPFLAGS  = -O3 -xHost -fp-model fast=2 -fma -ipo \
+               -Dfortran3 -Dnewc -I${MKLROOT}/include
+f90appflags  = -O3 -xHost -fp-model fast=2 -fma -ipo \
                -align array64byte -fpp -I${MKLROOT}/include
 f90          = ifx
 f77          = ifx
 CXX          = icpx
 CC           = icx
-CLINKER      = icpx
+CLINKER      = mpiicpx 
 
 Cu = nvcc
 CUDA_LIB_PATH = -L/usr/lib/cuda/lib64 -I/usr/include -I/usr/lib/cuda/include

AMSS_NCKU_source/misc.C

@@ -14,11 +14,7 @@ using namespace std;
 #include <string.h>
 #include <math.h>
 #endif
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "misc.h"
 #include "macrodef.h"

AMSS_NCKU_source/misc.h

@@ -24,11 +24,7 @@ using namespace std;
 #include <complex.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 namespace misc
 {

AMSS_NCKU_source/monitor.h

@@ -20,11 +20,7 @@ using namespace std;
 #endif
 #include <time.h>
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 class monitor
 {

AMSS_NCKU_source/mpi_stub.h

@@ -1,153 +0,0 @@
-#ifndef MPI_STUB_H
-#define MPI_STUB_H
-
-/*
- * MPI Stub Header — single-process shim for AMSS-NCKU ABE solver.
- * Provides all MPI types, constants, and functions used in the codebase
- * as no-ops or trivial implementations for nprocs=1, myrank=0.
- */
-
-#include <cstring>
-#include <cstdlib>
-#include <cstdio>
-#include <time.h>
-
-/* ── Types ─────────────────────────────────────────────────────────── */
-
-typedef int MPI_Comm;
-typedef int MPI_Datatype;
-typedef int MPI_Op;
-typedef int MPI_Request;
-typedef int MPI_Group;
-
-typedef struct MPI_Status {
-    int MPI_SOURCE;
-    int MPI_TAG;
-    int MPI_ERROR;
-} MPI_Status;
-
-/* ── Constants ─────────────────────────────────────────────────────── */
-
-#define MPI_COMM_WORLD  0
-
-#define MPI_INT              1
-#define MPI_DOUBLE           2
-#define MPI_DOUBLE_PRECISION 2
-#define MPI_DOUBLE_INT       3
-
-#define MPI_SUM    1
-#define MPI_MAX    2
-#define MPI_MAXLOC 3
-
-#define MPI_STATUS_IGNORE   ((MPI_Status *)0)
-#define MPI_STATUSES_IGNORE ((MPI_Status *)0)
-
-#define MPI_MAX_PROCESSOR_NAME 256
-
-/* ── Helper: sizeof for MPI_Datatype ──────────────────────────────── */
-
-static inline size_t mpi_stub_sizeof(MPI_Datatype type) {
-    switch (type) {
-        case MPI_INT:        return sizeof(int);
-        case MPI_DOUBLE:     return sizeof(double);
-        case MPI_DOUBLE_INT: return sizeof(double) + sizeof(int);
-        default:             return 0;
-    }
-}
-
-/* ── Init / Finalize ──────────────────────────────────────────────── */
-
-static inline int MPI_Init(int *, char ***) { return 0; }
-static inline int MPI_Finalize() { return 0; }
-
-/* ── Communicator queries ─────────────────────────────────────────── */
-
-static inline int MPI_Comm_rank(MPI_Comm, int *rank) { *rank = 0; return 0; }
-static inline int MPI_Comm_size(MPI_Comm, int *size) { *size = 1; return 0; }
-static inline int MPI_Comm_split(MPI_Comm comm, int, int, MPI_Comm *newcomm) {
-    *newcomm = comm;
-    return 0;
-}
-static inline int MPI_Comm_free(MPI_Comm *) { return 0; }
-
-/* ── Group operations ─────────────────────────────────────────────── */
-
-static inline int MPI_Comm_group(MPI_Comm, MPI_Group *group) {
-    *group = 0;
-    return 0;
-}
-static inline int MPI_Group_translate_ranks(MPI_Group, int n,
-        const int *ranks1, MPI_Group, int *ranks2) {
-    for (int i = 0; i < n; ++i) ranks2[i] = ranks1[i];
-    return 0;
-}
-static inline int MPI_Group_free(MPI_Group *) { return 0; }
-
-/* ── Collective operations ────────────────────────────────────────── */
-
-static inline int MPI_Allreduce(const void *sendbuf, void *recvbuf,
-        int count, MPI_Datatype datatype, MPI_Op, MPI_Comm) {
-    std::memcpy(recvbuf, sendbuf, count * mpi_stub_sizeof(datatype));
-    return 0;
-}
-
-static inline int MPI_Iallreduce(const void *sendbuf, void *recvbuf,
-        int count, MPI_Datatype datatype, MPI_Op, MPI_Comm,
-        MPI_Request *request) {
-    std::memcpy(recvbuf, sendbuf, count * mpi_stub_sizeof(datatype));
-    *request = 0;
-    return 0;
-}
-
-static inline int MPI_Bcast(void *, int, MPI_Datatype, int, MPI_Comm) {
-    return 0;
-}
-
-static inline int MPI_Barrier(MPI_Comm) { return 0; }
-
-/* ── Point-to-point (never reached with nprocs=1) ─────────────────── */
-
-static inline int MPI_Send(const void *, int, MPI_Datatype, int, int, MPI_Comm) {
-    return 0;
-}
-static inline int MPI_Recv(void *, int, MPI_Datatype, int, int, MPI_Comm, MPI_Status *) {
-    return 0;
-}
-static inline int MPI_Isend(const void *, int, MPI_Datatype, int, int, MPI_Comm,
-        MPI_Request *req) {
-    *req = 0;
-    return 0;
-}
-static inline int MPI_Irecv(void *, int, MPI_Datatype, int, int, MPI_Comm,
-        MPI_Request *req) {
-    *req = 0;
-    return 0;
-}
-
-/* ── Completion ───────────────────────────────────────────────────── */
-
-static inline int MPI_Wait(MPI_Request *, MPI_Status *) { return 0; }
-static inline int MPI_Waitall(int, MPI_Request *, MPI_Status *) { return 0; }
-
-/* ── Utility ──────────────────────────────────────────────────────── */
-
-static inline int MPI_Abort(MPI_Comm, int error_code) {
-    std::fprintf(stderr, "MPI_Abort called with error code %d\n", error_code);
-    std::exit(error_code);
-    return 0;
-}
-
-static inline double MPI_Wtime() {
-    struct timespec ts;
-    clock_gettime(CLOCK_MONOTONIC, &ts);
-    return (double)ts.tv_sec + (double)ts.tv_nsec * 1.0e-9;
-}
-
-static inline int MPI_Get_processor_name(char *name, int *resultlen) {
-    const char *stub_name = "localhost";
-    std::strcpy(name, stub_name);
-    *resultlen = (int)std::strlen(stub_name);
-    return 0;
-}
-
-#endif /* MPI_STUB_H */

AMSS_NCKU_source/parameters.h

@@ -24,11 +24,7 @@ using namespace std;
 #include <map.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 namespace parameters
 {

AMSS_NCKU_source/perf.h

@@ -30,11 +30,7 @@ using namespace std;
 #include <sys/time.h>
 #include <sys/resource.h>
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 /* Real time */
 #define TimerSignal SIGALRM

AMSS_NCKU_source/rungekutta4_rout.f90

@@ -109,33 +109,23 @@
 
   if( RK4 == 0 ) then
 
-   !$omp parallel workshare
    f1 = f0 + HLF * dT * f_rhs
-   !$omp end parallel workshare
 
   elseif(RK4 == 1 ) then
 
-   !$omp parallel workshare
    f_rhs = f_rhs + TWO * f1
-   !$omp end parallel workshare
-   !$omp parallel workshare
+
    f1 = f0 + HLF * dT * f1
-   !$omp end parallel workshare
 
   elseif(RK4 == 2 ) then
 
-   !$omp parallel workshare
    f_rhs = f_rhs + TWO * f1
-   !$omp end parallel workshare
-   !$omp parallel workshare
+
    f1 = f0 +       dT * f1
-   !$omp end parallel workshare
 
   elseif( RK4 == 3 ) then
-
-   !$omp parallel workshare
+ 
    f1 = f0 +F1o6 * dT *(f1 + f_rhs)
-   !$omp end parallel workshare
 
   else
 
@@ -144,7 +134,7 @@
 
   endif
 
-  return
+  return   
 
   end subroutine rungekutta4_rout
 !-----------------------------------------------------------------------------
@@ -225,19 +215,15 @@
 
   if( RK4 == 0 ) then
 
-   !$omp parallel workshare
    f1 = f0 + dT * f_rhs
-   !$omp end parallel workshare
 
   else
 
-   !$omp parallel workshare
    f1 = f0 + HLF * dT * (f1+f_rhs)
-   !$omp end parallel workshare
 
   endif
 
-  return
+  return   
 
   end subroutine icn_rout
 !~~~~~~~~~~~~~~~~~~  
@@ -253,10 +239,8 @@
   real*8, dimension(ex(1),ex(2),ex(3)),intent(in) ::f_rhs
   real*8, dimension(ex(1),ex(2),ex(3)),intent(out) ::f1
 
-  !$omp parallel workshare
    f1 = f0 + dT * f_rhs
-  !$omp end parallel workshare
 
-  return
+  return   
 
   end subroutine euler_rout

AMSS_NCKU_source/scalar_class.h

@@ -19,11 +19,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "cgh.h"
 #include "ShellPatch.h"

AMSS_NCKU_source/scalarwaves.C

@@ -18,11 +18,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "misc.h"
 #include "microdef.h"

AMSS_NCKU_source/setup.C

@@ -3,11 +3,7 @@
 #include <math.h>
 #include <string.h>
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "util_Table.h"
 #include "cctk.h"

AMSS_NCKU_source/surface_integral.C

@@ -20,11 +20,7 @@ using namespace std;
 #include <math.h>
 #include <map.h>
 #endif
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "misc.h"
 #include "cgh.h"
@@ -224,9 +220,16 @@ void surface_integral::surf_Wave(double rex, int lev, cgh *GH, var *Rpsi4, var *
     pox[2][n] = rex * nz_g[n];
   }
 
+  double *shellf;
+  shellf = new double[n_tot * InList];
+
+  GH->PatL[lev]->data->Interp_Points(DG_List, n_tot, pox, shellf, Symmetry);
+
   int mp, Lp, Nmin, Nmax;
+
   mp = n_tot / cpusize;
   Lp = n_tot - cpusize * mp;
+
   if (Lp > myrank)
   {
     Nmin = myrank * mp + myrank;
@@ -238,11 +241,6 @@ void surface_integral::surf_Wave(double rex, int lev, cgh *GH, var *Rpsi4, var *
     Nmax = Nmin + mp - 1;
   }
 
-  double *shellf;
-  shellf = new double[n_tot * InList];
-
-  GH->PatL[lev]->data->Interp_Points(DG_List, n_tot, pox, shellf, Symmetry, Nmin, Nmax);
-
   //|~~~~~> Integrate the dot product of Dphi with the surface normal.
 
   double *RP_out, *IP_out;
@@ -365,17 +363,8 @@ void surface_integral::surf_Wave(double rex, int lev, cgh *GH, var *Rpsi4, var *
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
   //|------= Free memory.
 
@@ -567,17 +556,8 @@ void surface_integral::surf_Wave(double rex, int lev, cgh *GH, var *Rpsi4, var *
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, Comm_here);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, Comm_here);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, Comm_here);
 
   //|------= Free memory.
 
@@ -755,17 +735,8 @@ void surface_integral::surf_Wave(double rex, int lev, ShellPatch *GH, var *Rpsi4
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
   //|------= Free memory.
 
@@ -1013,17 +984,8 @@ void surface_integral::surf_Wave(double rex, int lev, ShellPatch *GH,
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
   //|------= Free memory.
 
@@ -1457,17 +1419,8 @@ void surface_integral::surf_Wave(double rex, int lev, ShellPatch *GH,
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
   //|------= Free memory.
 
@@ -1901,17 +1854,8 @@ void surface_integral::surf_Wave(double rex, int lev, cgh *GH,
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
   //|------= Free memory.
 
@@ -2096,17 +2040,8 @@ void surface_integral::surf_Wave(double rex, int lev, NullShellPatch2 *GH, var *
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
   //|------= Free memory.
 
@@ -2291,17 +2226,8 @@ void surface_integral::surf_Wave(double rex, int lev, NullShellPatch *GH, var *R
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
   //|------= Free memory.
 
@@ -2388,9 +2314,25 @@ void surface_integral::surf_MassPAng(double rex, int lev, cgh *GH, var *chi, var
     pox[2][n] = rex * nz_g[n];
   }
 
+  double *shellf;
+  shellf = new double[n_tot * InList];
+
+  // we have assumed there is only one box on this level,
+  // so we do not need loop boxes
+  GH->PatL[lev]->data->Interp_Points(DG_List, n_tot, pox, shellf, Symmetry);
+
+  double Mass_out = 0;
+  double ang_outx, ang_outy, ang_outz;
+  double p_outx, p_outy, p_outz;
+  ang_outx = ang_outy = ang_outz = 0.0;
+  p_outx = p_outy = p_outz = 0.0;
+  const double f1o8 = 0.125;
+
   int mp, Lp, Nmin, Nmax;
+
   mp = n_tot / cpusize;
   Lp = n_tot - cpusize * mp;
+
   if (Lp > myrank)
   {
     Nmin = myrank * mp + myrank;
@@ -2402,20 +2344,6 @@ void surface_integral::surf_MassPAng(double rex, int lev, cgh *GH, var *chi, var
     Nmax = Nmin + mp - 1;
   }
 
-  double *shellf;
-  shellf = new double[n_tot * InList];
-
-  // we have assumed there is only one box on this level,
-  // so we do not need loop boxes
-  GH->PatL[lev]->data->Interp_Points(DG_List, n_tot, pox, shellf, Symmetry, Nmin, Nmax);
-
-  double Mass_out = 0;
-  double ang_outx, ang_outy, ang_outz;
-  double p_outx, p_outy, p_outz;
-  ang_outx = ang_outy = ang_outz = 0.0;
-  p_outx = p_outy = p_outz = 0.0;
-  const double f1o8 = 0.125;
-
   double Chi, Psi;
   double Gxx, Gxy, Gxz, Gyy, Gyz, Gzz;
   double gupxx, gupxy, gupxz, gupyy, gupyz, gupzz;
@@ -2536,13 +2464,15 @@ void surface_integral::surf_MassPAng(double rex, int lev, cgh *GH, var *chi, var
     }
   }
 
-  {
-    double scalar_out[7] = {Mass_out, ang_outx, ang_outy, ang_outz, p_outx, p_outy, p_outz};
-    double scalar_in[7];
-    MPI_Allreduce(scalar_out, scalar_in, 7, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    mass = scalar_in[0]; sx = scalar_in[1]; sy = scalar_in[2]; sz = scalar_in[3];
-    px = scalar_in[4]; py = scalar_in[5]; pz = scalar_in[6];
-  }
+  MPI_Allreduce(&Mass_out, &mass, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+
+  MPI_Allreduce(&ang_outx, &sx, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(&ang_outy, &sy, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(&ang_outz, &sz, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+
+  MPI_Allreduce(&p_outx, &px, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(&p_outy, &py, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(&p_outz, &pz, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
 #ifdef GaussInt
   mass = mass * rex * rex * dphi * factor;
@@ -2805,13 +2735,15 @@ void surface_integral::surf_MassPAng(double rex, int lev, cgh *GH, var *chi, var
     }
   }
 
-  {
-    double scalar_out[7] = {Mass_out, ang_outx, ang_outy, ang_outz, p_outx, p_outy, p_outz};
-    double scalar_in[7];
-    MPI_Allreduce(scalar_out, scalar_in, 7, MPI_DOUBLE, MPI_SUM, Comm_here);
-    mass = scalar_in[0]; sx = scalar_in[1]; sy = scalar_in[2]; sz = scalar_in[3];
-    px = scalar_in[4]; py = scalar_in[5]; pz = scalar_in[6];
-  }
+  MPI_Allreduce(&Mass_out, &mass, 1, MPI_DOUBLE, MPI_SUM, Comm_here);
+
+  MPI_Allreduce(&ang_outx, &sx, 1, MPI_DOUBLE, MPI_SUM, Comm_here);
+  MPI_Allreduce(&ang_outy, &sy, 1, MPI_DOUBLE, MPI_SUM, Comm_here);
+  MPI_Allreduce(&ang_outz, &sz, 1, MPI_DOUBLE, MPI_SUM, Comm_here);
+
+  MPI_Allreduce(&p_outx, &px, 1, MPI_DOUBLE, MPI_SUM, Comm_here);
+  MPI_Allreduce(&p_outy, &py, 1, MPI_DOUBLE, MPI_SUM, Comm_here);
+  MPI_Allreduce(&p_outz, &pz, 1, MPI_DOUBLE, MPI_SUM, Comm_here);
 
 #ifdef GaussInt
   mass = mass * rex * rex * dphi * factor;
@@ -3088,13 +3020,15 @@ void surface_integral::surf_MassPAng(double rex, int lev, ShellPatch *GH, var *c
     }
   }
 
-  {
-    double scalar_out[7] = {Mass_out, ang_outx, ang_outy, ang_outz, p_outx, p_outy, p_outz};
-    double scalar_in[7];
-    MPI_Allreduce(scalar_out, scalar_in, 7, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    mass = scalar_in[0]; sx = scalar_in[1]; sy = scalar_in[2]; sz = scalar_in[3];
-    px = scalar_in[4]; py = scalar_in[5]; pz = scalar_in[6];
-  }
+  MPI_Allreduce(&Mass_out, &mass, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+
+  MPI_Allreduce(&ang_outx, &sx, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(&ang_outy, &sy, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(&ang_outz, &sz, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+
+  MPI_Allreduce(&p_outx, &px, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(&p_outy, &py, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(&p_outz, &pz, 1, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
 #ifdef GaussInt
   mass = mass * rex * rex * dphi * factor;
@@ -3673,17 +3607,8 @@ void surface_integral::surf_Wave(double rex, cgh *GH, ShellPatch *SH,
   }
   //|------+  Communicate and sum the results from each processor.
 
-  {
-    double *RPIP_out = new double[2 * NN];
-    double *RPIP = new double[2 * NN];
-    memcpy(RPIP_out, RP_out, NN * sizeof(double));
-    memcpy(RPIP_out + NN, IP_out, NN * sizeof(double));
-    MPI_Allreduce(RPIP_out, RPIP, 2 * NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
-    memcpy(RP, RPIP, NN * sizeof(double));
-    memcpy(IP, RPIP + NN, NN * sizeof(double));
-    delete[] RPIP_out;
-    delete[] RPIP;
-  }
+  MPI_Allreduce(RP_out, RP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
+  MPI_Allreduce(IP_out, IP, NN, MPI_DOUBLE, MPI_SUM, MPI_COMM_WORLD);
 
   //|------= Free memory.
 

AMSS_NCKU_source/testNull.C

@@ -18,11 +18,7 @@ using namespace std;
 #include <math.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "misc.h"
 #include "macrodef.h"

AMSS_NCKU_source/testNull2.C

@@ -20,11 +20,7 @@ using namespace std;
 #include <map.h>
 #endif
 
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "misc.h"
 #include "macrodef.h"

AMSS_NCKU_source/var.C

@@ -9,11 +9,7 @@
 using namespace std;
 
 #include <time.h>
-#ifdef MPI_STUB
-#include "mpi_stub.h"
-#else
 #include <mpi.h>
-#endif
 
 #include "var.h"
 

makefile_and_run.py

@@ -10,31 +10,18 @@
 
 import AMSS_NCKU_Input as input_data
 import subprocess
-import time
-import os
 
-## OpenMP configuration for threaded Fortran kernels
-## OMP_NUM_THREADS: set to number of physical cores (not hyperthreads)
-## OMP_PROC_BIND: bind threads to cores to avoid migration overhead
-## OMP_STACKSIZE: each thread needs stack space for fh arrays (~3.6MB)
-if "OMP_NUM_THREADS" not in os.environ:
-    os.environ["OMP_NUM_THREADS"] = "96"
-os.environ["OMP_STACKSIZE"] = "16M"
-os.environ["OMP_PROC_BIND"] = "close"
-os.environ["OMP_PLACES"] = "cores"
 ## CPU core binding configuration using taskset
 ## taskset ensures all child processes inherit the CPU affinity mask
 ## This forces make and all compiler processes to use only nohz_full cores (4-55, 60-111)
 ## Format: taskset -c 4-55,60-111 ensures processes only run on these cores
+NUMACTL_CPU_BIND = "taskset -c 16-47,64-95"
 #NUMACTL_CPU_BIND = "taskset -c 0-111"
-#NUMACTL_CPU_BIND = "taskset -c 16-47,64-95"
-#NUMACTL_CPU_BIND = "taskset -c 8-15"
-NUMACTL_CPU_BIND = ""
 
 ## Build parallelism configuration
 ## Use nohz_full cores (4-55, 60-111) for compilation: 52 + 52 = 104 cores
 ## Set make -j to utilize available cores for faster builds
-BUILD_JOBS = 16
+BUILD_JOBS = 64
 
 
 ##################################################################
@@ -128,28 +115,28 @@ def run_ABE():
     print(                                                      )
 
     ## Define the command to run; cast other values to strings as needed
-
+    
     if (input_data.GPU_Calculation == "no"):
-        run_command         = NUMACTL_CPU_BIND + " ./ABE"
-        run_command_outfile = "ABE_out.log"
+        mpi_command         = NUMACTL_CPU_BIND + " mpirun -np " + str(input_data.MPI_processes) + " ./ABE"
+        mpi_command_outfile = "ABE_out.log"
     elif (input_data.GPU_Calculation == "yes"):
-        run_command         = NUMACTL_CPU_BIND + " ./ABEGPU"
-        run_command_outfile = "ABEGPU_out.log"
-
-    ## Execute the command and stream output
-    run_process = subprocess.Popen(run_command, shell=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True)
+        mpi_command         = NUMACTL_CPU_BIND + " mpirun -np " + str(input_data.MPI_processes) + " ./ABEGPU"
+        mpi_command_outfile = "ABEGPU_out.log"
+ 
+    ## Execute the MPI command and stream output
+    mpi_process = subprocess.Popen(mpi_command, shell=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True)
 
     ## Write ABE run output to file while printing to stdout
-    with open(run_command_outfile, 'w') as file0:
+    with open(mpi_command_outfile, 'w') as file0:  
         ## Read and print output lines; also write each line to file
-        for line in run_process.stdout:
+        for line in mpi_process.stdout:
             print(line, end='')  # stream output in real time
             file0.write(line)    # write the line to file
             file0.flush()        # flush to ensure each line is written immediately (optional)            
     file0.close()
 
     ## Wait for the process to finish
-    run_return_code = run_process.wait()
+    mpi_return_code = mpi_process.wait()
     
     print(                                           )
     print( " The ABE/ABEGPU simulation is finished " ) 
@@ -166,14 +153,13 @@ def run_ABE():
 ## Run the AMSS-NCKU TwoPuncture program TwoPunctureABE
 
 def run_TwoPunctureABE():
-    tp_time1=time.time()
+
     print(                                                          )
     print( " Running the AMSS-NCKU executable file TwoPunctureABE " ) 
     print(                                                          )
     
     ## Define the command to run
-    #TwoPuncture_command         = NUMACTL_CPU_BIND + " ./TwoPunctureABE"
-    TwoPuncture_command         = " ./TwoPunctureABE"
+    TwoPuncture_command         = NUMACTL_CPU_BIND + " ./TwoPunctureABE"
     TwoPuncture_command_outfile = "TwoPunctureABE_out.log"
 
     ## Execute the command with subprocess.Popen and stream output
@@ -194,9 +180,7 @@ def run_TwoPunctureABE():
     print(                                               )
     print( " The TwoPunctureABE simulation is finished " ) 
     print(                                               )
-    tp_time2=time.time()
-    et=tp_time2-tp_time1
-    print(f"Used time: {et}")
+    
     return
 
 ##################################################################

pgo_profile/PGO_Profile_Analysis.md

@@ -1,97 +0,0 @@
-# AMSS-NCKU PGO Profile Analysis Report
-
-## 1. Profiling Environment
-
-| Item | Value |
-|------|-------|
-| Compiler | Intel oneAPI DPC++/C++ 2025.3.0 (icpx/ifx) |
-| Instrumentation Flag | `-fprofile-instr-generate` |
-| Optimization Level (instrumented) | `-O2 -xHost -fma` |
-| MPI Processes | 1 (single process to avoid MPI+instrumentation deadlock) |
-| Profile File | `default_9725750769337483397_0.profraw` (327 KB) |
-| Merged Profile | `default.profdata` (394 KB) |
-| llvm-profdata | `/home/intel/oneapi/compiler/2025.3/bin/compiler/llvm-profdata` |
-
-## 2. Reduced Simulation Parameters (for profiling run)
-
-| Parameter | Production Value | Profiling Value |
-|-----------|-----------------|-----------------|
-| MPI_processes | 64 | 1 |
-| grid_level | 9 | 4 |
-| static_grid_level | 5 | 3 |
-| static_grid_number | 96 | 24 |
-| moving_grid_number | 48 | 16 |
-| largest_box_xyz_max | 320^3 | 160^3 |
-| Final_Evolution_Time | 1000.0 | 10.0 |
-| Evolution_Step_Number | 10,000,000 | 1,000 |
-| Detector_Number | 12 | 2 |
-
-## 3. Profile Summary
-
-| Metric | Value |
-|--------|-------|
-| Total instrumented functions | 1,392 |
-| Functions with non-zero counts | 117 (8.4%) |
-| Functions with zero counts | 1,275 (91.6%) |
-| Maximum function entry count | 386,459,248 |
-| Maximum internal block count | 370,477,680 |
-| Total block count | 4,198,023,118 |
-
-## 4. Top 20 Hotspot Functions
-
-| Rank | Total Count | Max Block Count | Function | Category |
-|------|------------|-----------------|----------|----------|
-| 1 | 1,241,601,732 | 370,477,680 | `polint_` | Interpolation |
-| 2 | 755,994,435 | 230,156,640 | `prolong3_` | Grid prolongation |
-| 3 | 667,964,095 | 3,697,792 | `compute_rhs_bssn_` | BSSN RHS evolution |
-| 4 | 539,736,051 | 386,459,248 | `symmetry_bd_` | Symmetry boundary |
-| 5 | 277,310,808 | 53,170,728 | `lopsided_` | Lopsided FD stencil |
-| 6 | 155,534,488 | 94,535,040 | `decide3d_` | 3D grid decision |
-| 7 | 119,267,712 | 19,266,048 | `rungekutta4_rout_` | RK4 time integrator |
-| 8 | 91,574,616 | 48,824,160 | `kodis_` | Kreiss-Oliger dissipation |
-| 9 | 67,555,389 | 43,243,680 | `fderivs_` | Finite differences |
-| 10 | 55,296,000 | 42,246,144 | `misc::fact(int)` | Factorial utility |
-| 11 | 43,191,071 | 27,663,328 | `fdderivs_` | 2nd-order FD derivatives |
-| 12 | 36,233,965 | 22,429,440 | `restrict3_` | Grid restriction |
-| 13 | 24,698,512 | 17,231,520 | `polin3_` | Polynomial interpolation |
-| 14 | 22,962,942 | 20,968,768 | `copy_` | Data copy |
-| 15 | 20,135,696 | 17,259,168 | `Ansorg::barycentric(...)` | Spectral interpolation |
-| 16 | 14,650,224 | 7,224,768 | `Ansorg::barycentric_omega(...)` | Spectral weights |
-| 17 | 13,242,296 | 2,871,920 | `global_interp_` | Global interpolation |
-| 18 | 12,672,000 | 7,734,528 | `sommerfeld_rout_` | Sommerfeld boundary |
-| 19 | 6,872,832 | 1,880,064 | `sommerfeld_routbam_` | Sommerfeld boundary (BAM) |
-| 20 | 5,709,900 | 2,809,632 | `l2normhelper_` | L2 norm computation |
-
-## 5. Hotspot Category Breakdown
-
-Top 20 functions account for ~98% of total execution counts:
-
-| Category | Functions | Combined Count | Share |
-|----------|-----------|---------------|-------|
-| Interpolation / Prolongation / Restriction | polint_, prolong3_, restrict3_, polin3_, global_interp_, Ansorg::* | ~2,093M | ~50% |
-| BSSN RHS + FD stencils | compute_rhs_bssn_, lopsided_, fderivs_, fdderivs_ | ~1,056M | ~25% |
-| Boundary conditions | symmetry_bd_, sommerfeld_rout_, sommerfeld_routbam_ | ~559M | ~13% |
-| Time integration | rungekutta4_rout_ | ~119M | ~3% |
-| Dissipation | kodis_ | ~92M | ~2% |
-| Utilities | misc::fact, decide3d_, copy_, l2normhelper_ | ~256M | ~6% |
-
-## 6. Conclusions
-
-1. **Profile data is valid**: 1,392 functions instrumented, 117 exercised with ~4.2 billion total counts.
-2. **Hotspot concentration is high**: Top 5 functions alone account for ~76% of all counts, which is ideal for PGO — the compiler has strong branch/layout optimization targets.
-3. **Fortran numerical kernels dominate**: `polint_`, `prolong3_`, `compute_rhs_bssn_`, `symmetry_bd_`, `lopsided_` are all Fortran routines in the inner evolution loop. PGO will optimize their branch prediction and basic block layout.
-4. **91.6% of functions have zero counts**: These are code paths for unused features (GPU, BSSN-EScalar, BSSN-EM, Z4C, etc.). PGO will deprioritize them, improving instruction cache utilization.
-5. **Profile is representative**: Despite the reduced grid size, the code path coverage matches production — the same kernels (RHS, prolongation, restriction, boundary) are exercised. PGO branch probabilities from this profile will transfer well to full-scale runs.
-
-## 7. PGO Phase 2 Usage
-
-To apply the profile, use the following flags in `makefile.inc`:
-
-```makefile
-CXXAPPFLAGS = -O3 -xHost -fp-model fast=2 -fma -ipo \
-              -fprofile-instr-use=/home/amss/AMSS-NCKU/pgo_profile/default.profdata \
-              -Dfortran3 -Dnewc -I${MKLROOT}/include
-f90appflags = -O3 -xHost -fp-model fast=2 -fma -ipo \
-              -fprofile-instr-use=/home/amss/AMSS-NCKU/pgo_profile/default.profdata \
-              -align array64byte -fpp -I${MKLROOT}/include
-```