| 1 | /*
|
|---|
| 2 | A simple 2D hydro code
|
|---|
| 3 | (C) Romain Teyssier : CEA/IRFU -- original F90 code
|
|---|
| 4 | (C) Pierre-Francois Lavallee : IDRIS -- original F90 code
|
|---|
| 5 | (C) Guillaume Colin de Verdiere : CEA/DAM -- for the C version
|
|---|
| 6 | */
|
|---|
| 7 | /*
|
|---|
| 8 |
|
|---|
| 9 | This software is governed by the CeCILL license under French law and
|
|---|
| 10 | abiding by the rules of distribution of free software. You can use,
|
|---|
| 11 | modify and/ or redistribute the software under the terms of the CeCILL
|
|---|
| 12 | license as circulated by CEA, CNRS and INRIA at the following URL
|
|---|
| 13 | "http://www.cecill.info".
|
|---|
| 14 |
|
|---|
| 15 | As a counterpart to the access to the source code and rights to copy,
|
|---|
| 16 | modify and redistribute granted by the license, users are provided only
|
|---|
| 17 | with a limited warranty and the software's author, the holder of the
|
|---|
| 18 | economic rights, and the successive licensors have only limited
|
|---|
| 19 | liability.
|
|---|
| 20 |
|
|---|
| 21 | In this respect, the user's attention is drawn to the risks associated
|
|---|
| 22 | with loading, using, modifying and/or developing or reproducing the
|
|---|
| 23 | software by the user in light of its specific status of free software,
|
|---|
| 24 | that may mean that it is complicated to manipulate, and that also
|
|---|
| 25 | therefore means that it is reserved for developers and experienced
|
|---|
| 26 | professionals having in-depth computer knowledge. Users are therefore
|
|---|
| 27 | encouraged to load and test the software's suitability as regards their
|
|---|
| 28 | requirements in conditions enabling the security of their systems and/or
|
|---|
| 29 | data to be ensured and, more generally, to use and operate it in the
|
|---|
| 30 | same conditions as regards security.
|
|---|
| 31 |
|
|---|
| 32 | The fact that you are presently reading this means that you have had
|
|---|
| 33 | knowledge of the CeCILL license and that you accept its terms.
|
|---|
| 34 |
|
|---|
| 35 | */
|
|---|
| 36 |
|
|---|
| 37 | #include <stdio.h>
|
|---|
| 38 | #include <string.h>
|
|---|
| 39 | #include <malloc.h>
|
|---|
| 40 | // #include <unistd.h>
|
|---|
| 41 | #include <math.h>
|
|---|
| 42 |
|
|---|
| 43 | #ifdef HMPP
|
|---|
| 44 | #undef HMPP
|
|---|
| 45 | #endif
|
|---|
| 46 |
|
|---|
| 47 | #include "parametres.h"
|
|---|
| 48 | #include "compute_deltat.h"
|
|---|
| 49 | #include "utils.h"
|
|---|
| 50 | #include "perfcnt.h"
|
|---|
| 51 | #include "equation_of_state.h"
|
|---|
| 52 |
|
|---|
| 53 | #define DABS(x) (real_t) fabs((x))
|
|---|
| 54 |
|
|---|
| 55 | inline void
|
|---|
| 56 | ComputeQEforRow(const int j,
|
|---|
| 57 | const real_t Hsmallr,
|
|---|
| 58 | const int Hnx,
|
|---|
| 59 | const int Hnxt,
|
|---|
| 60 | const int Hnyt,
|
|---|
| 61 | const int Hnxyt,
|
|---|
| 62 | const int Hnvar,
|
|---|
| 63 | const int slices, const int Hstep,
|
|---|
| 64 | real_t * uold,
|
|---|
| 65 | real_t q[Hnvar][Hstep][Hnxyt], real_t e[Hstep][Hnxyt]
|
|---|
| 66 | ) {
|
|---|
| 67 | int i, s;
|
|---|
| 68 |
|
|---|
| 69 | #define IHV(i, j, v) ((i) + Hnxt * ((j) + Hnyt * (v)))
|
|---|
| 70 |
|
|---|
| 71 | #pragma omp parallel for shared(q, e) private(s, i) COLLAPSE
|
|---|
| 72 | for (s = 0; s < slices; s++) {
|
|---|
| 73 | for (i = 0; i < Hnx; i++) {
|
|---|
| 74 | real_t eken;
|
|---|
| 75 | real_t tmp;
|
|---|
| 76 | int idxuID = IHV(i + ExtraLayer, j + s, ID);
|
|---|
| 77 | int idxuIU = IHV(i + ExtraLayer, j + s, IU);
|
|---|
| 78 | int idxuIV = IHV(i + ExtraLayer, j + s, IV);
|
|---|
| 79 | int idxuIP = IHV(i + ExtraLayer, j + s, IP);
|
|---|
| 80 | q[ID][s][i] = MAX(uold[idxuID], Hsmallr);
|
|---|
| 81 | q[IU][s][i] = uold[idxuIU] / q[ID][s][i];
|
|---|
| 82 | q[IV][s][i] = uold[idxuIV] / q[ID][s][i];
|
|---|
| 83 | eken = half * (Square(q[IU][s][i]) + Square(q[IV][s][i]));
|
|---|
| 84 | tmp = uold[idxuIP] / q[ID][s][i] - eken;
|
|---|
| 85 | q[IP][s][i] = tmp;
|
|---|
| 86 | e[s][i] = tmp;
|
|---|
| 87 | }
|
|---|
| 88 | }
|
|---|
| 89 | {
|
|---|
| 90 | int nops = slices * Hnx;
|
|---|
| 91 | FLOPS(5 * nops, 3 * nops, 1 * nops, 0 * nops);
|
|---|
| 92 | }
|
|---|
| 93 | #undef IHV
|
|---|
| 94 | #undef IHVW
|
|---|
| 95 | }
|
|---|
| 96 |
|
|---|
| 97 | // to force a parallel reduction with OpenMP
|
|---|
| 98 | #define WOMP
|
|---|
| 99 |
|
|---|
| 100 | inline void
|
|---|
| 101 | courantOnXY(real_t *cournox,
|
|---|
| 102 | real_t *cournoy,
|
|---|
| 103 | const int Hnx,
|
|---|
| 104 | const int Hnxyt,
|
|---|
| 105 | const int Hnvar, const int slices, const int Hstep, real_t c[Hstep][Hnxyt], real_t q[Hnvar][Hstep][Hnxyt],
|
|---|
| 106 | real_t *tmpm1,
|
|---|
| 107 | real_t *tmpm2
|
|---|
| 108 | )
|
|---|
| 109 | {
|
|---|
| 110 | #ifdef WOMP
|
|---|
| 111 | int s, i;
|
|---|
| 112 | // real_t maxValC = zero;
|
|---|
| 113 | real_t tmp1 = *cournox, tmp2 = *cournoy;
|
|---|
| 114 |
|
|---|
| 115 | #pragma omp parallel for shared(tmpm1, tmpm2) private(s,i) reduction(max:tmp1) reduction(max:tmp2)
|
|---|
| 116 | for (s = 0; s < slices; s++) {
|
|---|
| 117 | for (i = 0; i < Hnx; i++) {
|
|---|
| 118 | tmp1 = MAX(tmp1, c[s][i] + DABS(q[IU][s][i]));
|
|---|
| 119 | tmp2 = MAX(tmp2, c[s][i] + DABS(q[IV][s][i]));
|
|---|
| 120 | }
|
|---|
| 121 | }
|
|---|
| 122 | *cournox = tmp1;
|
|---|
| 123 | *cournoy = tmp2;
|
|---|
| 124 | {
|
|---|
| 125 | int nops = (slices) * Hnx;
|
|---|
| 126 | FLOPS(2 * nops, 0 * nops, 2 * nops, 0 * nops);
|
|---|
| 127 | }
|
|---|
| 128 | #else
|
|---|
| 129 | int i, s;
|
|---|
| 130 | real_t tmp1, tmp2;
|
|---|
| 131 | for (s = 0; s < slices; s++) {
|
|---|
| 132 | for (i = 0; i < Hnx; i++) {
|
|---|
| 133 | tmp1 = c[s][i] + DABS(q[IU][s][i]);
|
|---|
| 134 | tmp2 = c[s][i] + DABS(q[IV][s][i]);
|
|---|
| 135 | *cournox = MAX(*cournox, tmp1);
|
|---|
| 136 | *cournoy = MAX(*cournoy, tmp2);
|
|---|
| 137 | }
|
|---|
| 138 | }
|
|---|
| 139 | {
|
|---|
| 140 | int nops = (slices) * Hnx;
|
|---|
| 141 | FLOPS(2 * nops, 0 * nops, 5 * nops, 0 * nops);
|
|---|
| 142 | }
|
|---|
| 143 | #endif
|
|---|
| 144 | #undef IHVW
|
|---|
| 145 | }
|
|---|
| 146 |
|
|---|
| 147 | void compute_deltat_init_mem(const hydroparam_t H, hydrowork_t * Hw, hydrovarwork_t * Hvw)
|
|---|
| 148 | {
|
|---|
| 149 | Hvw->q = (real_t (*)) DMalloc(H.nvar * H.nxyt * H.nxystep);
|
|---|
| 150 | Hw->e = (real_t (*)) DMalloc( H.nxyt * H.nxystep);
|
|---|
| 151 | Hw->c = (real_t (*)) DMalloc( H.nxyt * H.nxystep);
|
|---|
| 152 | Hw->tmpm1 = (real_t *) DMalloc(H.nxystep);
|
|---|
| 153 | Hw->tmpm2 = (real_t *) DMalloc(H.nxystep);
|
|---|
| 154 |
|
|---|
| 155 | }
|
|---|
| 156 |
|
|---|
| 157 | void compute_deltat_clean_mem(const hydroparam_t H, hydrowork_t * Hw, hydrovarwork_t * Hvw)
|
|---|
| 158 | {
|
|---|
| 159 | DFree(&Hvw->q, H.nvar * H.nxyt * H.nxystep);
|
|---|
| 160 | DFree(&Hw->e, H.nxyt * H.nxystep);
|
|---|
| 161 | DFree(&Hw->c, H.nxyt * H.nxystep);
|
|---|
| 162 | DFree(&Hw->tmpm1, H.nxystep);
|
|---|
| 163 | DFree(&Hw->tmpm2, H.nxystep);
|
|---|
| 164 | }
|
|---|
| 165 |
|
|---|
| 166 | void
|
|---|
| 167 | compute_deltat(real_t *dt, const hydroparam_t H, hydrowork_t * Hw, hydrovar_t * Hv, hydrovarwork_t * Hvw) {
|
|---|
| 168 | real_t cournox, cournoy;
|
|---|
| 169 | int j, jend, slices, Hstep, Hmin, Hmax;
|
|---|
| 170 | real_t (*e)[H.nxyt];
|
|---|
| 171 | real_t (*c)[H.nxystep];
|
|---|
| 172 | real_t (*q)[H.nxystep][H.nxyt];
|
|---|
| 173 | WHERE("compute_deltat");
|
|---|
| 174 |
|
|---|
| 175 | // compute time step on grid interior
|
|---|
| 176 | cournox = zero;
|
|---|
| 177 | cournoy = zero;
|
|---|
| 178 |
|
|---|
| 179 | c = (real_t (*)[H.nxystep]) Hw->c;
|
|---|
| 180 | e = (real_t (*)[H.nxystep]) Hw->e;
|
|---|
| 181 | q = (real_t (*)[H.nxystep][H.nxyt]) Hvw->q;
|
|---|
| 182 |
|
|---|
| 183 | Hstep = H.nxystep;
|
|---|
| 184 | Hmin = H.jmin + ExtraLayer;
|
|---|
| 185 | Hmax = H.jmax - ExtraLayer;
|
|---|
| 186 | for (j = Hmin; j < Hmax; j += Hstep) {
|
|---|
| 187 | jend = j + Hstep;
|
|---|
| 188 | if (jend >= Hmax)
|
|---|
| 189 | jend = Hmax;
|
|---|
| 190 | slices = jend - j; // numbre of slices to compute
|
|---|
| 191 | ComputeQEforRow(j, H.smallr, H.nx, H.nxt, H.nyt, H.nxyt, H.nvar, slices, Hstep, Hv->uold, q, e);
|
|---|
| 192 | equation_of_state(0, H.nx, H.nxyt, H.nvar, H.smallc, H.gamma, slices, Hstep, e, q, c);
|
|---|
| 193 | courantOnXY(&cournox, &cournoy, H.nx, H.nxyt, H.nvar, slices, Hstep, c, q, Hw->tmpm1, Hw->tmpm2);
|
|---|
| 194 | // fprintf(stdout, "[%2d]\t%g %g %g %g\n", H.mype, cournox, cournoy, H.smallc, H.courant_factor);
|
|---|
| 195 | }
|
|---|
| 196 |
|
|---|
| 197 | *dt = H.courant_factor * H.dx / MAX(cournox, MAX(cournoy, H.smallc));
|
|---|
| 198 | FLOPS(1, 1, 2, 0);
|
|---|
| 199 | // fprintf(stdout, "[%2d]\t%g %g %g %g %g %g\n", H.mype, cournox, cournoy, H.smallc, H.courant_factor, H.dx, *dt);
|
|---|
| 200 | } // compute_deltat
|
|---|
| 201 |
|
|---|
| 202 | //EOF
|
|---|