1 #include <../src/mat/impls/aij/seq/aij.h> 2 #include <petsc/private/dmpleximpl.h> /*I "petscdmplex.h" I*/ 3 #include <petsclandau.h> /*I "petsclandau.h" I*/ 4 #include <petscts.h> 5 #include <petscdmforest.h> 6 #include <petscdmcomposite.h> 7 8 /* Landau collision operator */ 9 10 /* relativistic terms */ 11 #if defined(PETSC_USE_REAL_SINGLE) 12 #define SPEED_OF_LIGHT 2.99792458e8F 13 #define C_0(v0) (SPEED_OF_LIGHT / v0) /* needed for relativistic tensor on all architectures */ 14 #else 15 #define SPEED_OF_LIGHT 2.99792458e8 16 #define C_0(v0) (SPEED_OF_LIGHT / v0) /* needed for relativistic tensor on all architectures */ 17 #endif 18 19 #include "land_tensors.h" 20 21 #if defined(PETSC_HAVE_OPENMP) 22 #include <omp.h> 23 #endif 24 25 static PetscErrorCode LandauGPUMapsDestroy(void **ptr) 26 { 27 P4estVertexMaps *maps = (P4estVertexMaps *)*ptr; 28 29 PetscFunctionBegin; 30 // free device data 31 if (maps[0].deviceType != LANDAU_CPU) { 32 #if defined(PETSC_HAVE_KOKKOS) 33 if (maps[0].deviceType == LANDAU_KOKKOS) PetscCall(LandauKokkosDestroyMatMaps(maps, maps[0].numgrids)); // implies Kokkos does 34 #endif 35 } 36 // free host data 37 for (PetscInt grid = 0; grid < maps[0].numgrids; grid++) { 38 PetscCall(PetscFree(maps[grid].c_maps)); 39 PetscCall(PetscFree(maps[grid].gIdx)); 40 } 41 PetscCall(PetscFree(maps)); 42 PetscFunctionReturn(PETSC_SUCCESS); 43 } 44 static PetscErrorCode energy_f(PetscInt dim, PetscReal time, const PetscReal x[], PetscInt Nf_dummy, PetscScalar *u, void *actx) 45 { 46 PetscReal v2 = 0; 47 48 PetscFunctionBegin; 49 /* compute v^2 / 2 */ 50 for (PetscInt i = 0; i < dim; ++i) v2 += x[i] * x[i]; 51 /* evaluate the Maxwellian */ 52 u[0] = v2 / 2; 53 PetscFunctionReturn(PETSC_SUCCESS); 54 } 55 56 /* needs double */ 57 static PetscErrorCode gamma_m1_f(PetscInt dim, PetscReal time, const PetscReal x[], PetscInt Nf_dummy, PetscScalar *u, void *actx) 58 { 59 PetscReal *c2_0_arr = ((PetscReal *)actx); 60 double u2 = 0, c02 = (double)*c2_0_arr, xx; 61 62 PetscFunctionBegin; 63 /* compute u^2 / 2 */ 64 for (PetscInt i = 0; i < dim; ++i) u2 += x[i] * x[i]; 65 /* gamma - 1 = g_eps, for conditioning and we only take derivatives */ 66 xx = u2 / c02; 67 #if defined(PETSC_USE_DEBUG) 68 u[0] = PetscSqrtReal(1. + xx); 69 #else 70 u[0] = xx / (PetscSqrtReal(1. + xx) + 1.) - 1.; // better conditioned. -1 might help condition and only used for derivative 71 #endif 72 PetscFunctionReturn(PETSC_SUCCESS); 73 } 74 75 /* 76 LandauFormJacobian_Internal - Evaluates Jacobian matrix. 77 78 Input Parameters: 79 . globX - input vector 80 . actx - optional user-defined context 81 . dim - dimension 82 83 Output Parameter: 84 . J0acP - Jacobian matrix filled, not created 85 */ 86 static PetscErrorCode LandauFormJacobian_Internal(Vec a_X, Mat JacP, const PetscInt dim, PetscReal shift, void *a_ctx) 87 { 88 LandauCtx *ctx = (LandauCtx *)a_ctx; 89 PetscInt numCells[LANDAU_MAX_GRIDS], Nq, Nb; 90 PetscQuadrature quad; 91 PetscReal Eq_m[LANDAU_MAX_SPECIES]; // could be static data w/o quench (ex2) 92 PetscScalar *cellClosure = NULL; 93 const PetscScalar *xdata = NULL; 94 PetscDS prob; 95 PetscContainer container; 96 P4estVertexMaps *maps; 97 Mat subJ[LANDAU_MAX_GRIDS * LANDAU_MAX_BATCH_SZ]; 98 99 PetscFunctionBegin; 100 PetscValidHeaderSpecific(a_X, VEC_CLASSID, 1); 101 PetscValidHeaderSpecific(JacP, MAT_CLASSID, 2); 102 PetscAssertPointer(ctx, 5); 103 /* check for matrix container for GPU assembly. Support CPU assembly for debugging */ 104 PetscCheck(ctx->plex[0] != NULL, ctx->comm, PETSC_ERR_ARG_WRONG, "Plex not created"); 105 PetscCall(PetscLogEventBegin(ctx->events[10], 0, 0, 0, 0)); 106 PetscCall(DMGetDS(ctx->plex[0], &prob)); // same DS for all grids 107 PetscCall(PetscObjectQuery((PetscObject)JacP, "assembly_maps", (PetscObject *)&container)); 108 if (container) { 109 PetscCheck(ctx->gpu_assembly, ctx->comm, PETSC_ERR_ARG_WRONG, "maps but no GPU assembly"); 110 PetscCall(PetscContainerGetPointer(container, (void **)&maps)); 111 PetscCheck(maps, ctx->comm, PETSC_ERR_ARG_WRONG, "empty GPU matrix container"); 112 for (PetscInt i = 0; i < ctx->num_grids * ctx->batch_sz; i++) subJ[i] = NULL; 113 } else { 114 PetscCheck(!ctx->gpu_assembly, ctx->comm, PETSC_ERR_ARG_WRONG, "No maps but GPU assembly"); 115 for (PetscInt tid = 0; tid < ctx->batch_sz; tid++) { 116 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) PetscCall(DMCreateMatrix(ctx->plex[grid], &subJ[LAND_PACK_IDX(tid, grid)])); 117 } 118 maps = NULL; 119 } 120 // get dynamic data (Eq is odd, for quench and Spitzer test) for CPU assembly and raw data for Jacobian GPU assembly. Get host numCells[], Nq (yuck) 121 PetscCall(PetscFEGetQuadrature(ctx->fe[0], &quad)); 122 PetscCall(PetscQuadratureGetData(quad, NULL, NULL, &Nq, NULL, NULL)); 123 PetscCall(PetscFEGetDimension(ctx->fe[0], &Nb)); 124 PetscCheck(Nq <= LANDAU_MAX_NQND, ctx->comm, PETSC_ERR_ARG_WRONG, "Order too high. Nq = %" PetscInt_FMT " > LANDAU_MAX_NQND (%d)", Nq, LANDAU_MAX_NQND); 125 PetscCheck(Nb <= LANDAU_MAX_NQND, ctx->comm, PETSC_ERR_ARG_WRONG, "Order too high. Nb = %" PetscInt_FMT " > LANDAU_MAX_NQND (%d)", Nb, LANDAU_MAX_NQND); 126 // get metadata for collecting dynamic data 127 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 128 PetscInt cStart, cEnd; 129 PetscCheck(ctx->plex[grid] != NULL, ctx->comm, PETSC_ERR_ARG_WRONG, "Plex not created"); 130 PetscCall(DMPlexGetHeightStratum(ctx->plex[grid], 0, &cStart, &cEnd)); 131 numCells[grid] = cEnd - cStart; // grids can have different topology 132 } 133 PetscCall(PetscLogEventEnd(ctx->events[10], 0, 0, 0, 0)); 134 if (shift == 0) { /* create dynamic point data: f_alpha for closure of each cell (cellClosure[nbatch,ngrids,ncells[g],f[Nb,ns[g]]]) or xdata */ 135 DM pack; 136 PetscCall(VecGetDM(a_X, &pack)); 137 PetscCheck(pack, PETSC_COMM_SELF, PETSC_ERR_PLIB, "pack has no DM"); 138 PetscCall(PetscLogEventBegin(ctx->events[1], 0, 0, 0, 0)); 139 for (PetscInt fieldA = 0; fieldA < ctx->num_species; fieldA++) { 140 Eq_m[fieldA] = ctx->Ez * ctx->t_0 * ctx->charges[fieldA] / (ctx->v_0 * ctx->masses[fieldA]); /* normalize dimensionless */ 141 if (dim == 2) Eq_m[fieldA] *= 2 * PETSC_PI; /* add the 2pi term that is not in Landau */ 142 } 143 if (!ctx->gpu_assembly) { 144 Vec *locXArray, *globXArray; 145 PetscScalar *cellClosure_it; 146 PetscInt cellClosure_sz = 0, nDMs, Nf[LANDAU_MAX_GRIDS]; 147 PetscSection section[LANDAU_MAX_GRIDS], globsection[LANDAU_MAX_GRIDS]; 148 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 149 PetscCall(DMGetLocalSection(ctx->plex[grid], §ion[grid])); 150 PetscCall(DMGetGlobalSection(ctx->plex[grid], &globsection[grid])); 151 PetscCall(PetscSectionGetNumFields(section[grid], &Nf[grid])); 152 } 153 /* count cellClosure size */ 154 PetscCall(DMCompositeGetNumberDM(pack, &nDMs)); 155 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) cellClosure_sz += Nb * Nf[grid] * numCells[grid]; 156 PetscCall(PetscMalloc1(cellClosure_sz * ctx->batch_sz, &cellClosure)); 157 cellClosure_it = cellClosure; 158 PetscCall(PetscMalloc(sizeof(*locXArray) * nDMs, &locXArray)); 159 PetscCall(PetscMalloc(sizeof(*globXArray) * nDMs, &globXArray)); 160 PetscCall(DMCompositeGetLocalAccessArray(pack, a_X, nDMs, NULL, locXArray)); 161 PetscCall(DMCompositeGetAccessArray(pack, a_X, nDMs, NULL, globXArray)); 162 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { // OpenMP (once) 163 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 164 Vec locX = locXArray[LAND_PACK_IDX(b_id, grid)], globX = globXArray[LAND_PACK_IDX(b_id, grid)], locX2; 165 PetscInt cStart, cEnd, ei; 166 PetscCall(VecDuplicate(locX, &locX2)); 167 PetscCall(DMGlobalToLocalBegin(ctx->plex[grid], globX, INSERT_VALUES, locX2)); 168 PetscCall(DMGlobalToLocalEnd(ctx->plex[grid], globX, INSERT_VALUES, locX2)); 169 PetscCall(DMPlexGetHeightStratum(ctx->plex[grid], 0, &cStart, &cEnd)); 170 for (ei = cStart; ei < cEnd; ++ei) { 171 PetscScalar *coef = NULL; 172 PetscCall(DMPlexVecGetClosure(ctx->plex[grid], section[grid], locX2, ei, NULL, &coef)); 173 PetscCall(PetscMemcpy(cellClosure_it, coef, Nb * Nf[grid] * sizeof(*cellClosure_it))); /* change if LandauIPReal != PetscScalar */ 174 PetscCall(DMPlexVecRestoreClosure(ctx->plex[grid], section[grid], locX2, ei, NULL, &coef)); 175 cellClosure_it += Nb * Nf[grid]; 176 } 177 PetscCall(VecDestroy(&locX2)); 178 } 179 } 180 PetscCheck(cellClosure_it - cellClosure == cellClosure_sz * ctx->batch_sz, PETSC_COMM_SELF, PETSC_ERR_PLIB, "iteration wrong %" PetscCount_FMT " != cellClosure_sz = %" PetscInt_FMT, cellClosure_it - cellClosure, cellClosure_sz * ctx->batch_sz); 181 PetscCall(DMCompositeRestoreLocalAccessArray(pack, a_X, nDMs, NULL, locXArray)); 182 PetscCall(DMCompositeRestoreAccessArray(pack, a_X, nDMs, NULL, globXArray)); 183 PetscCall(PetscFree(locXArray)); 184 PetscCall(PetscFree(globXArray)); 185 xdata = NULL; 186 } else { 187 PetscMemType mtype; 188 if (ctx->jacobian_field_major_order) { // get data in batch ordering 189 PetscCall(VecScatterBegin(ctx->plex_batch, a_X, ctx->work_vec, INSERT_VALUES, SCATTER_FORWARD)); 190 PetscCall(VecScatterEnd(ctx->plex_batch, a_X, ctx->work_vec, INSERT_VALUES, SCATTER_FORWARD)); 191 PetscCall(VecGetArrayReadAndMemType(ctx->work_vec, &xdata, &mtype)); 192 } else { 193 PetscCall(VecGetArrayReadAndMemType(a_X, &xdata, &mtype)); 194 } 195 PetscCheck(mtype == PETSC_MEMTYPE_HOST || ctx->deviceType != LANDAU_CPU, ctx->comm, PETSC_ERR_ARG_WRONG, "CPU run with device data: use -mat_type aij"); 196 cellClosure = NULL; 197 } 198 PetscCall(PetscLogEventEnd(ctx->events[1], 0, 0, 0, 0)); 199 } else xdata = cellClosure = NULL; 200 201 /* do it */ 202 if (ctx->deviceType == LANDAU_KOKKOS) { 203 #if defined(PETSC_HAVE_KOKKOS) 204 PetscCall(LandauKokkosJacobian(ctx->plex, Nq, Nb, ctx->batch_sz, ctx->num_grids, numCells, Eq_m, cellClosure, xdata, &ctx->SData_d, shift, ctx->events, ctx->mat_offset, ctx->species_offset, subJ, JacP)); 205 #else 206 SETERRQ(ctx->comm, PETSC_ERR_ARG_WRONG, "-landau_device_type %s not built", "kokkos"); 207 #endif 208 } else { /* CPU version */ 209 PetscTabulation *Tf; // used for CPU and print info. Same on all grids and all species 210 PetscInt ip_offset[LANDAU_MAX_GRIDS + 1], ipf_offset[LANDAU_MAX_GRIDS + 1], elem_offset[LANDAU_MAX_GRIDS + 1], IPf_sz_glb, IPf_sz_tot, num_grids = ctx->num_grids, Nf[LANDAU_MAX_GRIDS]; 211 PetscReal *ff, *dudx, *dudy, *dudz, *invJ_a = (PetscReal *)ctx->SData_d.invJ, *xx = (PetscReal *)ctx->SData_d.x, *yy = (PetscReal *)ctx->SData_d.y, *zz = (PetscReal *)ctx->SData_d.z, *ww = (PetscReal *)ctx->SData_d.w; 212 PetscReal *nu_alpha = (PetscReal *)ctx->SData_d.alpha, *nu_beta = (PetscReal *)ctx->SData_d.beta, *invMass = (PetscReal *)ctx->SData_d.invMass; 213 PetscReal (*lambdas)[LANDAU_MAX_GRIDS][LANDAU_MAX_GRIDS] = (PetscReal (*)[LANDAU_MAX_GRIDS][LANDAU_MAX_GRIDS])ctx->SData_d.lambdas; 214 PetscSection section[LANDAU_MAX_GRIDS], globsection[LANDAU_MAX_GRIDS]; 215 PetscScalar *coo_vals = NULL; 216 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 217 PetscCall(DMGetLocalSection(ctx->plex[grid], §ion[grid])); 218 PetscCall(DMGetGlobalSection(ctx->plex[grid], &globsection[grid])); 219 PetscCall(PetscSectionGetNumFields(section[grid], &Nf[grid])); 220 } 221 /* count IPf size, etc */ 222 PetscCall(PetscDSGetTabulation(prob, &Tf)); // Bf, &Df same for all grids 223 const PetscReal *const BB = Tf[0]->T[0], *const DD = Tf[0]->T[1]; 224 ip_offset[0] = ipf_offset[0] = elem_offset[0] = 0; 225 for (PetscInt grid = 0; grid < num_grids; grid++) { 226 PetscInt nfloc = ctx->species_offset[grid + 1] - ctx->species_offset[grid]; 227 elem_offset[grid + 1] = elem_offset[grid] + numCells[grid]; 228 ip_offset[grid + 1] = ip_offset[grid] + numCells[grid] * Nq; 229 ipf_offset[grid + 1] = ipf_offset[grid] + Nq * nfloc * numCells[grid]; 230 } 231 IPf_sz_glb = ipf_offset[num_grids]; 232 IPf_sz_tot = IPf_sz_glb * ctx->batch_sz; 233 // prep COO 234 PetscCall(PetscMalloc1(ctx->SData_d.coo_size, &coo_vals)); // allocate every time? 235 if (shift == 0.0) { /* compute dynamic data f and df and init data for Jacobian */ 236 #if defined(PETSC_HAVE_THREADSAFETY) 237 double starttime, endtime; 238 starttime = MPI_Wtime(); 239 #endif 240 PetscCall(PetscLogEventBegin(ctx->events[8], 0, 0, 0, 0)); 241 PetscCall(PetscMalloc4(IPf_sz_tot, &ff, IPf_sz_tot, &dudx, IPf_sz_tot, &dudy, (dim == 3 ? IPf_sz_tot : 0), &dudz)); 242 // F df/dx 243 for (PetscInt tid = 0; tid < ctx->batch_sz * elem_offset[num_grids]; tid++) { // for each element 244 const PetscInt b_Nelem = elem_offset[num_grids], b_elem_idx = tid % b_Nelem, b_id = tid / b_Nelem; // b_id == OMP thd_id in batch 245 // find my grid: 246 PetscInt grid = 0; 247 while (b_elem_idx >= elem_offset[grid + 1]) grid++; // yuck search for grid 248 { 249 const PetscInt loc_nip = numCells[grid] * Nq, loc_Nf = ctx->species_offset[grid + 1] - ctx->species_offset[grid], loc_elem = b_elem_idx - elem_offset[grid]; 250 const PetscInt moffset = LAND_MOFFSET(b_id, grid, ctx->batch_sz, ctx->num_grids, ctx->mat_offset); //b_id*b_N + ctx->mat_offset[grid]; 251 PetscScalar *coef, coef_buff[LANDAU_MAX_SPECIES * LANDAU_MAX_NQND]; 252 PetscReal *invJe = &invJ_a[(ip_offset[grid] + loc_elem * Nq) * dim * dim]; // ingJ is static data on batch 0 253 PetscInt b, f, q; 254 if (cellClosure) { 255 coef = &cellClosure[b_id * IPf_sz_glb + ipf_offset[grid] + loc_elem * Nb * loc_Nf]; // this is const 256 } else { 257 coef = coef_buff; 258 for (f = 0; f < loc_Nf; ++f) { 259 LandauIdx *const Idxs = &maps[grid].gIdx[loc_elem][f][0]; 260 for (b = 0; b < Nb; ++b) { 261 PetscInt idx = Idxs[b]; 262 if (idx >= 0) { 263 coef[f * Nb + b] = xdata[idx + moffset]; 264 } else { 265 idx = -idx - 1; 266 coef[f * Nb + b] = 0; 267 for (q = 0; q < maps[grid].num_face; q++) { 268 PetscInt id = maps[grid].c_maps[idx][q].gid; 269 PetscScalar scale = maps[grid].c_maps[idx][q].scale; 270 coef[f * Nb + b] += scale * xdata[id + moffset]; 271 } 272 } 273 } 274 } 275 } 276 /* get f and df */ 277 for (PetscInt qi = 0; qi < Nq; qi++) { 278 const PetscReal *invJ = &invJe[qi * dim * dim]; 279 const PetscReal *Bq = &BB[qi * Nb]; 280 const PetscReal *Dq = &DD[qi * Nb * dim]; 281 PetscReal u_x[LANDAU_DIM]; 282 /* get f & df */ 283 for (f = 0; f < loc_Nf; ++f) { 284 const PetscInt idx = b_id * IPf_sz_glb + ipf_offset[grid] + f * loc_nip + loc_elem * Nq + qi; 285 PetscInt b, e; 286 PetscReal refSpaceDer[LANDAU_DIM]; 287 ff[idx] = 0.0; 288 for (PetscInt d = 0; d < LANDAU_DIM; ++d) refSpaceDer[d] = 0.0; 289 for (b = 0; b < Nb; ++b) { 290 const PetscInt cidx = b; 291 ff[idx] += Bq[cidx] * PetscRealPart(coef[f * Nb + cidx]); 292 for (PetscInt d = 0; d < dim; ++d) refSpaceDer[d] += Dq[cidx * dim + d] * PetscRealPart(coef[f * Nb + cidx]); 293 } 294 for (PetscInt d = 0; d < LANDAU_DIM; ++d) { 295 for (e = 0, u_x[d] = 0.0; e < LANDAU_DIM; ++e) u_x[d] += invJ[e * dim + d] * refSpaceDer[e]; 296 } 297 dudx[idx] = u_x[0]; 298 dudy[idx] = u_x[1]; 299 #if LANDAU_DIM == 3 300 dudz[idx] = u_x[2]; 301 #endif 302 } 303 } // q 304 } // grid 305 } // grid*batch 306 PetscCall(PetscLogEventEnd(ctx->events[8], 0, 0, 0, 0)); 307 #if defined(PETSC_HAVE_THREADSAFETY) 308 endtime = MPI_Wtime(); 309 if (ctx->stage) ctx->times[LANDAU_F_DF] += (endtime - starttime); 310 #endif 311 } // Jacobian setup 312 // assemble Jacobian (or mass) 313 for (PetscInt tid = 0; tid < ctx->batch_sz * elem_offset[num_grids]; tid++) { // for each element 314 const PetscInt b_Nelem = elem_offset[num_grids]; 315 const PetscInt glb_elem_idx = tid % b_Nelem, b_id = tid / b_Nelem; 316 PetscInt grid = 0; 317 #if defined(PETSC_HAVE_THREADSAFETY) 318 double starttime, endtime; 319 starttime = MPI_Wtime(); 320 #endif 321 while (glb_elem_idx >= elem_offset[grid + 1]) grid++; 322 { 323 const PetscInt loc_Nf = ctx->species_offset[grid + 1] - ctx->species_offset[grid], loc_elem = glb_elem_idx - elem_offset[grid]; 324 const PetscInt moffset = LAND_MOFFSET(b_id, grid, ctx->batch_sz, ctx->num_grids, ctx->mat_offset), totDim = loc_Nf * Nq, elemMatSize = totDim * totDim; 325 PetscScalar *elemMat; 326 const PetscReal *invJe = &invJ_a[(ip_offset[grid] + loc_elem * Nq) * dim * dim]; 327 PetscCall(PetscMalloc1(elemMatSize, &elemMat)); 328 PetscCall(PetscMemzero(elemMat, elemMatSize * sizeof(*elemMat))); 329 if (shift == 0.0) { // Jacobian 330 PetscCall(PetscLogEventBegin(ctx->events[4], 0, 0, 0, 0)); 331 } else { // mass 332 PetscCall(PetscLogEventBegin(ctx->events[16], 0, 0, 0, 0)); 333 } 334 for (PetscInt qj = 0; qj < Nq; ++qj) { 335 const PetscInt jpidx_glb = ip_offset[grid] + qj + loc_elem * Nq; 336 PetscReal g0[LANDAU_MAX_SPECIES], g2[LANDAU_MAX_SPECIES][LANDAU_DIM], g3[LANDAU_MAX_SPECIES][LANDAU_DIM][LANDAU_DIM]; // could make a LANDAU_MAX_SPECIES_GRID ~ number of ions - 1 337 PetscInt d, d2, dp, d3, IPf_idx; 338 if (shift == 0.0) { // Jacobian 339 const PetscReal *const invJj = &invJe[qj * dim * dim]; 340 PetscReal gg2[LANDAU_MAX_SPECIES][LANDAU_DIM], gg3[LANDAU_MAX_SPECIES][LANDAU_DIM][LANDAU_DIM], gg2_temp[LANDAU_DIM], gg3_temp[LANDAU_DIM][LANDAU_DIM]; 341 const PetscReal vj[3] = {xx[jpidx_glb], yy[jpidx_glb], zz ? zz[jpidx_glb] : 0}, wj = ww[jpidx_glb]; 342 // create g2 & g3 343 for (d = 0; d < LANDAU_DIM; d++) { // clear accumulation data D & K 344 gg2_temp[d] = 0; 345 for (d2 = 0; d2 < LANDAU_DIM; d2++) gg3_temp[d][d2] = 0; 346 } 347 /* inner beta reduction */ 348 IPf_idx = 0; 349 for (PetscInt grid_r = 0, f_off = 0, ipidx = 0; grid_r < ctx->num_grids; grid_r++, f_off = ctx->species_offset[grid_r]) { // IPf_idx += nip_loc_r*Nfloc_r 350 PetscInt nip_loc_r = numCells[grid_r] * Nq, Nfloc_r = Nf[grid_r]; 351 for (PetscInt ei_r = 0; ei_r < numCells[grid_r]; ++ei_r) { 352 for (PetscInt qi = 0; qi < Nq; qi++, ipidx++) { 353 const PetscReal wi = ww[ipidx], x = xx[ipidx], y = yy[ipidx]; 354 PetscReal temp1[3] = {0, 0, 0}, temp2 = 0; 355 #if LANDAU_DIM == 2 356 PetscReal Ud[2][2], Uk[2][2], mask = (PetscAbs(vj[0] - x) < 100 * PETSC_SQRT_MACHINE_EPSILON && PetscAbs(vj[1] - y) < 100 * PETSC_SQRT_MACHINE_EPSILON) ? 0. : 1.; 357 LandauTensor2D(vj, x, y, Ud, Uk, mask); 358 #else 359 PetscReal U[3][3], z = zz[ipidx], mask = (PetscAbs(vj[0] - x) < 100 * PETSC_SQRT_MACHINE_EPSILON && PetscAbs(vj[1] - y) < 100 * PETSC_SQRT_MACHINE_EPSILON && PetscAbs(vj[2] - z) < 100 * PETSC_SQRT_MACHINE_EPSILON) ? 0. : 1.; 360 if (ctx->use_relativistic_corrections) { 361 LandauTensor3DRelativistic(vj, x, y, z, U, mask, C_0(ctx->v_0)); 362 } else { 363 LandauTensor3D(vj, x, y, z, U, mask); 364 } 365 #endif 366 for (PetscInt f = 0; f < Nfloc_r; ++f) { 367 const PetscInt idx = b_id * IPf_sz_glb + ipf_offset[grid_r] + f * nip_loc_r + ei_r * Nq + qi; 368 369 temp1[0] += dudx[idx] * nu_beta[f + f_off] * invMass[f + f_off] * (*lambdas)[grid][grid_r]; 370 temp1[1] += dudy[idx] * nu_beta[f + f_off] * invMass[f + f_off] * (*lambdas)[grid][grid_r]; 371 #if LANDAU_DIM == 3 372 temp1[2] += dudz[idx] * nu_beta[f + f_off] * invMass[f + f_off] * (*lambdas)[grid][grid_r]; 373 #endif 374 temp2 += ff[idx] * nu_beta[f + f_off] * (*lambdas)[grid][grid_r]; 375 } 376 temp1[0] *= wi; 377 temp1[1] *= wi; 378 #if LANDAU_DIM == 3 379 temp1[2] *= wi; 380 #endif 381 temp2 *= wi; 382 #if LANDAU_DIM == 2 383 for (d2 = 0; d2 < 2; d2++) { 384 for (d3 = 0; d3 < 2; ++d3) { 385 /* K = U * grad(f): g2=e: i,A */ 386 gg2_temp[d2] += Uk[d2][d3] * temp1[d3]; 387 /* D = -U * (I \kron (fx)): g3=f: i,j,A */ 388 gg3_temp[d2][d3] += Ud[d2][d3] * temp2; 389 } 390 } 391 #else 392 for (d2 = 0; d2 < 3; ++d2) { 393 for (d3 = 0; d3 < 3; ++d3) { 394 /* K = U * grad(f): g2 = e: i,A */ 395 gg2_temp[d2] += U[d2][d3] * temp1[d3]; 396 /* D = -U * (I \kron (fx)): g3 = f: i,j,A */ 397 gg3_temp[d2][d3] += U[d2][d3] * temp2; 398 } 399 } 400 #endif 401 } // qi 402 } // ei_r 403 IPf_idx += nip_loc_r * Nfloc_r; 404 } /* grid_r - IPs */ 405 PetscCheck(IPf_idx == IPf_sz_glb, PETSC_COMM_SELF, PETSC_ERR_PLIB, "IPf_idx != IPf_sz %" PetscInt_FMT " %" PetscInt_FMT, IPf_idx, IPf_sz_glb); 406 // add alpha and put in gg2/3 407 for (PetscInt fieldA = 0, f_off = ctx->species_offset[grid]; fieldA < loc_Nf; ++fieldA) { 408 for (d2 = 0; d2 < LANDAU_DIM; d2++) { 409 gg2[fieldA][d2] = gg2_temp[d2] * nu_alpha[fieldA + f_off]; 410 for (d3 = 0; d3 < LANDAU_DIM; d3++) gg3[fieldA][d2][d3] = -gg3_temp[d2][d3] * nu_alpha[fieldA + f_off] * invMass[fieldA + f_off]; 411 } 412 } 413 /* add electric field term once per IP */ 414 for (PetscInt fieldA = 0, f_off = ctx->species_offset[grid]; fieldA < loc_Nf; ++fieldA) gg2[fieldA][LANDAU_DIM - 1] += Eq_m[fieldA + f_off]; 415 /* Jacobian transform - g2, g3 */ 416 for (PetscInt fieldA = 0; fieldA < loc_Nf; ++fieldA) { 417 for (d = 0; d < dim; ++d) { 418 g2[fieldA][d] = 0.0; 419 for (d2 = 0; d2 < dim; ++d2) { 420 g2[fieldA][d] += invJj[d * dim + d2] * gg2[fieldA][d2]; 421 g3[fieldA][d][d2] = 0.0; 422 for (d3 = 0; d3 < dim; ++d3) { 423 for (dp = 0; dp < dim; ++dp) g3[fieldA][d][d2] += invJj[d * dim + d3] * gg3[fieldA][d3][dp] * invJj[d2 * dim + dp]; 424 } 425 g3[fieldA][d][d2] *= wj; 426 } 427 g2[fieldA][d] *= wj; 428 } 429 } 430 } else { // mass 431 PetscReal wj = ww[jpidx_glb]; 432 /* Jacobian transform - g0 */ 433 for (PetscInt fieldA = 0; fieldA < loc_Nf; ++fieldA) { 434 if (dim == 2) { 435 g0[fieldA] = wj * shift * 2. * PETSC_PI; // move this to below and remove g0 436 } else { 437 g0[fieldA] = wj * shift; // move this to below and remove g0 438 } 439 } 440 } 441 /* FE matrix construction */ 442 { 443 PetscInt fieldA, d, f, d2, g; 444 const PetscReal *BJq = &BB[qj * Nb], *DIq = &DD[qj * Nb * dim]; 445 /* assemble - on the diagonal (I,I) */ 446 for (fieldA = 0; fieldA < loc_Nf; fieldA++) { 447 for (f = 0; f < Nb; f++) { 448 const PetscInt i = fieldA * Nb + f; /* Element matrix row */ 449 for (g = 0; g < Nb; ++g) { 450 const PetscInt j = fieldA * Nb + g; /* Element matrix column */ 451 const PetscInt fOff = i * totDim + j; 452 if (shift == 0.0) { 453 for (d = 0; d < dim; ++d) { 454 elemMat[fOff] += DIq[f * dim + d] * g2[fieldA][d] * BJq[g]; 455 for (d2 = 0; d2 < dim; ++d2) elemMat[fOff] += DIq[f * dim + d] * g3[fieldA][d][d2] * DIq[g * dim + d2]; 456 } 457 } else { // mass 458 elemMat[fOff] += BJq[f] * g0[fieldA] * BJq[g]; 459 } 460 } 461 } 462 } 463 } 464 } /* qj loop */ 465 if (shift == 0.0) { // Jacobian 466 PetscCall(PetscLogEventEnd(ctx->events[4], 0, 0, 0, 0)); 467 } else { 468 PetscCall(PetscLogEventEnd(ctx->events[16], 0, 0, 0, 0)); 469 } 470 #if defined(PETSC_HAVE_THREADSAFETY) 471 endtime = MPI_Wtime(); 472 if (ctx->stage) ctx->times[LANDAU_KERNEL] += (endtime - starttime); 473 #endif 474 /* assemble matrix */ 475 if (!container) { 476 PetscInt cStart; 477 PetscCall(PetscLogEventBegin(ctx->events[6], 0, 0, 0, 0)); 478 PetscCall(DMPlexGetHeightStratum(ctx->plex[grid], 0, &cStart, NULL)); 479 PetscCall(DMPlexMatSetClosure(ctx->plex[grid], section[grid], globsection[grid], subJ[LAND_PACK_IDX(b_id, grid)], loc_elem + cStart, elemMat, ADD_VALUES)); 480 PetscCall(PetscLogEventEnd(ctx->events[6], 0, 0, 0, 0)); 481 } else { // GPU like assembly for debugging 482 PetscInt fieldA, q, f, g, d, nr, nc, rows0[LANDAU_MAX_Q_FACE] = {0}, cols0[LANDAU_MAX_Q_FACE] = {0}, rows[LANDAU_MAX_Q_FACE], cols[LANDAU_MAX_Q_FACE]; 483 PetscScalar vals[LANDAU_MAX_Q_FACE * LANDAU_MAX_Q_FACE] = {0}, row_scale[LANDAU_MAX_Q_FACE] = {0}, col_scale[LANDAU_MAX_Q_FACE] = {0}; 484 LandauIdx *coo_elem_offsets = (LandauIdx *)ctx->SData_d.coo_elem_offsets, *coo_elem_fullNb = (LandauIdx *)ctx->SData_d.coo_elem_fullNb, (*coo_elem_point_offsets)[LANDAU_MAX_NQND + 1] = (LandauIdx(*)[LANDAU_MAX_NQND + 1]) ctx->SData_d.coo_elem_point_offsets; 485 /* assemble - from the diagonal (I,I) in this format for DMPlexMatSetClosure */ 486 for (fieldA = 0; fieldA < loc_Nf; fieldA++) { 487 LandauIdx *const Idxs = &maps[grid].gIdx[loc_elem][fieldA][0]; 488 for (f = 0; f < Nb; f++) { 489 PetscInt idx = Idxs[f]; 490 if (idx >= 0) { 491 nr = 1; 492 rows0[0] = idx; 493 row_scale[0] = 1.; 494 } else { 495 idx = -idx - 1; 496 for (q = 0, nr = 0; q < maps[grid].num_face; q++, nr++) { 497 if (maps[grid].c_maps[idx][q].gid < 0) break; 498 rows0[q] = maps[grid].c_maps[idx][q].gid; 499 row_scale[q] = maps[grid].c_maps[idx][q].scale; 500 } 501 } 502 for (g = 0; g < Nb; ++g) { 503 idx = Idxs[g]; 504 if (idx >= 0) { 505 nc = 1; 506 cols0[0] = idx; 507 col_scale[0] = 1.; 508 } else { 509 idx = -idx - 1; 510 nc = maps[grid].num_face; 511 for (q = 0, nc = 0; q < maps[grid].num_face; q++, nc++) { 512 if (maps[grid].c_maps[idx][q].gid < 0) break; 513 cols0[q] = maps[grid].c_maps[idx][q].gid; 514 col_scale[q] = maps[grid].c_maps[idx][q].scale; 515 } 516 } 517 const PetscInt i = fieldA * Nb + f; /* Element matrix row */ 518 const PetscInt j = fieldA * Nb + g; /* Element matrix column */ 519 const PetscScalar Aij = elemMat[i * totDim + j]; 520 if (coo_vals) { // mirror (i,j) in CreateStaticGPUData 521 const PetscInt fullNb = coo_elem_fullNb[glb_elem_idx], fullNb2 = fullNb * fullNb; 522 const PetscInt idx0 = b_id * coo_elem_offsets[elem_offset[num_grids]] + coo_elem_offsets[glb_elem_idx] + fieldA * fullNb2 + fullNb * coo_elem_point_offsets[glb_elem_idx][f] + nr * coo_elem_point_offsets[glb_elem_idx][g]; 523 for (PetscInt q = 0, idx2 = idx0; q < nr; q++) { 524 for (PetscInt d = 0; d < nc; d++, idx2++) coo_vals[idx2] = row_scale[q] * col_scale[d] * Aij; 525 } 526 } else { 527 for (q = 0; q < nr; q++) rows[q] = rows0[q] + moffset; 528 for (d = 0; d < nc; d++) cols[d] = cols0[d] + moffset; 529 for (q = 0; q < nr; q++) { 530 for (d = 0; d < nc; d++) vals[q * nc + d] = row_scale[q] * col_scale[d] * Aij; 531 } 532 PetscCall(MatSetValues(JacP, nr, rows, nc, cols, vals, ADD_VALUES)); 533 } 534 } 535 } 536 } 537 } 538 if (loc_elem == -1) { 539 PetscCall(PetscPrintf(ctx->comm, "CPU Element matrix\n")); 540 for (PetscInt d = 0; d < totDim; ++d) { 541 for (PetscInt f = 0; f < totDim; ++f) PetscCall(PetscPrintf(ctx->comm, " %12.5e", (double)PetscRealPart(elemMat[d * totDim + f]))); 542 PetscCall(PetscPrintf(ctx->comm, "\n")); 543 } 544 exit(12); 545 } 546 PetscCall(PetscFree(elemMat)); 547 } /* grid */ 548 } /* outer element & batch loop */ 549 if (shift == 0.0) { // mass 550 PetscCall(PetscFree4(ff, dudx, dudy, dudz)); 551 } 552 if (!container) { // 'CPU' assembly move nest matrix to global JacP 553 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { // OpenMP 554 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 555 const PetscInt moffset = LAND_MOFFSET(b_id, grid, ctx->batch_sz, ctx->num_grids, ctx->mat_offset); // b_id*b_N + ctx->mat_offset[grid]; 556 PetscInt nloc, nzl, colbuf[1024], row; 557 const PetscInt *cols; 558 const PetscScalar *vals; 559 Mat B = subJ[LAND_PACK_IDX(b_id, grid)]; 560 PetscCall(MatAssemblyBegin(B, MAT_FINAL_ASSEMBLY)); 561 PetscCall(MatAssemblyEnd(B, MAT_FINAL_ASSEMBLY)); 562 PetscCall(MatGetSize(B, &nloc, NULL)); 563 for (PetscInt i = 0; i < nloc; i++) { 564 PetscCall(MatGetRow(B, i, &nzl, &cols, &vals)); 565 PetscCheck(nzl <= 1024, PetscObjectComm((PetscObject)B), PETSC_ERR_PLIB, "Row too big: %" PetscInt_FMT, nzl); 566 for (PetscInt j = 0; j < nzl; j++) colbuf[j] = moffset + cols[j]; 567 row = moffset + i; 568 PetscCall(MatSetValues(JacP, 1, &row, nzl, colbuf, vals, ADD_VALUES)); 569 PetscCall(MatRestoreRow(B, i, &nzl, &cols, &vals)); 570 } 571 PetscCall(MatDestroy(&B)); 572 } 573 } 574 } 575 if (coo_vals) { 576 PetscCall(MatSetValuesCOO(JacP, coo_vals, ADD_VALUES)); 577 PetscCall(PetscFree(coo_vals)); 578 } 579 } /* CPU version */ 580 PetscCall(MatAssemblyBegin(JacP, MAT_FINAL_ASSEMBLY)); 581 PetscCall(MatAssemblyEnd(JacP, MAT_FINAL_ASSEMBLY)); 582 /* clean up */ 583 if (cellClosure) PetscCall(PetscFree(cellClosure)); 584 if (xdata) PetscCall(VecRestoreArrayReadAndMemType(a_X, &xdata)); 585 PetscFunctionReturn(PETSC_SUCCESS); 586 } 587 588 static PetscErrorCode GeometryDMLandau(DM base, PetscInt point, PetscInt dim, const PetscReal abc[], PetscReal xyz[], void *a_ctx) 589 { 590 PetscReal r = abc[0], z = abc[1]; 591 LandauCtx *ctx = (LandauCtx *)a_ctx; 592 593 PetscFunctionBegin; 594 if (ctx->sphere && dim == 3) { // make sphere: works for one AMR and Q2 595 PetscInt nzero = 0, idx = 0; 596 xyz[0] = r; 597 xyz[1] = z; 598 xyz[2] = abc[2]; 599 for (PetscInt i = 0; i < 3; i++) { 600 if (PetscAbs(xyz[i]) < PETSC_SQRT_MACHINE_EPSILON) nzero++; 601 else idx = i; 602 } 603 if (nzero == 2) xyz[idx] *= 1.732050807568877; // sqrt(3) 604 else if (nzero == 1) { 605 for (PetscInt i = 0; i < 3; i++) xyz[i] *= 1.224744871391589; // sqrt(3/2) 606 } 607 } else { 608 xyz[0] = r; 609 xyz[1] = z; 610 if (dim == 3) xyz[2] = abc[2]; 611 } 612 PetscFunctionReturn(PETSC_SUCCESS); 613 } 614 615 /* create DMComposite of meshes for each species group */ 616 static PetscErrorCode LandauDMCreateVMeshes(MPI_Comm comm_self, const PetscInt dim, const char prefix[], LandauCtx *ctx, DM pack) 617 { 618 PetscFunctionBegin; 619 { /* p4est, quads */ 620 /* Create plex mesh of Landau domain */ 621 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 622 PetscReal par_radius = ctx->radius_par[grid], perp_radius = ctx->radius_perp[grid]; 623 if (!ctx->sphere && !ctx->simplex) { // 2 or 3D (only 3D option) 624 PetscReal lo[] = {-perp_radius, -par_radius, -par_radius}, hi[] = {perp_radius, par_radius, par_radius}; 625 DMBoundaryType periodicity[3] = {DM_BOUNDARY_NONE, dim == 2 ? DM_BOUNDARY_NONE : DM_BOUNDARY_NONE, DM_BOUNDARY_NONE}; 626 if (dim == 2) lo[0] = 0; 627 else { 628 lo[1] = -perp_radius; 629 hi[1] = perp_radius; // 3D y is a perp 630 } 631 PetscCall(DMPlexCreateBoxMesh(comm_self, dim, PETSC_FALSE, ctx->cells0, lo, hi, periodicity, PETSC_TRUE, 0, PETSC_TRUE, &ctx->plex[grid])); // TODO: make composite and create dm[grid] here 632 PetscCall(DMLocalizeCoordinates(ctx->plex[grid])); /* needed for periodic */ 633 if (dim == 3) PetscCall(PetscObjectSetName((PetscObject)ctx->plex[grid], "cube")); 634 else PetscCall(PetscObjectSetName((PetscObject)ctx->plex[grid], "half-plane")); 635 } else if (dim == 2) { 636 size_t len; 637 PetscCall(PetscStrlen(ctx->filename, &len)); 638 if (len) { 639 Vec coords; 640 PetscScalar *x; 641 PetscInt N; 642 char str[] = "-dm_landau_view_file_0"; 643 str[21] += grid; 644 PetscCall(DMPlexCreateFromFile(comm_self, ctx->filename, "plexland.c", PETSC_TRUE, &ctx->plex[grid])); 645 PetscCall(DMPlexOrient(ctx->plex[grid])); 646 PetscCall(DMGetCoordinatesLocal(ctx->plex[grid], &coords)); 647 PetscCall(VecGetSize(coords, &N)); 648 PetscCall(VecGetArray(coords, &x)); 649 /* scale by domain size */ 650 for (PetscInt i = 0; i < N; i += 2) { 651 x[i + 0] *= ctx->radius_perp[grid]; 652 x[i + 1] *= ctx->radius_par[grid]; 653 } 654 PetscCall(VecRestoreArray(coords, &x)); 655 PetscCall(PetscObjectSetName((PetscObject)ctx->plex[grid], ctx->filename)); 656 PetscCall(PetscInfo(ctx->plex[grid], "%" PetscInt_FMT ") Read %s mesh file (%s)\n", grid, ctx->filename, str)); 657 PetscCall(DMViewFromOptions(ctx->plex[grid], NULL, str)); 658 } else { // simplex forces a sphere 659 PetscInt numCells = ctx->simplex ? 12 : 6, cell_size = ctx->simplex ? 3 : 4, j; 660 const PetscInt numVerts = 11; 661 PetscInt cellsT[][4] = { 662 {0, 1, 6, 5 }, 663 {1, 2, 7, 6 }, 664 {2, 3, 8, 7 }, 665 {3, 4, 9, 8 }, 666 {5, 6, 7, 10}, 667 {10, 7, 8, 9 } 668 }; 669 PetscInt cellsS[][3] = { 670 {0, 1, 6 }, 671 {1, 2, 6 }, 672 {6, 2, 7 }, 673 {7, 2, 8 }, 674 {8, 2, 3 }, 675 {8, 3, 4 }, 676 {0, 6, 5 }, 677 {5, 6, 7 }, 678 {5, 7, 10}, 679 {10, 7, 9 }, 680 {9, 7, 8 }, 681 {9, 8, 4 } 682 }; 683 const PetscInt *pcell = (const PetscInt *)(ctx->simplex ? &cellsS[0][0] : &cellsT[0][0]); 684 PetscReal coords[11][2], *flatCoords = &coords[0][0]; 685 PetscReal rad = ctx->radius[grid]; 686 for (j = 0; j < 5; j++) { // outside edge 687 PetscReal z, r, theta = -PETSC_PI / 2 + (j % 5) * PETSC_PI / 4; 688 r = rad * PetscCosReal(theta); 689 coords[j][0] = r; 690 z = rad * PetscSinReal(theta); 691 coords[j][1] = z; 692 } 693 coords[j][0] = 0; 694 coords[j++][1] = -rad * ctx->sphere_inner_radius_90degree[grid]; 695 coords[j][0] = rad * ctx->sphere_inner_radius_45degree[grid] * 0.707106781186548; 696 coords[j++][1] = -rad * ctx->sphere_inner_radius_45degree[grid] * 0.707106781186548; 697 coords[j][0] = rad * ctx->sphere_inner_radius_90degree[grid]; 698 coords[j++][1] = 0; 699 coords[j][0] = rad * ctx->sphere_inner_radius_45degree[grid] * 0.707106781186548; 700 coords[j++][1] = rad * ctx->sphere_inner_radius_45degree[grid] * 0.707106781186548; 701 coords[j][0] = 0; 702 coords[j++][1] = rad * ctx->sphere_inner_radius_90degree[grid]; 703 coords[j][0] = 0; 704 coords[j++][1] = 0; 705 PetscCall(DMPlexCreateFromCellListPetsc(comm_self, 2, numCells, numVerts, cell_size, ctx->interpolate, pcell, 2, flatCoords, &ctx->plex[grid])); 706 PetscCall(PetscObjectSetName((PetscObject)ctx->plex[grid], "semi-circle")); 707 PetscCall(PetscInfo(ctx->plex[grid], "\t%" PetscInt_FMT ") Make circle %s mesh\n", grid, ctx->simplex ? "simplex" : "tensor")); 708 } 709 } else { 710 PetscCheck(dim == 3 && ctx->sphere && !ctx->simplex, ctx->comm, PETSC_ERR_ARG_WRONG, "not: dim == 3 && ctx->sphere && !ctx->simplex"); 711 PetscReal rad = ctx->radius[grid] / 1.732050807568877, inner_rad = rad * ctx->sphere_inner_radius_45degree[grid], outer_rad = rad; 712 const PetscInt numCells = 7, cell_size = 8, numVerts = 16; 713 const PetscInt cells[][8] = { 714 {0, 3, 2, 1, 4, 5, 6, 7 }, 715 {0, 4, 5, 1, 8, 9, 13, 12}, 716 {1, 5, 6, 2, 9, 10, 14, 13}, 717 {2, 6, 7, 3, 10, 11, 15, 14}, 718 {0, 3, 7, 4, 8, 12, 15, 11}, 719 {0, 1, 2, 3, 8, 11, 10, 9 }, 720 {4, 7, 6, 5, 12, 13, 14, 15} 721 }; 722 PetscReal coords[16 /* numVerts */][3]; 723 for (PetscInt j = 0; j < 4; j++) { // inner edge, low 724 coords[j][0] = inner_rad * (j == 0 || j == 3 ? 1 : -1); 725 coords[j][1] = inner_rad * (j / 2 < 1 ? 1 : -1); 726 coords[j][2] = inner_rad * -1; 727 } 728 for (PetscInt j = 0, jj = 4; j < 4; j++, jj++) { // inner edge, hi 729 coords[jj][0] = inner_rad * (j == 0 || j == 3 ? 1 : -1); 730 coords[jj][1] = inner_rad * (j / 2 < 1 ? 1 : -1); 731 coords[jj][2] = inner_rad * 1; 732 } 733 for (PetscInt j = 0, jj = 8; j < 4; j++, jj++) { // outer edge, low 734 coords[jj][0] = outer_rad * (j == 0 || j == 3 ? 1 : -1); 735 coords[jj][1] = outer_rad * (j / 2 < 1 ? 1 : -1); 736 coords[jj][2] = outer_rad * -1; 737 } 738 for (PetscInt j = 0, jj = 12; j < 4; j++, jj++) { // outer edge, hi 739 coords[jj][0] = outer_rad * (j == 0 || j == 3 ? 1 : -1); 740 coords[jj][1] = outer_rad * (j / 2 < 1 ? 1 : -1); 741 coords[jj][2] = outer_rad * 1; 742 } 743 PetscCall(DMPlexCreateFromCellListPetsc(comm_self, 3, numCells, numVerts, cell_size, ctx->interpolate, (const PetscInt *)cells, 3, (const PetscReal *)coords, &ctx->plex[grid])); 744 PetscCall(PetscObjectSetName((PetscObject)ctx->plex[grid], "cubed sphere")); 745 PetscCall(PetscInfo(ctx->plex[grid], "\t%" PetscInt_FMT ") Make cubed sphere %s mesh\n", grid, ctx->simplex ? "simplex" : "tensor")); 746 } 747 PetscCall(DMSetFromOptions(ctx->plex[grid])); 748 } // grid loop 749 PetscCall(PetscObjectSetOptionsPrefix((PetscObject)pack, prefix)); 750 { /* convert to p4est (or whatever), wait for discretization to create pack */ 751 char convType[256]; 752 PetscBool flg; 753 754 PetscOptionsBegin(ctx->comm, prefix, "Mesh conversion options", "DMPLEX"); 755 PetscCall(PetscOptionsFList("-dm_landau_type", "Convert DMPlex to another format (p4est)", "plexland.c", DMList, DMPLEX, convType, 256, &flg)); 756 PetscOptionsEnd(); 757 if (flg) { 758 ctx->use_p4est = PETSC_TRUE; /* flag for Forest */ 759 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 760 DM dmforest; 761 PetscBool isForest; 762 763 PetscCall(DMConvert(ctx->plex[grid], convType, &dmforest)); 764 PetscCheck(dmforest, ctx->comm, PETSC_ERR_PLIB, "Convert failed?"); 765 PetscCall(PetscObjectSetOptionsPrefix((PetscObject)dmforest, prefix)); 766 PetscCall(DMIsForest(dmforest, &isForest)); 767 PetscCheck(isForest, ctx->comm, PETSC_ERR_PLIB, "Converted to non Forest?"); 768 if (ctx->sphere) PetscCall(DMForestSetBaseCoordinateMapping(dmforest, GeometryDMLandau, ctx)); 769 PetscCall(DMDestroy(&ctx->plex[grid])); 770 ctx->plex[grid] = dmforest; // Forest for adaptivity 771 } 772 } else ctx->use_p4est = PETSC_FALSE; /* flag for Forest */ 773 } 774 } /* non-file */ 775 PetscCall(DMSetDimension(pack, dim)); 776 PetscCall(PetscObjectSetName((PetscObject)pack, "Mesh")); 777 PetscCall(DMSetApplicationContext(pack, ctx)); 778 PetscFunctionReturn(PETSC_SUCCESS); 779 } 780 781 static PetscErrorCode SetupDS(DM pack, PetscInt dim, PetscInt grid, LandauCtx *ctx) 782 { 783 PetscInt ii, i0; 784 char buf[256]; 785 PetscSection section; 786 787 PetscFunctionBegin; 788 for (ii = ctx->species_offset[grid], i0 = 0; ii < ctx->species_offset[grid + 1]; ii++, i0++) { 789 if (ii == 0) PetscCall(PetscSNPrintf(buf, sizeof(buf), "e")); 790 else PetscCall(PetscSNPrintf(buf, sizeof(buf), "i%" PetscInt_FMT, ii)); 791 /* Setup Discretization - FEM */ 792 PetscCall(PetscFECreateDefault(PETSC_COMM_SELF, dim, 1, ctx->simplex, NULL, PETSC_DECIDE, &ctx->fe[ii])); 793 PetscCall(PetscObjectSetName((PetscObject)ctx->fe[ii], buf)); 794 PetscCall(DMSetField(ctx->plex[grid], i0, NULL, (PetscObject)ctx->fe[ii])); 795 } 796 PetscCall(DMCreateDS(ctx->plex[grid])); 797 PetscCall(DMGetLocalSection(ctx->plex[grid], §ion)); 798 for (PetscInt ii = ctx->species_offset[grid], i0 = 0; ii < ctx->species_offset[grid + 1]; ii++, i0++) { 799 if (ii == 0) PetscCall(PetscSNPrintf(buf, sizeof(buf), "se")); 800 else PetscCall(PetscSNPrintf(buf, sizeof(buf), "si%" PetscInt_FMT, ii)); 801 PetscCall(PetscSectionSetComponentName(section, i0, 0, buf)); 802 } 803 PetscFunctionReturn(PETSC_SUCCESS); 804 } 805 806 /* Define a Maxwellian function for testing out the operator. */ 807 808 /* Using cartesian velocity space coordinates, the particle */ 809 /* density, [1/m^3], is defined according to */ 810 811 /* $$ n=\int_{R^3} dv^3 \left(\frac{m}{2\pi T}\right)^{3/2}\exp [- mv^2/(2T)] $$ */ 812 813 /* Using some constant, c, we normalize the velocity vector into a */ 814 /* dimensionless variable according to v=c*x. Thus the density, $n$, becomes */ 815 816 /* $$ n=\int_{R^3} dx^3 \left(\frac{mc^2}{2\pi T}\right)^{3/2}\exp [- mc^2/(2T)*x^2] $$ */ 817 818 /* Defining $\theta=2T/mc^2$, we thus find that the probability density */ 819 /* for finding the particle within the interval in a box dx^3 around x is */ 820 821 /* f(x;\theta)=\left(\frac{1}{\pi\theta}\right)^{3/2} \exp [ -x^2/\theta ] */ 822 823 typedef struct { 824 PetscReal v_0; 825 PetscReal kT_m; 826 PetscReal n; 827 PetscReal shift; 828 } MaxwellianCtx; 829 830 static PetscErrorCode maxwellian(PetscInt dim, PetscReal time, const PetscReal x[], PetscInt Nf_dummy, PetscScalar *u, void *actx) 831 { 832 MaxwellianCtx *mctx = (MaxwellianCtx *)actx; 833 PetscInt i; 834 PetscReal v2 = 0, theta = 2 * mctx->kT_m / (mctx->v_0 * mctx->v_0), shift; /* theta = 2kT/mc^2 */ 835 836 PetscFunctionBegin; 837 /* compute the exponents, v^2 */ 838 for (i = 0; i < dim; ++i) v2 += x[i] * x[i]; 839 /* evaluate the Maxwellian */ 840 if (mctx->shift < 0) shift = -mctx->shift; 841 else { 842 u[0] = mctx->n * PetscPowReal(PETSC_PI * theta, -1.5) * (PetscExpReal(-v2 / theta)); 843 shift = mctx->shift; 844 } 845 if (shift != 0.) { 846 v2 = 0; 847 for (i = 0; i < dim - 1; ++i) v2 += x[i] * x[i]; 848 v2 += (x[dim - 1] - shift) * (x[dim - 1] - shift); 849 /* evaluate the shifted Maxwellian */ 850 u[0] += mctx->n * PetscPowReal(PETSC_PI * theta, -1.5) * (PetscExpReal(-v2 / theta)); 851 } 852 PetscFunctionReturn(PETSC_SUCCESS); 853 } 854 855 /*@ 856 DMPlexLandauAddMaxwellians - Add a Maxwellian distribution to a state 857 858 Collective 859 860 Input Parameters: 861 + dm - The mesh (local) 862 . time - Current time 863 . temps - Temperatures of each species (global) 864 . ns - Number density of each species (global) 865 . grid - index into current grid - just used for offset into `temp` and `ns` 866 . b_id - batch index 867 . n_batch - number of batches 868 - actx - Landau context 869 870 Output Parameter: 871 . X - The state (local to this grid) 872 873 Level: beginner 874 875 .seealso: `DMPlexLandauCreateVelocitySpace()` 876 @*/ 877 PetscErrorCode DMPlexLandauAddMaxwellians(DM dm, Vec X, PetscReal time, PetscReal temps[], PetscReal ns[], PetscInt grid, PetscInt b_id, PetscInt n_batch, void *actx) 878 { 879 LandauCtx *ctx = (LandauCtx *)actx; 880 PetscErrorCode (*initu[LANDAU_MAX_SPECIES])(PetscInt, PetscReal, const PetscReal[], PetscInt, PetscScalar[], void *); 881 PetscInt dim; 882 MaxwellianCtx *mctxs[LANDAU_MAX_SPECIES], data[LANDAU_MAX_SPECIES]; 883 884 PetscFunctionBegin; 885 PetscCall(DMGetDimension(dm, &dim)); 886 if (!ctx) PetscCall(DMGetApplicationContext(dm, &ctx)); 887 for (PetscInt ii = ctx->species_offset[grid], i0 = 0; ii < ctx->species_offset[grid + 1]; ii++, i0++) { 888 mctxs[i0] = &data[i0]; 889 data[i0].v_0 = ctx->v_0; // v_0 same for all grids 890 data[i0].kT_m = ctx->k * temps[ii] / ctx->masses[ii]; /* kT/m */ 891 data[i0].n = ns[ii]; 892 initu[i0] = maxwellian; 893 data[i0].shift = 0; 894 } 895 data[0].shift = ctx->electronShift; 896 /* need to make ADD_ALL_VALUES work - TODO */ 897 PetscCall(DMProjectFunction(dm, time, initu, (void **)mctxs, INSERT_ALL_VALUES, X)); 898 PetscFunctionReturn(PETSC_SUCCESS); 899 } 900 901 /* 902 LandauSetInitialCondition - Adds Maxwellians with context 903 904 Collective 905 906 Input Parameters: 907 . dm - The mesh 908 - grid - index into current grid - just used for offset into temp and ns 909 . b_id - batch index 910 - n_batch - number of batches 911 + actx - Landau context with T and n 912 913 Output Parameter: 914 . X - The state 915 916 Level: beginner 917 918 .seealso: `DMPlexLandauCreateVelocitySpace()`, `DMPlexLandauAddMaxwellians()` 919 */ 920 static PetscErrorCode LandauSetInitialCondition(DM dm, Vec X, PetscInt grid, PetscInt b_id, PetscInt n_batch, void *actx) 921 { 922 LandauCtx *ctx = (LandauCtx *)actx; 923 924 PetscFunctionBegin; 925 if (!ctx) PetscCall(DMGetApplicationContext(dm, &ctx)); 926 PetscCall(VecZeroEntries(X)); 927 PetscCall(DMPlexLandauAddMaxwellians(dm, X, 0.0, ctx->thermal_temps, ctx->n, grid, b_id, n_batch, ctx)); 928 PetscFunctionReturn(PETSC_SUCCESS); 929 } 930 931 // adapt a level once. Forest in/out 932 #if defined(PETSC_USE_INFO) 933 static const char *s_refine_names[] = {"RE", "Z1", "Origin", "Z2", "Uniform"}; 934 #endif 935 static PetscErrorCode adaptToleranceFEM(PetscFE fem, Vec sol, PetscInt type, PetscInt grid, LandauCtx *ctx, DM *newForest) 936 { 937 DM forest, plex, adaptedDM = NULL; 938 PetscDS prob; 939 PetscBool isForest; 940 PetscQuadrature quad; 941 PetscInt Nq, Nb, *Nb2, cStart, cEnd, c, dim, qj, k; 942 DMLabel adaptLabel = NULL; 943 944 PetscFunctionBegin; 945 forest = ctx->plex[grid]; 946 PetscCall(DMCreateDS(forest)); 947 PetscCall(DMGetDS(forest, &prob)); 948 PetscCall(DMGetDimension(forest, &dim)); 949 PetscCall(DMIsForest(forest, &isForest)); 950 PetscCheck(isForest, ctx->comm, PETSC_ERR_ARG_WRONG, "! Forest"); 951 PetscCall(DMConvert(forest, DMPLEX, &plex)); 952 PetscCall(DMPlexGetHeightStratum(plex, 0, &cStart, &cEnd)); 953 PetscCall(DMLabelCreate(PETSC_COMM_SELF, "adapt", &adaptLabel)); 954 PetscCall(PetscFEGetQuadrature(fem, &quad)); 955 PetscCall(PetscQuadratureGetData(quad, NULL, NULL, &Nq, NULL, NULL)); 956 PetscCheck(Nq <= LANDAU_MAX_NQND, ctx->comm, PETSC_ERR_ARG_WRONG, "Order too high. Nq = %" PetscInt_FMT " > LANDAU_MAX_NQND (%d)", Nq, LANDAU_MAX_NQND); 957 PetscCall(PetscFEGetDimension(ctx->fe[0], &Nb)); 958 PetscCall(PetscDSGetDimensions(prob, &Nb2)); 959 PetscCheck(Nb2[0] == Nb, ctx->comm, PETSC_ERR_ARG_WRONG, " Nb = %" PetscInt_FMT " != Nb (%" PetscInt_FMT ")", Nb, Nb2[0]); 960 PetscCheck(Nb <= LANDAU_MAX_NQND, ctx->comm, PETSC_ERR_ARG_WRONG, "Order too high. Nb = %" PetscInt_FMT " > LANDAU_MAX_NQND (%d)", Nb, LANDAU_MAX_NQND); 961 PetscCall(PetscInfo(sol, "%" PetscInt_FMT ") Refine phase: %s\n", grid, s_refine_names[type])); 962 if (type == 4) { 963 for (c = cStart; c < cEnd; c++) PetscCall(DMLabelSetValue(adaptLabel, c, DM_ADAPT_REFINE)); 964 } else if (type == 2) { 965 PetscInt rCellIdx[8], nr = 0, nrmax = (dim == 3) ? 8 : 2; 966 PetscReal minRad = PETSC_INFINITY, r; 967 for (c = cStart; c < cEnd; c++) { 968 PetscReal tt, v0[LANDAU_MAX_NQND * 3], J[LANDAU_MAX_NQND * 9], invJ[LANDAU_MAX_NQND * 9], detJ[LANDAU_MAX_NQND]; 969 PetscCall(DMPlexComputeCellGeometryFEM(plex, c, quad, v0, J, invJ, detJ)); 970 (void)J; 971 (void)invJ; 972 for (qj = 0; qj < Nq; ++qj) { 973 tt = PetscSqr(v0[dim * qj + 0]) + PetscSqr(v0[dim * qj + 1]) + PetscSqr((dim == 3) ? v0[dim * qj + 2] : 0); 974 r = PetscSqrtReal(tt); 975 if (r < minRad - PETSC_SQRT_MACHINE_EPSILON * 10.) { 976 minRad = r; 977 nr = 0; 978 rCellIdx[nr++] = c; 979 PetscCall(PetscInfo(sol, "\t\t%" PetscInt_FMT ") Found first inner r=%e, cell %" PetscInt_FMT ", qp %" PetscInt_FMT "/%" PetscInt_FMT "\n", grid, (double)r, c, qj + 1, Nq)); 980 } else if ((r - minRad) < PETSC_SQRT_MACHINE_EPSILON * 100. && nr < nrmax) { 981 for (k = 0; k < nr; k++) 982 if (c == rCellIdx[k]) break; 983 if (k == nr) { 984 rCellIdx[nr++] = c; 985 PetscCall(PetscInfo(sol, "\t\t\t%" PetscInt_FMT ") Found another inner r=%e, cell %" PetscInt_FMT ", qp %" PetscInt_FMT "/%" PetscInt_FMT ", d=%e\n", grid, (double)r, c, qj + 1, Nq, (double)(r - minRad))); 986 } 987 } 988 } 989 } 990 for (k = 0; k < nr; k++) PetscCall(DMLabelSetValue(adaptLabel, rCellIdx[k], DM_ADAPT_REFINE)); 991 PetscCall(PetscInfo(sol, "\t\t\t%" PetscInt_FMT ") Refined %" PetscInt_FMT " origin cells %" PetscInt_FMT ",%" PetscInt_FMT " r=%g\n", grid, nr, rCellIdx[0], rCellIdx[1], (double)minRad)); 992 } else if (type == 0 || type == 1 || type == 3) { /* refine along r=0 axis */ 993 PetscScalar *coef = NULL; 994 Vec coords; 995 PetscInt csize, Nv, d, nz, nrefined = 0; 996 DM cdm; 997 PetscSection cs; 998 PetscCall(DMGetCoordinatesLocal(forest, &coords)); 999 PetscCall(DMGetCoordinateDM(forest, &cdm)); 1000 PetscCall(DMGetLocalSection(cdm, &cs)); 1001 for (c = cStart; c < cEnd; c++) { 1002 PetscInt doit = 0, outside = 0; 1003 PetscCall(DMPlexVecGetClosure(cdm, cs, coords, c, &csize, &coef)); 1004 Nv = csize / dim; 1005 for (nz = d = 0; d < Nv; d++) { 1006 PetscReal z = PetscRealPart(coef[d * dim + (dim - 1)]), x = PetscSqr(PetscRealPart(coef[d * dim + 0])) + ((dim == 3) ? PetscSqr(PetscRealPart(coef[d * dim + 1])) : 0); 1007 x = PetscSqrtReal(x); 1008 if (type == 0) { 1009 if (ctx->re_radius > PETSC_SQRT_MACHINE_EPSILON && (z < -PETSC_MACHINE_EPSILON * 10. || z > ctx->re_radius + PETSC_MACHINE_EPSILON * 10.)) outside++; /* first pass don't refine bottom */ 1010 } else if (type == 1 && (z > ctx->vperp0_radius1 || z < -ctx->vperp0_radius1)) { 1011 outside++; /* don't refine outside electron refine radius */ 1012 PetscCall(PetscInfo(sol, "\t%" PetscInt_FMT ") (debug) found %s cells\n", grid, s_refine_names[type])); 1013 } else if (type == 3 && (z > ctx->vperp0_radius2 || z < -ctx->vperp0_radius2)) { 1014 outside++; /* refine r=0 cells on refinement front */ 1015 PetscCall(PetscInfo(sol, "\t%" PetscInt_FMT ") (debug) found %s cells\n", grid, s_refine_names[type])); 1016 } 1017 if (x < PETSC_MACHINE_EPSILON * 10. && (type != 0 || ctx->re_radius > PETSC_SQRT_MACHINE_EPSILON)) nz++; 1018 } 1019 PetscCall(DMPlexVecRestoreClosure(cdm, cs, coords, c, &csize, &coef)); 1020 if (doit || (outside < Nv && nz)) { 1021 PetscCall(DMLabelSetValue(adaptLabel, c, DM_ADAPT_REFINE)); 1022 nrefined++; 1023 } 1024 } 1025 PetscCall(PetscInfo(sol, "\t%" PetscInt_FMT ") Refined %" PetscInt_FMT " cells\n", grid, nrefined)); 1026 } 1027 PetscCall(DMDestroy(&plex)); 1028 PetscCall(DMAdaptLabel(forest, adaptLabel, &adaptedDM)); 1029 PetscCall(DMLabelDestroy(&adaptLabel)); 1030 *newForest = adaptedDM; 1031 if (adaptedDM) { 1032 if (isForest) PetscCall(DMForestSetAdaptivityForest(adaptedDM, NULL)); // ???? 1033 PetscCall(DMConvert(adaptedDM, DMPLEX, &plex)); 1034 PetscCall(DMPlexGetHeightStratum(plex, 0, &cStart, &cEnd)); 1035 PetscCall(PetscInfo(sol, "\t\t\t\t%" PetscInt_FMT ") %" PetscInt_FMT " cells, %" PetscInt_FMT " total quadrature points\n", grid, cEnd - cStart, Nq * (cEnd - cStart))); 1036 PetscCall(DMDestroy(&plex)); 1037 } else *newForest = NULL; 1038 PetscFunctionReturn(PETSC_SUCCESS); 1039 } 1040 1041 // forest goes in (ctx->plex[grid]), plex comes out 1042 static PetscErrorCode adapt(PetscInt grid, LandauCtx *ctx, Vec *uu) 1043 { 1044 PetscInt adaptIter; 1045 1046 PetscFunctionBegin; 1047 PetscInt type, limits[5] = {(grid == 0) ? ctx->numRERefine : 0, (grid == 0) ? ctx->nZRefine1 : 0, ctx->numAMRRefine[grid], (grid == 0) ? ctx->nZRefine2 : 0, ctx->postAMRRefine[grid]}; 1048 for (type = 0; type < 5; type++) { 1049 for (adaptIter = 0; adaptIter < limits[type]; adaptIter++) { 1050 DM newForest = NULL; 1051 PetscCall(adaptToleranceFEM(ctx->fe[0], *uu, type, grid, ctx, &newForest)); 1052 if (newForest) { 1053 PetscCall(DMDestroy(&ctx->plex[grid])); 1054 PetscCall(VecDestroy(uu)); 1055 PetscCall(DMCreateGlobalVector(newForest, uu)); 1056 PetscCall(LandauSetInitialCondition(newForest, *uu, grid, 0, 1, ctx)); 1057 ctx->plex[grid] = newForest; 1058 } else { 1059 PetscCall(PetscInfo(*uu, "No refinement\n")); 1060 } 1061 } 1062 } 1063 PetscCall(PetscObjectSetName((PetscObject)*uu, "uAMR")); 1064 PetscFunctionReturn(PETSC_SUCCESS); 1065 } 1066 1067 // make log(Lambdas) from NRL Plasma formulary 1068 static PetscErrorCode makeLambdas(LandauCtx *ctx) 1069 { 1070 PetscFunctionBegin; 1071 for (PetscInt gridi = 0; gridi < ctx->num_grids; gridi++) { 1072 PetscInt iii = ctx->species_offset[gridi]; 1073 PetscReal Ti_ev = (ctx->thermal_temps[iii] / 1.1604525e7) * 1000; // convert (back) to eV 1074 PetscReal ni = ctx->n[iii] * ctx->n_0; 1075 for (PetscInt gridj = gridi; gridj < ctx->num_grids; gridj++) { 1076 PetscInt jjj = ctx->species_offset[gridj]; 1077 PetscReal Zj = ctx->charges[jjj] / 1.6022e-19; 1078 if (gridi == 0) { 1079 if (gridj == 0) { // lam_ee 1080 ctx->lambdas[gridi][gridj] = 23.5 - PetscLogReal(PetscSqrtReal(ni) * PetscPowReal(Ti_ev, -1.25)) - PetscSqrtReal(1e-5 + PetscSqr(PetscLogReal(Ti_ev) - 2) / 16); 1081 } else { // lam_ei == lam_ie 1082 if (10 * Zj * Zj > Ti_ev) { 1083 ctx->lambdas[gridi][gridj] = ctx->lambdas[gridj][gridi] = 23 - PetscLogReal(PetscSqrtReal(ni) * Zj * PetscPowReal(Ti_ev, -1.5)); 1084 } else { 1085 ctx->lambdas[gridi][gridj] = ctx->lambdas[gridj][gridi] = 24 - PetscLogReal(PetscSqrtReal(ni) / Ti_ev); 1086 } 1087 } 1088 } else { // lam_ii' 1089 PetscReal mui = ctx->masses[iii] / 1.6720e-27, Zi = ctx->charges[iii] / 1.6022e-19; 1090 PetscReal Tj_ev = (ctx->thermal_temps[jjj] / 1.1604525e7) * 1000; // convert (back) to eV 1091 PetscReal muj = ctx->masses[jjj] / 1.6720e-27; 1092 PetscReal nj = ctx->n[jjj] * ctx->n_0; 1093 ctx->lambdas[gridi][gridj] = ctx->lambdas[gridj][gridi] = 23 - PetscLogReal(Zi * Zj * (mui + muj) / (mui * Tj_ev + muj * Ti_ev) * PetscSqrtReal(ni * Zi * Zi / Ti_ev + nj * Zj * Zj / Tj_ev)); 1094 } 1095 } 1096 } 1097 //PetscReal v0 = PetscSqrtReal(ctx->k * ctx->thermal_temps[iii] / ctx->masses[iii]); /* arbitrary units for non-dimensionalization: plasma formulary def */ 1098 PetscFunctionReturn(PETSC_SUCCESS); 1099 } 1100 1101 static PetscErrorCode ProcessOptions(LandauCtx *ctx, const char prefix[]) 1102 { 1103 PetscBool flg, fileflg; 1104 PetscInt ii, nt, nm, nc, num_species_grid[LANDAU_MAX_GRIDS], non_dim_grid; 1105 PetscReal lnLam = 10; 1106 DM dummy; 1107 1108 PetscFunctionBegin; 1109 PetscCall(DMCreate(ctx->comm, &dummy)); 1110 /* get options - initialize context */ 1111 ctx->verbose = 1; // should be 0 for silent compliance 1112 ctx->batch_sz = 1; 1113 ctx->batch_view_idx = 0; 1114 ctx->interpolate = PETSC_TRUE; 1115 ctx->gpu_assembly = PETSC_TRUE; 1116 ctx->norm_state = 0; 1117 ctx->electronShift = 0; 1118 ctx->M = NULL; 1119 ctx->J = NULL; 1120 /* geometry and grids */ 1121 ctx->sphere = PETSC_FALSE; 1122 ctx->use_p4est = PETSC_FALSE; 1123 ctx->simplex = PETSC_FALSE; 1124 for (PetscInt grid = 0; grid < LANDAU_MAX_GRIDS; grid++) { 1125 ctx->radius[grid] = 5.; /* thermal radius (velocity) */ 1126 ctx->radius_perp[grid] = 5.; /* thermal radius (velocity) */ 1127 ctx->radius_par[grid] = 5.; /* thermal radius (velocity) */ 1128 ctx->numAMRRefine[grid] = 0; 1129 ctx->postAMRRefine[grid] = 0; 1130 ctx->species_offset[grid + 1] = 1; // one species default 1131 num_species_grid[grid] = 0; 1132 ctx->plex[grid] = NULL; /* cache as expensive to Convert */ 1133 } 1134 ctx->species_offset[0] = 0; 1135 ctx->re_radius = 0.; 1136 ctx->vperp0_radius1 = 0; 1137 ctx->vperp0_radius2 = 0; 1138 ctx->nZRefine1 = 0; 1139 ctx->nZRefine2 = 0; 1140 ctx->numRERefine = 0; 1141 num_species_grid[0] = 1; // one species default 1142 /* species - [0] electrons, [1] one ion species eg, duetarium, [2] heavy impurity ion, ... */ 1143 ctx->charges[0] = -1; /* electron charge (MKS) */ 1144 ctx->masses[0] = 1 / 1835.469965278441013; /* temporary value in proton mass */ 1145 ctx->n[0] = 1; 1146 ctx->v_0 = 1; /* thermal velocity, we could start with a scale != 1 */ 1147 ctx->thermal_temps[0] = 1; 1148 /* constants, etc. */ 1149 ctx->epsilon0 = 8.8542e-12; /* permittivity of free space (MKS) F/m */ 1150 ctx->k = 1.38064852e-23; /* Boltzmann constant (MKS) J/K */ 1151 ctx->n_0 = 1.e20; /* typical plasma n, but could set it to 1 */ 1152 ctx->Ez = 0; 1153 for (PetscInt grid = 0; grid < LANDAU_NUM_TIMERS; grid++) ctx->times[grid] = 0; 1154 for (PetscInt ii = 0; ii < LANDAU_DIM; ii++) ctx->cells0[ii] = 2; 1155 if (LANDAU_DIM == 2) ctx->cells0[0] = 1; 1156 ctx->use_matrix_mass = PETSC_FALSE; 1157 ctx->use_relativistic_corrections = PETSC_FALSE; 1158 ctx->use_energy_tensor_trick = PETSC_FALSE; /* Use Eero's trick for energy conservation v --> grad(v^2/2) */ 1159 ctx->SData_d.w = NULL; 1160 ctx->SData_d.x = NULL; 1161 ctx->SData_d.y = NULL; 1162 ctx->SData_d.z = NULL; 1163 ctx->SData_d.invJ = NULL; 1164 ctx->jacobian_field_major_order = PETSC_FALSE; 1165 ctx->SData_d.coo_elem_offsets = NULL; 1166 ctx->SData_d.coo_elem_point_offsets = NULL; 1167 ctx->SData_d.coo_elem_fullNb = NULL; 1168 ctx->SData_d.coo_size = 0; 1169 PetscOptionsBegin(ctx->comm, prefix, "Options for Fokker-Plank-Landau collision operator", "none"); 1170 #if defined(PETSC_HAVE_KOKKOS) 1171 ctx->deviceType = LANDAU_KOKKOS; 1172 PetscCall(PetscStrncpy(ctx->filename, "kokkos", sizeof(ctx->filename))); 1173 #else 1174 ctx->deviceType = LANDAU_CPU; 1175 PetscCall(PetscStrncpy(ctx->filename, "cpu", sizeof(ctx->filename))); 1176 #endif 1177 PetscCall(PetscOptionsString("-dm_landau_device_type", "Use kernels on 'cpu' 'kokkos'", "plexland.c", ctx->filename, ctx->filename, sizeof(ctx->filename), NULL)); 1178 PetscCall(PetscStrcmp("cpu", ctx->filename, &flg)); 1179 if (flg) { 1180 ctx->deviceType = LANDAU_CPU; 1181 } else { 1182 PetscCall(PetscStrcmp("kokkos", ctx->filename, &flg)); 1183 PetscCheck(flg, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_device_type %s", ctx->filename); 1184 ctx->deviceType = LANDAU_KOKKOS; 1185 } 1186 ctx->filename[0] = '\0'; 1187 PetscCall(PetscOptionsString("-dm_landau_filename", "file to read mesh from", "plexland.c", ctx->filename, ctx->filename, sizeof(ctx->filename), &fileflg)); 1188 PetscCall(PetscOptionsReal("-dm_landau_electron_shift", "Shift in thermal velocity of electrons", "none", ctx->electronShift, &ctx->electronShift, NULL)); 1189 PetscCall(PetscOptionsInt("-dm_landau_verbose", "Level of verbosity output", "plexland.c", ctx->verbose, &ctx->verbose, NULL)); 1190 PetscCall(PetscOptionsInt("-dm_landau_batch_size", "Number of 'vertices' to batch", "ex2.c", ctx->batch_sz, &ctx->batch_sz, NULL)); 1191 PetscCheck(LANDAU_MAX_BATCH_SZ >= ctx->batch_sz, ctx->comm, PETSC_ERR_ARG_WRONG, "LANDAU_MAX_BATCH_SZ %d < ctx->batch_sz %" PetscInt_FMT, LANDAU_MAX_BATCH_SZ, ctx->batch_sz); 1192 PetscCall(PetscOptionsInt("-dm_landau_batch_view_idx", "Index of batch for diagnostics like plotting", "ex2.c", ctx->batch_view_idx, &ctx->batch_view_idx, NULL)); 1193 PetscCheck(ctx->batch_view_idx < ctx->batch_sz, ctx->comm, PETSC_ERR_ARG_WRONG, "-ctx->batch_view_idx %" PetscInt_FMT " > ctx->batch_sz %" PetscInt_FMT, ctx->batch_view_idx, ctx->batch_sz); 1194 PetscCall(PetscOptionsReal("-dm_landau_Ez", "Initial parallel electric field in unites of Conner-Hastie critical field", "plexland.c", ctx->Ez, &ctx->Ez, NULL)); 1195 PetscCall(PetscOptionsReal("-dm_landau_n_0", "Normalization constant for number density", "plexland.c", ctx->n_0, &ctx->n_0, NULL)); 1196 PetscCall(PetscOptionsBool("-dm_landau_use_mataxpy_mass", "Use fast but slightly fragile MATAXPY to add mass term", "plexland.c", ctx->use_matrix_mass, &ctx->use_matrix_mass, NULL)); 1197 PetscCall(PetscOptionsBool("-dm_landau_use_relativistic_corrections", "Use relativistic corrections", "plexland.c", ctx->use_relativistic_corrections, &ctx->use_relativistic_corrections, NULL)); 1198 PetscCall(PetscOptionsBool("-dm_landau_simplex", "Use simplex elements", "plexland.c", ctx->simplex, &ctx->simplex, NULL)); 1199 PetscCall(PetscOptionsBool("-dm_landau_sphere", "use sphere/semi-circle domain instead of rectangle", "plexland.c", ctx->sphere, &ctx->sphere, NULL)); 1200 if (LANDAU_DIM == 2 && ctx->use_relativistic_corrections) ctx->use_relativistic_corrections = PETSC_FALSE; // should warn 1201 PetscCall(PetscOptionsBool("-dm_landau_use_energy_tensor_trick", "Use Eero's trick of using grad(v^2/2) instead of v as args to Landau tensor to conserve energy with relativistic corrections and Q1 elements", "plexland.c", ctx->use_energy_tensor_trick, 1202 &ctx->use_energy_tensor_trick, NULL)); 1203 1204 /* get num species with temperature, set defaults */ 1205 for (ii = 1; ii < LANDAU_MAX_SPECIES; ii++) { 1206 ctx->thermal_temps[ii] = 1; 1207 ctx->charges[ii] = 1; 1208 ctx->masses[ii] = 1; 1209 ctx->n[ii] = 1; 1210 } 1211 nt = LANDAU_MAX_SPECIES; 1212 PetscCall(PetscOptionsRealArray("-dm_landau_thermal_temps", "Temperature of each species [e,i_0,i_1,...] in keV (must be set to set number of species)", "plexland.c", ctx->thermal_temps, &nt, &flg)); 1213 PetscCheck(flg, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_thermal_temps ,t1,t2,.. must be provided to set the number of species"); 1214 PetscCall(PetscInfo(dummy, "num_species set to number of thermal temps provided (%" PetscInt_FMT ")\n", nt)); 1215 ctx->num_species = nt; 1216 for (ii = 0; ii < ctx->num_species; ii++) ctx->thermal_temps[ii] *= 1.1604525e7; /* convert to Kelvin */ 1217 nm = LANDAU_MAX_SPECIES - 1; 1218 PetscCall(PetscOptionsRealArray("-dm_landau_ion_masses", "Mass of each species in units of proton mass [i_0=2,i_1=40...]", "plexland.c", &ctx->masses[1], &nm, &flg)); 1219 PetscCheck(!flg || nm == ctx->num_species - 1, ctx->comm, PETSC_ERR_ARG_WRONG, "num ion masses %" PetscInt_FMT " != num species %" PetscInt_FMT, nm, ctx->num_species - 1); 1220 nm = LANDAU_MAX_SPECIES; 1221 PetscCall(PetscOptionsRealArray("-dm_landau_n", "Number density of each species = n_s * n_0", "plexland.c", ctx->n, &nm, &flg)); 1222 PetscCheck(!flg || nm == ctx->num_species, ctx->comm, PETSC_ERR_ARG_WRONG, "wrong num n: %" PetscInt_FMT " != num species %" PetscInt_FMT, nm, ctx->num_species); 1223 for (ii = 0; ii < LANDAU_MAX_SPECIES; ii++) ctx->masses[ii] *= 1.6720e-27; /* scale by proton mass kg */ 1224 ctx->masses[0] = 9.10938356e-31; /* electron mass kg (should be about right already) */ 1225 nc = LANDAU_MAX_SPECIES - 1; 1226 PetscCall(PetscOptionsRealArray("-dm_landau_ion_charges", "Charge of each species in units of proton charge [i_0=2,i_1=18,...]", "plexland.c", &ctx->charges[1], &nc, &flg)); 1227 if (flg) PetscCheck(nc == ctx->num_species - 1, ctx->comm, PETSC_ERR_ARG_WRONG, "num charges %" PetscInt_FMT " != num species %" PetscInt_FMT, nc, ctx->num_species - 1); 1228 for (ii = 0; ii < LANDAU_MAX_SPECIES; ii++) ctx->charges[ii] *= 1.6022e-19; /* electron/proton charge (MKS) */ 1229 /* geometry and grids */ 1230 nt = LANDAU_MAX_GRIDS; 1231 PetscCall(PetscOptionsIntArray("-dm_landau_num_species_grid", "Number of species on each grid: [ 1, ....] or [S, 0 ....] for single grid", "plexland.c", num_species_grid, &nt, &flg)); 1232 if (flg) { 1233 ctx->num_grids = nt; 1234 for (ii = nt = 0; ii < ctx->num_grids; ii++) nt += num_species_grid[ii]; 1235 PetscCheck(ctx->num_species == nt, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_num_species_grid: sum %" PetscInt_FMT " != num_species = %" PetscInt_FMT ". %" PetscInt_FMT " grids (check that number of grids <= LANDAU_MAX_GRIDS = %d)", nt, ctx->num_species, 1236 ctx->num_grids, LANDAU_MAX_GRIDS); 1237 } else { 1238 if (ctx->num_species > LANDAU_MAX_GRIDS) { 1239 num_species_grid[0] = 1; 1240 num_species_grid[1] = ctx->num_species - 1; 1241 ctx->num_grids = 2; 1242 } else { 1243 ctx->num_grids = ctx->num_species; 1244 for (ii = 0; ii < ctx->num_grids; ii++) num_species_grid[ii] = 1; 1245 } 1246 } 1247 for (ctx->species_offset[0] = ii = 0; ii < ctx->num_grids; ii++) ctx->species_offset[ii + 1] = ctx->species_offset[ii] + num_species_grid[ii]; 1248 PetscCheck(ctx->species_offset[ctx->num_grids] == ctx->num_species, ctx->comm, PETSC_ERR_ARG_WRONG, "ctx->species_offset[ctx->num_grids] %" PetscInt_FMT " != ctx->num_species = %" PetscInt_FMT " ???????????", ctx->species_offset[ctx->num_grids], 1249 ctx->num_species); 1250 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1251 PetscInt iii = ctx->species_offset[grid]; // normalize with first (arbitrary) species on grid 1252 ctx->thermal_speed[grid] = PetscSqrtReal(ctx->k * ctx->thermal_temps[iii] / ctx->masses[iii]); /* arbitrary units for non-dimensionalization: plasma formulary def */ 1253 } 1254 // get lambdas here because we need them for t_0 etc 1255 PetscCall(PetscOptionsReal("-dm_landau_ln_lambda", "Universal cross section parameter. Default uses NRL formulas", "plexland.c", lnLam, &lnLam, &flg)); 1256 if (flg) { 1257 for (PetscInt grid = 0; grid < LANDAU_MAX_GRIDS; grid++) { 1258 for (PetscInt gridj = 0; gridj < LANDAU_MAX_GRIDS; gridj++) ctx->lambdas[gridj][grid] = lnLam; /* cross section ratio large - small angle collisions */ 1259 } 1260 } else { 1261 PetscCall(makeLambdas(ctx)); 1262 } 1263 non_dim_grid = 0; 1264 PetscCall(PetscOptionsInt("-dm_landau_normalization_grid", "Index of grid to use for setting v_0, m_0, t_0. (Not recommended)", "plexland.c", non_dim_grid, &non_dim_grid, &flg)); 1265 if (non_dim_grid != 0) PetscCall(PetscInfo(dummy, "Normalization grid set to %" PetscInt_FMT ", but non-default not well verified\n", non_dim_grid)); 1266 PetscCheck(non_dim_grid >= 0 && non_dim_grid < ctx->num_species, ctx->comm, PETSC_ERR_ARG_WRONG, "Normalization grid wrong: %" PetscInt_FMT, non_dim_grid); 1267 ctx->v_0 = ctx->thermal_speed[non_dim_grid]; /* arbitrary units for non dimensionalization: global mean velocity in 1D of electrons */ 1268 ctx->m_0 = ctx->masses[non_dim_grid]; /* arbitrary reference mass, electrons */ 1269 ctx->t_0 = 8 * PETSC_PI * PetscSqr(ctx->epsilon0 * ctx->m_0 / PetscSqr(ctx->charges[non_dim_grid])) / ctx->lambdas[non_dim_grid][non_dim_grid] / ctx->n_0 * PetscPowReal(ctx->v_0, 3); /* note, this t_0 makes nu[non_dim_grid,non_dim_grid]=1 */ 1270 /* domain */ 1271 nt = LANDAU_MAX_GRIDS; 1272 PetscCall(PetscOptionsRealArray("-dm_landau_domain_radius", "Phase space size in units of thermal velocity of grid", "plexland.c", ctx->radius, &nt, &flg)); 1273 if (flg) { 1274 PetscCheck(nt >= ctx->num_grids, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_domain_radius: given %" PetscInt_FMT " radius != number grids %" PetscInt_FMT, nt, ctx->num_grids); 1275 while (nt--) ctx->radius_par[nt] = ctx->radius_perp[nt] = ctx->radius[nt]; 1276 } else { 1277 nt = LANDAU_MAX_GRIDS; 1278 PetscCall(PetscOptionsRealArray("-dm_landau_domain_max_par", "Parallel velocity domain size in units of thermal velocity of grid", "plexland.c", ctx->radius_par, &nt, &flg)); 1279 if (flg) PetscCheck(nt >= ctx->num_grids, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_domain_max_par: given %" PetscInt_FMT " radius != number grids %" PetscInt_FMT, nt, ctx->num_grids); 1280 PetscCall(PetscOptionsRealArray("-dm_landau_domain_max_perp", "Perpendicular velocity domain size in units of thermal velocity of grid", "plexland.c", ctx->radius_perp, &nt, &flg)); 1281 if (flg) PetscCheck(nt >= ctx->num_grids, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_domain_max_perp: given %" PetscInt_FMT " radius != number grids %" PetscInt_FMT, nt, ctx->num_grids); 1282 } 1283 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1284 if (flg && ctx->radius[grid] <= 0) { /* negative is ratio of c - need to set par and perp with this -- todo */ 1285 if (ctx->radius[grid] == 0) ctx->radius[grid] = 0.75; 1286 else ctx->radius[grid] = -ctx->radius[grid]; 1287 ctx->radius[grid] = ctx->radius[grid] * SPEED_OF_LIGHT / ctx->v_0; // use any species on grid to normalize (v_0 same for all on grid) 1288 PetscCall(PetscInfo(dummy, "Change domain radius to %g for grid %" PetscInt_FMT "\n", (double)ctx->radius[grid], grid)); 1289 } 1290 ctx->radius[grid] *= ctx->thermal_speed[grid] / ctx->v_0; // scale domain by thermal radius relative to v_0 1291 ctx->radius_perp[grid] *= ctx->thermal_speed[grid] / ctx->v_0; // scale domain by thermal radius relative to v_0 1292 ctx->radius_par[grid] *= ctx->thermal_speed[grid] / ctx->v_0; // scale domain by thermal radius relative to v_0 1293 } 1294 /* amr parameters */ 1295 if (!fileflg) { 1296 nt = LANDAU_MAX_GRIDS; 1297 PetscCall(PetscOptionsIntArray("-dm_landau_amr_levels_max", "Number of AMR levels of refinement around origin, after (RE) refinements along z", "plexland.c", ctx->numAMRRefine, &nt, &flg)); 1298 PetscCheck(!flg || nt >= ctx->num_grids, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_amr_levels_max: given %" PetscInt_FMT " != number grids %" PetscInt_FMT, nt, ctx->num_grids); 1299 nt = LANDAU_MAX_GRIDS; 1300 PetscCall(PetscOptionsIntArray("-dm_landau_amr_post_refine", "Number of levels to uniformly refine after AMR", "plexland.c", ctx->postAMRRefine, &nt, &flg)); 1301 for (ii = 1; ii < ctx->num_grids; ii++) ctx->postAMRRefine[ii] = ctx->postAMRRefine[0]; // all grids the same now 1302 PetscCall(PetscOptionsInt("-dm_landau_amr_re_levels", "Number of levels to refine along v_perp=0, z>0", "plexland.c", ctx->numRERefine, &ctx->numRERefine, &flg)); 1303 PetscCall(PetscOptionsInt("-dm_landau_amr_z_refine_pre", "Number of levels to refine along v_perp=0 before origin refine", "plexland.c", ctx->nZRefine1, &ctx->nZRefine1, &flg)); 1304 PetscCall(PetscOptionsInt("-dm_landau_amr_z_refine_post", "Number of levels to refine along v_perp=0 after origin refine", "plexland.c", ctx->nZRefine2, &ctx->nZRefine2, &flg)); 1305 PetscCall(PetscOptionsReal("-dm_landau_re_radius", "velocity range to refine on positive (z>0) r=0 axis for runaways", "plexland.c", ctx->re_radius, &ctx->re_radius, &flg)); 1306 PetscCall(PetscOptionsReal("-dm_landau_z_radius_pre", "velocity range to refine r=0 axis (for electrons)", "plexland.c", ctx->vperp0_radius1, &ctx->vperp0_radius1, &flg)); 1307 PetscCall(PetscOptionsReal("-dm_landau_z_radius_post", "velocity range to refine r=0 axis (for electrons) after origin AMR", "plexland.c", ctx->vperp0_radius2, &ctx->vperp0_radius2, &flg)); 1308 /* spherical domain */ 1309 if (ctx->sphere || ctx->simplex) { 1310 ctx->sphere_uniform_normal = PETSC_FALSE; 1311 PetscCall(PetscOptionsBool("-dm_landau_sphere_uniform_normal", "Scaling of circle radius to get uniform particles per cell with Maxwellians (not used)", "plexland.c", ctx->sphere_uniform_normal, &ctx->sphere_uniform_normal, NULL)); 1312 if (!ctx->sphere_uniform_normal) { // true 1313 nt = LANDAU_MAX_GRIDS; 1314 PetscCall(PetscOptionsRealArray("-dm_landau_sphere_inner_radius_90degree_scale", "Scaling of radius for inner circle on 90 degree grid", "plexland.c", ctx->sphere_inner_radius_90degree, &nt, &flg)); 1315 if (flg && nt < ctx->num_grids) { 1316 for (PetscInt grid = nt; grid < ctx->num_grids; grid++) ctx->sphere_inner_radius_90degree[grid] = ctx->sphere_inner_radius_90degree[0]; 1317 } else if (!flg || nt == 0) { 1318 if (LANDAU_DIM == 2) { 1319 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) ctx->sphere_inner_radius_90degree[grid] = 0.4; // optimized for R=5, Q4, AMR=0 1320 } else { 1321 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) ctx->sphere_inner_radius_90degree[grid] = 0.577 * 0.40; 1322 } 1323 } 1324 nt = LANDAU_MAX_GRIDS; 1325 PetscCall(PetscOptionsRealArray("-dm_landau_sphere_inner_radius_45degree_scale", "Scaling of radius for inner circle on 45 degree grid", "plexland.c", ctx->sphere_inner_radius_45degree, &nt, &flg)); 1326 if (flg && nt < ctx->num_grids) { 1327 for (PetscInt grid = nt; grid < ctx->num_grids; grid++) ctx->sphere_inner_radius_45degree[grid] = ctx->sphere_inner_radius_45degree[0]; 1328 } else if (!flg || nt == 0) { 1329 if (LANDAU_DIM == 2) { 1330 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) ctx->sphere_inner_radius_45degree[grid] = 0.45; // optimized for R=5, Q4, AMR=0 1331 } else { 1332 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) ctx->sphere_inner_radius_45degree[grid] = 0.4; // 3D sphere 1333 } 1334 } 1335 if (ctx->sphere) PetscCall(PetscInfo(ctx->plex[0], "sphere : , 45 degree scaling = %g; 90 degree scaling = %g\n", (double)ctx->sphere_inner_radius_45degree[0], (double)ctx->sphere_inner_radius_90degree[0])); 1336 } else { 1337 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1338 switch (ctx->numAMRRefine[grid]) { 1339 case 0: 1340 case 1: 1341 case 2: 1342 case 3: 1343 default: 1344 if (LANDAU_DIM == 2) { 1345 ctx->sphere_inner_radius_90degree[grid] = 0.40; 1346 ctx->sphere_inner_radius_45degree[grid] = 0.45; 1347 } else { 1348 ctx->sphere_inner_radius_45degree[grid] = 0.25; 1349 } 1350 } 1351 } 1352 } 1353 } else { 1354 nt = LANDAU_DIM; 1355 PetscCall(PetscOptionsIntArray("-dm_landau_num_cells", "Number of cells in each dimension of base grid", "plexland.c", ctx->cells0, &nt, &flg)); 1356 } 1357 } 1358 /* processing options */ 1359 PetscCall(PetscOptionsBool("-dm_landau_gpu_assembly", "Assemble Jacobian on GPU", "plexland.c", ctx->gpu_assembly, &ctx->gpu_assembly, NULL)); 1360 PetscCall(PetscOptionsBool("-dm_landau_jacobian_field_major_order", "Reorder Jacobian for GPU assembly with field major, or block diagonal, ordering (DEPRECATED)", "plexland.c", ctx->jacobian_field_major_order, &ctx->jacobian_field_major_order, NULL)); 1361 if (ctx->jacobian_field_major_order) PetscCheck(ctx->gpu_assembly, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_jacobian_field_major_order requires -dm_landau_gpu_assembly"); 1362 PetscCheck(!ctx->jacobian_field_major_order, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_jacobian_field_major_order DEPRECATED"); 1363 PetscOptionsEnd(); 1364 1365 for (ii = ctx->num_species; ii < LANDAU_MAX_SPECIES; ii++) ctx->masses[ii] = ctx->thermal_temps[ii] = ctx->charges[ii] = 0; 1366 if (ctx->verbose != 0) { 1367 PetscReal pmassunit = PetscRealConstant(1.6720e-27); 1368 1369 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "masses: e=%10.3e; ions in proton mass units: %10.3e %10.3e ...\n", (double)ctx->masses[0], (double)(ctx->masses[1] / pmassunit), (double)(ctx->num_species > 2 ? ctx->masses[2] / pmassunit : 0))); 1370 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "charges: e=%10.3e; charges in elementary units: %10.3e %10.3e\n", (double)ctx->charges[0], (double)(-ctx->charges[1] / ctx->charges[0]), (double)(ctx->num_species > 2 ? -ctx->charges[2] / ctx->charges[0] : 0))); 1371 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "n: e: %10.3e i: %10.3e %10.3e\n", (double)ctx->n[0], (double)ctx->n[1], (double)(ctx->num_species > 2 ? ctx->n[2] : 0))); 1372 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "thermal T (K): e=%10.3e i=%10.3e %10.3e. Normalization grid %" PetscInt_FMT ": v_0=%10.3e (%10.3ec) n_0=%10.3e t_0=%10.3e %" PetscInt_FMT " batched, view batch %" PetscInt_FMT "\n", (double)ctx->thermal_temps[0], 1373 (double)ctx->thermal_temps[1], (double)((ctx->num_species > 2) ? ctx->thermal_temps[2] : 0), non_dim_grid, (double)ctx->v_0, (double)(ctx->v_0 / SPEED_OF_LIGHT), (double)ctx->n_0, (double)ctx->t_0, ctx->batch_sz, ctx->batch_view_idx)); 1374 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "Domain radius (AMR levels) grid %d: par=%10.3e perp=%10.3e (%" PetscInt_FMT ") ", 0, (double)ctx->radius_par[0], (double)ctx->radius_perp[0], ctx->numAMRRefine[0])); 1375 for (ii = 1; ii < ctx->num_grids; ii++) PetscCall(PetscPrintf(PETSC_COMM_WORLD, ", %" PetscInt_FMT ": par=%10.3e perp=%10.3e (%" PetscInt_FMT ") ", ii, (double)ctx->radius_par[ii], (double)ctx->radius_perp[ii], ctx->numAMRRefine[ii])); 1376 if (ctx->use_relativistic_corrections) PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\nUse relativistic corrections\n")); 1377 else PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\n")); 1378 } 1379 PetscCall(DMDestroy(&dummy)); 1380 { 1381 PetscMPIInt rank; 1382 PetscCallMPI(MPI_Comm_rank(PETSC_COMM_WORLD, &rank)); 1383 ctx->stage = 0; 1384 PetscCall(PetscLogEventRegister("Landau Create", DM_CLASSID, &ctx->events[13])); /* 13 */ 1385 PetscCall(PetscLogEventRegister(" GPU ass. setup", DM_CLASSID, &ctx->events[2])); /* 2 */ 1386 PetscCall(PetscLogEventRegister(" Build matrix", DM_CLASSID, &ctx->events[12])); /* 12 */ 1387 PetscCall(PetscLogEventRegister(" Assembly maps", DM_CLASSID, &ctx->events[15])); /* 15 */ 1388 PetscCall(PetscLogEventRegister("Landau Mass mat", DM_CLASSID, &ctx->events[14])); /* 14 */ 1389 PetscCall(PetscLogEventRegister("Landau Operator", DM_CLASSID, &ctx->events[11])); /* 11 */ 1390 PetscCall(PetscLogEventRegister("Landau Jacobian", DM_CLASSID, &ctx->events[0])); /* 0 */ 1391 PetscCall(PetscLogEventRegister("Landau Mass", DM_CLASSID, &ctx->events[9])); /* 9 */ 1392 PetscCall(PetscLogEventRegister(" Preamble", DM_CLASSID, &ctx->events[10])); /* 10 */ 1393 PetscCall(PetscLogEventRegister(" static IP Data", DM_CLASSID, &ctx->events[7])); /* 7 */ 1394 PetscCall(PetscLogEventRegister(" dynamic IP-Jac", DM_CLASSID, &ctx->events[1])); /* 1 */ 1395 PetscCall(PetscLogEventRegister(" Kernel-init", DM_CLASSID, &ctx->events[3])); /* 3 */ 1396 PetscCall(PetscLogEventRegister(" Jac-f-df (GPU)", DM_CLASSID, &ctx->events[8])); /* 8 */ 1397 PetscCall(PetscLogEventRegister(" J Kernel (GPU)", DM_CLASSID, &ctx->events[4])); /* 4 */ 1398 PetscCall(PetscLogEventRegister(" M Kernel (GPU)", DM_CLASSID, &ctx->events[16])); /* 16 */ 1399 PetscCall(PetscLogEventRegister(" Copy to CPU", DM_CLASSID, &ctx->events[5])); /* 5 */ 1400 PetscCall(PetscLogEventRegister(" CPU assemble", DM_CLASSID, &ctx->events[6])); /* 6 */ 1401 1402 if (rank) { /* turn off output stuff for duplicate runs - do we need to add the prefix to all this? */ 1403 PetscCall(PetscOptionsClearValue(NULL, "-snes_converged_reason")); 1404 PetscCall(PetscOptionsClearValue(NULL, "-ksp_converged_reason")); 1405 PetscCall(PetscOptionsClearValue(NULL, "-snes_monitor")); 1406 PetscCall(PetscOptionsClearValue(NULL, "-ksp_monitor")); 1407 PetscCall(PetscOptionsClearValue(NULL, "-ts_monitor")); 1408 PetscCall(PetscOptionsClearValue(NULL, "-ts_view")); 1409 PetscCall(PetscOptionsClearValue(NULL, "-ts_adapt_monitor")); 1410 PetscCall(PetscOptionsClearValue(NULL, "-dm_landau_amr_dm_view")); 1411 PetscCall(PetscOptionsClearValue(NULL, "-dm_landau_amr_vec_view")); 1412 PetscCall(PetscOptionsClearValue(NULL, "-dm_landau_mass_dm_view")); 1413 PetscCall(PetscOptionsClearValue(NULL, "-dm_landau_mass_view")); 1414 PetscCall(PetscOptionsClearValue(NULL, "-dm_landau_jacobian_view")); 1415 PetscCall(PetscOptionsClearValue(NULL, "-dm_landau_mat_view")); 1416 PetscCall(PetscOptionsClearValue(NULL, "-pc_bjkokkos_ksp_converged_reason")); 1417 PetscCall(PetscOptionsClearValue(NULL, "-pc_bjkokkos_ksp_monitor")); 1418 PetscCall(PetscOptionsClearValue(NULL, "-")); 1419 PetscCall(PetscOptionsClearValue(NULL, "-info")); 1420 } 1421 } 1422 PetscFunctionReturn(PETSC_SUCCESS); 1423 } 1424 1425 static PetscErrorCode CreateStaticData(PetscInt dim, IS grid_batch_is_inv[], LandauCtx *ctx) 1426 { 1427 PetscSection section[LANDAU_MAX_GRIDS], globsection[LANDAU_MAX_GRIDS]; 1428 PetscQuadrature quad; 1429 const PetscReal *quadWeights; 1430 PetscReal invMass[LANDAU_MAX_SPECIES], nu_alpha[LANDAU_MAX_SPECIES], nu_beta[LANDAU_MAX_SPECIES]; 1431 PetscInt numCells[LANDAU_MAX_GRIDS], Nq, Nb, Nf[LANDAU_MAX_GRIDS], ncellsTot = 0, MAP_BF_SIZE = 64 * LANDAU_DIM * LANDAU_DIM * LANDAU_MAX_Q_FACE * LANDAU_MAX_SPECIES; 1432 PetscTabulation *Tf; 1433 PetscDS prob; 1434 1435 PetscFunctionBegin; 1436 PetscCall(PetscFEGetDimension(ctx->fe[0], &Nb)); 1437 PetscCheck(Nb <= LANDAU_MAX_NQND, ctx->comm, PETSC_ERR_ARG_WRONG, "Order too high. Nb = %" PetscInt_FMT " > LANDAU_MAX_NQND (%d)", Nb, LANDAU_MAX_NQND); 1438 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1439 for (PetscInt ii = ctx->species_offset[grid]; ii < ctx->species_offset[grid + 1]; ii++) { 1440 invMass[ii] = ctx->m_0 / ctx->masses[ii]; 1441 nu_alpha[ii] = PetscSqr(ctx->charges[ii] / ctx->m_0) * ctx->m_0 / ctx->masses[ii]; 1442 nu_beta[ii] = PetscSqr(ctx->charges[ii] / ctx->epsilon0) / (8 * PETSC_PI) * ctx->t_0 * ctx->n_0 / PetscPowReal(ctx->v_0, 3); 1443 } 1444 } 1445 if (ctx->verbose == 4) { 1446 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "nu_alpha: ")); 1447 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1448 PetscInt iii = ctx->species_offset[grid]; 1449 for (PetscInt ii = iii; ii < ctx->species_offset[grid + 1]; ii++) PetscCall(PetscPrintf(PETSC_COMM_WORLD, " %e", (double)nu_alpha[ii])); 1450 } 1451 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\nnu_beta: ")); 1452 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1453 PetscInt iii = ctx->species_offset[grid]; 1454 for (PetscInt ii = iii; ii < ctx->species_offset[grid + 1]; ii++) PetscCall(PetscPrintf(PETSC_COMM_WORLD, " %e", (double)nu_beta[ii])); 1455 } 1456 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\nnu_alpha[i]*nu_beta[j]*lambda[i][j]:\n")); 1457 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1458 PetscInt iii = ctx->species_offset[grid]; 1459 for (PetscInt ii = iii; ii < ctx->species_offset[grid + 1]; ii++) { 1460 for (PetscInt gridj = 0; gridj < ctx->num_grids; gridj++) { 1461 PetscInt jjj = ctx->species_offset[gridj]; 1462 for (PetscInt jj = jjj; jj < ctx->species_offset[gridj + 1]; jj++) PetscCall(PetscPrintf(PETSC_COMM_WORLD, " %14.9e", (double)(nu_alpha[ii] * nu_beta[jj] * ctx->lambdas[grid][gridj]))); 1463 } 1464 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\n")); 1465 } 1466 } 1467 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "lambda[i][j]:\n")); 1468 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1469 PetscInt iii = ctx->species_offset[grid]; 1470 for (PetscInt ii = iii; ii < ctx->species_offset[grid + 1]; ii++) { 1471 for (PetscInt gridj = 0; gridj < ctx->num_grids; gridj++) { 1472 PetscInt jjj = ctx->species_offset[gridj]; 1473 for (PetscInt jj = jjj; jj < ctx->species_offset[gridj + 1]; jj++) PetscCall(PetscPrintf(PETSC_COMM_WORLD, " %14.9e", (double)ctx->lambdas[grid][gridj])); 1474 } 1475 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\n")); 1476 } 1477 } 1478 } 1479 PetscCall(DMGetDS(ctx->plex[0], &prob)); // same DS for all grids 1480 PetscCall(PetscDSGetTabulation(prob, &Tf)); // Bf, &Df same for all grids 1481 /* DS, Tab and quad is same on all grids */ 1482 PetscCheck(ctx->plex[0], ctx->comm, PETSC_ERR_ARG_WRONG, "Plex not created"); 1483 PetscCall(PetscFEGetQuadrature(ctx->fe[0], &quad)); 1484 PetscCall(PetscQuadratureGetData(quad, NULL, NULL, &Nq, NULL, &quadWeights)); 1485 PetscCheck(Nq <= LANDAU_MAX_NQND, ctx->comm, PETSC_ERR_ARG_WRONG, "Order too high. Nq = %" PetscInt_FMT " > LANDAU_MAX_NQND (%d)", Nq, LANDAU_MAX_NQND); 1486 /* setup each grid */ 1487 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1488 PetscInt cStart, cEnd; 1489 PetscCheck(ctx->plex[grid] != NULL, ctx->comm, PETSC_ERR_ARG_WRONG, "Plex not created"); 1490 PetscCall(DMPlexGetHeightStratum(ctx->plex[grid], 0, &cStart, &cEnd)); 1491 numCells[grid] = cEnd - cStart; // grids can have different topology 1492 PetscCall(DMGetLocalSection(ctx->plex[grid], §ion[grid])); 1493 PetscCall(DMGetGlobalSection(ctx->plex[grid], &globsection[grid])); 1494 PetscCall(PetscSectionGetNumFields(section[grid], &Nf[grid])); 1495 ncellsTot += numCells[grid]; 1496 } 1497 /* create GPU assembly data */ 1498 if (ctx->gpu_assembly) { /* we need GPU object with GPU assembly */ 1499 PetscContainer container; 1500 PetscScalar *elemMatrix, *elMat; 1501 pointInterpolationP4est(*pointMaps)[LANDAU_MAX_Q_FACE]; 1502 P4estVertexMaps *maps; 1503 const PetscInt *plex_batch = NULL, elMatSz = Nb * Nb * ctx->num_species * ctx->num_species; 1504 LandauIdx *coo_elem_offsets = NULL, *coo_elem_fullNb = NULL, (*coo_elem_point_offsets)[LANDAU_MAX_NQND + 1] = NULL; 1505 /* create GPU assembly data */ 1506 PetscCall(PetscInfo(ctx->plex[0], "Make GPU maps %d\n", 1)); 1507 PetscCall(PetscLogEventBegin(ctx->events[2], 0, 0, 0, 0)); 1508 PetscCall(PetscMalloc(sizeof(*maps) * ctx->num_grids, &maps)); 1509 PetscCall(PetscMalloc(sizeof(*pointMaps) * MAP_BF_SIZE, &pointMaps)); 1510 PetscCall(PetscMalloc(sizeof(*elemMatrix) * elMatSz, &elemMatrix)); 1511 1512 { // setup COO assembly -- put COO metadata directly in ctx->SData_d 1513 PetscCall(PetscMalloc3(ncellsTot + 1, &coo_elem_offsets, ncellsTot, &coo_elem_fullNb, ncellsTot, &coo_elem_point_offsets)); // array of integer pointers 1514 coo_elem_offsets[0] = 0; // finish later 1515 PetscCall(PetscInfo(ctx->plex[0], "COO initialization, %" PetscInt_FMT " cells\n", ncellsTot)); 1516 ctx->SData_d.coo_n_cellsTot = ncellsTot; 1517 ctx->SData_d.coo_elem_offsets = (void *)coo_elem_offsets; 1518 ctx->SData_d.coo_elem_fullNb = (void *)coo_elem_fullNb; 1519 ctx->SData_d.coo_elem_point_offsets = (void *)coo_elem_point_offsets; 1520 } 1521 1522 ctx->SData_d.coo_max_fullnb = 0; 1523 for (PetscInt grid = 0, glb_elem_idx = 0; grid < ctx->num_grids; grid++) { 1524 PetscInt cStart, cEnd, Nfloc = Nf[grid], totDim = Nfloc * Nb; 1525 if (grid_batch_is_inv[grid]) PetscCall(ISGetIndices(grid_batch_is_inv[grid], &plex_batch)); 1526 PetscCheck(!plex_batch, ctx->comm, PETSC_ERR_ARG_WRONG, "-dm_landau_jacobian_field_major_order DEPRECATED"); 1527 PetscCall(DMPlexGetHeightStratum(ctx->plex[grid], 0, &cStart, &cEnd)); 1528 // make maps 1529 maps[grid].d_self = NULL; 1530 maps[grid].num_elements = numCells[grid]; 1531 maps[grid].num_face = (PetscInt)(pow(Nq, 1. / ((double)dim)) + .001); // Q 1532 maps[grid].num_face = (PetscInt)(pow(maps[grid].num_face, (double)(dim - 1)) + .001); // Q^2 1533 maps[grid].num_reduced = 0; 1534 maps[grid].deviceType = ctx->deviceType; 1535 maps[grid].numgrids = ctx->num_grids; 1536 // count reduced and get 1537 PetscCall(PetscMalloc(maps[grid].num_elements * sizeof(*maps[grid].gIdx), &maps[grid].gIdx)); 1538 for (PetscInt ej = cStart, eidx = 0; ej < cEnd; ++ej, ++eidx, glb_elem_idx++) { 1539 if (coo_elem_offsets) coo_elem_offsets[glb_elem_idx + 1] = coo_elem_offsets[glb_elem_idx]; // start with last one, then add 1540 for (PetscInt fieldA = 0; fieldA < Nf[grid]; fieldA++) { 1541 PetscInt fullNb = 0; 1542 for (PetscInt q = 0; q < Nb; ++q) { 1543 PetscInt numindices, *indices; 1544 PetscScalar *valuesOrig = elMat = elemMatrix; 1545 PetscCall(PetscArrayzero(elMat, totDim * totDim)); 1546 elMat[(fieldA * Nb + q) * totDim + fieldA * Nb + q] = 1; 1547 PetscCall(DMPlexGetClosureIndices(ctx->plex[grid], section[grid], globsection[grid], ej, PETSC_TRUE, &numindices, &indices, NULL, &elMat)); 1548 if (ctx->simplex) { 1549 PetscCheck(numindices == Nb, ctx->comm, PETSC_ERR_ARG_WRONG, "numindices != Nb numindices=%" PetscInt_FMT " Nb=%" PetscInt_FMT, numindices, Nb); 1550 for (PetscInt q = 0; q < numindices; ++q) maps[grid].gIdx[eidx][fieldA][q] = indices[q]; 1551 fullNb++; 1552 } else { 1553 for (PetscInt f = 0; f < numindices; ++f) { // look for a non-zero on the diagonal (is this too complicated for simplices?) 1554 if (PetscAbs(PetscRealPart(elMat[f * numindices + f])) > PETSC_MACHINE_EPSILON) { 1555 // found it 1556 if (PetscAbs(PetscRealPart(elMat[f * numindices + f] - 1.)) < PETSC_MACHINE_EPSILON) { // normal vertex 1.0 1557 if (plex_batch) { 1558 maps[grid].gIdx[eidx][fieldA][q] = plex_batch[indices[f]]; 1559 } else { 1560 maps[grid].gIdx[eidx][fieldA][q] = indices[f]; 1561 } 1562 fullNb++; 1563 } else { //found a constraint 1564 PetscInt jj = 0; 1565 PetscReal sum = 0; 1566 const PetscInt ff = f; 1567 maps[grid].gIdx[eidx][fieldA][q] = -maps[grid].num_reduced - 1; // store (-)index: id = -(idx+1): idx = -id - 1 1568 PetscCheck(!ctx->simplex, ctx->comm, PETSC_ERR_ARG_WRONG, "No constraints with simplex"); 1569 do { // constraints are continuous in Plex - exploit that here 1570 PetscInt ii; // get 'scale' 1571 for (ii = 0, pointMaps[maps[grid].num_reduced][jj].scale = 0; ii < maps[grid].num_face; ii++) { // sum row of outer product to recover vector value 1572 if (ff + ii < numindices) { // 3D has Q and Q^2 interps so might run off end. We could test that elMat[f*numindices + ff + ii] > 0, and break if not 1573 pointMaps[maps[grid].num_reduced][jj].scale += PetscRealPart(elMat[f * numindices + ff + ii]); 1574 } 1575 } 1576 sum += pointMaps[maps[grid].num_reduced][jj].scale; // diagnostic 1577 // get 'gid' 1578 if (pointMaps[maps[grid].num_reduced][jj].scale == 0) pointMaps[maps[grid].num_reduced][jj].gid = -1; // 3D has Q and Q^2 interps 1579 else { 1580 if (plex_batch) { 1581 pointMaps[maps[grid].num_reduced][jj].gid = plex_batch[indices[f]]; 1582 } else { 1583 pointMaps[maps[grid].num_reduced][jj].gid = indices[f]; 1584 } 1585 fullNb++; 1586 } 1587 } while (++jj < maps[grid].num_face && ++f < numindices); // jj is incremented if we hit the end 1588 while (jj < maps[grid].num_face) { 1589 pointMaps[maps[grid].num_reduced][jj].scale = 0; 1590 pointMaps[maps[grid].num_reduced][jj].gid = -1; 1591 jj++; 1592 } 1593 if (PetscAbs(sum - 1.0) > 10 * PETSC_MACHINE_EPSILON) { // debug 1594 PetscInt d, f; 1595 PetscReal tmp = 0; 1596 PetscCall( 1597 PetscPrintf(PETSC_COMM_SELF, "\t\t%" PetscInt_FMT ".%" PetscInt_FMT ".%" PetscInt_FMT ") ERROR total I = %22.16e (LANDAU_MAX_Q_FACE=%d, #face=%" PetscInt_FMT ")\n", eidx, q, fieldA, (double)sum, LANDAU_MAX_Q_FACE, maps[grid].num_face)); 1598 for (d = 0, tmp = 0; d < numindices; ++d) { 1599 if (tmp != 0 && PetscAbs(tmp - 1.0) > 10 * PETSC_MACHINE_EPSILON) PetscCall(PetscPrintf(PETSC_COMM_WORLD, "%3" PetscInt_FMT ") %3" PetscInt_FMT ": ", d, indices[d])); 1600 for (f = 0; f < numindices; ++f) tmp += PetscRealPart(elMat[d * numindices + f]); 1601 if (tmp != 0) PetscCall(PetscPrintf(ctx->comm, " | %22.16e\n", (double)tmp)); 1602 } 1603 } 1604 maps[grid].num_reduced++; 1605 PetscCheck(maps[grid].num_reduced < MAP_BF_SIZE, PETSC_COMM_SELF, PETSC_ERR_PLIB, "maps[grid].num_reduced %" PetscInt_FMT " > %" PetscInt_FMT, maps[grid].num_reduced, MAP_BF_SIZE); 1606 } 1607 break; 1608 } 1609 } 1610 } // !simplex 1611 // cleanup 1612 PetscCall(DMPlexRestoreClosureIndices(ctx->plex[grid], section[grid], globsection[grid], ej, PETSC_TRUE, &numindices, &indices, NULL, &elMat)); 1613 if (elMat != valuesOrig) PetscCall(DMRestoreWorkArray(ctx->plex[grid], numindices * numindices, MPIU_SCALAR, &elMat)); 1614 } 1615 { // setup COO assembly 1616 coo_elem_offsets[glb_elem_idx + 1] += fullNb * fullNb; // one species block, adds a block for each species, on this element in this grid 1617 if (fieldA == 0) { // cache full Nb for this element, on this grid per species 1618 coo_elem_fullNb[glb_elem_idx] = fullNb; 1619 if (fullNb > ctx->SData_d.coo_max_fullnb) ctx->SData_d.coo_max_fullnb = fullNb; 1620 } else PetscCheck(coo_elem_fullNb[glb_elem_idx] == fullNb, PETSC_COMM_SELF, PETSC_ERR_PLIB, "full element size change with species %" PetscInt_FMT " %" PetscInt_FMT, coo_elem_fullNb[glb_elem_idx], fullNb); 1621 } 1622 } // field 1623 } // cell 1624 // allocate and copy point data maps[grid].gIdx[eidx][field][q] 1625 PetscCall(PetscMalloc(maps[grid].num_reduced * sizeof(*maps[grid].c_maps), &maps[grid].c_maps)); 1626 for (PetscInt ej = 0; ej < maps[grid].num_reduced; ++ej) { 1627 for (PetscInt q = 0; q < maps[grid].num_face; ++q) { 1628 maps[grid].c_maps[ej][q].scale = pointMaps[ej][q].scale; 1629 maps[grid].c_maps[ej][q].gid = pointMaps[ej][q].gid; 1630 } 1631 } 1632 #if defined(PETSC_HAVE_KOKKOS) 1633 if (ctx->deviceType == LANDAU_KOKKOS) PetscCall(LandauKokkosCreateMatMaps(maps, pointMaps, Nf, grid)); // implies Kokkos does 1634 #endif 1635 if (plex_batch) { 1636 PetscCall(ISRestoreIndices(grid_batch_is_inv[grid], &plex_batch)); 1637 PetscCall(ISDestroy(&grid_batch_is_inv[grid])); // we are done with this 1638 } 1639 } /* grids */ 1640 // finish COO 1641 { // setup COO assembly 1642 PetscInt *oor, *ooc; 1643 ctx->SData_d.coo_size = coo_elem_offsets[ncellsTot] * ctx->batch_sz; 1644 PetscCall(PetscMalloc2(ctx->SData_d.coo_size, &oor, ctx->SData_d.coo_size, &ooc)); 1645 for (PetscInt i = 0; i < ctx->SData_d.coo_size; i++) oor[i] = ooc[i] = -1; 1646 // get 1647 for (PetscInt grid = 0, glb_elem_idx = 0; grid < ctx->num_grids; grid++) { 1648 for (PetscInt ej = 0; ej < numCells[grid]; ++ej, glb_elem_idx++) { 1649 const PetscInt fullNb = coo_elem_fullNb[glb_elem_idx]; 1650 const LandauIdx *const Idxs = &maps[grid].gIdx[ej][0][0]; // just use field-0 maps, They should be the same but this is just for COO storage 1651 coo_elem_point_offsets[glb_elem_idx][0] = 0; 1652 for (PetscInt f = 0, cnt2 = 0; f < Nb; f++) { 1653 PetscInt idx = Idxs[f]; 1654 coo_elem_point_offsets[glb_elem_idx][f + 1] = coo_elem_point_offsets[glb_elem_idx][f]; // start at last 1655 if (idx >= 0) { 1656 cnt2++; 1657 coo_elem_point_offsets[glb_elem_idx][f + 1]++; // inc 1658 } else { 1659 idx = -idx - 1; 1660 for (PetscInt q = 0; q < maps[grid].num_face; q++) { 1661 if (maps[grid].c_maps[idx][q].gid < 0) break; 1662 cnt2++; 1663 coo_elem_point_offsets[glb_elem_idx][f + 1]++; // inc 1664 } 1665 } 1666 PetscCheck(cnt2 <= fullNb, PETSC_COMM_SELF, PETSC_ERR_PLIB, "wrong count %" PetscInt_FMT " < %" PetscInt_FMT, fullNb, cnt2); 1667 } 1668 PetscCheck(coo_elem_point_offsets[glb_elem_idx][Nb] == fullNb, PETSC_COMM_SELF, PETSC_ERR_PLIB, "coo_elem_point_offsets size %" PetscInt_FMT " != fullNb=%" PetscInt_FMT, coo_elem_point_offsets[glb_elem_idx][Nb], fullNb); 1669 } 1670 } 1671 // set 1672 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { 1673 for (PetscInt grid = 0, glb_elem_idx = 0; grid < ctx->num_grids; grid++) { 1674 const PetscInt moffset = LAND_MOFFSET(b_id, grid, ctx->batch_sz, ctx->num_grids, ctx->mat_offset); 1675 for (PetscInt ej = 0; ej < numCells[grid]; ++ej, glb_elem_idx++) { 1676 const PetscInt fullNb = coo_elem_fullNb[glb_elem_idx], fullNb2 = fullNb * fullNb; 1677 // set (i,j) 1678 for (PetscInt fieldA = 0; fieldA < Nf[grid]; fieldA++) { 1679 const LandauIdx *const Idxs = &maps[grid].gIdx[ej][fieldA][0]; 1680 PetscInt rows[LANDAU_MAX_Q_FACE], cols[LANDAU_MAX_Q_FACE]; 1681 for (PetscInt f = 0; f < Nb; ++f) { 1682 const PetscInt nr = coo_elem_point_offsets[glb_elem_idx][f + 1] - coo_elem_point_offsets[glb_elem_idx][f]; 1683 if (nr == 1) rows[0] = Idxs[f]; 1684 else { 1685 const PetscInt idx = -Idxs[f] - 1; 1686 for (PetscInt q = 0; q < nr; q++) rows[q] = maps[grid].c_maps[idx][q].gid; 1687 } 1688 for (PetscInt g = 0; g < Nb; ++g) { 1689 const PetscInt nc = coo_elem_point_offsets[glb_elem_idx][g + 1] - coo_elem_point_offsets[glb_elem_idx][g]; 1690 if (nc == 1) cols[0] = Idxs[g]; 1691 else { 1692 const PetscInt idx = -Idxs[g] - 1; 1693 for (PetscInt q = 0; q < nc; q++) cols[q] = maps[grid].c_maps[idx][q].gid; 1694 } 1695 const PetscInt idx0 = b_id * coo_elem_offsets[ncellsTot] + coo_elem_offsets[glb_elem_idx] + fieldA * fullNb2 + fullNb * coo_elem_point_offsets[glb_elem_idx][f] + nr * coo_elem_point_offsets[glb_elem_idx][g]; 1696 for (PetscInt q = 0, idx = idx0; q < nr; q++) { 1697 for (PetscInt d = 0; d < nc; d++, idx++) { 1698 oor[idx] = rows[q] + moffset; 1699 ooc[idx] = cols[d] + moffset; 1700 } 1701 } 1702 } 1703 } 1704 } 1705 } // cell 1706 } // grid 1707 } // batch 1708 PetscCall(MatSetPreallocationCOO(ctx->J, ctx->SData_d.coo_size, oor, ooc)); 1709 PetscCall(PetscFree2(oor, ooc)); 1710 } 1711 PetscCall(PetscFree(pointMaps)); 1712 PetscCall(PetscFree(elemMatrix)); 1713 PetscCall(PetscContainerCreate(PETSC_COMM_SELF, &container)); 1714 PetscCall(PetscContainerSetPointer(container, (void *)maps)); 1715 PetscCall(PetscContainerSetCtxDestroy(container, LandauGPUMapsDestroy)); 1716 PetscCall(PetscObjectCompose((PetscObject)ctx->J, "assembly_maps", (PetscObject)container)); 1717 PetscCall(PetscContainerDestroy(&container)); 1718 PetscCall(PetscLogEventEnd(ctx->events[2], 0, 0, 0, 0)); 1719 } // end GPU assembly 1720 { /* create static point data, Jacobian called first, only one vertex copy */ 1721 PetscReal *invJe, *ww, *xx, *yy, *zz = NULL, *invJ_a; 1722 PetscInt outer_ipidx, outer_ej, grid, nip_glb = 0; 1723 PetscFE fe; 1724 PetscCall(PetscLogEventBegin(ctx->events[7], 0, 0, 0, 0)); 1725 PetscCall(PetscInfo(ctx->plex[0], "Initialize static data\n")); 1726 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) nip_glb += Nq * numCells[grid]; 1727 /* collect f data, first time is for Jacobian, but make mass now */ 1728 if (ctx->verbose != 0) { 1729 PetscInt ncells = 0, N; 1730 MatInfo info; 1731 PetscCall(MatGetInfo(ctx->J, MAT_LOCAL, &info)); 1732 PetscCall(MatGetSize(ctx->J, &N, NULL)); 1733 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) ncells += numCells[grid]; 1734 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "%d) %s %" PetscInt_FMT " IPs, %" PetscInt_FMT " cells total, Nb=%" PetscInt_FMT ", Nq=%" PetscInt_FMT ", dim=%" PetscInt_FMT ", Tab: Nb=%" PetscInt_FMT " Nf=%" PetscInt_FMT " Np=%" PetscInt_FMT " cdim=%" PetscInt_FMT " N=%" PetscInt_FMT " nnz= %" PetscInt_FMT "\n", 0, "FormLandau", nip_glb, ncells, Nb, Nq, dim, Nb, 1735 ctx->num_species, Nb, dim, N, (PetscInt)info.nz_used)); 1736 } 1737 PetscCall(PetscMalloc4(nip_glb, &ww, nip_glb, &xx, nip_glb, &yy, nip_glb * dim * dim, &invJ_a)); 1738 if (dim == 3) PetscCall(PetscMalloc1(nip_glb, &zz)); 1739 if (ctx->use_energy_tensor_trick) { 1740 PetscCall(PetscFECreateDefault(PETSC_COMM_SELF, dim, 1, ctx->simplex, NULL, PETSC_DECIDE, &fe)); 1741 PetscCall(PetscObjectSetName((PetscObject)fe, "energy")); 1742 } 1743 /* init each grids static data - no batch */ 1744 for (grid = 0, outer_ipidx = 0, outer_ej = 0; grid < ctx->num_grids; grid++) { // OpenMP (once) 1745 Vec v2_2 = NULL; // projected function: v^2/2 for non-relativistic, gamma... for relativistic 1746 PetscSection e_section; 1747 DM dmEnergy; 1748 PetscInt cStart, cEnd, ej; 1749 1750 PetscCall(DMPlexGetHeightStratum(ctx->plex[grid], 0, &cStart, &cEnd)); 1751 // prep energy trick, get v^2 / 2 vector 1752 if (ctx->use_energy_tensor_trick) { 1753 PetscErrorCode (*energyf[1])(PetscInt, PetscReal, const PetscReal[], PetscInt, PetscScalar[], void *) = {ctx->use_relativistic_corrections ? gamma_m1_f : energy_f}; 1754 Vec glob_v2; 1755 PetscReal *c2_0[1], data[1] = {PetscSqr(C_0(ctx->v_0))}; 1756 1757 PetscCall(DMClone(ctx->plex[grid], &dmEnergy)); 1758 PetscCall(PetscObjectSetName((PetscObject)dmEnergy, "energy")); 1759 PetscCall(DMSetField(dmEnergy, 0, NULL, (PetscObject)fe)); 1760 PetscCall(DMCreateDS(dmEnergy)); 1761 PetscCall(DMGetLocalSection(dmEnergy, &e_section)); 1762 PetscCall(DMGetGlobalVector(dmEnergy, &glob_v2)); 1763 PetscCall(PetscObjectSetName((PetscObject)glob_v2, "trick")); 1764 c2_0[0] = &data[0]; 1765 PetscCall(DMProjectFunction(dmEnergy, 0., energyf, (void **)c2_0, INSERT_ALL_VALUES, glob_v2)); 1766 PetscCall(DMGetLocalVector(dmEnergy, &v2_2)); 1767 PetscCall(VecZeroEntries(v2_2)); /* zero BCs so don't set */ 1768 PetscCall(DMGlobalToLocalBegin(dmEnergy, glob_v2, INSERT_VALUES, v2_2)); 1769 PetscCall(DMGlobalToLocalEnd(dmEnergy, glob_v2, INSERT_VALUES, v2_2)); 1770 PetscCall(DMViewFromOptions(dmEnergy, NULL, "-energy_dm_view")); 1771 PetscCall(VecViewFromOptions(glob_v2, NULL, "-energy_vec_view")); 1772 PetscCall(DMRestoreGlobalVector(dmEnergy, &glob_v2)); 1773 } 1774 /* append part of the IP data for each grid */ 1775 for (ej = 0; ej < numCells[grid]; ++ej, ++outer_ej) { 1776 PetscScalar *coefs = NULL; 1777 PetscReal vj[LANDAU_MAX_NQND * LANDAU_DIM], detJj[LANDAU_MAX_NQND], Jdummy[LANDAU_MAX_NQND * LANDAU_DIM * LANDAU_DIM], c0 = C_0(ctx->v_0), c02 = PetscSqr(c0); 1778 invJe = invJ_a + outer_ej * Nq * dim * dim; 1779 PetscCall(DMPlexComputeCellGeometryFEM(ctx->plex[grid], ej + cStart, quad, vj, Jdummy, invJe, detJj)); 1780 if (ctx->use_energy_tensor_trick) PetscCall(DMPlexVecGetClosure(dmEnergy, e_section, v2_2, ej + cStart, NULL, &coefs)); 1781 /* create static point data */ 1782 for (PetscInt qj = 0; qj < Nq; qj++, outer_ipidx++) { 1783 const PetscInt gidx = outer_ipidx; 1784 const PetscReal *invJ = &invJe[qj * dim * dim]; 1785 ww[gidx] = detJj[qj] * quadWeights[qj]; 1786 if (dim == 2) ww[gidx] *= vj[qj * dim + 0]; /* cylindrical coordinate, w/o 2pi */ 1787 // get xx, yy, zz 1788 if (ctx->use_energy_tensor_trick) { 1789 double refSpaceDer[3], eGradPhi[3]; 1790 const PetscReal *const DD = Tf[0]->T[1]; 1791 const PetscReal *Dq = &DD[qj * Nb * dim]; 1792 for (PetscInt d = 0; d < 3; ++d) refSpaceDer[d] = eGradPhi[d] = 0.0; 1793 for (PetscInt b = 0; b < Nb; ++b) { 1794 for (PetscInt d = 0; d < dim; ++d) refSpaceDer[d] += Dq[b * dim + d] * PetscRealPart(coefs[b]); 1795 } 1796 xx[gidx] = 1e10; 1797 if (ctx->use_relativistic_corrections) { 1798 double dg2_c2 = 0; 1799 //for (PetscInt d = 0; d < dim; ++d) refSpaceDer[d] *= c02; 1800 for (PetscInt d = 0; d < dim; ++d) dg2_c2 += PetscSqr(refSpaceDer[d]); 1801 dg2_c2 *= (double)c02; 1802 if (dg2_c2 >= .999) { 1803 xx[gidx] = vj[qj * dim + 0]; /* coordinate */ 1804 yy[gidx] = vj[qj * dim + 1]; 1805 if (dim == 3) zz[gidx] = vj[qj * dim + 2]; 1806 PetscCall(PetscPrintf(ctx->comm, "Error: %12.5e %" PetscInt_FMT ".%" PetscInt_FMT ") dg2/c02 = %12.5e x= %12.5e %12.5e %12.5e\n", (double)PetscSqrtReal(xx[gidx] * xx[gidx] + yy[gidx] * yy[gidx] + zz[gidx] * zz[gidx]), ej, qj, dg2_c2, (double)xx[gidx], (double)yy[gidx], (double)zz[gidx])); 1807 } else { 1808 PetscReal fact = c02 / PetscSqrtReal(1. - dg2_c2); 1809 for (PetscInt d = 0; d < dim; ++d) refSpaceDer[d] *= fact; 1810 // could test with other point u' that (grad - grad') * U (refSpaceDer, refSpaceDer') == 0 1811 } 1812 } 1813 if (xx[gidx] == 1e10) { 1814 for (PetscInt d = 0; d < dim; ++d) { 1815 for (PetscInt e = 0; e < dim; ++e) eGradPhi[d] += invJ[e * dim + d] * refSpaceDer[e]; 1816 } 1817 xx[gidx] = eGradPhi[0]; 1818 yy[gidx] = eGradPhi[1]; 1819 if (dim == 3) zz[gidx] = eGradPhi[2]; 1820 } 1821 } else { 1822 xx[gidx] = vj[qj * dim + 0]; /* coordinate */ 1823 yy[gidx] = vj[qj * dim + 1]; 1824 if (dim == 3) zz[gidx] = vj[qj * dim + 2]; 1825 } 1826 } /* q */ 1827 if (ctx->use_energy_tensor_trick) PetscCall(DMPlexVecRestoreClosure(dmEnergy, e_section, v2_2, ej + cStart, NULL, &coefs)); 1828 } /* ej */ 1829 if (ctx->use_energy_tensor_trick) { 1830 PetscCall(DMRestoreLocalVector(dmEnergy, &v2_2)); 1831 PetscCall(DMDestroy(&dmEnergy)); 1832 } 1833 } /* grid */ 1834 if (ctx->use_energy_tensor_trick) PetscCall(PetscFEDestroy(&fe)); 1835 /* cache static data */ 1836 if (ctx->deviceType == LANDAU_KOKKOS) { 1837 #if defined(PETSC_HAVE_KOKKOS) 1838 PetscCall(LandauKokkosStaticDataSet(ctx->plex[0], Nq, Nb, ctx->batch_sz, ctx->num_grids, numCells, ctx->species_offset, ctx->mat_offset, nu_alpha, nu_beta, invMass, (PetscReal *)ctx->lambdas, invJ_a, xx, yy, zz, ww, &ctx->SData_d)); 1839 /* free */ 1840 PetscCall(PetscFree4(ww, xx, yy, invJ_a)); 1841 if (dim == 3) PetscCall(PetscFree(zz)); 1842 #else 1843 SETERRQ(ctx->comm, PETSC_ERR_ARG_WRONG, "-landau_device_type kokkos not built"); 1844 #endif 1845 } else { /* CPU version, just copy in, only use part */ 1846 PetscReal *nu_alpha_p = (PetscReal *)ctx->SData_d.alpha, *nu_beta_p = (PetscReal *)ctx->SData_d.beta, *invMass_p = (PetscReal *)ctx->SData_d.invMass, *lambdas_p = NULL; // why set these ? 1847 ctx->SData_d.w = (void *)ww; 1848 ctx->SData_d.x = (void *)xx; 1849 ctx->SData_d.y = (void *)yy; 1850 ctx->SData_d.z = (void *)zz; 1851 ctx->SData_d.invJ = (void *)invJ_a; 1852 PetscCall(PetscMalloc4(ctx->num_species, &nu_alpha_p, ctx->num_species, &nu_beta_p, ctx->num_species, &invMass_p, LANDAU_MAX_GRIDS * LANDAU_MAX_GRIDS, &lambdas_p)); 1853 for (PetscInt ii = 0; ii < ctx->num_species; ii++) { 1854 nu_alpha_p[ii] = nu_alpha[ii]; 1855 nu_beta_p[ii] = nu_beta[ii]; 1856 invMass_p[ii] = invMass[ii]; 1857 } 1858 ctx->SData_d.alpha = (void *)nu_alpha_p; 1859 ctx->SData_d.beta = (void *)nu_beta_p; 1860 ctx->SData_d.invMass = (void *)invMass_p; 1861 ctx->SData_d.lambdas = (void *)lambdas_p; 1862 for (PetscInt grid = 0; grid < LANDAU_MAX_GRIDS; grid++) { 1863 PetscReal (*lambdas)[LANDAU_MAX_GRIDS][LANDAU_MAX_GRIDS] = (PetscReal (*)[LANDAU_MAX_GRIDS][LANDAU_MAX_GRIDS])ctx->SData_d.lambdas; 1864 for (PetscInt gridj = 0; gridj < LANDAU_MAX_GRIDS; gridj++) (*lambdas)[grid][gridj] = ctx->lambdas[grid][gridj]; 1865 } 1866 } 1867 PetscCall(PetscLogEventEnd(ctx->events[7], 0, 0, 0, 0)); 1868 } // initialize 1869 PetscFunctionReturn(PETSC_SUCCESS); 1870 } 1871 1872 /* < v, u > */ 1873 static void g0_1(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, PetscReal u_tShift, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar g0[]) 1874 { 1875 g0[0] = 1.; 1876 } 1877 1878 /* < v, u > */ 1879 static void g0_fake(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, PetscReal u_tShift, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar g0[]) 1880 { 1881 static double ttt = 1e-12; 1882 g0[0] = ttt++; 1883 } 1884 1885 /* < v, u > */ 1886 static void g0_r(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, PetscReal u_tShift, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar g0[]) 1887 { 1888 g0[0] = 2. * PETSC_PI * x[0]; 1889 } 1890 1891 /* 1892 LandauCreateJacobianMatrix - creates ctx->J with without real data. Hard to keep sparse. 1893 - Like DMPlexLandauCreateMassMatrix. Should remove one and combine 1894 - has old support for field major ordering 1895 */ 1896 static PetscErrorCode LandauCreateJacobianMatrix(MPI_Comm comm, Vec X, IS grid_batch_is_inv[LANDAU_MAX_GRIDS], LandauCtx *ctx) 1897 { 1898 PetscInt *idxs = NULL; 1899 Mat subM[LANDAU_MAX_GRIDS]; 1900 1901 PetscFunctionBegin; 1902 if (!ctx->gpu_assembly) { /* we need GPU object with GPU assembly */ 1903 PetscFunctionReturn(PETSC_SUCCESS); 1904 } 1905 // get the RCM for this grid to separate out species into blocks -- create 'idxs' & 'ctx->batch_is' -- not used 1906 if (ctx->gpu_assembly && ctx->jacobian_field_major_order) PetscCall(PetscMalloc1(ctx->mat_offset[ctx->num_grids] * ctx->batch_sz, &idxs)); 1907 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1908 const PetscInt *values, n = ctx->mat_offset[grid + 1] - ctx->mat_offset[grid]; 1909 Mat gMat; 1910 DM massDM; 1911 PetscDS prob; 1912 Vec tvec; 1913 // get "mass" matrix for reordering 1914 PetscCall(DMClone(ctx->plex[grid], &massDM)); 1915 PetscCall(DMCopyFields(ctx->plex[grid], PETSC_DETERMINE, PETSC_DETERMINE, massDM)); 1916 PetscCall(DMCreateDS(massDM)); 1917 PetscCall(DMGetDS(massDM, &prob)); 1918 for (PetscInt ix = 0, ii = ctx->species_offset[grid]; ii < ctx->species_offset[grid + 1]; ii++, ix++) PetscCall(PetscDSSetJacobian(prob, ix, ix, g0_fake, NULL, NULL, NULL)); 1919 PetscCall(PetscOptionsInsertString(NULL, "-dm_preallocate_only")); // this trick is need to both sparsify the matrix and avoid runtime error 1920 PetscCall(DMCreateMatrix(massDM, &gMat)); 1921 PetscCall(PetscOptionsInsertString(NULL, "-dm_preallocate_only false")); 1922 PetscCall(MatSetOption(gMat, MAT_STRUCTURALLY_SYMMETRIC, PETSC_TRUE)); 1923 PetscCall(MatSetOption(gMat, MAT_IGNORE_ZERO_ENTRIES, PETSC_TRUE)); 1924 PetscCall(DMCreateLocalVector(ctx->plex[grid], &tvec)); 1925 PetscCall(DMPlexSNESComputeJacobianFEM(massDM, tvec, gMat, gMat, ctx)); 1926 PetscCall(MatViewFromOptions(gMat, NULL, "-dm_landau_reorder_mat_view")); 1927 PetscCall(DMDestroy(&massDM)); 1928 PetscCall(VecDestroy(&tvec)); 1929 subM[grid] = gMat; 1930 if (ctx->gpu_assembly && ctx->jacobian_field_major_order) { 1931 MatOrderingType rtype = MATORDERINGRCM; 1932 IS isrow, isicol; 1933 PetscCall(MatGetOrdering(gMat, rtype, &isrow, &isicol)); 1934 PetscCall(ISInvertPermutation(isrow, PETSC_DECIDE, &grid_batch_is_inv[grid])); 1935 PetscCall(ISGetIndices(isrow, &values)); 1936 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { // add batch size DMs for this species grid 1937 #if !defined(LANDAU_SPECIES_MAJOR) 1938 PetscInt N = ctx->mat_offset[ctx->num_grids], n0 = ctx->mat_offset[grid] + b_id * N; 1939 for (PetscInt ii = 0; ii < n; ++ii) idxs[n0 + ii] = values[ii] + n0; 1940 #else 1941 PetscInt n0 = ctx->mat_offset[grid] * ctx->batch_sz + b_id * n; 1942 for (PetscInt ii = 0; ii < n; ++ii) idxs[n0 + ii] = values[ii] + n0; 1943 #endif 1944 } 1945 PetscCall(ISRestoreIndices(isrow, &values)); 1946 PetscCall(ISDestroy(&isrow)); 1947 PetscCall(ISDestroy(&isicol)); 1948 } 1949 } 1950 if (ctx->gpu_assembly && ctx->jacobian_field_major_order) PetscCall(ISCreateGeneral(comm, ctx->mat_offset[ctx->num_grids] * ctx->batch_sz, idxs, PETSC_OWN_POINTER, &ctx->batch_is)); 1951 // get a block matrix 1952 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 1953 Mat B = subM[grid]; 1954 PetscInt nloc, nzl, *colbuf, row, COL_BF_SIZE = 1024; 1955 PetscCall(PetscMalloc(sizeof(*colbuf) * COL_BF_SIZE, &colbuf)); 1956 PetscCall(MatGetSize(B, &nloc, NULL)); 1957 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { 1958 const PetscInt moffset = LAND_MOFFSET(b_id, grid, ctx->batch_sz, ctx->num_grids, ctx->mat_offset); 1959 const PetscInt *cols; 1960 const PetscScalar *vals; 1961 for (PetscInt i = 0; i < nloc; i++) { 1962 PetscCall(MatGetRow(B, i, &nzl, NULL, NULL)); 1963 if (nzl > COL_BF_SIZE) { 1964 PetscCall(PetscFree(colbuf)); 1965 PetscCall(PetscInfo(ctx->plex[grid], "Realloc buffer %" PetscInt_FMT " to %" PetscInt_FMT " (row size %" PetscInt_FMT ") \n", COL_BF_SIZE, 2 * COL_BF_SIZE, nzl)); 1966 COL_BF_SIZE = nzl; 1967 PetscCall(PetscMalloc(sizeof(*colbuf) * COL_BF_SIZE, &colbuf)); 1968 } 1969 PetscCall(MatGetRow(B, i, &nzl, &cols, &vals)); 1970 for (PetscInt j = 0; j < nzl; j++) colbuf[j] = cols[j] + moffset; 1971 row = i + moffset; 1972 PetscCall(MatSetValues(ctx->J, 1, &row, nzl, colbuf, vals, INSERT_VALUES)); 1973 PetscCall(MatRestoreRow(B, i, &nzl, &cols, &vals)); 1974 } 1975 } 1976 PetscCall(PetscFree(colbuf)); 1977 } 1978 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) PetscCall(MatDestroy(&subM[grid])); 1979 PetscCall(MatAssemblyBegin(ctx->J, MAT_FINAL_ASSEMBLY)); 1980 PetscCall(MatAssemblyEnd(ctx->J, MAT_FINAL_ASSEMBLY)); 1981 1982 // debug 1983 PetscCall(MatViewFromOptions(ctx->J, NULL, "-dm_landau_mat_view")); 1984 if (ctx->gpu_assembly && ctx->jacobian_field_major_order) { 1985 Mat mat_block_order; 1986 PetscCall(MatCreateSubMatrix(ctx->J, ctx->batch_is, ctx->batch_is, MAT_INITIAL_MATRIX, &mat_block_order)); // use MatPermute 1987 PetscCall(MatViewFromOptions(mat_block_order, NULL, "-dm_landau_mat_view")); 1988 PetscCall(MatDestroy(&mat_block_order)); 1989 PetscCall(VecScatterCreate(X, ctx->batch_is, X, NULL, &ctx->plex_batch)); 1990 PetscCall(VecDuplicate(X, &ctx->work_vec)); 1991 } 1992 PetscFunctionReturn(PETSC_SUCCESS); 1993 } 1994 1995 PetscErrorCode DMPlexLandauCreateMassMatrix(DM pack, Mat *Amat); 1996 /*@C 1997 DMPlexLandauCreateVelocitySpace - Create a `DMPLEX` velocity space mesh 1998 1999 Collective 2000 2001 Input Parameters: 2002 + comm - The MPI communicator 2003 . dim - velocity space dimension (2 for axisymmetric, 3 for full 3X + 3V solver) 2004 - prefix - prefix for options (not tested) 2005 2006 Output Parameters: 2007 + pack - The `DM` object representing the mesh 2008 . X - A vector (user destroys) 2009 - J - Optional matrix (object destroys) 2010 2011 Level: beginner 2012 2013 .seealso: `DMPlexCreate()`, `DMPlexLandauDestroyVelocitySpace()` 2014 @*/ 2015 PetscErrorCode DMPlexLandauCreateVelocitySpace(MPI_Comm comm, PetscInt dim, const char prefix[], Vec *X, Mat *J, DM *pack) 2016 { 2017 LandauCtx *ctx; 2018 Vec Xsub[LANDAU_MAX_GRIDS]; 2019 IS grid_batch_is_inv[LANDAU_MAX_GRIDS]; 2020 2021 PetscFunctionBegin; 2022 PetscCheck(dim == 2 || dim == 3, PETSC_COMM_SELF, PETSC_ERR_PLIB, "Only 2D and 3D supported"); 2023 PetscCheck(LANDAU_DIM == dim, PETSC_COMM_SELF, PETSC_ERR_PLIB, "dim %" PetscInt_FMT " != LANDAU_DIM %d", dim, LANDAU_DIM); 2024 PetscCall(PetscNew(&ctx)); 2025 ctx->comm = comm; /* used for diagnostics and global errors */ 2026 /* process options */ 2027 PetscCall(ProcessOptions(ctx, prefix)); 2028 if (dim == 2) ctx->use_relativistic_corrections = PETSC_FALSE; 2029 /* Create Mesh */ 2030 PetscCall(DMCompositeCreate(PETSC_COMM_SELF, pack)); 2031 PetscCall(PetscLogEventBegin(ctx->events[13], 0, 0, 0, 0)); 2032 PetscCall(PetscLogEventBegin(ctx->events[15], 0, 0, 0, 0)); 2033 PetscCall(LandauDMCreateVMeshes(PETSC_COMM_SELF, dim, prefix, ctx, *pack)); // creates grids (Forest of AMR) 2034 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 2035 /* create FEM */ 2036 PetscCall(SetupDS(ctx->plex[grid], dim, grid, ctx)); 2037 /* set initial state */ 2038 PetscCall(DMCreateGlobalVector(ctx->plex[grid], &Xsub[grid])); 2039 PetscCall(PetscObjectSetName((PetscObject)Xsub[grid], "u_orig")); 2040 /* initial static refinement, no solve */ 2041 PetscCall(LandauSetInitialCondition(ctx->plex[grid], Xsub[grid], grid, 0, 1, ctx)); 2042 /* forest refinement - forest goes in (if forest), plex comes out */ 2043 if (ctx->use_p4est) { 2044 DM plex; 2045 PetscCall(adapt(grid, ctx, &Xsub[grid])); // forest goes in, plex comes out 2046 if (grid == 0) { 2047 PetscCall(DMViewFromOptions(ctx->plex[grid], NULL, "-dm_landau_amr_dm_view")); // need to differentiate - todo 2048 PetscCall(VecViewFromOptions(Xsub[grid], NULL, "-dm_landau_amr_vec_view")); 2049 } 2050 // convert to plex, all done with this level 2051 PetscCall(DMConvert(ctx->plex[grid], DMPLEX, &plex)); 2052 PetscCall(DMDestroy(&ctx->plex[grid])); 2053 ctx->plex[grid] = plex; 2054 } 2055 #if !defined(LANDAU_SPECIES_MAJOR) 2056 PetscCall(DMCompositeAddDM(*pack, ctx->plex[grid])); 2057 #else 2058 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { // add batch size DMs for this species grid 2059 PetscCall(DMCompositeAddDM(*pack, ctx->plex[grid])); 2060 } 2061 #endif 2062 PetscCall(DMSetApplicationContext(ctx->plex[grid], ctx)); 2063 } 2064 #if !defined(LANDAU_SPECIES_MAJOR) 2065 // stack the batched DMs, could do it all here!!! b_id=0 2066 for (PetscInt b_id = 1; b_id < ctx->batch_sz; b_id++) { 2067 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) PetscCall(DMCompositeAddDM(*pack, ctx->plex[grid])); 2068 } 2069 #endif 2070 // create ctx->mat_offset 2071 ctx->mat_offset[0] = 0; 2072 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 2073 PetscInt n; 2074 PetscCall(VecGetLocalSize(Xsub[grid], &n)); 2075 ctx->mat_offset[grid + 1] = ctx->mat_offset[grid] + n; 2076 } 2077 // creat DM & Jac 2078 PetscCall(DMSetApplicationContext(*pack, ctx)); 2079 PetscCall(PetscOptionsInsertString(NULL, "-dm_preallocate_only")); 2080 PetscCall(DMCreateMatrix(*pack, &ctx->J)); 2081 PetscCall(PetscOptionsInsertString(NULL, "-dm_preallocate_only false")); 2082 PetscCall(MatSetOption(ctx->J, MAT_STRUCTURALLY_SYMMETRIC, PETSC_TRUE)); 2083 PetscCall(MatSetOption(ctx->J, MAT_IGNORE_ZERO_ENTRIES, PETSC_TRUE)); 2084 PetscCall(PetscObjectSetName((PetscObject)ctx->J, "Jac")); 2085 // construct initial conditions in X 2086 PetscCall(DMCreateGlobalVector(*pack, X)); 2087 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 2088 PetscInt n; 2089 PetscCall(VecGetLocalSize(Xsub[grid], &n)); 2090 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { 2091 PetscScalar const *values; 2092 const PetscInt moffset = LAND_MOFFSET(b_id, grid, ctx->batch_sz, ctx->num_grids, ctx->mat_offset); 2093 PetscCall(LandauSetInitialCondition(ctx->plex[grid], Xsub[grid], grid, b_id, ctx->batch_sz, ctx)); 2094 PetscCall(VecGetArrayRead(Xsub[grid], &values)); // Drop whole grid in Plex ordering 2095 for (PetscInt i = 0, idx = moffset; i < n; i++, idx++) PetscCall(VecSetValue(*X, idx, values[i], INSERT_VALUES)); 2096 PetscCall(VecRestoreArrayRead(Xsub[grid], &values)); 2097 } 2098 } 2099 // cleanup 2100 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) PetscCall(VecDestroy(&Xsub[grid])); 2101 /* check for correct matrix type */ 2102 if (ctx->gpu_assembly) { /* we need GPU object with GPU assembly */ 2103 PetscBool flg; 2104 if (ctx->deviceType == LANDAU_KOKKOS) { 2105 PetscCall(PetscObjectTypeCompareAny((PetscObject)ctx->J, &flg, MATSEQAIJKOKKOS, MATMPIAIJKOKKOS, MATAIJKOKKOS, "")); 2106 #if defined(PETSC_HAVE_KOKKOS) 2107 PetscCheck(flg, ctx->comm, PETSC_ERR_ARG_WRONG, "must use '-dm_mat_type aijkokkos -dm_vec_type kokkos' for GPU assembly and Kokkos or use '-dm_landau_device_type cpu'"); 2108 #else 2109 PetscCheck(flg, ctx->comm, PETSC_ERR_ARG_WRONG, "must configure with '--download-kokkos-kernels' for GPU assembly and Kokkos or use '-dm_landau_device_type cpu'"); 2110 #endif 2111 } 2112 } 2113 PetscCall(PetscLogEventEnd(ctx->events[15], 0, 0, 0, 0)); 2114 2115 // create field major ordering 2116 ctx->work_vec = NULL; 2117 ctx->plex_batch = NULL; 2118 ctx->batch_is = NULL; 2119 for (PetscInt i = 0; i < LANDAU_MAX_GRIDS; i++) grid_batch_is_inv[i] = NULL; 2120 PetscCall(PetscLogEventBegin(ctx->events[12], 0, 0, 0, 0)); 2121 PetscCall(LandauCreateJacobianMatrix(comm, *X, grid_batch_is_inv, ctx)); 2122 PetscCall(PetscLogEventEnd(ctx->events[12], 0, 0, 0, 0)); 2123 2124 // create AMR GPU assembly maps and static GPU data 2125 PetscCall(CreateStaticData(dim, grid_batch_is_inv, ctx)); 2126 2127 PetscCall(PetscLogEventEnd(ctx->events[13], 0, 0, 0, 0)); 2128 2129 // create mass matrix 2130 PetscCall(DMPlexLandauCreateMassMatrix(*pack, NULL)); 2131 2132 if (J) *J = ctx->J; 2133 2134 if (ctx->gpu_assembly && ctx->jacobian_field_major_order) { 2135 PetscContainer container; 2136 // cache ctx for KSP with batch/field major Jacobian ordering -ksp_type gmres/etc -dm_landau_jacobian_field_major_order 2137 PetscCall(PetscContainerCreate(PETSC_COMM_SELF, &container)); 2138 PetscCall(PetscContainerSetPointer(container, (void *)ctx)); 2139 PetscCall(PetscObjectCompose((PetscObject)ctx->J, "LandauCtx", (PetscObject)container)); 2140 PetscCall(PetscContainerDestroy(&container)); 2141 // batch solvers need to map -- can batch solvers work 2142 PetscCall(PetscContainerCreate(PETSC_COMM_SELF, &container)); 2143 PetscCall(PetscContainerSetPointer(container, (void *)ctx->plex_batch)); 2144 PetscCall(PetscObjectCompose((PetscObject)ctx->J, "plex_batch_is", (PetscObject)container)); 2145 PetscCall(PetscContainerDestroy(&container)); 2146 } 2147 // for batch solvers 2148 { 2149 PetscContainer container; 2150 PetscInt *pNf; 2151 PetscCall(PetscContainerCreate(PETSC_COMM_SELF, &container)); 2152 PetscCall(PetscMalloc1(sizeof(*pNf), &pNf)); 2153 *pNf = ctx->batch_sz; 2154 PetscCall(PetscContainerSetPointer(container, (void *)pNf)); 2155 PetscCall(PetscContainerSetCtxDestroy(container, PetscCtxDestroyDefault)); 2156 PetscCall(PetscObjectCompose((PetscObject)ctx->J, "batch size", (PetscObject)container)); 2157 PetscCall(PetscContainerDestroy(&container)); 2158 } 2159 PetscFunctionReturn(PETSC_SUCCESS); 2160 } 2161 2162 /*@C 2163 DMPlexLandauAccess - Access to the distribution function with user callback 2164 2165 Collective 2166 2167 Input Parameters: 2168 + pack - the `DMCOMPOSITE` 2169 . func - call back function 2170 - user_ctx - user context 2171 2172 Input/Output Parameter: 2173 . X - Vector to data to 2174 2175 Level: advanced 2176 2177 .seealso: `DMPlexLandauCreateVelocitySpace()` 2178 @*/ 2179 PetscErrorCode DMPlexLandauAccess(DM pack, Vec X, PetscErrorCode (*func)(DM, Vec, PetscInt, PetscInt, PetscInt, void *), void *user_ctx) 2180 { 2181 LandauCtx *ctx; 2182 2183 PetscFunctionBegin; 2184 PetscCall(DMGetApplicationContext(pack, &ctx)); // uses ctx->num_grids; ctx->plex[grid]; ctx->batch_sz; ctx->mat_offset 2185 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 2186 PetscInt dim, n; 2187 PetscCall(DMGetDimension(pack, &dim)); 2188 for (PetscInt sp = ctx->species_offset[grid], i0 = 0; sp < ctx->species_offset[grid + 1]; sp++, i0++) { 2189 Vec vec; 2190 PetscInt vf[1] = {i0}; 2191 IS vis; 2192 DM vdm; 2193 PetscCall(DMCreateSubDM(ctx->plex[grid], 1, vf, &vis, &vdm)); 2194 PetscCall(DMSetApplicationContext(vdm, ctx)); // the user might want this 2195 PetscCall(DMCreateGlobalVector(vdm, &vec)); 2196 PetscCall(VecGetSize(vec, &n)); 2197 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { 2198 const PetscInt moffset = LAND_MOFFSET(b_id, grid, ctx->batch_sz, ctx->num_grids, ctx->mat_offset); 2199 PetscCall(VecZeroEntries(vec)); 2200 /* Add your data with 'dm' for species 'sp' to 'vec' */ 2201 PetscCall(func(vdm, vec, i0, grid, b_id, user_ctx)); 2202 /* add to global */ 2203 PetscScalar const *values; 2204 const PetscInt *offsets; 2205 PetscCall(VecGetArrayRead(vec, &values)); 2206 PetscCall(ISGetIndices(vis, &offsets)); 2207 for (PetscInt i = 0; i < n; i++) PetscCall(VecSetValue(X, moffset + offsets[i], values[i], ADD_VALUES)); 2208 PetscCall(VecRestoreArrayRead(vec, &values)); 2209 PetscCall(ISRestoreIndices(vis, &offsets)); 2210 } // batch 2211 PetscCall(VecDestroy(&vec)); 2212 PetscCall(ISDestroy(&vis)); 2213 PetscCall(DMDestroy(&vdm)); 2214 } 2215 } // grid 2216 PetscFunctionReturn(PETSC_SUCCESS); 2217 } 2218 2219 /*@ 2220 DMPlexLandauDestroyVelocitySpace - Destroy a `DMPLEX` velocity space mesh 2221 2222 Collective 2223 2224 Input/Output Parameters: 2225 . dm - the `DM` to destroy 2226 2227 Level: beginner 2228 2229 .seealso: `DMPlexLandauCreateVelocitySpace()` 2230 @*/ 2231 PetscErrorCode DMPlexLandauDestroyVelocitySpace(DM *dm) 2232 { 2233 LandauCtx *ctx; 2234 2235 PetscFunctionBegin; 2236 PetscCall(DMGetApplicationContext(*dm, &ctx)); 2237 PetscCall(MatDestroy(&ctx->M)); 2238 PetscCall(MatDestroy(&ctx->J)); 2239 for (PetscInt ii = 0; ii < ctx->num_species; ii++) PetscCall(PetscFEDestroy(&ctx->fe[ii])); 2240 PetscCall(ISDestroy(&ctx->batch_is)); 2241 PetscCall(VecDestroy(&ctx->work_vec)); 2242 PetscCall(VecScatterDestroy(&ctx->plex_batch)); 2243 if (ctx->deviceType == LANDAU_KOKKOS) { 2244 #if defined(PETSC_HAVE_KOKKOS) 2245 PetscCall(LandauKokkosStaticDataClear(&ctx->SData_d)); 2246 #else 2247 SETERRQ(ctx->comm, PETSC_ERR_ARG_WRONG, "-landau_device_type %s not built", "kokkos"); 2248 #endif 2249 } else { 2250 if (ctx->SData_d.x) { /* in a CPU run */ 2251 PetscReal *invJ = (PetscReal *)ctx->SData_d.invJ, *xx = (PetscReal *)ctx->SData_d.x, *yy = (PetscReal *)ctx->SData_d.y, *zz = (PetscReal *)ctx->SData_d.z, *ww = (PetscReal *)ctx->SData_d.w; 2252 LandauIdx *coo_elem_offsets = (LandauIdx *)ctx->SData_d.coo_elem_offsets, *coo_elem_fullNb = (LandauIdx *)ctx->SData_d.coo_elem_fullNb, (*coo_elem_point_offsets)[LANDAU_MAX_NQND + 1] = (LandauIdx(*)[LANDAU_MAX_NQND + 1]) ctx->SData_d.coo_elem_point_offsets; 2253 PetscCall(PetscFree4(ww, xx, yy, invJ)); 2254 if (zz) PetscCall(PetscFree(zz)); 2255 if (coo_elem_offsets) PetscCall(PetscFree3(coo_elem_offsets, coo_elem_fullNb, coo_elem_point_offsets)); // could be NULL 2256 PetscCall(PetscFree4(ctx->SData_d.alpha, ctx->SData_d.beta, ctx->SData_d.invMass, ctx->SData_d.lambdas)); 2257 } 2258 } 2259 2260 if (ctx->times[LANDAU_MATRIX_TOTAL] > 0) { // OMP timings 2261 PetscCall(PetscPrintf(ctx->comm, "TSStep N 1.0 %10.3e\n", ctx->times[LANDAU_EX2_TSSOLVE])); 2262 PetscCall(PetscPrintf(ctx->comm, "2: Solve: %10.3e with %" PetscInt_FMT " threads\n", ctx->times[LANDAU_EX2_TSSOLVE] - ctx->times[LANDAU_MATRIX_TOTAL], ctx->batch_sz)); 2263 PetscCall(PetscPrintf(ctx->comm, "3: Landau: %10.3e\n", ctx->times[LANDAU_MATRIX_TOTAL])); 2264 PetscCall(PetscPrintf(ctx->comm, "Landau Jacobian %" PetscInt_FMT " 1.0 %10.3e\n", (PetscInt)ctx->times[LANDAU_JACOBIAN_COUNT], ctx->times[LANDAU_JACOBIAN])); 2265 PetscCall(PetscPrintf(ctx->comm, "Landau Operator N 1.0 %10.3e\n", ctx->times[LANDAU_OPERATOR])); 2266 PetscCall(PetscPrintf(ctx->comm, "Landau Mass N 1.0 %10.3e\n", ctx->times[LANDAU_MASS])); 2267 PetscCall(PetscPrintf(ctx->comm, " Jac-f-df (GPU) N 1.0 %10.3e\n", ctx->times[LANDAU_F_DF])); 2268 PetscCall(PetscPrintf(ctx->comm, " Kernel (GPU) N 1.0 %10.3e\n", ctx->times[LANDAU_KERNEL])); 2269 PetscCall(PetscPrintf(ctx->comm, "MatLUFactorNum X 1.0 %10.3e\n", ctx->times[KSP_FACTOR])); 2270 PetscCall(PetscPrintf(ctx->comm, "MatSolve X 1.0 %10.3e\n", ctx->times[KSP_SOLVE])); 2271 } 2272 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) PetscCall(DMDestroy(&ctx->plex[grid])); 2273 PetscCall(PetscFree(ctx)); 2274 PetscCall(DMDestroy(dm)); 2275 PetscFunctionReturn(PETSC_SUCCESS); 2276 } 2277 2278 /* < v, ru > */ 2279 static void f0_s_den(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar *f0) 2280 { 2281 PetscInt ii = (PetscInt)PetscRealPart(constants[0]); 2282 f0[0] = u[ii]; 2283 } 2284 2285 /* < v, ru > */ 2286 static void f0_s_mom(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar *f0) 2287 { 2288 PetscInt ii = (PetscInt)PetscRealPart(constants[0]), jj = (PetscInt)PetscRealPart(constants[1]); 2289 f0[0] = x[jj] * u[ii]; /* x momentum */ 2290 } 2291 2292 static void f0_s_v2(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar *f0) 2293 { 2294 PetscInt i, ii = (PetscInt)PetscRealPart(constants[0]); 2295 double tmp1 = 0.; 2296 for (i = 0; i < dim; ++i) tmp1 += x[i] * x[i]; 2297 f0[0] = tmp1 * u[ii]; 2298 } 2299 2300 static PetscErrorCode gamma_n_f(PetscInt dim, PetscReal time, const PetscReal x[], PetscInt Nf, PetscScalar *u, void *actx) 2301 { 2302 const PetscReal *c2_0_arr = ((PetscReal *)actx); 2303 const PetscReal c02 = c2_0_arr[0]; 2304 2305 PetscFunctionBegin; 2306 for (PetscInt s = 0; s < Nf; s++) { 2307 PetscReal tmp1 = 0.; 2308 for (PetscInt i = 0; i < dim; ++i) tmp1 += x[i] * x[i]; 2309 #if defined(PETSC_USE_DEBUG) 2310 u[s] = PetscSqrtReal(1. + tmp1 / c02); // u[0] = PetscSqrtReal(1. + xx); 2311 #else 2312 { 2313 PetscReal xx = tmp1 / c02; 2314 u[s] = xx / (PetscSqrtReal(1. + xx) + 1.); // better conditioned = xx/(PetscSqrtReal(1. + xx) + 1.) 2315 } 2316 #endif 2317 } 2318 PetscFunctionReturn(PETSC_SUCCESS); 2319 } 2320 2321 /* < v, ru > */ 2322 static void f0_s_rden(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar *f0) 2323 { 2324 PetscInt ii = (PetscInt)PetscRealPart(constants[0]); 2325 f0[0] = 2. * PETSC_PI * x[0] * u[ii]; 2326 } 2327 2328 /* < v, ru > */ 2329 static void f0_s_rmom(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar *f0) 2330 { 2331 PetscInt ii = (PetscInt)PetscRealPart(constants[0]); 2332 f0[0] = 2. * PETSC_PI * x[0] * x[1] * u[ii]; 2333 } 2334 2335 static void f0_s_rv2(PetscInt dim, PetscInt Nf, PetscInt NfAux, const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[], const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[], PetscReal t, const PetscReal x[], PetscInt numConstants, const PetscScalar constants[], PetscScalar *f0) 2336 { 2337 PetscInt ii = (PetscInt)PetscRealPart(constants[0]); 2338 f0[0] = 2. * PETSC_PI * x[0] * (x[0] * x[0] + x[1] * x[1]) * u[ii]; 2339 } 2340 2341 /*@ 2342 DMPlexLandauPrintNorms - collects moments and prints them 2343 2344 Collective 2345 2346 Input Parameters: 2347 + X - the state 2348 - stepi - current step to print 2349 2350 Level: beginner 2351 2352 .seealso: `DMPlexLandauCreateVelocitySpace()` 2353 @*/ 2354 PetscErrorCode DMPlexLandauPrintNorms(Vec X, PetscInt stepi) 2355 { 2356 LandauCtx *ctx; 2357 PetscDS prob; 2358 DM pack; 2359 PetscInt cStart, cEnd, dim, ii, i0, nDMs; 2360 PetscScalar xmomentumtot = 0, ymomentumtot = 0, zmomentumtot = 0, energytot = 0, densitytot = 0, tt[LANDAU_MAX_SPECIES]; 2361 PetscScalar xmomentum[LANDAU_MAX_SPECIES], ymomentum[LANDAU_MAX_SPECIES], zmomentum[LANDAU_MAX_SPECIES], energy[LANDAU_MAX_SPECIES], density[LANDAU_MAX_SPECIES]; 2362 Vec *globXArray; 2363 2364 PetscFunctionBegin; 2365 PetscCall(VecGetDM(X, &pack)); 2366 PetscCheck(pack, PETSC_COMM_SELF, PETSC_ERR_PLIB, "Vector has no DM"); 2367 PetscCall(DMGetDimension(pack, &dim)); 2368 PetscCheck(dim == 2 || dim == 3, PETSC_COMM_SELF, PETSC_ERR_PLIB, "dim %" PetscInt_FMT " not in [2,3]", dim); 2369 PetscCall(DMGetApplicationContext(pack, &ctx)); 2370 PetscCheck(ctx, PETSC_COMM_SELF, PETSC_ERR_PLIB, "no context"); 2371 /* print momentum and energy */ 2372 PetscCall(DMCompositeGetNumberDM(pack, &nDMs)); 2373 PetscCheck(nDMs == ctx->num_grids * ctx->batch_sz, PETSC_COMM_WORLD, PETSC_ERR_PLIB, "#DM wrong %" PetscInt_FMT " %" PetscInt_FMT, nDMs, ctx->num_grids * ctx->batch_sz); 2374 PetscCall(PetscMalloc(sizeof(*globXArray) * nDMs, &globXArray)); 2375 PetscCall(DMCompositeGetAccessArray(pack, X, nDMs, NULL, globXArray)); 2376 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 2377 Vec Xloc = globXArray[LAND_PACK_IDX(ctx->batch_view_idx, grid)]; 2378 PetscCall(DMGetDS(ctx->plex[grid], &prob)); 2379 for (ii = ctx->species_offset[grid], i0 = 0; ii < ctx->species_offset[grid + 1]; ii++, i0++) { 2380 PetscScalar user[2] = {(PetscScalar)i0, ctx->charges[ii]}; 2381 PetscCall(PetscDSSetConstants(prob, 2, user)); 2382 if (dim == 2) { /* 2/3X + 3V (cylindrical coordinates) */ 2383 PetscCall(PetscDSSetObjective(prob, 0, &f0_s_rden)); 2384 PetscCall(DMPlexComputeIntegralFEM(ctx->plex[grid], Xloc, tt, ctx)); 2385 density[ii] = tt[0] * ctx->n_0 * ctx->charges[ii]; 2386 PetscCall(PetscDSSetObjective(prob, 0, &f0_s_rmom)); 2387 PetscCall(DMPlexComputeIntegralFEM(ctx->plex[grid], Xloc, tt, ctx)); 2388 zmomentum[ii] = tt[0] * ctx->n_0 * ctx->v_0 * ctx->masses[ii]; 2389 PetscCall(PetscDSSetObjective(prob, 0, &f0_s_rv2)); 2390 PetscCall(DMPlexComputeIntegralFEM(ctx->plex[grid], Xloc, tt, ctx)); 2391 energy[ii] = tt[0] * 0.5 * ctx->n_0 * ctx->v_0 * ctx->v_0 * ctx->masses[ii]; 2392 zmomentumtot += zmomentum[ii]; 2393 energytot += energy[ii]; 2394 densitytot += density[ii]; 2395 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "%3" PetscInt_FMT ") species-%" PetscInt_FMT ": charge density= %20.13e z-momentum= %20.13e energy= %20.13e", stepi, ii, (double)PetscRealPart(density[ii]), (double)PetscRealPart(zmomentum[ii]), (double)PetscRealPart(energy[ii]))); 2396 } else { /* 2/3Xloc + 3V */ 2397 PetscCall(PetscDSSetObjective(prob, 0, &f0_s_den)); 2398 PetscCall(DMPlexComputeIntegralFEM(ctx->plex[grid], Xloc, tt, ctx)); 2399 density[ii] = tt[0] * ctx->n_0 * ctx->charges[ii]; 2400 PetscCall(PetscDSSetObjective(prob, 0, &f0_s_mom)); 2401 user[1] = 0; 2402 PetscCall(PetscDSSetConstants(prob, 2, user)); 2403 PetscCall(DMPlexComputeIntegralFEM(ctx->plex[grid], Xloc, tt, ctx)); 2404 xmomentum[ii] = tt[0] * ctx->n_0 * ctx->v_0 * ctx->masses[ii]; 2405 user[1] = 1; 2406 PetscCall(PetscDSSetConstants(prob, 2, user)); 2407 PetscCall(DMPlexComputeIntegralFEM(ctx->plex[grid], Xloc, tt, ctx)); 2408 ymomentum[ii] = tt[0] * ctx->n_0 * ctx->v_0 * ctx->masses[ii]; 2409 user[1] = 2; 2410 PetscCall(PetscDSSetConstants(prob, 2, user)); 2411 PetscCall(DMPlexComputeIntegralFEM(ctx->plex[grid], Xloc, tt, ctx)); 2412 zmomentum[ii] = tt[0] * ctx->n_0 * ctx->v_0 * ctx->masses[ii]; 2413 if (ctx->use_relativistic_corrections) { 2414 /* gamma * M * f */ 2415 if (ii == 0 && grid == 0) { // do all at once 2416 Vec Mf, globGamma, *globMfArray, *globGammaArray; 2417 PetscErrorCode (*gammaf[1])(PetscInt, PetscReal, const PetscReal[], PetscInt, PetscScalar[], void *) = {gamma_n_f}; 2418 PetscReal *c2_0[1], data[1]; 2419 2420 PetscCall(VecDuplicate(X, &globGamma)); 2421 PetscCall(VecDuplicate(X, &Mf)); 2422 PetscCall(PetscMalloc(sizeof(*globMfArray) * nDMs, &globMfArray)); 2423 PetscCall(PetscMalloc(sizeof(*globMfArray) * nDMs, &globGammaArray)); 2424 /* M * f */ 2425 PetscCall(MatMult(ctx->M, X, Mf)); 2426 /* gamma */ 2427 PetscCall(DMCompositeGetAccessArray(pack, globGamma, nDMs, NULL, globGammaArray)); 2428 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { // yes a grid loop in a grid loop to print nice, need to fix for batching 2429 Vec v1 = globGammaArray[LAND_PACK_IDX(ctx->batch_view_idx, grid)]; 2430 data[0] = PetscSqr(C_0(ctx->v_0)); 2431 c2_0[0] = &data[0]; 2432 PetscCall(DMProjectFunction(ctx->plex[grid], 0., gammaf, (void **)c2_0, INSERT_ALL_VALUES, v1)); 2433 } 2434 PetscCall(DMCompositeRestoreAccessArray(pack, globGamma, nDMs, NULL, globGammaArray)); 2435 /* gamma * Mf */ 2436 PetscCall(DMCompositeGetAccessArray(pack, globGamma, nDMs, NULL, globGammaArray)); 2437 PetscCall(DMCompositeGetAccessArray(pack, Mf, nDMs, NULL, globMfArray)); 2438 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { // yes a grid loop in a grid loop to print nice 2439 PetscInt Nf = ctx->species_offset[grid + 1] - ctx->species_offset[grid], N, bs; 2440 Vec Mfsub = globMfArray[LAND_PACK_IDX(ctx->batch_view_idx, grid)], Gsub = globGammaArray[LAND_PACK_IDX(ctx->batch_view_idx, grid)], v1, v2; 2441 // get each component 2442 PetscCall(VecGetSize(Mfsub, &N)); 2443 PetscCall(VecCreate(ctx->comm, &v1)); 2444 PetscCall(VecSetSizes(v1, PETSC_DECIDE, N / Nf)); 2445 PetscCall(VecCreate(ctx->comm, &v2)); 2446 PetscCall(VecSetSizes(v2, PETSC_DECIDE, N / Nf)); 2447 PetscCall(VecSetFromOptions(v1)); // ??? 2448 PetscCall(VecSetFromOptions(v2)); 2449 // get each component 2450 PetscCall(VecGetBlockSize(Gsub, &bs)); 2451 PetscCheck(bs == Nf, PETSC_COMM_SELF, PETSC_ERR_PLIB, "bs %" PetscInt_FMT " != num_species %" PetscInt_FMT " in Gsub", bs, Nf); 2452 PetscCall(VecGetBlockSize(Mfsub, &bs)); 2453 PetscCheck(bs == Nf, PETSC_COMM_SELF, PETSC_ERR_PLIB, "bs %" PetscInt_FMT " != num_species %" PetscInt_FMT, bs, Nf); 2454 for (PetscInt i = 0, ix = ctx->species_offset[grid]; i < Nf; i++, ix++) { 2455 PetscScalar val; 2456 PetscCall(VecStrideGather(Gsub, i, v1, INSERT_VALUES)); // this is not right -- TODO 2457 PetscCall(VecStrideGather(Mfsub, i, v2, INSERT_VALUES)); 2458 PetscCall(VecDot(v1, v2, &val)); 2459 energy[ix] = PetscRealPart(val) * ctx->n_0 * ctx->v_0 * ctx->v_0 * ctx->masses[ix]; 2460 } 2461 PetscCall(VecDestroy(&v1)); 2462 PetscCall(VecDestroy(&v2)); 2463 } /* grids */ 2464 PetscCall(DMCompositeRestoreAccessArray(pack, globGamma, nDMs, NULL, globGammaArray)); 2465 PetscCall(DMCompositeRestoreAccessArray(pack, Mf, nDMs, NULL, globMfArray)); 2466 PetscCall(PetscFree(globGammaArray)); 2467 PetscCall(PetscFree(globMfArray)); 2468 PetscCall(VecDestroy(&globGamma)); 2469 PetscCall(VecDestroy(&Mf)); 2470 } 2471 } else { 2472 PetscCall(PetscDSSetObjective(prob, 0, &f0_s_v2)); 2473 PetscCall(DMPlexComputeIntegralFEM(ctx->plex[grid], Xloc, tt, ctx)); 2474 energy[ii] = 0.5 * tt[0] * ctx->n_0 * ctx->v_0 * ctx->v_0 * ctx->masses[ii]; 2475 } 2476 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "%3" PetscInt_FMT ") species %" PetscInt_FMT ": density=%20.13e, x-momentum=%20.13e, y-momentum=%20.13e, z-momentum=%20.13e, energy=%21.13e", stepi, ii, (double)PetscRealPart(density[ii]), (double)PetscRealPart(xmomentum[ii]), (double)PetscRealPart(ymomentum[ii]), (double)PetscRealPart(zmomentum[ii]), (double)PetscRealPart(energy[ii]))); 2477 xmomentumtot += xmomentum[ii]; 2478 ymomentumtot += ymomentum[ii]; 2479 zmomentumtot += zmomentum[ii]; 2480 energytot += energy[ii]; 2481 densitytot += density[ii]; 2482 } 2483 if (ctx->num_species > 1) PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\n")); 2484 } 2485 } 2486 PetscCall(DMCompositeRestoreAccessArray(pack, X, nDMs, NULL, globXArray)); 2487 PetscCall(PetscFree(globXArray)); 2488 /* totals */ 2489 PetscCall(DMPlexGetHeightStratum(ctx->plex[0], 0, &cStart, &cEnd)); 2490 if (ctx->num_species > 1) { 2491 if (dim == 2) { 2492 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\t%3" PetscInt_FMT ") Total: charge density=%21.13e, momentum=%21.13e, energy=%21.13e (m_i[0]/m_e = %g, %" PetscInt_FMT " cells on electron grid)", stepi, (double)PetscRealPart(densitytot), (double)PetscRealPart(zmomentumtot), (double)PetscRealPart(energytot), 2493 (double)(ctx->masses[1] / ctx->masses[0]), cEnd - cStart)); 2494 } else { 2495 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\t%3" PetscInt_FMT ") Total: charge density=%21.13e, x-momentum=%21.13e, y-momentum=%21.13e, z-momentum=%21.13e, energy=%21.13e (m_i[0]/m_e = %g, %" PetscInt_FMT " cells)", stepi, (double)PetscRealPart(densitytot), (double)PetscRealPart(xmomentumtot), (double)PetscRealPart(ymomentumtot), (double)PetscRealPart(zmomentumtot), (double)PetscRealPart(energytot), 2496 (double)(ctx->masses[1] / ctx->masses[0]), cEnd - cStart)); 2497 } 2498 } else PetscCall(PetscPrintf(PETSC_COMM_WORLD, " -- %" PetscInt_FMT " cells", cEnd - cStart)); 2499 PetscCall(PetscPrintf(PETSC_COMM_WORLD, "\n")); 2500 PetscFunctionReturn(PETSC_SUCCESS); 2501 } 2502 2503 /*@ 2504 DMPlexLandauCreateMassMatrix - Create mass matrix for Landau in Plex space (not field major order of Jacobian) 2505 - puts mass matrix into ctx->M 2506 2507 Collective 2508 2509 Input Parameter: 2510 . pack - the `DM` object. Puts matrix in Landau context M field 2511 2512 Output Parameter: 2513 . Amat - The mass matrix (optional), mass matrix is added to the `DM` context 2514 2515 Level: beginner 2516 2517 .seealso: `DMPlexLandauCreateVelocitySpace()` 2518 @*/ 2519 PetscErrorCode DMPlexLandauCreateMassMatrix(DM pack, Mat *Amat) 2520 { 2521 DM mass_pack, massDM[LANDAU_MAX_GRIDS]; 2522 PetscDS prob; 2523 PetscInt ii, dim, N1 = 1, N2; 2524 LandauCtx *ctx; 2525 Mat packM, subM[LANDAU_MAX_GRIDS]; 2526 2527 PetscFunctionBegin; 2528 PetscValidHeaderSpecific(pack, DM_CLASSID, 1); 2529 if (Amat) PetscAssertPointer(Amat, 2); 2530 PetscCall(DMGetApplicationContext(pack, &ctx)); 2531 PetscCheck(ctx, PETSC_COMM_SELF, PETSC_ERR_PLIB, "no context"); 2532 PetscCall(PetscLogEventBegin(ctx->events[14], 0, 0, 0, 0)); 2533 PetscCall(DMGetDimension(pack, &dim)); 2534 PetscCall(DMCompositeCreate(PetscObjectComm((PetscObject)pack), &mass_pack)); 2535 /* create pack mass matrix */ 2536 for (PetscInt grid = 0, ix = 0; grid < ctx->num_grids; grid++) { 2537 PetscCall(DMClone(ctx->plex[grid], &massDM[grid])); 2538 PetscCall(DMCopyFields(ctx->plex[grid], PETSC_DETERMINE, PETSC_DETERMINE, massDM[grid])); 2539 PetscCall(DMCreateDS(massDM[grid])); 2540 PetscCall(DMGetDS(massDM[grid], &prob)); 2541 for (ix = 0, ii = ctx->species_offset[grid]; ii < ctx->species_offset[grid + 1]; ii++, ix++) { 2542 if (dim == 3) PetscCall(PetscDSSetJacobian(prob, ix, ix, g0_1, NULL, NULL, NULL)); 2543 else PetscCall(PetscDSSetJacobian(prob, ix, ix, g0_r, NULL, NULL, NULL)); 2544 } 2545 #if !defined(LANDAU_SPECIES_MAJOR) 2546 PetscCall(DMCompositeAddDM(mass_pack, massDM[grid])); 2547 #else 2548 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { // add batch size DMs for this species grid 2549 PetscCall(DMCompositeAddDM(mass_pack, massDM[grid])); 2550 } 2551 #endif 2552 PetscCall(DMCreateMatrix(massDM[grid], &subM[grid])); 2553 } 2554 #if !defined(LANDAU_SPECIES_MAJOR) 2555 // stack the batched DMs 2556 for (PetscInt b_id = 1; b_id < ctx->batch_sz; b_id++) { 2557 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) PetscCall(DMCompositeAddDM(mass_pack, massDM[grid])); 2558 } 2559 #endif 2560 PetscCall(PetscOptionsInsertString(NULL, "-dm_preallocate_only")); 2561 PetscCall(DMCreateMatrix(mass_pack, &packM)); 2562 PetscCall(PetscOptionsInsertString(NULL, "-dm_preallocate_only false")); 2563 PetscCall(MatSetOption(packM, MAT_STRUCTURALLY_SYMMETRIC, PETSC_TRUE)); 2564 PetscCall(MatSetOption(packM, MAT_IGNORE_ZERO_ENTRIES, PETSC_TRUE)); 2565 PetscCall(DMDestroy(&mass_pack)); 2566 /* make mass matrix for each block */ 2567 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 2568 Vec locX; 2569 DM plex = massDM[grid]; 2570 PetscCall(DMGetLocalVector(plex, &locX)); 2571 /* Mass matrix is independent of the input, so no need to fill locX */ 2572 PetscCall(DMPlexSNESComputeJacobianFEM(plex, locX, subM[grid], subM[grid], ctx)); 2573 PetscCall(DMRestoreLocalVector(plex, &locX)); 2574 PetscCall(DMDestroy(&massDM[grid])); 2575 } 2576 PetscCall(MatGetSize(ctx->J, &N1, NULL)); 2577 PetscCall(MatGetSize(packM, &N2, NULL)); 2578 PetscCheck(N1 == N2, PetscObjectComm((PetscObject)pack), PETSC_ERR_PLIB, "Incorrect matrix sizes: |Jacobian| = %" PetscInt_FMT ", |Mass|=%" PetscInt_FMT, N1, N2); 2579 /* assemble block diagonals */ 2580 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) { 2581 Mat B = subM[grid]; 2582 PetscInt nloc, nzl, *colbuf, COL_BF_SIZE = 1024, row; 2583 PetscCall(PetscMalloc(sizeof(*colbuf) * COL_BF_SIZE, &colbuf)); 2584 PetscCall(MatGetSize(B, &nloc, NULL)); 2585 for (PetscInt b_id = 0; b_id < ctx->batch_sz; b_id++) { 2586 const PetscInt moffset = LAND_MOFFSET(b_id, grid, ctx->batch_sz, ctx->num_grids, ctx->mat_offset); 2587 const PetscInt *cols; 2588 const PetscScalar *vals; 2589 for (PetscInt i = 0; i < nloc; i++) { 2590 PetscCall(MatGetRow(B, i, &nzl, NULL, NULL)); 2591 if (nzl > COL_BF_SIZE) { 2592 PetscCall(PetscFree(colbuf)); 2593 PetscCall(PetscInfo(pack, "Realloc buffer %" PetscInt_FMT " to %" PetscInt_FMT " (row size %" PetscInt_FMT ") \n", COL_BF_SIZE, 2 * COL_BF_SIZE, nzl)); 2594 COL_BF_SIZE = nzl; 2595 PetscCall(PetscMalloc(sizeof(*colbuf) * COL_BF_SIZE, &colbuf)); 2596 } 2597 PetscCall(MatGetRow(B, i, &nzl, &cols, &vals)); 2598 for (PetscInt j = 0; j < nzl; j++) colbuf[j] = cols[j] + moffset; 2599 row = i + moffset; 2600 PetscCall(MatSetValues(packM, 1, &row, nzl, colbuf, vals, INSERT_VALUES)); 2601 PetscCall(MatRestoreRow(B, i, &nzl, &cols, &vals)); 2602 } 2603 } 2604 PetscCall(PetscFree(colbuf)); 2605 } 2606 // cleanup 2607 for (PetscInt grid = 0; grid < ctx->num_grids; grid++) PetscCall(MatDestroy(&subM[grid])); 2608 PetscCall(MatAssemblyBegin(packM, MAT_FINAL_ASSEMBLY)); 2609 PetscCall(MatAssemblyEnd(packM, MAT_FINAL_ASSEMBLY)); 2610 PetscCall(PetscObjectSetName((PetscObject)packM, "mass")); 2611 PetscCall(MatViewFromOptions(packM, NULL, "-dm_landau_mass_view")); 2612 ctx->M = packM; 2613 if (Amat) *Amat = packM; 2614 PetscCall(PetscLogEventEnd(ctx->events[14], 0, 0, 0, 0)); 2615 PetscFunctionReturn(PETSC_SUCCESS); 2616 } 2617 2618 /*@ 2619 DMPlexLandauIFunction - `TS` residual calculation, confusingly this computes the Jacobian w/o mass 2620 2621 Collective 2622 2623 Input Parameters: 2624 + ts - The time stepping context 2625 . time_dummy - current time (not used) 2626 . X - Current state 2627 . X_t - Time derivative of current state 2628 - actx - Landau context 2629 2630 Output Parameter: 2631 . F - The residual 2632 2633 Level: beginner 2634 2635 .seealso: `DMPlexLandauCreateVelocitySpace()`, `DMPlexLandauIJacobian()` 2636 @*/ 2637 PetscErrorCode DMPlexLandauIFunction(TS ts, PetscReal time_dummy, Vec X, Vec X_t, Vec F, void *actx) 2638 { 2639 LandauCtx *ctx = (LandauCtx *)actx; 2640 PetscInt dim; 2641 DM pack; 2642 #if defined(PETSC_HAVE_THREADSAFETY) 2643 double starttime, endtime; 2644 #endif 2645 PetscObjectState state; 2646 2647 PetscFunctionBegin; 2648 PetscCall(TSGetDM(ts, &pack)); 2649 PetscCall(DMGetApplicationContext(pack, &ctx)); 2650 PetscCheck(ctx, PETSC_COMM_SELF, PETSC_ERR_PLIB, "no context"); 2651 if (ctx->stage) PetscCall(PetscLogStagePush(ctx->stage)); 2652 PetscCall(PetscLogEventBegin(ctx->events[11], 0, 0, 0, 0)); 2653 PetscCall(PetscLogEventBegin(ctx->events[0], 0, 0, 0, 0)); 2654 #if defined(PETSC_HAVE_THREADSAFETY) 2655 starttime = MPI_Wtime(); 2656 #endif 2657 PetscCall(DMGetDimension(pack, &dim)); 2658 PetscCall(PetscObjectStateGet((PetscObject)ctx->J, &state)); 2659 if (state != ctx->norm_state) { 2660 PetscCall(MatZeroEntries(ctx->J)); 2661 PetscCall(LandauFormJacobian_Internal(X, ctx->J, dim, 0.0, (void *)ctx)); 2662 PetscCall(MatViewFromOptions(ctx->J, NULL, "-dm_landau_jacobian_view")); 2663 PetscCall(PetscObjectStateGet((PetscObject)ctx->J, &state)); 2664 ctx->norm_state = state; 2665 } else { 2666 PetscCall(PetscInfo(ts, "WARNING Skip forming Jacobian, has not changed %" PetscInt64_FMT "\n", state)); 2667 } 2668 /* mat vec for op */ 2669 PetscCall(MatMult(ctx->J, X, F)); /* C*f */ 2670 /* add time term */ 2671 if (X_t) PetscCall(MatMultAdd(ctx->M, X_t, F, F)); 2672 #if defined(PETSC_HAVE_THREADSAFETY) 2673 if (ctx->stage) { 2674 endtime = MPI_Wtime(); 2675 ctx->times[LANDAU_OPERATOR] += (endtime - starttime); 2676 ctx->times[LANDAU_JACOBIAN] += (endtime - starttime); 2677 ctx->times[LANDAU_MATRIX_TOTAL] += (endtime - starttime); 2678 ctx->times[LANDAU_JACOBIAN_COUNT] += 1; 2679 } 2680 #endif 2681 PetscCall(PetscLogEventEnd(ctx->events[0], 0, 0, 0, 0)); 2682 PetscCall(PetscLogEventEnd(ctx->events[11], 0, 0, 0, 0)); 2683 if (ctx->stage) PetscCall(PetscLogStagePop()); 2684 PetscFunctionReturn(PETSC_SUCCESS); 2685 } 2686 2687 /*@ 2688 DMPlexLandauIJacobian - `TS` Jacobian construction, confusingly this adds mass 2689 2690 Collective 2691 2692 Input Parameters: 2693 + ts - The time stepping context 2694 . time_dummy - current time (not used) 2695 . X - Current state 2696 . U_tdummy - Time derivative of current state (not used) 2697 . shift - shift for du/dt term 2698 - actx - Landau context 2699 2700 Output Parameters: 2701 + Amat - Jacobian 2702 - Pmat - same as Amat 2703 2704 Level: beginner 2705 2706 .seealso: `DMPlexLandauCreateVelocitySpace()`, `DMPlexLandauIFunction()` 2707 @*/ 2708 PetscErrorCode DMPlexLandauIJacobian(TS ts, PetscReal time_dummy, Vec X, Vec U_tdummy, PetscReal shift, Mat Amat, Mat Pmat, void *actx) 2709 { 2710 LandauCtx *ctx = NULL; 2711 PetscInt dim; 2712 DM pack; 2713 #if defined(PETSC_HAVE_THREADSAFETY) 2714 double starttime, endtime; 2715 #endif 2716 PetscObjectState state; 2717 2718 PetscFunctionBegin; 2719 PetscCall(TSGetDM(ts, &pack)); 2720 PetscCall(DMGetApplicationContext(pack, &ctx)); 2721 PetscCheck(ctx, PETSC_COMM_SELF, PETSC_ERR_PLIB, "no context"); 2722 PetscCheck(Amat == Pmat && Amat == ctx->J, ctx->comm, PETSC_ERR_PLIB, "Amat!=Pmat || Amat!=ctx->J"); 2723 PetscCall(DMGetDimension(pack, &dim)); 2724 /* get collision Jacobian into A */ 2725 if (ctx->stage) PetscCall(PetscLogStagePush(ctx->stage)); 2726 PetscCall(PetscLogEventBegin(ctx->events[11], 0, 0, 0, 0)); 2727 PetscCall(PetscLogEventBegin(ctx->events[9], 0, 0, 0, 0)); 2728 #if defined(PETSC_HAVE_THREADSAFETY) 2729 starttime = MPI_Wtime(); 2730 #endif 2731 PetscCheck(shift != 0.0, ctx->comm, PETSC_ERR_PLIB, "zero shift"); 2732 PetscCall(PetscObjectStateGet((PetscObject)ctx->J, &state)); 2733 PetscCheck(state == ctx->norm_state, ctx->comm, PETSC_ERR_PLIB, "wrong state, %" PetscInt64_FMT " %" PetscInt64_FMT, ctx->norm_state, state); 2734 if (!ctx->use_matrix_mass) { 2735 PetscCall(LandauFormJacobian_Internal(X, ctx->J, dim, shift, (void *)ctx)); 2736 } else { /* add mass */ 2737 PetscCall(MatAXPY(Pmat, shift, ctx->M, SAME_NONZERO_PATTERN)); 2738 } 2739 #if defined(PETSC_HAVE_THREADSAFETY) 2740 if (ctx->stage) { 2741 endtime = MPI_Wtime(); 2742 ctx->times[LANDAU_OPERATOR] += (endtime - starttime); 2743 ctx->times[LANDAU_MASS] += (endtime - starttime); 2744 ctx->times[LANDAU_MATRIX_TOTAL] += (endtime - starttime); 2745 } 2746 #endif 2747 PetscCall(PetscLogEventEnd(ctx->events[9], 0, 0, 0, 0)); 2748 PetscCall(PetscLogEventEnd(ctx->events[11], 0, 0, 0, 0)); 2749 if (ctx->stage) PetscCall(PetscLogStagePop()); 2750 PetscFunctionReturn(PETSC_SUCCESS); 2751 } 2752