1 #include <../src/mat/impls/sell/mpi/mpisell.h> 2 #include <../src/mat/impls/aij/mpi/mpiaij.h> 3 #include <../src/mat/impls/baij/mpi/mpibaij.h> 4 #include <petsc/private/isimpl.h> 5 6 PetscErrorCode MatFDColoringApply_BAIJ(Mat J,MatFDColoring coloring,Vec x1,void *sctx) 7 { 8 PetscErrorCode (*f)(void*,Vec,Vec,void*)=(PetscErrorCode (*)(void*,Vec,Vec,void*))coloring->f; 9 PetscErrorCode ierr; 10 PetscInt k,cstart,cend,l,row,col,nz,spidx,i,j; 11 PetscScalar dx=0.0,*w3_array,*dy_i,*dy=coloring->dy; 12 PetscScalar *vscale_array; 13 const PetscScalar *xx; 14 PetscReal epsilon=coloring->error_rel,umin=coloring->umin,unorm; 15 Vec w1=coloring->w1,w2=coloring->w2,w3,vscale=coloring->vscale; 16 void *fctx=coloring->fctx; 17 PetscInt ctype=coloring->ctype,nxloc,nrows_k; 18 PetscScalar *valaddr; 19 MatEntry *Jentry=coloring->matentry; 20 MatEntry2 *Jentry2=coloring->matentry2; 21 const PetscInt ncolors=coloring->ncolors,*ncolumns=coloring->ncolumns,*nrows=coloring->nrows; 22 PetscInt bs=J->rmap->bs; 23 24 PetscFunctionBegin; 25 ierr = VecBindToCPU(x1,PETSC_TRUE);CHKERRQ(ierr); 26 /* (1) Set w1 = F(x1) */ 27 if (!coloring->fset) { 28 ierr = PetscLogEventBegin(MAT_FDColoringFunction,coloring,0,0,0);CHKERRQ(ierr); 29 ierr = (*f)(sctx,x1,w1,fctx);CHKERRQ(ierr); 30 ierr = PetscLogEventEnd(MAT_FDColoringFunction,coloring,0,0,0);CHKERRQ(ierr); 31 } else { 32 coloring->fset = PETSC_FALSE; 33 } 34 35 /* (2) Compute vscale = 1./dx - the local scale factors, including ghost points */ 36 ierr = VecGetLocalSize(x1,&nxloc);CHKERRQ(ierr); 37 if (coloring->htype[0] == 'w') { 38 /* vscale = dx is a constant scalar */ 39 ierr = VecNorm(x1,NORM_2,&unorm);CHKERRQ(ierr); 40 dx = 1.0/(PetscSqrtReal(1.0 + unorm)*epsilon); 41 } else { 42 ierr = VecGetArrayRead(x1,&xx);CHKERRQ(ierr); 43 ierr = VecGetArray(vscale,&vscale_array);CHKERRQ(ierr); 44 for (col=0; col<nxloc; col++) { 45 dx = xx[col]; 46 if (PetscAbsScalar(dx) < umin) { 47 if (PetscRealPart(dx) >= 0.0) dx = umin; 48 else if (PetscRealPart(dx) < 0.0) dx = -umin; 49 } 50 dx *= epsilon; 51 vscale_array[col] = 1.0/dx; 52 } 53 ierr = VecRestoreArrayRead(x1,&xx);CHKERRQ(ierr); 54 ierr = VecRestoreArray(vscale,&vscale_array);CHKERRQ(ierr); 55 } 56 if (ctype == IS_COLORING_GLOBAL && coloring->htype[0] == 'd') { 57 ierr = VecGhostUpdateBegin(vscale,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 58 ierr = VecGhostUpdateEnd(vscale,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 59 } 60 61 /* (3) Loop over each color */ 62 if (!coloring->w3) { 63 ierr = VecDuplicate(x1,&coloring->w3);CHKERRQ(ierr); 64 /* Vec is used intensively in particular piece of scalar CPU code; won't benefit from bouncing back and forth to the GPU */ 65 ierr = VecBindToCPU(coloring->w3,PETSC_TRUE);CHKERRQ(ierr); 66 ierr = PetscLogObjectParent((PetscObject)coloring,(PetscObject)coloring->w3);CHKERRQ(ierr); 67 } 68 w3 = coloring->w3; 69 70 ierr = VecGetOwnershipRange(x1,&cstart,&cend);CHKERRQ(ierr); /* used by ghosted vscale */ 71 if (vscale) { 72 ierr = VecGetArray(vscale,&vscale_array);CHKERRQ(ierr); 73 } 74 nz = 0; 75 for (k=0; k<ncolors; k++) { 76 coloring->currentcolor = k; 77 78 /* 79 (3-1) Loop over each column associated with color 80 adding the perturbation to the vector w3 = x1 + dx. 81 */ 82 ierr = VecCopy(x1,w3);CHKERRQ(ierr); 83 dy_i = dy; 84 for (i=0; i<bs; i++) { /* Loop over a block of columns */ 85 ierr = VecGetArray(w3,&w3_array);CHKERRQ(ierr); 86 if (ctype == IS_COLORING_GLOBAL) w3_array -= cstart; /* shift pointer so global index can be used */ 87 if (coloring->htype[0] == 'w') { 88 for (l=0; l<ncolumns[k]; l++) { 89 col = i + bs*coloring->columns[k][l]; /* local column (in global index!) of the matrix we are probing for */ 90 w3_array[col] += 1.0/dx; 91 if (i) w3_array[col-1] -= 1.0/dx; /* resume original w3[col-1] */ 92 } 93 } else { /* htype == 'ds' */ 94 vscale_array -= cstart; /* shift pointer so global index can be used */ 95 for (l=0; l<ncolumns[k]; l++) { 96 col = i + bs*coloring->columns[k][l]; /* local column (in global index!) of the matrix we are probing for */ 97 w3_array[col] += 1.0/vscale_array[col]; 98 if (i) w3_array[col-1] -= 1.0/vscale_array[col-1]; /* resume original w3[col-1] */ 99 } 100 vscale_array += cstart; 101 } 102 if (ctype == IS_COLORING_GLOBAL) w3_array += cstart; 103 ierr = VecRestoreArray(w3,&w3_array);CHKERRQ(ierr); 104 105 /* 106 (3-2) Evaluate function at w3 = x1 + dx (here dx is a vector of perturbations) 107 w2 = F(x1 + dx) - F(x1) 108 */ 109 ierr = PetscLogEventBegin(MAT_FDColoringFunction,0,0,0,0);CHKERRQ(ierr); 110 ierr = VecPlaceArray(w2,dy_i);CHKERRQ(ierr); /* place w2 to the array dy_i */ 111 ierr = (*f)(sctx,w3,w2,fctx);CHKERRQ(ierr); 112 ierr = PetscLogEventEnd(MAT_FDColoringFunction,0,0,0,0);CHKERRQ(ierr); 113 ierr = VecAXPY(w2,-1.0,w1);CHKERRQ(ierr); 114 ierr = VecResetArray(w2);CHKERRQ(ierr); 115 dy_i += nxloc; /* points to dy+i*nxloc */ 116 } 117 118 /* 119 (3-3) Loop over rows of vector, putting results into Jacobian matrix 120 */ 121 nrows_k = nrows[k]; 122 if (coloring->htype[0] == 'w') { 123 for (l=0; l<nrows_k; l++) { 124 row = bs*Jentry2[nz].row; /* local row index */ 125 valaddr = Jentry2[nz++].valaddr; 126 spidx = 0; 127 dy_i = dy; 128 for (i=0; i<bs; i++) { /* column of the block */ 129 for (j=0; j<bs; j++) { /* row of the block */ 130 valaddr[spidx++] = dy_i[row+j]*dx; 131 } 132 dy_i += nxloc; /* points to dy+i*nxloc */ 133 } 134 } 135 } else { /* htype == 'ds' */ 136 for (l=0; l<nrows_k; l++) { 137 row = bs*Jentry[nz].row; /* local row index */ 138 col = bs*Jentry[nz].col; /* local column index */ 139 valaddr = Jentry[nz++].valaddr; 140 spidx = 0; 141 dy_i = dy; 142 for (i=0; i<bs; i++) { /* column of the block */ 143 for (j=0; j<bs; j++) { /* row of the block */ 144 valaddr[spidx++] = dy_i[row+j]*vscale_array[col+i]; 145 } 146 dy_i += nxloc; /* points to dy+i*nxloc */ 147 } 148 } 149 } 150 } 151 ierr = MatAssemblyBegin(J,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 152 ierr = MatAssemblyEnd(J,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 153 if (vscale) { 154 ierr = VecRestoreArray(vscale,&vscale_array);CHKERRQ(ierr); 155 } 156 157 coloring->currentcolor = -1; 158 ierr = VecBindToCPU(x1,PETSC_FALSE);CHKERRQ(ierr); 159 PetscFunctionReturn(0); 160 } 161 162 /* this is declared PETSC_EXTERN because it is used by MatFDColoringUseDM() which is in the DM library */ 163 PetscErrorCode MatFDColoringApply_AIJ(Mat J,MatFDColoring coloring,Vec x1,void *sctx) 164 { 165 PetscErrorCode (*f)(void*,Vec,Vec,void*) = (PetscErrorCode (*)(void*,Vec,Vec,void*))coloring->f; 166 PetscErrorCode ierr; 167 PetscInt k,cstart,cend,l,row,col,nz; 168 PetscScalar dx=0.0,*y,*w3_array; 169 const PetscScalar *xx; 170 PetscScalar *vscale_array; 171 PetscReal epsilon=coloring->error_rel,umin=coloring->umin,unorm; 172 Vec w1=coloring->w1,w2=coloring->w2,w3,vscale=coloring->vscale; 173 void *fctx=coloring->fctx; 174 ISColoringType ctype=coloring->ctype; 175 PetscInt nxloc,nrows_k; 176 MatEntry *Jentry=coloring->matentry; 177 MatEntry2 *Jentry2=coloring->matentry2; 178 const PetscInt ncolors=coloring->ncolors,*ncolumns=coloring->ncolumns,*nrows=coloring->nrows; 179 PetscBool alreadyboundtocpu; 180 181 PetscFunctionBegin; 182 ierr = VecBoundToCPU(x1,&alreadyboundtocpu);CHKERRQ(ierr); 183 ierr = VecBindToCPU(x1,PETSC_TRUE);CHKERRQ(ierr); 184 if ((ctype == IS_COLORING_LOCAL) && (J->ops->fdcoloringapply == MatFDColoringApply_AIJ)) SETERRQ(PetscObjectComm((PetscObject)J),PETSC_ERR_SUP,"Must call MatColoringUseDM() with IS_COLORING_LOCAL"); 185 /* (1) Set w1 = F(x1) */ 186 if (!coloring->fset) { 187 ierr = PetscLogEventBegin(MAT_FDColoringFunction,0,0,0,0);CHKERRQ(ierr); 188 ierr = (*f)(sctx,x1,w1,fctx);CHKERRQ(ierr); 189 ierr = PetscLogEventEnd(MAT_FDColoringFunction,0,0,0,0);CHKERRQ(ierr); 190 } else { 191 coloring->fset = PETSC_FALSE; 192 } 193 194 /* (2) Compute vscale = 1./dx - the local scale factors, including ghost points */ 195 if (coloring->htype[0] == 'w') { 196 /* vscale = 1./dx is a constant scalar */ 197 ierr = VecNorm(x1,NORM_2,&unorm);CHKERRQ(ierr); 198 dx = 1.0/(PetscSqrtReal(1.0 + unorm)*epsilon); 199 } else { 200 ierr = VecGetLocalSize(x1,&nxloc);CHKERRQ(ierr); 201 ierr = VecGetArrayRead(x1,&xx);CHKERRQ(ierr); 202 ierr = VecGetArray(vscale,&vscale_array);CHKERRQ(ierr); 203 for (col=0; col<nxloc; col++) { 204 dx = xx[col]; 205 if (PetscAbsScalar(dx) < umin) { 206 if (PetscRealPart(dx) >= 0.0) dx = umin; 207 else if (PetscRealPart(dx) < 0.0) dx = -umin; 208 } 209 dx *= epsilon; 210 vscale_array[col] = 1.0/dx; 211 } 212 ierr = VecRestoreArrayRead(x1,&xx);CHKERRQ(ierr); 213 ierr = VecRestoreArray(vscale,&vscale_array);CHKERRQ(ierr); 214 } 215 if (ctype == IS_COLORING_GLOBAL && coloring->htype[0] == 'd') { 216 ierr = VecGhostUpdateBegin(vscale,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 217 ierr = VecGhostUpdateEnd(vscale,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 218 } 219 220 /* (3) Loop over each color */ 221 if (!coloring->w3) { 222 ierr = VecDuplicate(x1,&coloring->w3);CHKERRQ(ierr); 223 ierr = PetscLogObjectParent((PetscObject)coloring,(PetscObject)coloring->w3);CHKERRQ(ierr); 224 } 225 w3 = coloring->w3; 226 227 ierr = VecGetOwnershipRange(x1,&cstart,&cend);CHKERRQ(ierr); /* used by ghosted vscale */ 228 if (vscale) { 229 ierr = VecGetArray(vscale,&vscale_array);CHKERRQ(ierr); 230 } 231 nz = 0; 232 233 if (coloring->bcols > 1) { /* use blocked insertion of Jentry */ 234 PetscInt i,m=J->rmap->n,nbcols,bcols=coloring->bcols; 235 PetscScalar *dy=coloring->dy,*dy_k; 236 237 nbcols = 0; 238 for (k=0; k<ncolors; k+=bcols) { 239 240 /* 241 (3-1) Loop over each column associated with color 242 adding the perturbation to the vector w3 = x1 + dx. 243 */ 244 245 dy_k = dy; 246 if (k + bcols > ncolors) bcols = ncolors - k; 247 for (i=0; i<bcols; i++) { 248 coloring->currentcolor = k+i; 249 250 ierr = VecCopy(x1,w3);CHKERRQ(ierr); 251 ierr = VecGetArray(w3,&w3_array);CHKERRQ(ierr); 252 if (ctype == IS_COLORING_GLOBAL) w3_array -= cstart; /* shift pointer so global index can be used */ 253 if (coloring->htype[0] == 'w') { 254 for (l=0; l<ncolumns[k+i]; l++) { 255 col = coloring->columns[k+i][l]; /* local column (in global index!) of the matrix we are probing for */ 256 w3_array[col] += 1.0/dx; 257 } 258 } else { /* htype == 'ds' */ 259 vscale_array -= cstart; /* shift pointer so global index can be used */ 260 for (l=0; l<ncolumns[k+i]; l++) { 261 col = coloring->columns[k+i][l]; /* local column (in global index!) of the matrix we are probing for */ 262 w3_array[col] += 1.0/vscale_array[col]; 263 } 264 vscale_array += cstart; 265 } 266 if (ctype == IS_COLORING_GLOBAL) w3_array += cstart; 267 ierr = VecRestoreArray(w3,&w3_array);CHKERRQ(ierr); 268 269 /* 270 (3-2) Evaluate function at w3 = x1 + dx (here dx is a vector of perturbations) 271 w2 = F(x1 + dx) - F(x1) 272 */ 273 ierr = PetscLogEventBegin(MAT_FDColoringFunction,0,0,0,0);CHKERRQ(ierr); 274 ierr = VecPlaceArray(w2,dy_k);CHKERRQ(ierr); /* place w2 to the array dy_i */ 275 ierr = (*f)(sctx,w3,w2,fctx);CHKERRQ(ierr); 276 ierr = PetscLogEventEnd(MAT_FDColoringFunction,0,0,0,0);CHKERRQ(ierr); 277 ierr = VecAXPY(w2,-1.0,w1);CHKERRQ(ierr); 278 ierr = VecResetArray(w2);CHKERRQ(ierr); 279 dy_k += m; /* points to dy+i*nxloc */ 280 } 281 282 /* 283 (3-3) Loop over block rows of vector, putting results into Jacobian matrix 284 */ 285 nrows_k = nrows[nbcols++]; 286 287 if (coloring->htype[0] == 'w') { 288 for (l=0; l<nrows_k; l++) { 289 row = Jentry2[nz].row; /* local row index */ 290 /* The 'useless' ifdef is due to a bug in NVIDIA nvc 21.11, which triggers a segfault on this line. We write it in 291 another way, and it seems work. See https://lists.mcs.anl.gov/pipermail/petsc-users/2021-December/045158.html 292 */ 293 #if defined(PETSC_USE_COMPLEX) 294 PetscScalar *tmp = Jentry2[nz].valaddr; 295 *tmp = dy[row]*dx; 296 #else 297 *(Jentry2[nz].valaddr) = dy[row]*dx; 298 #endif 299 nz++; 300 } 301 } else { /* htype == 'ds' */ 302 for (l=0; l<nrows_k; l++) { 303 row = Jentry[nz].row; /* local row index */ 304 #if defined(PETSC_USE_COMPLEX) /* See https://lists.mcs.anl.gov/pipermail/petsc-users/2021-December/045158.html */ 305 PetscScalar *tmp = Jentry[nz].valaddr; 306 *tmp = dy[row]*vscale_array[Jentry[nz].col]; 307 #else 308 *(Jentry[nz].valaddr) = dy[row]*vscale_array[Jentry[nz].col]; 309 #endif 310 nz++; 311 } 312 } 313 } 314 } else { /* bcols == 1 */ 315 for (k=0; k<ncolors; k++) { 316 coloring->currentcolor = k; 317 318 /* 319 (3-1) Loop over each column associated with color 320 adding the perturbation to the vector w3 = x1 + dx. 321 */ 322 ierr = VecCopy(x1,w3);CHKERRQ(ierr); 323 ierr = VecGetArray(w3,&w3_array);CHKERRQ(ierr); 324 if (ctype == IS_COLORING_GLOBAL) w3_array -= cstart; /* shift pointer so global index can be used */ 325 if (coloring->htype[0] == 'w') { 326 for (l=0; l<ncolumns[k]; l++) { 327 col = coloring->columns[k][l]; /* local column (in global index!) of the matrix we are probing for */ 328 w3_array[col] += 1.0/dx; 329 } 330 } else { /* htype == 'ds' */ 331 vscale_array -= cstart; /* shift pointer so global index can be used */ 332 for (l=0; l<ncolumns[k]; l++) { 333 col = coloring->columns[k][l]; /* local column (in global index!) of the matrix we are probing for */ 334 w3_array[col] += 1.0/vscale_array[col]; 335 } 336 vscale_array += cstart; 337 } 338 if (ctype == IS_COLORING_GLOBAL) w3_array += cstart; 339 ierr = VecRestoreArray(w3,&w3_array);CHKERRQ(ierr); 340 341 /* 342 (3-2) Evaluate function at w3 = x1 + dx (here dx is a vector of perturbations) 343 w2 = F(x1 + dx) - F(x1) 344 */ 345 ierr = PetscLogEventBegin(MAT_FDColoringFunction,0,0,0,0);CHKERRQ(ierr); 346 ierr = (*f)(sctx,w3,w2,fctx);CHKERRQ(ierr); 347 ierr = PetscLogEventEnd(MAT_FDColoringFunction,0,0,0,0);CHKERRQ(ierr); 348 ierr = VecAXPY(w2,-1.0,w1);CHKERRQ(ierr); 349 350 /* 351 (3-3) Loop over rows of vector, putting results into Jacobian matrix 352 */ 353 nrows_k = nrows[k]; 354 ierr = VecGetArray(w2,&y);CHKERRQ(ierr); 355 if (coloring->htype[0] == 'w') { 356 for (l=0; l<nrows_k; l++) { 357 row = Jentry2[nz].row; /* local row index */ 358 #if defined(PETSC_USE_COMPLEX) /* See https://lists.mcs.anl.gov/pipermail/petsc-users/2021-December/045158.html */ 359 PetscScalar *tmp = Jentry2[nz].valaddr; 360 *tmp = y[row]*dx; 361 #else 362 *(Jentry2[nz].valaddr) = y[row]*dx; 363 #endif 364 nz++; 365 } 366 } else { /* htype == 'ds' */ 367 for (l=0; l<nrows_k; l++) { 368 row = Jentry[nz].row; /* local row index */ 369 #if defined(PETSC_USE_COMPLEX) /* See https://lists.mcs.anl.gov/pipermail/petsc-users/2021-December/045158.html */ 370 PetscScalar *tmp = Jentry[nz].valaddr; 371 *tmp = y[row]*vscale_array[Jentry[nz].col]; 372 #else 373 *(Jentry[nz].valaddr) = y[row]*vscale_array[Jentry[nz].col]; 374 #endif 375 nz++; 376 } 377 } 378 ierr = VecRestoreArray(w2,&y);CHKERRQ(ierr); 379 } 380 } 381 382 #if defined(PETSC_HAVE_VIENNACL) || defined(PETSC_HAVE_CUDA) 383 if (J->offloadmask != PETSC_OFFLOAD_UNALLOCATED) J->offloadmask = PETSC_OFFLOAD_CPU; 384 #endif 385 ierr = MatAssemblyBegin(J,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 386 ierr = MatAssemblyEnd(J,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 387 if (vscale) { 388 ierr = VecRestoreArray(vscale,&vscale_array);CHKERRQ(ierr); 389 } 390 coloring->currentcolor = -1; 391 if (!alreadyboundtocpu) {ierr = VecBindToCPU(x1,PETSC_FALSE);CHKERRQ(ierr);} 392 PetscFunctionReturn(0); 393 } 394 395 PetscErrorCode MatFDColoringSetUp_MPIXAIJ(Mat mat,ISColoring iscoloring,MatFDColoring c) 396 { 397 PetscErrorCode ierr; 398 PetscMPIInt size,*ncolsonproc,*disp,nn; 399 PetscInt i,n,nrows,nrows_i,j,k,m,ncols,col,*rowhit,cstart,cend,colb; 400 const PetscInt *is,*A_ci,*A_cj,*B_ci,*B_cj,*row=NULL,*ltog=NULL; 401 PetscInt nis=iscoloring->n,nctot,*cols,tmp = 0; 402 ISLocalToGlobalMapping map=mat->cmap->mapping; 403 PetscInt ctype=c->ctype,*spidxA,*spidxB,nz,bs,bs2,spidx; 404 Mat A,B; 405 PetscScalar *A_val,*B_val,**valaddrhit; 406 MatEntry *Jentry; 407 MatEntry2 *Jentry2; 408 PetscBool isBAIJ,isSELL; 409 PetscInt bcols=c->bcols; 410 #if defined(PETSC_USE_CTABLE) 411 PetscTable colmap=NULL; 412 #else 413 PetscInt *colmap=NULL; /* local col number of off-diag col */ 414 #endif 415 416 PetscFunctionBegin; 417 if (ctype == IS_COLORING_LOCAL) { 418 if (!map) SETERRQ(PetscObjectComm((PetscObject)mat),PETSC_ERR_ARG_INCOMP,"When using ghosted differencing matrix must have local to global mapping provided with MatSetLocalToGlobalMapping"); 419 ierr = ISLocalToGlobalMappingGetIndices(map,<og);CHKERRQ(ierr); 420 } 421 422 ierr = MatGetBlockSize(mat,&bs);CHKERRQ(ierr); 423 ierr = PetscObjectBaseTypeCompare((PetscObject)mat,MATMPIBAIJ,&isBAIJ);CHKERRQ(ierr); 424 ierr = PetscObjectTypeCompare((PetscObject)mat,MATMPISELL,&isSELL);CHKERRQ(ierr); 425 if (isBAIJ) { 426 Mat_MPIBAIJ *baij=(Mat_MPIBAIJ*)mat->data; 427 Mat_SeqBAIJ *spA,*spB; 428 A = baij->A; spA = (Mat_SeqBAIJ*)A->data; A_val = spA->a; 429 B = baij->B; spB = (Mat_SeqBAIJ*)B->data; B_val = spB->a; 430 nz = spA->nz + spB->nz; /* total nonzero entries of mat */ 431 if (!baij->colmap) { 432 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 433 } 434 colmap = baij->colmap; 435 ierr = MatGetColumnIJ_SeqBAIJ_Color(A,0,PETSC_FALSE,PETSC_FALSE,&ncols,&A_ci,&A_cj,&spidxA,NULL);CHKERRQ(ierr); 436 ierr = MatGetColumnIJ_SeqBAIJ_Color(B,0,PETSC_FALSE,PETSC_FALSE,&ncols,&B_ci,&B_cj,&spidxB,NULL);CHKERRQ(ierr); 437 438 if (ctype == IS_COLORING_GLOBAL && c->htype[0] == 'd') { /* create vscale for storing dx */ 439 PetscInt *garray; 440 ierr = PetscMalloc1(B->cmap->n,&garray);CHKERRQ(ierr); 441 for (i=0; i<baij->B->cmap->n/bs; i++) { 442 for (j=0; j<bs; j++) { 443 garray[i*bs+j] = bs*baij->garray[i]+j; 444 } 445 } 446 ierr = VecCreateGhost(PetscObjectComm((PetscObject)mat),mat->cmap->n,PETSC_DETERMINE,B->cmap->n,garray,&c->vscale);CHKERRQ(ierr); 447 ierr = VecBindToCPU(c->vscale,PETSC_TRUE);CHKERRQ(ierr); 448 ierr = PetscFree(garray);CHKERRQ(ierr); 449 } 450 } else if (isSELL) { 451 Mat_MPISELL *sell=(Mat_MPISELL*)mat->data; 452 Mat_SeqSELL *spA,*spB; 453 A = sell->A; spA = (Mat_SeqSELL*)A->data; A_val = spA->val; 454 B = sell->B; spB = (Mat_SeqSELL*)B->data; B_val = spB->val; 455 nz = spA->nz + spB->nz; /* total nonzero entries of mat */ 456 if (!sell->colmap) { 457 /* Allow access to data structures of local part of matrix 458 - creates aij->colmap which maps global column number to local number in part B */ 459 ierr = MatCreateColmap_MPISELL_Private(mat);CHKERRQ(ierr); 460 } 461 colmap = sell->colmap; 462 ierr = MatGetColumnIJ_SeqSELL_Color(A,0,PETSC_FALSE,PETSC_FALSE,&ncols,&A_ci,&A_cj,&spidxA,NULL);CHKERRQ(ierr); 463 ierr = MatGetColumnIJ_SeqSELL_Color(B,0,PETSC_FALSE,PETSC_FALSE,&ncols,&B_ci,&B_cj,&spidxB,NULL);CHKERRQ(ierr); 464 465 bs = 1; /* only bs=1 is supported for non MPIBAIJ matrix */ 466 467 if (ctype == IS_COLORING_GLOBAL && c->htype[0] == 'd') { /* create vscale for storing dx */ 468 ierr = VecCreateGhost(PetscObjectComm((PetscObject)mat),mat->cmap->n,PETSC_DETERMINE,B->cmap->n,sell->garray,&c->vscale);CHKERRQ(ierr); 469 ierr = VecBindToCPU(c->vscale,PETSC_TRUE);CHKERRQ(ierr); 470 } 471 } else { 472 Mat_MPIAIJ *aij=(Mat_MPIAIJ*)mat->data; 473 Mat_SeqAIJ *spA,*spB; 474 A = aij->A; spA = (Mat_SeqAIJ*)A->data; A_val = spA->a; 475 B = aij->B; spB = (Mat_SeqAIJ*)B->data; B_val = spB->a; 476 nz = spA->nz + spB->nz; /* total nonzero entries of mat */ 477 if (!aij->colmap) { 478 /* Allow access to data structures of local part of matrix 479 - creates aij->colmap which maps global column number to local number in part B */ 480 ierr = MatCreateColmap_MPIAIJ_Private(mat);CHKERRQ(ierr); 481 } 482 colmap = aij->colmap; 483 ierr = MatGetColumnIJ_SeqAIJ_Color(A,0,PETSC_FALSE,PETSC_FALSE,&ncols,&A_ci,&A_cj,&spidxA,NULL);CHKERRQ(ierr); 484 ierr = MatGetColumnIJ_SeqAIJ_Color(B,0,PETSC_FALSE,PETSC_FALSE,&ncols,&B_ci,&B_cj,&spidxB,NULL);CHKERRQ(ierr); 485 486 bs = 1; /* only bs=1 is supported for non MPIBAIJ matrix */ 487 488 if (ctype == IS_COLORING_GLOBAL && c->htype[0] == 'd') { /* create vscale for storing dx */ 489 ierr = VecCreateGhost(PetscObjectComm((PetscObject)mat),mat->cmap->n,PETSC_DETERMINE,B->cmap->n,aij->garray,&c->vscale);CHKERRQ(ierr); 490 ierr = VecBindToCPU(c->vscale,PETSC_TRUE);CHKERRQ(ierr); 491 } 492 } 493 494 m = mat->rmap->n/bs; 495 cstart = mat->cmap->rstart/bs; 496 cend = mat->cmap->rend/bs; 497 498 ierr = PetscMalloc2(nis,&c->ncolumns,nis,&c->columns);CHKERRQ(ierr); 499 ierr = PetscMalloc1(nis,&c->nrows);CHKERRQ(ierr); 500 ierr = PetscLogObjectMemory((PetscObject)c,3*nis*sizeof(PetscInt));CHKERRQ(ierr); 501 502 if (c->htype[0] == 'd') { 503 ierr = PetscMalloc1(nz,&Jentry);CHKERRQ(ierr); 504 ierr = PetscLogObjectMemory((PetscObject)c,nz*sizeof(MatEntry));CHKERRQ(ierr); 505 c->matentry = Jentry; 506 } else if (c->htype[0] == 'w') { 507 ierr = PetscMalloc1(nz,&Jentry2);CHKERRQ(ierr); 508 ierr = PetscLogObjectMemory((PetscObject)c,nz*sizeof(MatEntry2));CHKERRQ(ierr); 509 c->matentry2 = Jentry2; 510 } else SETERRQ(PetscObjectComm((PetscObject)mat),PETSC_ERR_SUP,"htype is not supported"); 511 512 ierr = PetscMalloc2(m+1,&rowhit,m+1,&valaddrhit);CHKERRQ(ierr); 513 nz = 0; 514 ierr = ISColoringGetIS(iscoloring,PETSC_OWN_POINTER, PETSC_IGNORE,&c->isa);CHKERRQ(ierr); 515 516 if (ctype == IS_COLORING_GLOBAL) { 517 ierr = MPI_Comm_size(PetscObjectComm((PetscObject)mat),&size);CHKERRMPI(ierr); 518 ierr = PetscMalloc2(size,&ncolsonproc,size,&disp);CHKERRQ(ierr); 519 } 520 521 for (i=0; i<nis; i++) { /* for each local color */ 522 ierr = ISGetLocalSize(c->isa[i],&n);CHKERRQ(ierr); 523 ierr = ISGetIndices(c->isa[i],&is);CHKERRQ(ierr); 524 525 c->ncolumns[i] = n; /* local number of columns of this color on this process */ 526 c->columns[i] = (PetscInt*)is; 527 528 if (ctype == IS_COLORING_GLOBAL) { 529 /* Determine nctot, the total (parallel) number of columns of this color */ 530 /* ncolsonproc[j]: local ncolumns on proc[j] of this color */ 531 ierr = PetscMPIIntCast(n,&nn);CHKERRQ(ierr); 532 ierr = MPI_Allgather(&nn,1,MPI_INT,ncolsonproc,1,MPI_INT,PetscObjectComm((PetscObject)mat));CHKERRMPI(ierr); 533 nctot = 0; for (j=0; j<size; j++) nctot += ncolsonproc[j]; 534 if (!nctot) { 535 ierr = PetscInfo(mat,"Coloring of matrix has some unneeded colors with no corresponding rows\n");CHKERRQ(ierr); 536 } 537 538 disp[0] = 0; 539 for (j=1; j<size; j++) { 540 disp[j] = disp[j-1] + ncolsonproc[j-1]; 541 } 542 543 /* Get cols, the complete list of columns for this color on each process */ 544 ierr = PetscMalloc1(nctot+1,&cols);CHKERRQ(ierr); 545 ierr = MPI_Allgatherv((void*)is,n,MPIU_INT,cols,ncolsonproc,disp,MPIU_INT,PetscObjectComm((PetscObject)mat));CHKERRMPI(ierr); 546 } else if (ctype == IS_COLORING_LOCAL) { 547 /* Determine local number of columns of this color on this process, including ghost points */ 548 nctot = n; 549 cols = (PetscInt*)is; 550 } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Not provided for this MatFDColoring type"); 551 552 /* Mark all rows affect by these columns */ 553 ierr = PetscArrayzero(rowhit,m);CHKERRQ(ierr); 554 bs2 = bs*bs; 555 nrows_i = 0; 556 for (j=0; j<nctot; j++) { /* loop over columns*/ 557 if (ctype == IS_COLORING_LOCAL) { 558 col = ltog[cols[j]]; 559 } else { 560 col = cols[j]; 561 } 562 if (col >= cstart && col < cend) { /* column is in A, diagonal block of mat */ 563 tmp = A_ci[col-cstart]; 564 row = A_cj + tmp; 565 nrows = A_ci[col-cstart+1] - tmp; 566 nrows_i += nrows; 567 568 /* loop over columns of A marking them in rowhit */ 569 for (k=0; k<nrows; k++) { 570 /* set valaddrhit for part A */ 571 spidx = bs2*spidxA[tmp + k]; 572 valaddrhit[*row] = &A_val[spidx]; 573 rowhit[*row++] = col - cstart + 1; /* local column index */ 574 } 575 } else { /* column is in B, off-diagonal block of mat */ 576 #if defined(PETSC_USE_CTABLE) 577 ierr = PetscTableFind(colmap,col+1,&colb);CHKERRQ(ierr); 578 colb--; 579 #else 580 colb = colmap[col] - 1; /* local column index */ 581 #endif 582 if (colb == -1) { 583 nrows = 0; 584 } else { 585 colb = colb/bs; 586 tmp = B_ci[colb]; 587 row = B_cj + tmp; 588 nrows = B_ci[colb+1] - tmp; 589 } 590 nrows_i += nrows; 591 /* loop over columns of B marking them in rowhit */ 592 for (k=0; k<nrows; k++) { 593 /* set valaddrhit for part B */ 594 spidx = bs2*spidxB[tmp + k]; 595 valaddrhit[*row] = &B_val[spidx]; 596 rowhit[*row++] = colb + 1 + cend - cstart; /* local column index */ 597 } 598 } 599 } 600 c->nrows[i] = nrows_i; 601 602 if (c->htype[0] == 'd') { 603 for (j=0; j<m; j++) { 604 if (rowhit[j]) { 605 Jentry[nz].row = j; /* local row index */ 606 Jentry[nz].col = rowhit[j] - 1; /* local column index */ 607 Jentry[nz].valaddr = valaddrhit[j]; /* address of mat value for this entry */ 608 nz++; 609 } 610 } 611 } else { /* c->htype == 'wp' */ 612 for (j=0; j<m; j++) { 613 if (rowhit[j]) { 614 Jentry2[nz].row = j; /* local row index */ 615 Jentry2[nz].valaddr = valaddrhit[j]; /* address of mat value for this entry */ 616 nz++; 617 } 618 } 619 } 620 if (ctype == IS_COLORING_GLOBAL) { 621 ierr = PetscFree(cols);CHKERRQ(ierr); 622 } 623 } 624 if (ctype == IS_COLORING_GLOBAL) { 625 ierr = PetscFree2(ncolsonproc,disp);CHKERRQ(ierr); 626 } 627 628 if (bcols > 1) { /* reorder Jentry for faster MatFDColoringApply() */ 629 ierr = MatFDColoringSetUpBlocked_AIJ_Private(mat,c,nz);CHKERRQ(ierr); 630 } 631 632 if (isBAIJ) { 633 ierr = MatRestoreColumnIJ_SeqBAIJ_Color(A,0,PETSC_FALSE,PETSC_FALSE,&ncols,&A_ci,&A_cj,&spidxA,NULL);CHKERRQ(ierr); 634 ierr = MatRestoreColumnIJ_SeqBAIJ_Color(B,0,PETSC_FALSE,PETSC_FALSE,&ncols,&B_ci,&B_cj,&spidxB,NULL);CHKERRQ(ierr); 635 ierr = PetscMalloc1(bs*mat->rmap->n,&c->dy);CHKERRQ(ierr); 636 } else if (isSELL) { 637 ierr = MatRestoreColumnIJ_SeqSELL_Color(A,0,PETSC_FALSE,PETSC_FALSE,&ncols,&A_ci,&A_cj,&spidxA,NULL);CHKERRQ(ierr); 638 ierr = MatRestoreColumnIJ_SeqSELL_Color(B,0,PETSC_FALSE,PETSC_FALSE,&ncols,&B_ci,&B_cj,&spidxB,NULL);CHKERRQ(ierr); 639 } else { 640 ierr = MatRestoreColumnIJ_SeqAIJ_Color(A,0,PETSC_FALSE,PETSC_FALSE,&ncols,&A_ci,&A_cj,&spidxA,NULL);CHKERRQ(ierr); 641 ierr = MatRestoreColumnIJ_SeqAIJ_Color(B,0,PETSC_FALSE,PETSC_FALSE,&ncols,&B_ci,&B_cj,&spidxB,NULL);CHKERRQ(ierr); 642 } 643 644 ierr = ISColoringRestoreIS(iscoloring,PETSC_OWN_POINTER,&c->isa);CHKERRQ(ierr); 645 ierr = PetscFree2(rowhit,valaddrhit);CHKERRQ(ierr); 646 647 if (ctype == IS_COLORING_LOCAL) { 648 ierr = ISLocalToGlobalMappingRestoreIndices(map,<og);CHKERRQ(ierr); 649 } 650 ierr = PetscInfo3(c,"ncolors %D, brows %D and bcols %D are used.\n",c->ncolors,c->brows,c->bcols);CHKERRQ(ierr); 651 PetscFunctionReturn(0); 652 } 653 654 PetscErrorCode MatFDColoringCreate_MPIXAIJ(Mat mat,ISColoring iscoloring,MatFDColoring c) 655 { 656 PetscErrorCode ierr; 657 PetscInt bs,nis=iscoloring->n,m=mat->rmap->n; 658 PetscBool isBAIJ,isSELL; 659 660 PetscFunctionBegin; 661 /* set default brows and bcols for speedup inserting the dense matrix into sparse Jacobian; 662 bcols is chosen s.t. dy-array takes 50% of memory space as mat */ 663 ierr = MatGetBlockSize(mat,&bs);CHKERRQ(ierr); 664 ierr = PetscObjectBaseTypeCompare((PetscObject)mat,MATMPIBAIJ,&isBAIJ);CHKERRQ(ierr); 665 ierr = PetscObjectTypeCompare((PetscObject)mat,MATMPISELL,&isSELL);CHKERRQ(ierr); 666 if (isBAIJ || m == 0) { 667 c->brows = m; 668 c->bcols = 1; 669 } else if (isSELL) { 670 /* bcols is chosen s.t. dy-array takes 50% of local memory space as mat */ 671 Mat_MPISELL *sell=(Mat_MPISELL*)mat->data; 672 Mat_SeqSELL *spA,*spB; 673 Mat A,B; 674 PetscInt nz,brows,bcols; 675 PetscReal mem; 676 677 bs = 1; /* only bs=1 is supported for MPISELL matrix */ 678 679 A = sell->A; spA = (Mat_SeqSELL*)A->data; 680 B = sell->B; spB = (Mat_SeqSELL*)B->data; 681 nz = spA->nz + spB->nz; /* total local nonzero entries of mat */ 682 mem = nz*(sizeof(PetscScalar) + sizeof(PetscInt)) + 3*m*sizeof(PetscInt); 683 bcols = (PetscInt)(0.5*mem /(m*sizeof(PetscScalar))); 684 brows = 1000/bcols; 685 if (bcols > nis) bcols = nis; 686 if (brows == 0 || brows > m) brows = m; 687 c->brows = brows; 688 c->bcols = bcols; 689 } else { /* mpiaij matrix */ 690 /* bcols is chosen s.t. dy-array takes 50% of local memory space as mat */ 691 Mat_MPIAIJ *aij=(Mat_MPIAIJ*)mat->data; 692 Mat_SeqAIJ *spA,*spB; 693 Mat A,B; 694 PetscInt nz,brows,bcols; 695 PetscReal mem; 696 697 bs = 1; /* only bs=1 is supported for MPIAIJ matrix */ 698 699 A = aij->A; spA = (Mat_SeqAIJ*)A->data; 700 B = aij->B; spB = (Mat_SeqAIJ*)B->data; 701 nz = spA->nz + spB->nz; /* total local nonzero entries of mat */ 702 mem = nz*(sizeof(PetscScalar) + sizeof(PetscInt)) + 3*m*sizeof(PetscInt); 703 bcols = (PetscInt)(0.5*mem /(m*sizeof(PetscScalar))); 704 brows = 1000/bcols; 705 if (bcols > nis) bcols = nis; 706 if (brows == 0 || brows > m) brows = m; 707 c->brows = brows; 708 c->bcols = bcols; 709 } 710 711 c->M = mat->rmap->N/bs; /* set the global rows and columns and local rows */ 712 c->N = mat->cmap->N/bs; 713 c->m = mat->rmap->n/bs; 714 c->rstart = mat->rmap->rstart/bs; 715 c->ncolors = nis; 716 PetscFunctionReturn(0); 717 } 718 719 /*@C 720 721 MatFDColoringSetValues - takes a matrix in compressed color format and enters the matrix into a PETSc Mat 722 723 Collective on J 724 725 Input Parameters: 726 + J - the sparse matrix 727 . coloring - created with MatFDColoringCreate() and a local coloring 728 - y - column major storage of matrix values with one color of values per column, the number of rows of y should match 729 the number of local rows of J and the number of columns is the number of colors. 730 731 Level: intermediate 732 733 Notes: the matrix in compressed color format may come from an Automatic Differentiation code 734 735 The code will be slightly faster if MatFDColoringSetBlockSize(coloring,PETSC_DEFAULT,nc); is called immediately after creating the coloring 736 737 .seealso: MatFDColoringCreate(), ISColoring, ISColoringCreate(), ISColoringSetType(), IS_COLORING_LOCAL, MatFDColoringSetBlockSize() 738 739 @*/ 740 PetscErrorCode MatFDColoringSetValues(Mat J,MatFDColoring coloring,const PetscScalar *y) 741 { 742 PetscErrorCode ierr; 743 MatEntry2 *Jentry2; 744 PetscInt row,i,nrows_k,l,ncolors,nz = 0,bcols,nbcols = 0; 745 const PetscInt *nrows; 746 PetscBool eq; 747 748 PetscFunctionBegin; 749 PetscValidHeaderSpecific(J,MAT_CLASSID,1); 750 PetscValidHeaderSpecific(coloring,MAT_FDCOLORING_CLASSID,2); 751 ierr = PetscObjectCompareId((PetscObject)J,coloring->matid,&eq);CHKERRQ(ierr); 752 if (!eq) SETERRQ(PetscObjectComm((PetscObject)J),PETSC_ERR_ARG_WRONG,"Matrix used with MatFDColoringSetValues() must be that used with MatFDColoringCreate()"); 753 Jentry2 = coloring->matentry2; 754 nrows = coloring->nrows; 755 ncolors = coloring->ncolors; 756 bcols = coloring->bcols; 757 758 for (i=0; i<ncolors; i+=bcols) { 759 nrows_k = nrows[nbcols++]; 760 for (l=0; l<nrows_k; l++) { 761 row = Jentry2[nz].row; /* local row index */ 762 *(Jentry2[nz++].valaddr) = y[row]; 763 } 764 y += bcols*coloring->m; 765 } 766 ierr = MatAssemblyBegin(J,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 767 ierr = MatAssemblyEnd(J,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 768 PetscFunctionReturn(0); 769 } 770