1 2 #include <../src/mat/impls/baij/mpi/mpibaij.h> /*I "petscmat.h" I*/ 3 4 #include <petscblaslapack.h> 5 #include <petscsf.h> 6 7 #undef __FUNCT__ 8 #define __FUNCT__ "MatGetRowMaxAbs_MPIBAIJ" 9 PetscErrorCode MatGetRowMaxAbs_MPIBAIJ(Mat A,Vec v,PetscInt idx[]) 10 { 11 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 12 PetscErrorCode ierr; 13 PetscInt i,*idxb = 0; 14 PetscScalar *va,*vb; 15 Vec vtmp; 16 17 PetscFunctionBegin; 18 ierr = MatGetRowMaxAbs(a->A,v,idx);CHKERRQ(ierr); 19 ierr = VecGetArray(v,&va);CHKERRQ(ierr); 20 if (idx) { 21 for (i=0; i<A->rmap->n; i++) { 22 if (PetscAbsScalar(va[i])) idx[i] += A->cmap->rstart; 23 } 24 } 25 26 ierr = VecCreateSeq(PETSC_COMM_SELF,A->rmap->n,&vtmp);CHKERRQ(ierr); 27 if (idx) {ierr = PetscMalloc1(A->rmap->n,&idxb);CHKERRQ(ierr);} 28 ierr = MatGetRowMaxAbs(a->B,vtmp,idxb);CHKERRQ(ierr); 29 ierr = VecGetArray(vtmp,&vb);CHKERRQ(ierr); 30 31 for (i=0; i<A->rmap->n; i++) { 32 if (PetscAbsScalar(va[i]) < PetscAbsScalar(vb[i])) { 33 va[i] = vb[i]; 34 if (idx) idx[i] = A->cmap->bs*a->garray[idxb[i]/A->cmap->bs] + (idxb[i] % A->cmap->bs); 35 } 36 } 37 38 ierr = VecRestoreArray(v,&va);CHKERRQ(ierr); 39 ierr = VecRestoreArray(vtmp,&vb);CHKERRQ(ierr); 40 ierr = PetscFree(idxb);CHKERRQ(ierr); 41 ierr = VecDestroy(&vtmp);CHKERRQ(ierr); 42 PetscFunctionReturn(0); 43 } 44 45 #undef __FUNCT__ 46 #define __FUNCT__ "MatStoreValues_MPIBAIJ" 47 PetscErrorCode MatStoreValues_MPIBAIJ(Mat mat) 48 { 49 Mat_MPIBAIJ *aij = (Mat_MPIBAIJ*)mat->data; 50 PetscErrorCode ierr; 51 52 PetscFunctionBegin; 53 ierr = MatStoreValues(aij->A);CHKERRQ(ierr); 54 ierr = MatStoreValues(aij->B);CHKERRQ(ierr); 55 PetscFunctionReturn(0); 56 } 57 58 #undef __FUNCT__ 59 #define __FUNCT__ "MatRetrieveValues_MPIBAIJ" 60 PetscErrorCode MatRetrieveValues_MPIBAIJ(Mat mat) 61 { 62 Mat_MPIBAIJ *aij = (Mat_MPIBAIJ*)mat->data; 63 PetscErrorCode ierr; 64 65 PetscFunctionBegin; 66 ierr = MatRetrieveValues(aij->A);CHKERRQ(ierr); 67 ierr = MatRetrieveValues(aij->B);CHKERRQ(ierr); 68 PetscFunctionReturn(0); 69 } 70 71 /* 72 Local utility routine that creates a mapping from the global column 73 number to the local number in the off-diagonal part of the local 74 storage of the matrix. This is done in a non scalable way since the 75 length of colmap equals the global matrix length. 76 */ 77 #undef __FUNCT__ 78 #define __FUNCT__ "MatCreateColmap_MPIBAIJ_Private" 79 PetscErrorCode MatCreateColmap_MPIBAIJ_Private(Mat mat) 80 { 81 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 82 Mat_SeqBAIJ *B = (Mat_SeqBAIJ*)baij->B->data; 83 PetscErrorCode ierr; 84 PetscInt nbs = B->nbs,i,bs=mat->rmap->bs; 85 86 PetscFunctionBegin; 87 #if defined(PETSC_USE_CTABLE) 88 ierr = PetscTableCreate(baij->nbs,baij->Nbs+1,&baij->colmap);CHKERRQ(ierr); 89 for (i=0; i<nbs; i++) { 90 ierr = PetscTableAdd(baij->colmap,baij->garray[i]+1,i*bs+1,INSERT_VALUES);CHKERRQ(ierr); 91 } 92 #else 93 ierr = PetscMalloc1(baij->Nbs+1,&baij->colmap);CHKERRQ(ierr); 94 ierr = PetscLogObjectMemory((PetscObject)mat,baij->Nbs*sizeof(PetscInt));CHKERRQ(ierr); 95 ierr = PetscMemzero(baij->colmap,baij->Nbs*sizeof(PetscInt));CHKERRQ(ierr); 96 for (i=0; i<nbs; i++) baij->colmap[baij->garray[i]] = i*bs+1; 97 #endif 98 PetscFunctionReturn(0); 99 } 100 101 #define MatSetValues_SeqBAIJ_A_Private(row,col,value,addv,orow,ocol) \ 102 { \ 103 \ 104 brow = row/bs; \ 105 rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; \ 106 rmax = aimax[brow]; nrow = ailen[brow]; \ 107 bcol = col/bs; \ 108 ridx = row % bs; cidx = col % bs; \ 109 low = 0; high = nrow; \ 110 while (high-low > 3) { \ 111 t = (low+high)/2; \ 112 if (rp[t] > bcol) high = t; \ 113 else low = t; \ 114 } \ 115 for (_i=low; _i<high; _i++) { \ 116 if (rp[_i] > bcol) break; \ 117 if (rp[_i] == bcol) { \ 118 bap = ap + bs2*_i + bs*cidx + ridx; \ 119 if (addv == ADD_VALUES) *bap += value; \ 120 else *bap = value; \ 121 goto a_noinsert; \ 122 } \ 123 } \ 124 if (a->nonew == 1) goto a_noinsert; \ 125 if (a->nonew == -1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero at global row/column (%D, %D) into matrix", orow, ocol); \ 126 MatSeqXAIJReallocateAIJ(A,a->mbs,bs2,nrow,brow,bcol,rmax,aa,ai,aj,rp,ap,aimax,a->nonew,MatScalar); \ 127 N = nrow++ - 1; \ 128 /* shift up all the later entries in this row */ \ 129 for (ii=N; ii>=_i; ii--) { \ 130 rp[ii+1] = rp[ii]; \ 131 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 132 } \ 133 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr); } \ 134 rp[_i] = bcol; \ 135 ap[bs2*_i + bs*cidx + ridx] = value; \ 136 a_noinsert:; \ 137 ailen[brow] = nrow; \ 138 } 139 140 #define MatSetValues_SeqBAIJ_B_Private(row,col,value,addv,orow,ocol) \ 141 { \ 142 brow = row/bs; \ 143 rp = bj + bi[brow]; ap = ba + bs2*bi[brow]; \ 144 rmax = bimax[brow]; nrow = bilen[brow]; \ 145 bcol = col/bs; \ 146 ridx = row % bs; cidx = col % bs; \ 147 low = 0; high = nrow; \ 148 while (high-low > 3) { \ 149 t = (low+high)/2; \ 150 if (rp[t] > bcol) high = t; \ 151 else low = t; \ 152 } \ 153 for (_i=low; _i<high; _i++) { \ 154 if (rp[_i] > bcol) break; \ 155 if (rp[_i] == bcol) { \ 156 bap = ap + bs2*_i + bs*cidx + ridx; \ 157 if (addv == ADD_VALUES) *bap += value; \ 158 else *bap = value; \ 159 goto b_noinsert; \ 160 } \ 161 } \ 162 if (b->nonew == 1) goto b_noinsert; \ 163 if (b->nonew == -1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero at global row/column (%D, %D) into matrix", orow, ocol); \ 164 MatSeqXAIJReallocateAIJ(B,b->mbs,bs2,nrow,brow,bcol,rmax,ba,bi,bj,rp,ap,bimax,b->nonew,MatScalar); \ 165 N = nrow++ - 1; \ 166 /* shift up all the later entries in this row */ \ 167 for (ii=N; ii>=_i; ii--) { \ 168 rp[ii+1] = rp[ii]; \ 169 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 170 } \ 171 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr);} \ 172 rp[_i] = bcol; \ 173 ap[bs2*_i + bs*cidx + ridx] = value; \ 174 b_noinsert:; \ 175 bilen[brow] = nrow; \ 176 } 177 178 #undef __FUNCT__ 179 #define __FUNCT__ "MatSetValues_MPIBAIJ" 180 PetscErrorCode MatSetValues_MPIBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 181 { 182 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 183 MatScalar value; 184 PetscBool roworiented = baij->roworiented; 185 PetscErrorCode ierr; 186 PetscInt i,j,row,col; 187 PetscInt rstart_orig=mat->rmap->rstart; 188 PetscInt rend_orig =mat->rmap->rend,cstart_orig=mat->cmap->rstart; 189 PetscInt cend_orig =mat->cmap->rend,bs=mat->rmap->bs; 190 191 /* Some Variables required in the macro */ 192 Mat A = baij->A; 193 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)(A)->data; 194 PetscInt *aimax=a->imax,*ai=a->i,*ailen=a->ilen,*aj=a->j; 195 MatScalar *aa =a->a; 196 197 Mat B = baij->B; 198 Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(B)->data; 199 PetscInt *bimax=b->imax,*bi=b->i,*bilen=b->ilen,*bj=b->j; 200 MatScalar *ba =b->a; 201 202 PetscInt *rp,ii,nrow,_i,rmax,N,brow,bcol; 203 PetscInt low,high,t,ridx,cidx,bs2=a->bs2; 204 MatScalar *ap,*bap; 205 206 PetscFunctionBegin; 207 for (i=0; i<m; i++) { 208 if (im[i] < 0) continue; 209 #if defined(PETSC_USE_DEBUG) 210 if (im[i] >= mat->rmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",im[i],mat->rmap->N-1); 211 #endif 212 if (im[i] >= rstart_orig && im[i] < rend_orig) { 213 row = im[i] - rstart_orig; 214 for (j=0; j<n; j++) { 215 if (in[j] >= cstart_orig && in[j] < cend_orig) { 216 col = in[j] - cstart_orig; 217 if (roworiented) value = v[i*n+j]; 218 else value = v[i+j*m]; 219 MatSetValues_SeqBAIJ_A_Private(row,col,value,addv,im[i],in[j]); 220 /* ierr = MatSetValues_SeqBAIJ(baij->A,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 221 } else if (in[j] < 0) continue; 222 #if defined(PETSC_USE_DEBUG) 223 else if (in[j] >= mat->cmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[j],mat->cmap->N-1); 224 #endif 225 else { 226 if (mat->was_assembled) { 227 if (!baij->colmap) { 228 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 229 } 230 #if defined(PETSC_USE_CTABLE) 231 ierr = PetscTableFind(baij->colmap,in[j]/bs + 1,&col);CHKERRQ(ierr); 232 col = col - 1; 233 #else 234 col = baij->colmap[in[j]/bs] - 1; 235 #endif 236 if (col < 0 && !((Mat_SeqBAIJ*)(baij->B->data))->nonew) { 237 ierr = MatDisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 238 col = in[j]; 239 /* Reinitialize the variables required by MatSetValues_SeqBAIJ_B_Private() */ 240 B = baij->B; 241 b = (Mat_SeqBAIJ*)(B)->data; 242 bimax=b->imax;bi=b->i;bilen=b->ilen;bj=b->j; 243 ba =b->a; 244 } else if (col < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", im[i], in[j]); 245 else col += in[j]%bs; 246 } else col = in[j]; 247 if (roworiented) value = v[i*n+j]; 248 else value = v[i+j*m]; 249 MatSetValues_SeqBAIJ_B_Private(row,col,value,addv,im[i],in[j]); 250 /* ierr = MatSetValues_SeqBAIJ(baij->B,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 251 } 252 } 253 } else { 254 if (mat->nooffprocentries) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Setting off process row %D even though MatSetOption(,MAT_NO_OFF_PROC_ENTRIES,PETSC_TRUE) was set",im[i]); 255 if (!baij->donotstash) { 256 mat->assembled = PETSC_FALSE; 257 if (roworiented) { 258 ierr = MatStashValuesRow_Private(&mat->stash,im[i],n,in,v+i*n,PETSC_FALSE);CHKERRQ(ierr); 259 } else { 260 ierr = MatStashValuesCol_Private(&mat->stash,im[i],n,in,v+i,m,PETSC_FALSE);CHKERRQ(ierr); 261 } 262 } 263 } 264 } 265 PetscFunctionReturn(0); 266 } 267 268 #undef __FUNCT__ 269 #define __FUNCT__ "MatSetValuesBlocked_SeqBAIJ_Inlined" 270 PETSC_STATIC_INLINE PetscErrorCode MatSetValuesBlocked_SeqBAIJ_Inlined(Mat A,PetscInt row,PetscInt col,const PetscScalar v[],InsertMode is,PetscInt orow,PetscInt ocol) 271 { 272 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data; 273 PetscInt *rp,low,high,t,ii,jj,nrow,i,rmax,N; 274 PetscInt *imax=a->imax,*ai=a->i,*ailen=a->ilen; 275 PetscErrorCode ierr; 276 PetscInt *aj =a->j,nonew=a->nonew,bs2=a->bs2,bs=A->rmap->bs; 277 PetscBool roworiented=a->roworiented; 278 const PetscScalar *value = v; 279 MatScalar *ap,*aa = a->a,*bap; 280 281 PetscFunctionBegin; 282 rp = aj + ai[row]; 283 ap = aa + bs2*ai[row]; 284 rmax = imax[row]; 285 nrow = ailen[row]; 286 low = 0; 287 high = nrow; 288 value = v; 289 low = 0; 290 high = nrow; 291 while (high-low > 7) { 292 t = (low+high)/2; 293 if (rp[t] > col) high = t; 294 else low = t; 295 } 296 for (i=low; i<high; i++) { 297 if (rp[i] > col) break; 298 if (rp[i] == col) { 299 bap = ap + bs2*i; 300 if (roworiented) { 301 if (is == ADD_VALUES) { 302 for (ii=0; ii<bs; ii++) { 303 for (jj=ii; jj<bs2; jj+=bs) { 304 bap[jj] += *value++; 305 } 306 } 307 } else { 308 for (ii=0; ii<bs; ii++) { 309 for (jj=ii; jj<bs2; jj+=bs) { 310 bap[jj] = *value++; 311 } 312 } 313 } 314 } else { 315 if (is == ADD_VALUES) { 316 for (ii=0; ii<bs; ii++,value+=bs) { 317 for (jj=0; jj<bs; jj++) { 318 bap[jj] += value[jj]; 319 } 320 bap += bs; 321 } 322 } else { 323 for (ii=0; ii<bs; ii++,value+=bs) { 324 for (jj=0; jj<bs; jj++) { 325 bap[jj] = value[jj]; 326 } 327 bap += bs; 328 } 329 } 330 } 331 goto noinsert2; 332 } 333 } 334 if (nonew == 1) goto noinsert2; 335 if (nonew == -1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new global block indexed nonzero block (%D, %D) in the matrix", orow, ocol); 336 MatSeqXAIJReallocateAIJ(A,a->mbs,bs2,nrow,row,col,rmax,aa,ai,aj,rp,ap,imax,nonew,MatScalar); 337 N = nrow++ - 1; high++; 338 /* shift up all the later entries in this row */ 339 for (ii=N; ii>=i; ii--) { 340 rp[ii+1] = rp[ii]; 341 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); 342 } 343 if (N >= i) { 344 ierr = PetscMemzero(ap+bs2*i,bs2*sizeof(MatScalar));CHKERRQ(ierr); 345 } 346 rp[i] = col; 347 bap = ap + bs2*i; 348 if (roworiented) { 349 for (ii=0; ii<bs; ii++) { 350 for (jj=ii; jj<bs2; jj+=bs) { 351 bap[jj] = *value++; 352 } 353 } 354 } else { 355 for (ii=0; ii<bs; ii++) { 356 for (jj=0; jj<bs; jj++) { 357 *bap++ = *value++; 358 } 359 } 360 } 361 noinsert2:; 362 ailen[row] = nrow; 363 PetscFunctionReturn(0); 364 } 365 366 #undef __FUNCT__ 367 #define __FUNCT__ "MatSetValuesBlocked_MPIBAIJ" 368 /* 369 This routine should be optimized so that the block copy at ** Here a copy is required ** below is not needed 370 by passing additional stride information into the MatSetValuesBlocked_SeqBAIJ_Inlined() routine 371 */ 372 PetscErrorCode MatSetValuesBlocked_MPIBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 373 { 374 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 375 const PetscScalar *value; 376 MatScalar *barray = baij->barray; 377 PetscBool roworiented = baij->roworiented; 378 PetscErrorCode ierr; 379 PetscInt i,j,ii,jj,row,col,rstart=baij->rstartbs; 380 PetscInt rend=baij->rendbs,cstart=baij->cstartbs,stepval; 381 PetscInt cend=baij->cendbs,bs=mat->rmap->bs,bs2=baij->bs2; 382 383 PetscFunctionBegin; 384 if (!barray) { 385 ierr = PetscMalloc1(bs2,&barray);CHKERRQ(ierr); 386 baij->barray = barray; 387 } 388 389 if (roworiented) stepval = (n-1)*bs; 390 else stepval = (m-1)*bs; 391 392 for (i=0; i<m; i++) { 393 if (im[i] < 0) continue; 394 #if defined(PETSC_USE_DEBUG) 395 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Block indexed row too large %D max %D",im[i],baij->Mbs-1); 396 #endif 397 if (im[i] >= rstart && im[i] < rend) { 398 row = im[i] - rstart; 399 for (j=0; j<n; j++) { 400 /* If NumCol = 1 then a copy is not required */ 401 if ((roworiented) && (n == 1)) { 402 barray = (MatScalar*)v + i*bs2; 403 } else if ((!roworiented) && (m == 1)) { 404 barray = (MatScalar*)v + j*bs2; 405 } else { /* Here a copy is required */ 406 if (roworiented) { 407 value = v + (i*(stepval+bs) + j)*bs; 408 } else { 409 value = v + (j*(stepval+bs) + i)*bs; 410 } 411 for (ii=0; ii<bs; ii++,value+=bs+stepval) { 412 for (jj=0; jj<bs; jj++) barray[jj] = value[jj]; 413 barray += bs; 414 } 415 barray -= bs2; 416 } 417 418 if (in[j] >= cstart && in[j] < cend) { 419 col = in[j] - cstart; 420 ierr = MatSetValuesBlocked_SeqBAIJ_Inlined(baij->A,row,col,barray,addv,im[i],in[j]);CHKERRQ(ierr); 421 } else if (in[j] < 0) continue; 422 #if defined(PETSC_USE_DEBUG) 423 else if (in[j] >= baij->Nbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Block indexed column too large %D max %D",in[j],baij->Nbs-1); 424 #endif 425 else { 426 if (mat->was_assembled) { 427 if (!baij->colmap) { 428 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 429 } 430 431 #if defined(PETSC_USE_DEBUG) 432 #if defined(PETSC_USE_CTABLE) 433 { PetscInt data; 434 ierr = PetscTableFind(baij->colmap,in[j]+1,&data);CHKERRQ(ierr); 435 if ((data - 1) % bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Incorrect colmap"); 436 } 437 #else 438 if ((baij->colmap[in[j]] - 1) % bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Incorrect colmap"); 439 #endif 440 #endif 441 #if defined(PETSC_USE_CTABLE) 442 ierr = PetscTableFind(baij->colmap,in[j]+1,&col);CHKERRQ(ierr); 443 col = (col - 1)/bs; 444 #else 445 col = (baij->colmap[in[j]] - 1)/bs; 446 #endif 447 if (col < 0 && !((Mat_SeqBAIJ*)(baij->B->data))->nonew) { 448 ierr = MatDisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 449 col = in[j]; 450 } else if (col < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new blocked indexed nonzero block (%D, %D) into matrix",im[i],in[j]); 451 } else col = in[j]; 452 ierr = MatSetValuesBlocked_SeqBAIJ_Inlined(baij->B,row,col,barray,addv,im[i],in[j]);CHKERRQ(ierr); 453 } 454 } 455 } else { 456 if (mat->nooffprocentries) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Setting off process block indexed row %D even though MatSetOption(,MAT_NO_OFF_PROC_ENTRIES,PETSC_TRUE) was set",im[i]); 457 if (!baij->donotstash) { 458 if (roworiented) { 459 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 460 } else { 461 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 462 } 463 } 464 } 465 } 466 PetscFunctionReturn(0); 467 } 468 469 #define HASH_KEY 0.6180339887 470 #define HASH(size,key,tmp) (tmp = (key)*HASH_KEY,(PetscInt)((size)*(tmp-(PetscInt)tmp))) 471 /* #define HASH(size,key) ((PetscInt)((size)*fmod(((key)*HASH_KEY),1))) */ 472 /* #define HASH(size,key,tmp) ((PetscInt)((size)*fmod(((key)*HASH_KEY),1))) */ 473 #undef __FUNCT__ 474 #define __FUNCT__ "MatSetValues_MPIBAIJ_HT" 475 PetscErrorCode MatSetValues_MPIBAIJ_HT(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 476 { 477 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 478 PetscBool roworiented = baij->roworiented; 479 PetscErrorCode ierr; 480 PetscInt i,j,row,col; 481 PetscInt rstart_orig=mat->rmap->rstart; 482 PetscInt rend_orig =mat->rmap->rend,Nbs=baij->Nbs; 483 PetscInt h1,key,size=baij->ht_size,bs=mat->rmap->bs,*HT=baij->ht,idx; 484 PetscReal tmp; 485 MatScalar **HD = baij->hd,value; 486 #if defined(PETSC_USE_DEBUG) 487 PetscInt total_ct=baij->ht_total_ct,insert_ct=baij->ht_insert_ct; 488 #endif 489 490 PetscFunctionBegin; 491 for (i=0; i<m; i++) { 492 #if defined(PETSC_USE_DEBUG) 493 if (im[i] < 0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative row"); 494 if (im[i] >= mat->rmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",im[i],mat->rmap->N-1); 495 #endif 496 row = im[i]; 497 if (row >= rstart_orig && row < rend_orig) { 498 for (j=0; j<n; j++) { 499 col = in[j]; 500 if (roworiented) value = v[i*n+j]; 501 else value = v[i+j*m]; 502 /* Look up PetscInto the Hash Table */ 503 key = (row/bs)*Nbs+(col/bs)+1; 504 h1 = HASH(size,key,tmp); 505 506 507 idx = h1; 508 #if defined(PETSC_USE_DEBUG) 509 insert_ct++; 510 total_ct++; 511 if (HT[idx] != key) { 512 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++,total_ct++) ; 513 if (idx == size) { 514 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++,total_ct++) ; 515 if (idx == h1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"(%D,%D) has no entry in the hash table", row, col); 516 } 517 } 518 #else 519 if (HT[idx] != key) { 520 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++) ; 521 if (idx == size) { 522 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++) ; 523 if (idx == h1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"(%D,%D) has no entry in the hash table", row, col); 524 } 525 } 526 #endif 527 /* A HASH table entry is found, so insert the values at the correct address */ 528 if (addv == ADD_VALUES) *(HD[idx]+ (col % bs)*bs + (row % bs)) += value; 529 else *(HD[idx]+ (col % bs)*bs + (row % bs)) = value; 530 } 531 } else if (!baij->donotstash) { 532 if (roworiented) { 533 ierr = MatStashValuesRow_Private(&mat->stash,im[i],n,in,v+i*n,PETSC_FALSE);CHKERRQ(ierr); 534 } else { 535 ierr = MatStashValuesCol_Private(&mat->stash,im[i],n,in,v+i,m,PETSC_FALSE);CHKERRQ(ierr); 536 } 537 } 538 } 539 #if defined(PETSC_USE_DEBUG) 540 baij->ht_total_ct = total_ct; 541 baij->ht_insert_ct = insert_ct; 542 #endif 543 PetscFunctionReturn(0); 544 } 545 546 #undef __FUNCT__ 547 #define __FUNCT__ "MatSetValuesBlocked_MPIBAIJ_HT" 548 PetscErrorCode MatSetValuesBlocked_MPIBAIJ_HT(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 549 { 550 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 551 PetscBool roworiented = baij->roworiented; 552 PetscErrorCode ierr; 553 PetscInt i,j,ii,jj,row,col; 554 PetscInt rstart=baij->rstartbs; 555 PetscInt rend =mat->rmap->rend,stepval,bs=mat->rmap->bs,bs2=baij->bs2,nbs2=n*bs2; 556 PetscInt h1,key,size=baij->ht_size,idx,*HT=baij->ht,Nbs=baij->Nbs; 557 PetscReal tmp; 558 MatScalar **HD = baij->hd,*baij_a; 559 const PetscScalar *v_t,*value; 560 #if defined(PETSC_USE_DEBUG) 561 PetscInt total_ct=baij->ht_total_ct,insert_ct=baij->ht_insert_ct; 562 #endif 563 564 PetscFunctionBegin; 565 if (roworiented) stepval = (n-1)*bs; 566 else stepval = (m-1)*bs; 567 568 for (i=0; i<m; i++) { 569 #if defined(PETSC_USE_DEBUG) 570 if (im[i] < 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %D",im[i]); 571 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",im[i],baij->Mbs-1); 572 #endif 573 row = im[i]; 574 v_t = v + i*nbs2; 575 if (row >= rstart && row < rend) { 576 for (j=0; j<n; j++) { 577 col = in[j]; 578 579 /* Look up into the Hash Table */ 580 key = row*Nbs+col+1; 581 h1 = HASH(size,key,tmp); 582 583 idx = h1; 584 #if defined(PETSC_USE_DEBUG) 585 total_ct++; 586 insert_ct++; 587 if (HT[idx] != key) { 588 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++,total_ct++) ; 589 if (idx == size) { 590 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++,total_ct++) ; 591 if (idx == h1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"(%D,%D) has no entry in the hash table", row, col); 592 } 593 } 594 #else 595 if (HT[idx] != key) { 596 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++) ; 597 if (idx == size) { 598 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++) ; 599 if (idx == h1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"(%D,%D) has no entry in the hash table", row, col); 600 } 601 } 602 #endif 603 baij_a = HD[idx]; 604 if (roworiented) { 605 /*value = v + i*(stepval+bs)*bs + j*bs;*/ 606 /* value = v + (i*(stepval+bs)+j)*bs; */ 607 value = v_t; 608 v_t += bs; 609 if (addv == ADD_VALUES) { 610 for (ii=0; ii<bs; ii++,value+=stepval) { 611 for (jj=ii; jj<bs2; jj+=bs) { 612 baij_a[jj] += *value++; 613 } 614 } 615 } else { 616 for (ii=0; ii<bs; ii++,value+=stepval) { 617 for (jj=ii; jj<bs2; jj+=bs) { 618 baij_a[jj] = *value++; 619 } 620 } 621 } 622 } else { 623 value = v + j*(stepval+bs)*bs + i*bs; 624 if (addv == ADD_VALUES) { 625 for (ii=0; ii<bs; ii++,value+=stepval,baij_a+=bs) { 626 for (jj=0; jj<bs; jj++) { 627 baij_a[jj] += *value++; 628 } 629 } 630 } else { 631 for (ii=0; ii<bs; ii++,value+=stepval,baij_a+=bs) { 632 for (jj=0; jj<bs; jj++) { 633 baij_a[jj] = *value++; 634 } 635 } 636 } 637 } 638 } 639 } else { 640 if (!baij->donotstash) { 641 if (roworiented) { 642 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 643 } else { 644 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 645 } 646 } 647 } 648 } 649 #if defined(PETSC_USE_DEBUG) 650 baij->ht_total_ct = total_ct; 651 baij->ht_insert_ct = insert_ct; 652 #endif 653 PetscFunctionReturn(0); 654 } 655 656 #undef __FUNCT__ 657 #define __FUNCT__ "MatGetValues_MPIBAIJ" 658 PetscErrorCode MatGetValues_MPIBAIJ(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],PetscScalar v[]) 659 { 660 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 661 PetscErrorCode ierr; 662 PetscInt bs = mat->rmap->bs,i,j,bsrstart = mat->rmap->rstart,bsrend = mat->rmap->rend; 663 PetscInt bscstart = mat->cmap->rstart,bscend = mat->cmap->rend,row,col,data; 664 665 PetscFunctionBegin; 666 for (i=0; i<m; i++) { 667 if (idxm[i] < 0) continue; /* SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %D",idxm[i]);*/ 668 if (idxm[i] >= mat->rmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",idxm[i],mat->rmap->N-1); 669 if (idxm[i] >= bsrstart && idxm[i] < bsrend) { 670 row = idxm[i] - bsrstart; 671 for (j=0; j<n; j++) { 672 if (idxn[j] < 0) continue; /* SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative column: %D",idxn[j]); */ 673 if (idxn[j] >= mat->cmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",idxn[j],mat->cmap->N-1); 674 if (idxn[j] >= bscstart && idxn[j] < bscend) { 675 col = idxn[j] - bscstart; 676 ierr = MatGetValues_SeqBAIJ(baij->A,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 677 } else { 678 if (!baij->colmap) { 679 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 680 } 681 #if defined(PETSC_USE_CTABLE) 682 ierr = PetscTableFind(baij->colmap,idxn[j]/bs+1,&data);CHKERRQ(ierr); 683 data--; 684 #else 685 data = baij->colmap[idxn[j]/bs]-1; 686 #endif 687 if ((data < 0) || (baij->garray[data/bs] != idxn[j]/bs)) *(v+i*n+j) = 0.0; 688 else { 689 col = data + idxn[j]%bs; 690 ierr = MatGetValues_SeqBAIJ(baij->B,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 691 } 692 } 693 } 694 } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only local values currently supported"); 695 } 696 PetscFunctionReturn(0); 697 } 698 699 #undef __FUNCT__ 700 #define __FUNCT__ "MatNorm_MPIBAIJ" 701 PetscErrorCode MatNorm_MPIBAIJ(Mat mat,NormType type,PetscReal *nrm) 702 { 703 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 704 Mat_SeqBAIJ *amat = (Mat_SeqBAIJ*)baij->A->data,*bmat = (Mat_SeqBAIJ*)baij->B->data; 705 PetscErrorCode ierr; 706 PetscInt i,j,bs2=baij->bs2,bs=baij->A->rmap->bs,nz,row,col; 707 PetscReal sum = 0.0; 708 MatScalar *v; 709 710 PetscFunctionBegin; 711 if (baij->size == 1) { 712 ierr = MatNorm(baij->A,type,nrm);CHKERRQ(ierr); 713 } else { 714 if (type == NORM_FROBENIUS) { 715 v = amat->a; 716 nz = amat->nz*bs2; 717 for (i=0; i<nz; i++) { 718 sum += PetscRealPart(PetscConj(*v)*(*v)); v++; 719 } 720 v = bmat->a; 721 nz = bmat->nz*bs2; 722 for (i=0; i<nz; i++) { 723 sum += PetscRealPart(PetscConj(*v)*(*v)); v++; 724 } 725 ierr = MPI_Allreduce(&sum,nrm,1,MPIU_REAL,MPIU_SUM,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 726 *nrm = PetscSqrtReal(*nrm); 727 } else if (type == NORM_1) { /* max column sum */ 728 PetscReal *tmp,*tmp2; 729 PetscInt *jj,*garray=baij->garray,cstart=baij->rstartbs; 730 ierr = PetscMalloc2(mat->cmap->N,&tmp,mat->cmap->N,&tmp2);CHKERRQ(ierr); 731 ierr = PetscMemzero(tmp,mat->cmap->N*sizeof(PetscReal));CHKERRQ(ierr); 732 v = amat->a; jj = amat->j; 733 for (i=0; i<amat->nz; i++) { 734 for (j=0; j<bs; j++) { 735 col = bs*(cstart + *jj) + j; /* column index */ 736 for (row=0; row<bs; row++) { 737 tmp[col] += PetscAbsScalar(*v); v++; 738 } 739 } 740 jj++; 741 } 742 v = bmat->a; jj = bmat->j; 743 for (i=0; i<bmat->nz; i++) { 744 for (j=0; j<bs; j++) { 745 col = bs*garray[*jj] + j; 746 for (row=0; row<bs; row++) { 747 tmp[col] += PetscAbsScalar(*v); v++; 748 } 749 } 750 jj++; 751 } 752 ierr = MPI_Allreduce(tmp,tmp2,mat->cmap->N,MPIU_REAL,MPIU_SUM,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 753 *nrm = 0.0; 754 for (j=0; j<mat->cmap->N; j++) { 755 if (tmp2[j] > *nrm) *nrm = tmp2[j]; 756 } 757 ierr = PetscFree2(tmp,tmp2);CHKERRQ(ierr); 758 } else if (type == NORM_INFINITY) { /* max row sum */ 759 PetscReal *sums; 760 ierr = PetscMalloc1(bs,&sums);CHKERRQ(ierr); 761 sum = 0.0; 762 for (j=0; j<amat->mbs; j++) { 763 for (row=0; row<bs; row++) sums[row] = 0.0; 764 v = amat->a + bs2*amat->i[j]; 765 nz = amat->i[j+1]-amat->i[j]; 766 for (i=0; i<nz; i++) { 767 for (col=0; col<bs; col++) { 768 for (row=0; row<bs; row++) { 769 sums[row] += PetscAbsScalar(*v); v++; 770 } 771 } 772 } 773 v = bmat->a + bs2*bmat->i[j]; 774 nz = bmat->i[j+1]-bmat->i[j]; 775 for (i=0; i<nz; i++) { 776 for (col=0; col<bs; col++) { 777 for (row=0; row<bs; row++) { 778 sums[row] += PetscAbsScalar(*v); v++; 779 } 780 } 781 } 782 for (row=0; row<bs; row++) { 783 if (sums[row] > sum) sum = sums[row]; 784 } 785 } 786 ierr = MPI_Allreduce(&sum,nrm,1,MPIU_REAL,MPIU_MAX,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 787 ierr = PetscFree(sums);CHKERRQ(ierr); 788 } else SETERRQ(PetscObjectComm((PetscObject)mat),PETSC_ERR_SUP,"No support for this norm yet"); 789 } 790 PetscFunctionReturn(0); 791 } 792 793 /* 794 Creates the hash table, and sets the table 795 This table is created only once. 796 If new entried need to be added to the matrix 797 then the hash table has to be destroyed and 798 recreated. 799 */ 800 #undef __FUNCT__ 801 #define __FUNCT__ "MatCreateHashTable_MPIBAIJ_Private" 802 PetscErrorCode MatCreateHashTable_MPIBAIJ_Private(Mat mat,PetscReal factor) 803 { 804 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 805 Mat A = baij->A,B=baij->B; 806 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data,*b=(Mat_SeqBAIJ*)B->data; 807 PetscInt i,j,k,nz=a->nz+b->nz,h1,*ai=a->i,*aj=a->j,*bi=b->i,*bj=b->j; 808 PetscErrorCode ierr; 809 PetscInt ht_size,bs2=baij->bs2,rstart=baij->rstartbs; 810 PetscInt cstart=baij->cstartbs,*garray=baij->garray,row,col,Nbs=baij->Nbs; 811 PetscInt *HT,key; 812 MatScalar **HD; 813 PetscReal tmp; 814 #if defined(PETSC_USE_INFO) 815 PetscInt ct=0,max=0; 816 #endif 817 818 PetscFunctionBegin; 819 if (baij->ht) PetscFunctionReturn(0); 820 821 baij->ht_size = (PetscInt)(factor*nz); 822 ht_size = baij->ht_size; 823 824 /* Allocate Memory for Hash Table */ 825 ierr = PetscCalloc2(ht_size,&baij->hd,ht_size,&baij->ht);CHKERRQ(ierr); 826 HD = baij->hd; 827 HT = baij->ht; 828 829 /* Loop Over A */ 830 for (i=0; i<a->mbs; i++) { 831 for (j=ai[i]; j<ai[i+1]; j++) { 832 row = i+rstart; 833 col = aj[j]+cstart; 834 835 key = row*Nbs + col + 1; 836 h1 = HASH(ht_size,key,tmp); 837 for (k=0; k<ht_size; k++) { 838 if (!HT[(h1+k)%ht_size]) { 839 HT[(h1+k)%ht_size] = key; 840 HD[(h1+k)%ht_size] = a->a + j*bs2; 841 break; 842 #if defined(PETSC_USE_INFO) 843 } else { 844 ct++; 845 #endif 846 } 847 } 848 #if defined(PETSC_USE_INFO) 849 if (k> max) max = k; 850 #endif 851 } 852 } 853 /* Loop Over B */ 854 for (i=0; i<b->mbs; i++) { 855 for (j=bi[i]; j<bi[i+1]; j++) { 856 row = i+rstart; 857 col = garray[bj[j]]; 858 key = row*Nbs + col + 1; 859 h1 = HASH(ht_size,key,tmp); 860 for (k=0; k<ht_size; k++) { 861 if (!HT[(h1+k)%ht_size]) { 862 HT[(h1+k)%ht_size] = key; 863 HD[(h1+k)%ht_size] = b->a + j*bs2; 864 break; 865 #if defined(PETSC_USE_INFO) 866 } else { 867 ct++; 868 #endif 869 } 870 } 871 #if defined(PETSC_USE_INFO) 872 if (k> max) max = k; 873 #endif 874 } 875 } 876 877 /* Print Summary */ 878 #if defined(PETSC_USE_INFO) 879 for (i=0,j=0; i<ht_size; i++) { 880 if (HT[i]) j++; 881 } 882 ierr = PetscInfo2(mat,"Average Search = %5.2f,max search = %D\n",(!j)? 0.0:((PetscReal)(ct+j))/j,max);CHKERRQ(ierr); 883 #endif 884 PetscFunctionReturn(0); 885 } 886 887 #undef __FUNCT__ 888 #define __FUNCT__ "MatAssemblyBegin_MPIBAIJ" 889 PetscErrorCode MatAssemblyBegin_MPIBAIJ(Mat mat,MatAssemblyType mode) 890 { 891 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 892 PetscErrorCode ierr; 893 PetscInt nstash,reallocs; 894 InsertMode addv; 895 896 PetscFunctionBegin; 897 if (baij->donotstash || mat->nooffprocentries) PetscFunctionReturn(0); 898 899 /* make sure all processors are either in INSERTMODE or ADDMODE */ 900 ierr = MPI_Allreduce((PetscEnum*)&mat->insertmode,(PetscEnum*)&addv,1,MPIU_ENUM,MPI_BOR,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 901 if (addv == (ADD_VALUES|INSERT_VALUES)) SETERRQ(PetscObjectComm((PetscObject)mat),PETSC_ERR_ARG_WRONGSTATE,"Some processors inserted others added"); 902 mat->insertmode = addv; /* in case this processor had no cache */ 903 904 ierr = MatStashScatterBegin_Private(mat,&mat->stash,mat->rmap->range);CHKERRQ(ierr); 905 ierr = MatStashScatterBegin_Private(mat,&mat->bstash,baij->rangebs);CHKERRQ(ierr); 906 ierr = MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);CHKERRQ(ierr); 907 ierr = PetscInfo2(mat,"Stash has %D entries,uses %D mallocs.\n",nstash,reallocs);CHKERRQ(ierr); 908 ierr = MatStashGetInfo_Private(&mat->bstash,&nstash,&reallocs);CHKERRQ(ierr); 909 ierr = PetscInfo2(mat,"Block-Stash has %D entries, uses %D mallocs.\n",nstash,reallocs);CHKERRQ(ierr); 910 PetscFunctionReturn(0); 911 } 912 913 #undef __FUNCT__ 914 #define __FUNCT__ "MatAssemblyEnd_MPIBAIJ" 915 PetscErrorCode MatAssemblyEnd_MPIBAIJ(Mat mat,MatAssemblyType mode) 916 { 917 Mat_MPIBAIJ *baij=(Mat_MPIBAIJ*)mat->data; 918 Mat_SeqBAIJ *a =(Mat_SeqBAIJ*)baij->A->data; 919 PetscErrorCode ierr; 920 PetscInt i,j,rstart,ncols,flg,bs2=baij->bs2; 921 PetscInt *row,*col; 922 PetscBool r1,r2,r3,other_disassembled; 923 MatScalar *val; 924 InsertMode addv = mat->insertmode; 925 PetscMPIInt n; 926 927 PetscFunctionBegin; 928 /* do not use 'b=(Mat_SeqBAIJ*)baij->B->data' as B can be reset in disassembly */ 929 if (!baij->donotstash && !mat->nooffprocentries) { 930 while (1) { 931 ierr = MatStashScatterGetMesg_Private(&mat->stash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 932 if (!flg) break; 933 934 for (i=0; i<n;) { 935 /* Now identify the consecutive vals belonging to the same row */ 936 for (j=i,rstart=row[j]; j<n; j++) { 937 if (row[j] != rstart) break; 938 } 939 if (j < n) ncols = j-i; 940 else ncols = n-i; 941 /* Now assemble all these values with a single function call */ 942 ierr = MatSetValues_MPIBAIJ(mat,1,row+i,ncols,col+i,val+i,addv);CHKERRQ(ierr); 943 i = j; 944 } 945 } 946 ierr = MatStashScatterEnd_Private(&mat->stash);CHKERRQ(ierr); 947 /* Now process the block-stash. Since the values are stashed column-oriented, 948 set the roworiented flag to column oriented, and after MatSetValues() 949 restore the original flags */ 950 r1 = baij->roworiented; 951 r2 = a->roworiented; 952 r3 = ((Mat_SeqBAIJ*)baij->B->data)->roworiented; 953 954 baij->roworiented = PETSC_FALSE; 955 a->roworiented = PETSC_FALSE; 956 957 (((Mat_SeqBAIJ*)baij->B->data))->roworiented = PETSC_FALSE; /* b->roworiented */ 958 while (1) { 959 ierr = MatStashScatterGetMesg_Private(&mat->bstash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 960 if (!flg) break; 961 962 for (i=0; i<n;) { 963 /* Now identify the consecutive vals belonging to the same row */ 964 for (j=i,rstart=row[j]; j<n; j++) { 965 if (row[j] != rstart) break; 966 } 967 if (j < n) ncols = j-i; 968 else ncols = n-i; 969 ierr = MatSetValuesBlocked_MPIBAIJ(mat,1,row+i,ncols,col+i,val+i*bs2,addv);CHKERRQ(ierr); 970 i = j; 971 } 972 } 973 ierr = MatStashScatterEnd_Private(&mat->bstash);CHKERRQ(ierr); 974 975 baij->roworiented = r1; 976 a->roworiented = r2; 977 978 ((Mat_SeqBAIJ*)baij->B->data)->roworiented = r3; /* b->roworiented */ 979 } 980 981 ierr = MatAssemblyBegin(baij->A,mode);CHKERRQ(ierr); 982 ierr = MatAssemblyEnd(baij->A,mode);CHKERRQ(ierr); 983 984 /* determine if any processor has disassembled, if so we must 985 also disassemble ourselfs, in order that we may reassemble. */ 986 /* 987 if nonzero structure of submatrix B cannot change then we know that 988 no processor disassembled thus we can skip this stuff 989 */ 990 if (!((Mat_SeqBAIJ*)baij->B->data)->nonew) { 991 ierr = MPI_Allreduce(&mat->was_assembled,&other_disassembled,1,MPIU_BOOL,MPI_PROD,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 992 if (mat->was_assembled && !other_disassembled) { 993 ierr = MatDisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 994 } 995 } 996 997 if (!mat->was_assembled && mode == MAT_FINAL_ASSEMBLY) { 998 ierr = MatSetUpMultiply_MPIBAIJ(mat);CHKERRQ(ierr); 999 } 1000 ierr = MatAssemblyBegin(baij->B,mode);CHKERRQ(ierr); 1001 ierr = MatAssemblyEnd(baij->B,mode);CHKERRQ(ierr); 1002 1003 #if defined(PETSC_USE_INFO) 1004 if (baij->ht && mode== MAT_FINAL_ASSEMBLY) { 1005 ierr = PetscInfo1(mat,"Average Hash Table Search in MatSetValues = %5.2f\n",((PetscReal)baij->ht_total_ct)/baij->ht_insert_ct);CHKERRQ(ierr); 1006 1007 baij->ht_total_ct = 0; 1008 baij->ht_insert_ct = 0; 1009 } 1010 #endif 1011 if (baij->ht_flag && !baij->ht && mode == MAT_FINAL_ASSEMBLY) { 1012 ierr = MatCreateHashTable_MPIBAIJ_Private(mat,baij->ht_fact);CHKERRQ(ierr); 1013 1014 mat->ops->setvalues = MatSetValues_MPIBAIJ_HT; 1015 mat->ops->setvaluesblocked = MatSetValuesBlocked_MPIBAIJ_HT; 1016 } 1017 1018 ierr = PetscFree2(baij->rowvalues,baij->rowindices);CHKERRQ(ierr); 1019 1020 baij->rowvalues = 0; 1021 1022 /* if no new nonzero locations are allowed in matrix then only set the matrix state the first time through */ 1023 if ((!mat->was_assembled && mode == MAT_FINAL_ASSEMBLY) || !((Mat_SeqBAIJ*)(baij->A->data))->nonew) { 1024 PetscObjectState state = baij->A->nonzerostate + baij->B->nonzerostate; 1025 ierr = MPI_Allreduce(&state,&mat->nonzerostate,1,MPIU_INT64,MPI_SUM,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1026 } 1027 PetscFunctionReturn(0); 1028 } 1029 1030 extern PetscErrorCode MatView_SeqBAIJ(Mat,PetscViewer); 1031 #include <petscdraw.h> 1032 #undef __FUNCT__ 1033 #define __FUNCT__ "MatView_MPIBAIJ_ASCIIorDraworSocket" 1034 static PetscErrorCode MatView_MPIBAIJ_ASCIIorDraworSocket(Mat mat,PetscViewer viewer) 1035 { 1036 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1037 PetscErrorCode ierr; 1038 PetscMPIInt rank = baij->rank; 1039 PetscInt bs = mat->rmap->bs; 1040 PetscBool iascii,isdraw; 1041 PetscViewer sviewer; 1042 PetscViewerFormat format; 1043 1044 PetscFunctionBegin; 1045 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&iascii);CHKERRQ(ierr); 1046 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERDRAW,&isdraw);CHKERRQ(ierr); 1047 if (iascii) { 1048 ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr); 1049 if (format == PETSC_VIEWER_ASCII_INFO_DETAIL) { 1050 MatInfo info; 1051 ierr = MPI_Comm_rank(PetscObjectComm((PetscObject)mat),&rank);CHKERRQ(ierr); 1052 ierr = MatGetInfo(mat,MAT_LOCAL,&info);CHKERRQ(ierr); 1053 ierr = PetscViewerASCIIPushSynchronized(viewer);CHKERRQ(ierr); 1054 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] Local rows %D nz %D nz alloced %D bs %D mem %D\n", 1055 rank,mat->rmap->n,(PetscInt)info.nz_used,(PetscInt)info.nz_allocated,mat->rmap->bs,(PetscInt)info.memory);CHKERRQ(ierr); 1056 ierr = MatGetInfo(baij->A,MAT_LOCAL,&info);CHKERRQ(ierr); 1057 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] on-diagonal part: nz %D \n",rank,(PetscInt)info.nz_used);CHKERRQ(ierr); 1058 ierr = MatGetInfo(baij->B,MAT_LOCAL,&info);CHKERRQ(ierr); 1059 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] off-diagonal part: nz %D \n",rank,(PetscInt)info.nz_used);CHKERRQ(ierr); 1060 ierr = PetscViewerFlush(viewer);CHKERRQ(ierr); 1061 ierr = PetscViewerASCIIPopSynchronized(viewer);CHKERRQ(ierr); 1062 ierr = PetscViewerASCIIPrintf(viewer,"Information on VecScatter used in matrix-vector product: \n");CHKERRQ(ierr); 1063 ierr = VecScatterView(baij->Mvctx,viewer);CHKERRQ(ierr); 1064 PetscFunctionReturn(0); 1065 } else if (format == PETSC_VIEWER_ASCII_INFO) { 1066 ierr = PetscViewerASCIIPrintf(viewer," block size is %D\n",bs);CHKERRQ(ierr); 1067 PetscFunctionReturn(0); 1068 } else if (format == PETSC_VIEWER_ASCII_FACTOR_INFO) { 1069 PetscFunctionReturn(0); 1070 } 1071 } 1072 1073 if (isdraw) { 1074 PetscDraw draw; 1075 PetscBool isnull; 1076 ierr = PetscViewerDrawGetDraw(viewer,0,&draw);CHKERRQ(ierr); 1077 ierr = PetscDrawIsNull(draw,&isnull);CHKERRQ(ierr); if (isnull) PetscFunctionReturn(0); 1078 } 1079 1080 { 1081 /* assemble the entire matrix onto first processor. */ 1082 Mat A; 1083 Mat_SeqBAIJ *Aloc; 1084 PetscInt M = mat->rmap->N,N = mat->cmap->N,*ai,*aj,col,i,j,k,*rvals,mbs = baij->mbs; 1085 MatScalar *a; 1086 const char *matname; 1087 1088 /* Here we are creating a temporary matrix, so will assume MPIBAIJ is acceptable */ 1089 /* Perhaps this should be the type of mat? */ 1090 ierr = MatCreate(PetscObjectComm((PetscObject)mat),&A);CHKERRQ(ierr); 1091 if (!rank) { 1092 ierr = MatSetSizes(A,M,N,M,N);CHKERRQ(ierr); 1093 } else { 1094 ierr = MatSetSizes(A,0,0,M,N);CHKERRQ(ierr); 1095 } 1096 ierr = MatSetType(A,MATMPIBAIJ);CHKERRQ(ierr); 1097 ierr = MatMPIBAIJSetPreallocation(A,mat->rmap->bs,0,NULL,0,NULL);CHKERRQ(ierr); 1098 ierr = MatSetOption(A,MAT_NEW_NONZERO_LOCATION_ERR,PETSC_FALSE);CHKERRQ(ierr); 1099 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)A);CHKERRQ(ierr); 1100 1101 /* copy over the A part */ 1102 Aloc = (Mat_SeqBAIJ*)baij->A->data; 1103 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1104 ierr = PetscMalloc1(bs,&rvals);CHKERRQ(ierr); 1105 1106 for (i=0; i<mbs; i++) { 1107 rvals[0] = bs*(baij->rstartbs + i); 1108 for (j=1; j<bs; j++) rvals[j] = rvals[j-1] + 1; 1109 for (j=ai[i]; j<ai[i+1]; j++) { 1110 col = (baij->cstartbs+aj[j])*bs; 1111 for (k=0; k<bs; k++) { 1112 ierr = MatSetValues_MPIBAIJ(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 1113 col++; a += bs; 1114 } 1115 } 1116 } 1117 /* copy over the B part */ 1118 Aloc = (Mat_SeqBAIJ*)baij->B->data; 1119 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1120 for (i=0; i<mbs; i++) { 1121 rvals[0] = bs*(baij->rstartbs + i); 1122 for (j=1; j<bs; j++) rvals[j] = rvals[j-1] + 1; 1123 for (j=ai[i]; j<ai[i+1]; j++) { 1124 col = baij->garray[aj[j]]*bs; 1125 for (k=0; k<bs; k++) { 1126 ierr = MatSetValues_MPIBAIJ(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 1127 col++; a += bs; 1128 } 1129 } 1130 } 1131 ierr = PetscFree(rvals);CHKERRQ(ierr); 1132 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1133 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1134 /* 1135 Everyone has to call to draw the matrix since the graphics waits are 1136 synchronized across all processors that share the PetscDraw object 1137 */ 1138 ierr = PetscViewerGetSubViewer(viewer,PETSC_COMM_SELF,&sviewer);CHKERRQ(ierr); 1139 ierr = PetscObjectGetName((PetscObject)mat,&matname);CHKERRQ(ierr); 1140 if (!rank) { 1141 ierr = PetscObjectSetName((PetscObject)((Mat_MPIBAIJ*)(A->data))->A,matname);CHKERRQ(ierr); 1142 ierr = MatView_SeqBAIJ(((Mat_MPIBAIJ*)(A->data))->A,sviewer);CHKERRQ(ierr); 1143 } 1144 ierr = PetscViewerRestoreSubViewer(viewer,PETSC_COMM_SELF,&sviewer);CHKERRQ(ierr); 1145 ierr = PetscViewerFlush(viewer);CHKERRQ(ierr); 1146 ierr = MatDestroy(&A);CHKERRQ(ierr); 1147 } 1148 PetscFunctionReturn(0); 1149 } 1150 1151 #undef __FUNCT__ 1152 #define __FUNCT__ "MatView_MPIBAIJ_Binary" 1153 static PetscErrorCode MatView_MPIBAIJ_Binary(Mat mat,PetscViewer viewer) 1154 { 1155 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)mat->data; 1156 Mat_SeqBAIJ *A = (Mat_SeqBAIJ*)a->A->data; 1157 Mat_SeqBAIJ *B = (Mat_SeqBAIJ*)a->B->data; 1158 PetscErrorCode ierr; 1159 PetscInt i,*row_lens,*crow_lens,bs = mat->rmap->bs,j,k,bs2=a->bs2,header[4],nz,rlen; 1160 PetscInt *range=0,nzmax,*column_indices,cnt,col,*garray = a->garray,cstart = mat->cmap->rstart/bs,len,pcnt,l,ll; 1161 int fd; 1162 PetscScalar *column_values; 1163 FILE *file; 1164 PetscMPIInt rank,size,tag = ((PetscObject)viewer)->tag; 1165 PetscInt message_count,flowcontrolcount; 1166 1167 PetscFunctionBegin; 1168 ierr = MPI_Comm_rank(PetscObjectComm((PetscObject)mat),&rank);CHKERRQ(ierr); 1169 ierr = MPI_Comm_size(PetscObjectComm((PetscObject)mat),&size);CHKERRQ(ierr); 1170 nz = bs2*(A->nz + B->nz); 1171 rlen = mat->rmap->n; 1172 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 1173 if (!rank) { 1174 header[0] = MAT_FILE_CLASSID; 1175 header[1] = mat->rmap->N; 1176 header[2] = mat->cmap->N; 1177 1178 ierr = MPI_Reduce(&nz,&header[3],1,MPIU_INT,MPI_SUM,0,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1179 ierr = PetscBinaryWrite(fd,header,4,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1180 /* get largest number of rows any processor has */ 1181 range = mat->rmap->range; 1182 for (i=1; i<size; i++) { 1183 rlen = PetscMax(rlen,range[i+1] - range[i]); 1184 } 1185 } else { 1186 ierr = MPI_Reduce(&nz,0,1,MPIU_INT,MPI_SUM,0,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1187 } 1188 1189 ierr = PetscMalloc1(rlen/bs,&crow_lens);CHKERRQ(ierr); 1190 /* compute lengths of each row */ 1191 for (i=0; i<a->mbs; i++) { 1192 crow_lens[i] = A->i[i+1] - A->i[i] + B->i[i+1] - B->i[i]; 1193 } 1194 /* store the row lengths to the file */ 1195 ierr = PetscViewerFlowControlStart(viewer,&message_count,&flowcontrolcount);CHKERRQ(ierr); 1196 if (!rank) { 1197 MPI_Status status; 1198 ierr = PetscMalloc1(rlen,&row_lens);CHKERRQ(ierr); 1199 rlen = (range[1] - range[0])/bs; 1200 for (i=0; i<rlen; i++) { 1201 for (j=0; j<bs; j++) { 1202 row_lens[i*bs+j] = bs*crow_lens[i]; 1203 } 1204 } 1205 ierr = PetscBinaryWrite(fd,row_lens,bs*rlen,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1206 for (i=1; i<size; i++) { 1207 rlen = (range[i+1] - range[i])/bs; 1208 ierr = PetscViewerFlowControlStepMaster(viewer,i,&message_count,flowcontrolcount);CHKERRQ(ierr); 1209 ierr = MPI_Recv(crow_lens,rlen,MPIU_INT,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1210 for (k=0; k<rlen; k++) { 1211 for (j=0; j<bs; j++) { 1212 row_lens[k*bs+j] = bs*crow_lens[k]; 1213 } 1214 } 1215 ierr = PetscBinaryWrite(fd,row_lens,bs*rlen,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1216 } 1217 ierr = PetscViewerFlowControlEndMaster(viewer,&message_count);CHKERRQ(ierr); 1218 ierr = PetscFree(row_lens);CHKERRQ(ierr); 1219 } else { 1220 ierr = PetscViewerFlowControlStepWorker(viewer,rank,&message_count);CHKERRQ(ierr); 1221 ierr = MPI_Send(crow_lens,mat->rmap->n/bs,MPIU_INT,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1222 ierr = PetscViewerFlowControlEndWorker(viewer,&message_count);CHKERRQ(ierr); 1223 } 1224 ierr = PetscFree(crow_lens);CHKERRQ(ierr); 1225 1226 /* load up the local column indices. Include for all rows not just one for each block row since process 0 does not have the 1227 information needed to make it for each row from a block row. This does require more communication but still not more than 1228 the communication needed for the nonzero values */ 1229 nzmax = nz; /* space a largest processor needs */ 1230 ierr = MPI_Reduce(&nz,&nzmax,1,MPIU_INT,MPI_MAX,0,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1231 ierr = PetscMalloc1(nzmax,&column_indices);CHKERRQ(ierr); 1232 cnt = 0; 1233 for (i=0; i<a->mbs; i++) { 1234 pcnt = cnt; 1235 for (j=B->i[i]; j<B->i[i+1]; j++) { 1236 if ((col = garray[B->j[j]]) > cstart) break; 1237 for (l=0; l<bs; l++) { 1238 column_indices[cnt++] = bs*col+l; 1239 } 1240 } 1241 for (k=A->i[i]; k<A->i[i+1]; k++) { 1242 for (l=0; l<bs; l++) { 1243 column_indices[cnt++] = bs*(A->j[k] + cstart)+l; 1244 } 1245 } 1246 for (; j<B->i[i+1]; j++) { 1247 for (l=0; l<bs; l++) { 1248 column_indices[cnt++] = bs*garray[B->j[j]]+l; 1249 } 1250 } 1251 len = cnt - pcnt; 1252 for (k=1; k<bs; k++) { 1253 ierr = PetscMemcpy(&column_indices[cnt],&column_indices[pcnt],len*sizeof(PetscInt));CHKERRQ(ierr); 1254 cnt += len; 1255 } 1256 } 1257 if (cnt != nz) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_LIB,"Internal PETSc error: cnt = %D nz = %D",cnt,nz); 1258 1259 /* store the columns to the file */ 1260 ierr = PetscViewerFlowControlStart(viewer,&message_count,&flowcontrolcount);CHKERRQ(ierr); 1261 if (!rank) { 1262 MPI_Status status; 1263 ierr = PetscBinaryWrite(fd,column_indices,nz,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1264 for (i=1; i<size; i++) { 1265 ierr = PetscViewerFlowControlStepMaster(viewer,i,&message_count,flowcontrolcount);CHKERRQ(ierr); 1266 ierr = MPI_Recv(&cnt,1,MPIU_INT,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1267 ierr = MPI_Recv(column_indices,cnt,MPIU_INT,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1268 ierr = PetscBinaryWrite(fd,column_indices,cnt,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1269 } 1270 ierr = PetscViewerFlowControlEndMaster(viewer,&message_count);CHKERRQ(ierr); 1271 } else { 1272 ierr = PetscViewerFlowControlStepWorker(viewer,rank,&message_count);CHKERRQ(ierr); 1273 ierr = MPI_Send(&cnt,1,MPIU_INT,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1274 ierr = MPI_Send(column_indices,cnt,MPIU_INT,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1275 ierr = PetscViewerFlowControlEndWorker(viewer,&message_count);CHKERRQ(ierr); 1276 } 1277 ierr = PetscFree(column_indices);CHKERRQ(ierr); 1278 1279 /* load up the numerical values */ 1280 ierr = PetscMalloc1(nzmax,&column_values);CHKERRQ(ierr); 1281 cnt = 0; 1282 for (i=0; i<a->mbs; i++) { 1283 rlen = bs*(B->i[i+1] - B->i[i] + A->i[i+1] - A->i[i]); 1284 for (j=B->i[i]; j<B->i[i+1]; j++) { 1285 if (garray[B->j[j]] > cstart) break; 1286 for (l=0; l<bs; l++) { 1287 for (ll=0; ll<bs; ll++) { 1288 column_values[cnt + l*rlen + ll] = B->a[bs2*j+l+bs*ll]; 1289 } 1290 } 1291 cnt += bs; 1292 } 1293 for (k=A->i[i]; k<A->i[i+1]; k++) { 1294 for (l=0; l<bs; l++) { 1295 for (ll=0; ll<bs; ll++) { 1296 column_values[cnt + l*rlen + ll] = A->a[bs2*k+l+bs*ll]; 1297 } 1298 } 1299 cnt += bs; 1300 } 1301 for (; j<B->i[i+1]; j++) { 1302 for (l=0; l<bs; l++) { 1303 for (ll=0; ll<bs; ll++) { 1304 column_values[cnt + l*rlen + ll] = B->a[bs2*j+l+bs*ll]; 1305 } 1306 } 1307 cnt += bs; 1308 } 1309 cnt += (bs-1)*rlen; 1310 } 1311 if (cnt != nz) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Internal PETSc error: cnt = %D nz = %D",cnt,nz); 1312 1313 /* store the column values to the file */ 1314 ierr = PetscViewerFlowControlStart(viewer,&message_count,&flowcontrolcount);CHKERRQ(ierr); 1315 if (!rank) { 1316 MPI_Status status; 1317 ierr = PetscBinaryWrite(fd,column_values,nz,PETSC_SCALAR,PETSC_TRUE);CHKERRQ(ierr); 1318 for (i=1; i<size; i++) { 1319 ierr = PetscViewerFlowControlStepMaster(viewer,i,&message_count,flowcontrolcount);CHKERRQ(ierr); 1320 ierr = MPI_Recv(&cnt,1,MPIU_INT,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1321 ierr = MPI_Recv(column_values,cnt,MPIU_SCALAR,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1322 ierr = PetscBinaryWrite(fd,column_values,cnt,PETSC_SCALAR,PETSC_TRUE);CHKERRQ(ierr); 1323 } 1324 ierr = PetscViewerFlowControlEndMaster(viewer,&message_count);CHKERRQ(ierr); 1325 } else { 1326 ierr = PetscViewerFlowControlStepWorker(viewer,rank,&message_count);CHKERRQ(ierr); 1327 ierr = MPI_Send(&nz,1,MPIU_INT,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1328 ierr = MPI_Send(column_values,nz,MPIU_SCALAR,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1329 ierr = PetscViewerFlowControlEndWorker(viewer,&message_count);CHKERRQ(ierr); 1330 } 1331 ierr = PetscFree(column_values);CHKERRQ(ierr); 1332 1333 ierr = PetscViewerBinaryGetInfoPointer(viewer,&file);CHKERRQ(ierr); 1334 if (file) { 1335 fprintf(file,"-matload_block_size %d\n",(int)mat->rmap->bs); 1336 } 1337 PetscFunctionReturn(0); 1338 } 1339 1340 #undef __FUNCT__ 1341 #define __FUNCT__ "MatView_MPIBAIJ" 1342 PetscErrorCode MatView_MPIBAIJ(Mat mat,PetscViewer viewer) 1343 { 1344 PetscErrorCode ierr; 1345 PetscBool iascii,isdraw,issocket,isbinary; 1346 1347 PetscFunctionBegin; 1348 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&iascii);CHKERRQ(ierr); 1349 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERDRAW,&isdraw);CHKERRQ(ierr); 1350 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERSOCKET,&issocket);CHKERRQ(ierr); 1351 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERBINARY,&isbinary);CHKERRQ(ierr); 1352 if (iascii || isdraw || issocket) { 1353 ierr = MatView_MPIBAIJ_ASCIIorDraworSocket(mat,viewer);CHKERRQ(ierr); 1354 } else if (isbinary) { 1355 ierr = MatView_MPIBAIJ_Binary(mat,viewer);CHKERRQ(ierr); 1356 } 1357 PetscFunctionReturn(0); 1358 } 1359 1360 #undef __FUNCT__ 1361 #define __FUNCT__ "MatDestroy_MPIBAIJ" 1362 PetscErrorCode MatDestroy_MPIBAIJ(Mat mat) 1363 { 1364 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1365 PetscErrorCode ierr; 1366 1367 PetscFunctionBegin; 1368 #if defined(PETSC_USE_LOG) 1369 PetscLogObjectState((PetscObject)mat,"Rows=%D,Cols=%D",mat->rmap->N,mat->cmap->N); 1370 #endif 1371 ierr = MatStashDestroy_Private(&mat->stash);CHKERRQ(ierr); 1372 ierr = MatStashDestroy_Private(&mat->bstash);CHKERRQ(ierr); 1373 ierr = MatDestroy(&baij->A);CHKERRQ(ierr); 1374 ierr = MatDestroy(&baij->B);CHKERRQ(ierr); 1375 #if defined(PETSC_USE_CTABLE) 1376 ierr = PetscTableDestroy(&baij->colmap);CHKERRQ(ierr); 1377 #else 1378 ierr = PetscFree(baij->colmap);CHKERRQ(ierr); 1379 #endif 1380 ierr = PetscFree(baij->garray);CHKERRQ(ierr); 1381 ierr = VecDestroy(&baij->lvec);CHKERRQ(ierr); 1382 ierr = VecScatterDestroy(&baij->Mvctx);CHKERRQ(ierr); 1383 ierr = PetscFree2(baij->rowvalues,baij->rowindices);CHKERRQ(ierr); 1384 ierr = PetscFree(baij->barray);CHKERRQ(ierr); 1385 ierr = PetscFree2(baij->hd,baij->ht);CHKERRQ(ierr); 1386 ierr = PetscFree(baij->rangebs);CHKERRQ(ierr); 1387 ierr = PetscFree(mat->data);CHKERRQ(ierr); 1388 1389 ierr = PetscObjectChangeTypeName((PetscObject)mat,0);CHKERRQ(ierr); 1390 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatStoreValues_C",NULL);CHKERRQ(ierr); 1391 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatRetrieveValues_C",NULL);CHKERRQ(ierr); 1392 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatGetDiagonalBlock_C",NULL);CHKERRQ(ierr); 1393 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatMPIBAIJSetPreallocation_C",NULL);CHKERRQ(ierr); 1394 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatMPIBAIJSetPreallocationCSR_C",NULL);CHKERRQ(ierr); 1395 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatDiagonalScaleLocal_C",NULL);CHKERRQ(ierr); 1396 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatSetHashTableFactor_C",NULL);CHKERRQ(ierr); 1397 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatConvert_mpibaij_mpisbaij_C",NULL);CHKERRQ(ierr); 1398 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatConvert_mpibaij_mpibstrm_C",NULL);CHKERRQ(ierr); 1399 PetscFunctionReturn(0); 1400 } 1401 1402 #undef __FUNCT__ 1403 #define __FUNCT__ "MatMult_MPIBAIJ" 1404 PetscErrorCode MatMult_MPIBAIJ(Mat A,Vec xx,Vec yy) 1405 { 1406 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1407 PetscErrorCode ierr; 1408 PetscInt nt; 1409 1410 PetscFunctionBegin; 1411 ierr = VecGetLocalSize(xx,&nt);CHKERRQ(ierr); 1412 if (nt != A->cmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx"); 1413 ierr = VecGetLocalSize(yy,&nt);CHKERRQ(ierr); 1414 if (nt != A->rmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Incompatible parition of A and yy"); 1415 ierr = VecScatterBegin(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1416 ierr = (*a->A->ops->mult)(a->A,xx,yy);CHKERRQ(ierr); 1417 ierr = VecScatterEnd(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1418 ierr = (*a->B->ops->multadd)(a->B,a->lvec,yy,yy);CHKERRQ(ierr); 1419 PetscFunctionReturn(0); 1420 } 1421 1422 #undef __FUNCT__ 1423 #define __FUNCT__ "MatMultAdd_MPIBAIJ" 1424 PetscErrorCode MatMultAdd_MPIBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 1425 { 1426 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1427 PetscErrorCode ierr; 1428 1429 PetscFunctionBegin; 1430 ierr = VecScatterBegin(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1431 ierr = (*a->A->ops->multadd)(a->A,xx,yy,zz);CHKERRQ(ierr); 1432 ierr = VecScatterEnd(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1433 ierr = (*a->B->ops->multadd)(a->B,a->lvec,zz,zz);CHKERRQ(ierr); 1434 PetscFunctionReturn(0); 1435 } 1436 1437 #undef __FUNCT__ 1438 #define __FUNCT__ "MatMultTranspose_MPIBAIJ" 1439 PetscErrorCode MatMultTranspose_MPIBAIJ(Mat A,Vec xx,Vec yy) 1440 { 1441 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1442 PetscErrorCode ierr; 1443 PetscBool merged; 1444 1445 PetscFunctionBegin; 1446 ierr = VecScatterGetMerged(a->Mvctx,&merged);CHKERRQ(ierr); 1447 /* do nondiagonal part */ 1448 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 1449 if (!merged) { 1450 /* send it on its way */ 1451 ierr = VecScatterBegin(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1452 /* do local part */ 1453 ierr = (*a->A->ops->multtranspose)(a->A,xx,yy);CHKERRQ(ierr); 1454 /* receive remote parts: note this assumes the values are not actually */ 1455 /* inserted in yy until the next line */ 1456 ierr = VecScatterEnd(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1457 } else { 1458 /* do local part */ 1459 ierr = (*a->A->ops->multtranspose)(a->A,xx,yy);CHKERRQ(ierr); 1460 /* send it on its way */ 1461 ierr = VecScatterBegin(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1462 /* values actually were received in the Begin() but we need to call this nop */ 1463 ierr = VecScatterEnd(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1464 } 1465 PetscFunctionReturn(0); 1466 } 1467 1468 #undef __FUNCT__ 1469 #define __FUNCT__ "MatMultTransposeAdd_MPIBAIJ" 1470 PetscErrorCode MatMultTransposeAdd_MPIBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 1471 { 1472 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1473 PetscErrorCode ierr; 1474 1475 PetscFunctionBegin; 1476 /* do nondiagonal part */ 1477 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 1478 /* send it on its way */ 1479 ierr = VecScatterBegin(a->Mvctx,a->lvec,zz,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1480 /* do local part */ 1481 ierr = (*a->A->ops->multtransposeadd)(a->A,xx,yy,zz);CHKERRQ(ierr); 1482 /* receive remote parts: note this assumes the values are not actually */ 1483 /* inserted in yy until the next line, which is true for my implementation*/ 1484 /* but is not perhaps always true. */ 1485 ierr = VecScatterEnd(a->Mvctx,a->lvec,zz,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1486 PetscFunctionReturn(0); 1487 } 1488 1489 /* 1490 This only works correctly for square matrices where the subblock A->A is the 1491 diagonal block 1492 */ 1493 #undef __FUNCT__ 1494 #define __FUNCT__ "MatGetDiagonal_MPIBAIJ" 1495 PetscErrorCode MatGetDiagonal_MPIBAIJ(Mat A,Vec v) 1496 { 1497 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1498 PetscErrorCode ierr; 1499 1500 PetscFunctionBegin; 1501 if (A->rmap->N != A->cmap->N) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Supports only square matrix where A->A is diag block"); 1502 ierr = MatGetDiagonal(a->A,v);CHKERRQ(ierr); 1503 PetscFunctionReturn(0); 1504 } 1505 1506 #undef __FUNCT__ 1507 #define __FUNCT__ "MatScale_MPIBAIJ" 1508 PetscErrorCode MatScale_MPIBAIJ(Mat A,PetscScalar aa) 1509 { 1510 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1511 PetscErrorCode ierr; 1512 1513 PetscFunctionBegin; 1514 ierr = MatScale(a->A,aa);CHKERRQ(ierr); 1515 ierr = MatScale(a->B,aa);CHKERRQ(ierr); 1516 PetscFunctionReturn(0); 1517 } 1518 1519 #undef __FUNCT__ 1520 #define __FUNCT__ "MatGetRow_MPIBAIJ" 1521 PetscErrorCode MatGetRow_MPIBAIJ(Mat matin,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v) 1522 { 1523 Mat_MPIBAIJ *mat = (Mat_MPIBAIJ*)matin->data; 1524 PetscScalar *vworkA,*vworkB,**pvA,**pvB,*v_p; 1525 PetscErrorCode ierr; 1526 PetscInt bs = matin->rmap->bs,bs2 = mat->bs2,i,*cworkA,*cworkB,**pcA,**pcB; 1527 PetscInt nztot,nzA,nzB,lrow,brstart = matin->rmap->rstart,brend = matin->rmap->rend; 1528 PetscInt *cmap,*idx_p,cstart = mat->cstartbs; 1529 1530 PetscFunctionBegin; 1531 if (row < brstart || row >= brend) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only local rows"); 1532 if (mat->getrowactive) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Already active"); 1533 mat->getrowactive = PETSC_TRUE; 1534 1535 if (!mat->rowvalues && (idx || v)) { 1536 /* 1537 allocate enough space to hold information from the longest row. 1538 */ 1539 Mat_SeqBAIJ *Aa = (Mat_SeqBAIJ*)mat->A->data,*Ba = (Mat_SeqBAIJ*)mat->B->data; 1540 PetscInt max = 1,mbs = mat->mbs,tmp; 1541 for (i=0; i<mbs; i++) { 1542 tmp = Aa->i[i+1] - Aa->i[i] + Ba->i[i+1] - Ba->i[i]; 1543 if (max < tmp) max = tmp; 1544 } 1545 ierr = PetscMalloc2(max*bs2,&mat->rowvalues,max*bs2,&mat->rowindices);CHKERRQ(ierr); 1546 } 1547 lrow = row - brstart; 1548 1549 pvA = &vworkA; pcA = &cworkA; pvB = &vworkB; pcB = &cworkB; 1550 if (!v) {pvA = 0; pvB = 0;} 1551 if (!idx) {pcA = 0; if (!v) pcB = 0;} 1552 ierr = (*mat->A->ops->getrow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1553 ierr = (*mat->B->ops->getrow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1554 nztot = nzA + nzB; 1555 1556 cmap = mat->garray; 1557 if (v || idx) { 1558 if (nztot) { 1559 /* Sort by increasing column numbers, assuming A and B already sorted */ 1560 PetscInt imark = -1; 1561 if (v) { 1562 *v = v_p = mat->rowvalues; 1563 for (i=0; i<nzB; i++) { 1564 if (cmap[cworkB[i]/bs] < cstart) v_p[i] = vworkB[i]; 1565 else break; 1566 } 1567 imark = i; 1568 for (i=0; i<nzA; i++) v_p[imark+i] = vworkA[i]; 1569 for (i=imark; i<nzB; i++) v_p[nzA+i] = vworkB[i]; 1570 } 1571 if (idx) { 1572 *idx = idx_p = mat->rowindices; 1573 if (imark > -1) { 1574 for (i=0; i<imark; i++) { 1575 idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs; 1576 } 1577 } else { 1578 for (i=0; i<nzB; i++) { 1579 if (cmap[cworkB[i]/bs] < cstart) idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs; 1580 else break; 1581 } 1582 imark = i; 1583 } 1584 for (i=0; i<nzA; i++) idx_p[imark+i] = cstart*bs + cworkA[i]; 1585 for (i=imark; i<nzB; i++) idx_p[nzA+i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ; 1586 } 1587 } else { 1588 if (idx) *idx = 0; 1589 if (v) *v = 0; 1590 } 1591 } 1592 *nz = nztot; 1593 ierr = (*mat->A->ops->restorerow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1594 ierr = (*mat->B->ops->restorerow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1595 PetscFunctionReturn(0); 1596 } 1597 1598 #undef __FUNCT__ 1599 #define __FUNCT__ "MatRestoreRow_MPIBAIJ" 1600 PetscErrorCode MatRestoreRow_MPIBAIJ(Mat mat,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v) 1601 { 1602 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1603 1604 PetscFunctionBegin; 1605 if (!baij->getrowactive) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"MatGetRow not called"); 1606 baij->getrowactive = PETSC_FALSE; 1607 PetscFunctionReturn(0); 1608 } 1609 1610 #undef __FUNCT__ 1611 #define __FUNCT__ "MatZeroEntries_MPIBAIJ" 1612 PetscErrorCode MatZeroEntries_MPIBAIJ(Mat A) 1613 { 1614 Mat_MPIBAIJ *l = (Mat_MPIBAIJ*)A->data; 1615 PetscErrorCode ierr; 1616 1617 PetscFunctionBegin; 1618 ierr = MatZeroEntries(l->A);CHKERRQ(ierr); 1619 ierr = MatZeroEntries(l->B);CHKERRQ(ierr); 1620 PetscFunctionReturn(0); 1621 } 1622 1623 #undef __FUNCT__ 1624 #define __FUNCT__ "MatGetInfo_MPIBAIJ" 1625 PetscErrorCode MatGetInfo_MPIBAIJ(Mat matin,MatInfoType flag,MatInfo *info) 1626 { 1627 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)matin->data; 1628 Mat A = a->A,B = a->B; 1629 PetscErrorCode ierr; 1630 PetscReal isend[5],irecv[5]; 1631 1632 PetscFunctionBegin; 1633 info->block_size = (PetscReal)matin->rmap->bs; 1634 1635 ierr = MatGetInfo(A,MAT_LOCAL,info);CHKERRQ(ierr); 1636 1637 isend[0] = info->nz_used; isend[1] = info->nz_allocated; isend[2] = info->nz_unneeded; 1638 isend[3] = info->memory; isend[4] = info->mallocs; 1639 1640 ierr = MatGetInfo(B,MAT_LOCAL,info);CHKERRQ(ierr); 1641 1642 isend[0] += info->nz_used; isend[1] += info->nz_allocated; isend[2] += info->nz_unneeded; 1643 isend[3] += info->memory; isend[4] += info->mallocs; 1644 1645 if (flag == MAT_LOCAL) { 1646 info->nz_used = isend[0]; 1647 info->nz_allocated = isend[1]; 1648 info->nz_unneeded = isend[2]; 1649 info->memory = isend[3]; 1650 info->mallocs = isend[4]; 1651 } else if (flag == MAT_GLOBAL_MAX) { 1652 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPIU_MAX,PetscObjectComm((PetscObject)matin));CHKERRQ(ierr); 1653 1654 info->nz_used = irecv[0]; 1655 info->nz_allocated = irecv[1]; 1656 info->nz_unneeded = irecv[2]; 1657 info->memory = irecv[3]; 1658 info->mallocs = irecv[4]; 1659 } else if (flag == MAT_GLOBAL_SUM) { 1660 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPIU_SUM,PetscObjectComm((PetscObject)matin));CHKERRQ(ierr); 1661 1662 info->nz_used = irecv[0]; 1663 info->nz_allocated = irecv[1]; 1664 info->nz_unneeded = irecv[2]; 1665 info->memory = irecv[3]; 1666 info->mallocs = irecv[4]; 1667 } else SETERRQ1(PetscObjectComm((PetscObject)matin),PETSC_ERR_ARG_WRONG,"Unknown MatInfoType argument %d",(int)flag); 1668 info->fill_ratio_given = 0; /* no parallel LU/ILU/Cholesky */ 1669 info->fill_ratio_needed = 0; 1670 info->factor_mallocs = 0; 1671 PetscFunctionReturn(0); 1672 } 1673 1674 #undef __FUNCT__ 1675 #define __FUNCT__ "MatSetOption_MPIBAIJ" 1676 PetscErrorCode MatSetOption_MPIBAIJ(Mat A,MatOption op,PetscBool flg) 1677 { 1678 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1679 PetscErrorCode ierr; 1680 1681 PetscFunctionBegin; 1682 switch (op) { 1683 case MAT_NEW_NONZERO_LOCATIONS: 1684 case MAT_NEW_NONZERO_ALLOCATION_ERR: 1685 case MAT_UNUSED_NONZERO_LOCATION_ERR: 1686 case MAT_KEEP_NONZERO_PATTERN: 1687 case MAT_NEW_NONZERO_LOCATION_ERR: 1688 ierr = MatSetOption(a->A,op,flg);CHKERRQ(ierr); 1689 ierr = MatSetOption(a->B,op,flg);CHKERRQ(ierr); 1690 break; 1691 case MAT_ROW_ORIENTED: 1692 a->roworiented = flg; 1693 1694 ierr = MatSetOption(a->A,op,flg);CHKERRQ(ierr); 1695 ierr = MatSetOption(a->B,op,flg);CHKERRQ(ierr); 1696 break; 1697 case MAT_NEW_DIAGONALS: 1698 ierr = PetscInfo1(A,"Option %s ignored\n",MatOptions[op]);CHKERRQ(ierr); 1699 break; 1700 case MAT_IGNORE_OFF_PROC_ENTRIES: 1701 a->donotstash = flg; 1702 break; 1703 case MAT_USE_HASH_TABLE: 1704 a->ht_flag = flg; 1705 break; 1706 case MAT_SYMMETRIC: 1707 case MAT_STRUCTURALLY_SYMMETRIC: 1708 case MAT_HERMITIAN: 1709 case MAT_SYMMETRY_ETERNAL: 1710 ierr = MatSetOption(a->A,op,flg);CHKERRQ(ierr); 1711 break; 1712 default: 1713 SETERRQ1(PetscObjectComm((PetscObject)A),PETSC_ERR_SUP,"unknown option %d",op); 1714 } 1715 PetscFunctionReturn(0); 1716 } 1717 1718 #undef __FUNCT__ 1719 #define __FUNCT__ "MatTranspose_MPIBAIJ" 1720 PetscErrorCode MatTranspose_MPIBAIJ(Mat A,MatReuse reuse,Mat *matout) 1721 { 1722 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)A->data; 1723 Mat_SeqBAIJ *Aloc; 1724 Mat B; 1725 PetscErrorCode ierr; 1726 PetscInt M =A->rmap->N,N=A->cmap->N,*ai,*aj,i,*rvals,j,k,col; 1727 PetscInt bs=A->rmap->bs,mbs=baij->mbs; 1728 MatScalar *a; 1729 1730 PetscFunctionBegin; 1731 if (reuse == MAT_REUSE_MATRIX && A == *matout && M != N) SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_SIZ,"Square matrix only for in-place"); 1732 if (reuse == MAT_INITIAL_MATRIX || *matout == A) { 1733 ierr = MatCreate(PetscObjectComm((PetscObject)A),&B);CHKERRQ(ierr); 1734 ierr = MatSetSizes(B,A->cmap->n,A->rmap->n,N,M);CHKERRQ(ierr); 1735 ierr = MatSetType(B,((PetscObject)A)->type_name);CHKERRQ(ierr); 1736 /* Do not know preallocation information, but must set block size */ 1737 ierr = MatMPIBAIJSetPreallocation(B,A->rmap->bs,PETSC_DECIDE,NULL,PETSC_DECIDE,NULL);CHKERRQ(ierr); 1738 } else { 1739 B = *matout; 1740 } 1741 1742 /* copy over the A part */ 1743 Aloc = (Mat_SeqBAIJ*)baij->A->data; 1744 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1745 ierr = PetscMalloc1(bs,&rvals);CHKERRQ(ierr); 1746 1747 for (i=0; i<mbs; i++) { 1748 rvals[0] = bs*(baij->rstartbs + i); 1749 for (j=1; j<bs; j++) rvals[j] = rvals[j-1] + 1; 1750 for (j=ai[i]; j<ai[i+1]; j++) { 1751 col = (baij->cstartbs+aj[j])*bs; 1752 for (k=0; k<bs; k++) { 1753 ierr = MatSetValues_MPIBAIJ(B,1,&col,bs,rvals,a,INSERT_VALUES);CHKERRQ(ierr); 1754 1755 col++; a += bs; 1756 } 1757 } 1758 } 1759 /* copy over the B part */ 1760 Aloc = (Mat_SeqBAIJ*)baij->B->data; 1761 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1762 for (i=0; i<mbs; i++) { 1763 rvals[0] = bs*(baij->rstartbs + i); 1764 for (j=1; j<bs; j++) rvals[j] = rvals[j-1] + 1; 1765 for (j=ai[i]; j<ai[i+1]; j++) { 1766 col = baij->garray[aj[j]]*bs; 1767 for (k=0; k<bs; k++) { 1768 ierr = MatSetValues_MPIBAIJ(B,1,&col,bs,rvals,a,INSERT_VALUES);CHKERRQ(ierr); 1769 col++; 1770 a += bs; 1771 } 1772 } 1773 } 1774 ierr = PetscFree(rvals);CHKERRQ(ierr); 1775 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1776 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1777 1778 if (reuse == MAT_INITIAL_MATRIX || *matout != A) *matout = B; 1779 else { 1780 ierr = MatHeaderMerge(A,B);CHKERRQ(ierr); 1781 } 1782 PetscFunctionReturn(0); 1783 } 1784 1785 #undef __FUNCT__ 1786 #define __FUNCT__ "MatDiagonalScale_MPIBAIJ" 1787 PetscErrorCode MatDiagonalScale_MPIBAIJ(Mat mat,Vec ll,Vec rr) 1788 { 1789 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1790 Mat a = baij->A,b = baij->B; 1791 PetscErrorCode ierr; 1792 PetscInt s1,s2,s3; 1793 1794 PetscFunctionBegin; 1795 ierr = MatGetLocalSize(mat,&s2,&s3);CHKERRQ(ierr); 1796 if (rr) { 1797 ierr = VecGetLocalSize(rr,&s1);CHKERRQ(ierr); 1798 if (s1!=s3) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"right vector non-conforming local size"); 1799 /* Overlap communication with computation. */ 1800 ierr = VecScatterBegin(baij->Mvctx,rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1801 } 1802 if (ll) { 1803 ierr = VecGetLocalSize(ll,&s1);CHKERRQ(ierr); 1804 if (s1!=s2) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"left vector non-conforming local size"); 1805 ierr = (*b->ops->diagonalscale)(b,ll,NULL);CHKERRQ(ierr); 1806 } 1807 /* scale the diagonal block */ 1808 ierr = (*a->ops->diagonalscale)(a,ll,rr);CHKERRQ(ierr); 1809 1810 if (rr) { 1811 /* Do a scatter end and then right scale the off-diagonal block */ 1812 ierr = VecScatterEnd(baij->Mvctx,rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1813 ierr = (*b->ops->diagonalscale)(b,NULL,baij->lvec);CHKERRQ(ierr); 1814 } 1815 PetscFunctionReturn(0); 1816 } 1817 1818 #undef __FUNCT__ 1819 #define __FUNCT__ "MatZeroRows_MPIBAIJ" 1820 PetscErrorCode MatZeroRows_MPIBAIJ(Mat A,PetscInt N,const PetscInt rows[],PetscScalar diag,Vec x,Vec b) 1821 { 1822 Mat_MPIBAIJ *l = (Mat_MPIBAIJ *) A->data; 1823 PetscInt *owners = A->rmap->range; 1824 PetscInt n = A->rmap->n; 1825 PetscSF sf; 1826 PetscInt *lrows; 1827 PetscSFNode *rrows; 1828 PetscInt r, p = 0, len = 0; 1829 PetscErrorCode ierr; 1830 1831 PetscFunctionBegin; 1832 /* Create SF where leaves are input rows and roots are owned rows */ 1833 ierr = PetscMalloc1(n, &lrows);CHKERRQ(ierr); 1834 for (r = 0; r < n; ++r) lrows[r] = -1; 1835 if (!A->nooffproczerorows) {ierr = PetscMalloc1(N, &rrows);CHKERRQ(ierr);} 1836 for (r = 0; r < N; ++r) { 1837 const PetscInt idx = rows[r]; 1838 if (idx < 0 || A->rmap->N <= idx) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row %D out of range [0,%D)",idx,A->rmap->N); 1839 if (idx < owners[p] || owners[p+1] <= idx) { /* short-circuit the search if the last p owns this row too */ 1840 ierr = PetscLayoutFindOwner(A->rmap,idx,&p);CHKERRQ(ierr); 1841 } 1842 if (A->nooffproczerorows) { 1843 if (p != l->rank) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"MAT_NO_OFF_PROC_ZERO_ROWS set, but row %D is not owned by rank %d",idx,l->rank); 1844 lrows[len++] = idx - owners[p]; 1845 } else { 1846 rrows[r].rank = p; 1847 rrows[r].index = rows[r] - owners[p]; 1848 } 1849 } 1850 if (!A->nooffproczerorows) { 1851 ierr = PetscSFCreate(PetscObjectComm((PetscObject) A), &sf);CHKERRQ(ierr); 1852 ierr = PetscSFSetGraph(sf, n, N, NULL, PETSC_OWN_POINTER, rrows, PETSC_OWN_POINTER);CHKERRQ(ierr); 1853 /* Collect flags for rows to be zeroed */ 1854 ierr = PetscSFReduceBegin(sf, MPIU_INT, (PetscInt *) rows, lrows, MPI_LOR);CHKERRQ(ierr); 1855 ierr = PetscSFReduceEnd(sf, MPIU_INT, (PetscInt *) rows, lrows, MPI_LOR);CHKERRQ(ierr); 1856 ierr = PetscSFDestroy(&sf);CHKERRQ(ierr); 1857 /* Compress and put in row numbers */ 1858 for (r = 0; r < n; ++r) if (lrows[r] >= 0) lrows[len++] = r; 1859 } 1860 /* fix right hand side if needed */ 1861 if (x && b) { 1862 const PetscScalar *xx; 1863 PetscScalar *bb; 1864 1865 ierr = VecGetArrayRead(x,&xx);CHKERRQ(ierr); 1866 ierr = VecGetArray(b,&bb);CHKERRQ(ierr); 1867 for (r = 0; r < len; ++r) bb[lrows[r]] = diag*xx[lrows[r]]; 1868 ierr = VecRestoreArrayRead(x,&xx);CHKERRQ(ierr); 1869 ierr = VecRestoreArray(b,&bb);CHKERRQ(ierr); 1870 } 1871 1872 /* actually zap the local rows */ 1873 /* 1874 Zero the required rows. If the "diagonal block" of the matrix 1875 is square and the user wishes to set the diagonal we use separate 1876 code so that MatSetValues() is not called for each diagonal allocating 1877 new memory, thus calling lots of mallocs and slowing things down. 1878 1879 */ 1880 /* must zero l->B before l->A because the (diag) case below may put values into l->B*/ 1881 ierr = MatZeroRows_SeqBAIJ(l->B,len,lrows,0.0,NULL,NULL);CHKERRQ(ierr); 1882 if ((diag != 0.0) && (l->A->rmap->N == l->A->cmap->N)) { 1883 ierr = MatZeroRows_SeqBAIJ(l->A,len,lrows,diag,NULL,NULL);CHKERRQ(ierr); 1884 } else if (diag != 0.0) { 1885 ierr = MatZeroRows_SeqBAIJ(l->A,len,lrows,0.0,0,0);CHKERRQ(ierr); 1886 if (((Mat_SeqBAIJ*)l->A->data)->nonew) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"MatZeroRows() on rectangular matrices cannot be used with the Mat options \n\ 1887 MAT_NEW_NONZERO_LOCATIONS,MAT_NEW_NONZERO_LOCATION_ERR,MAT_NEW_NONZERO_ALLOCATION_ERR"); 1888 for (r = 0; r < len; ++r) { 1889 const PetscInt row = lrows[r] + A->rmap->rstart; 1890 ierr = MatSetValues(A,1,&row,1,&row,&diag,INSERT_VALUES);CHKERRQ(ierr); 1891 } 1892 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1893 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1894 } else { 1895 ierr = MatZeroRows_SeqBAIJ(l->A,len,lrows,0.0,NULL,NULL);CHKERRQ(ierr); 1896 } 1897 ierr = PetscFree(lrows);CHKERRQ(ierr); 1898 1899 /* only change matrix nonzero state if pattern was allowed to be changed */ 1900 if (!((Mat_SeqBAIJ*)(l->A->data))->keepnonzeropattern) { 1901 PetscObjectState state = l->A->nonzerostate + l->B->nonzerostate; 1902 ierr = MPI_Allreduce(&state,&A->nonzerostate,1,MPIU_INT64,MPI_SUM,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 1903 } 1904 PetscFunctionReturn(0); 1905 } 1906 1907 #undef __FUNCT__ 1908 #define __FUNCT__ "MatZeroRowsColumns_MPIBAIJ" 1909 PetscErrorCode MatZeroRowsColumns_MPIBAIJ(Mat A,PetscInt N,const PetscInt rows[],PetscScalar diag,Vec x,Vec b) 1910 { 1911 Mat_MPIBAIJ *l = (Mat_MPIBAIJ*)A->data; 1912 PetscErrorCode ierr; 1913 PetscMPIInt n = A->rmap->n; 1914 PetscInt i,j,k,r,p = 0,len = 0,row,col,count; 1915 PetscInt *lrows,*owners = A->rmap->range; 1916 PetscSFNode *rrows; 1917 PetscSF sf; 1918 const PetscScalar *xx; 1919 PetscScalar *bb,*mask; 1920 Vec xmask,lmask; 1921 Mat_SeqBAIJ *baij = (Mat_SeqBAIJ*)l->B->data; 1922 PetscInt bs = A->rmap->bs, bs2 = baij->bs2; 1923 PetscScalar *aa; 1924 1925 PetscFunctionBegin; 1926 /* Create SF where leaves are input rows and roots are owned rows */ 1927 ierr = PetscMalloc1(n, &lrows);CHKERRQ(ierr); 1928 for (r = 0; r < n; ++r) lrows[r] = -1; 1929 ierr = PetscMalloc1(N, &rrows);CHKERRQ(ierr); 1930 for (r = 0; r < N; ++r) { 1931 const PetscInt idx = rows[r]; 1932 if (idx < 0 || A->rmap->N <= idx) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row %D out of range [0,%D)",idx,A->rmap->N); 1933 if (idx < owners[p] || owners[p+1] <= idx) { /* short-circuit the search if the last p owns this row too */ 1934 ierr = PetscLayoutFindOwner(A->rmap,idx,&p);CHKERRQ(ierr); 1935 } 1936 rrows[r].rank = p; 1937 rrows[r].index = rows[r] - owners[p]; 1938 } 1939 ierr = PetscSFCreate(PetscObjectComm((PetscObject) A), &sf);CHKERRQ(ierr); 1940 ierr = PetscSFSetGraph(sf, n, N, NULL, PETSC_OWN_POINTER, rrows, PETSC_OWN_POINTER);CHKERRQ(ierr); 1941 /* Collect flags for rows to be zeroed */ 1942 ierr = PetscSFReduceBegin(sf, MPIU_INT, (PetscInt *) rows, lrows, MPI_LOR);CHKERRQ(ierr); 1943 ierr = PetscSFReduceEnd(sf, MPIU_INT, (PetscInt *) rows, lrows, MPI_LOR);CHKERRQ(ierr); 1944 ierr = PetscSFDestroy(&sf);CHKERRQ(ierr); 1945 /* Compress and put in row numbers */ 1946 for (r = 0; r < n; ++r) if (lrows[r] >= 0) lrows[len++] = r; 1947 /* zero diagonal part of matrix */ 1948 ierr = MatZeroRowsColumns(l->A,len,lrows,diag,x,b);CHKERRQ(ierr); 1949 /* handle off diagonal part of matrix */ 1950 ierr = MatCreateVecs(A,&xmask,NULL);CHKERRQ(ierr); 1951 ierr = VecDuplicate(l->lvec,&lmask);CHKERRQ(ierr); 1952 ierr = VecGetArray(xmask,&bb);CHKERRQ(ierr); 1953 for (i=0; i<len; i++) bb[lrows[i]] = 1; 1954 ierr = VecRestoreArray(xmask,&bb);CHKERRQ(ierr); 1955 ierr = VecScatterBegin(l->Mvctx,xmask,lmask,ADD_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1956 ierr = VecScatterEnd(l->Mvctx,xmask,lmask,ADD_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1957 ierr = VecDestroy(&xmask);CHKERRQ(ierr); 1958 if (x) { 1959 ierr = VecScatterBegin(l->Mvctx,x,l->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1960 ierr = VecScatterEnd(l->Mvctx,x,l->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1961 ierr = VecGetArrayRead(l->lvec,&xx);CHKERRQ(ierr); 1962 ierr = VecGetArray(b,&bb);CHKERRQ(ierr); 1963 } 1964 ierr = VecGetArray(lmask,&mask);CHKERRQ(ierr); 1965 /* remove zeroed rows of off diagonal matrix */ 1966 for (i = 0; i < len; ++i) { 1967 row = lrows[i]; 1968 count = (baij->i[row/bs +1] - baij->i[row/bs])*bs; 1969 aa = ((MatScalar*)(baij->a)) + baij->i[row/bs]*bs2 + (row%bs); 1970 for (k = 0; k < count; ++k) { 1971 aa[0] = 0.0; 1972 aa += bs; 1973 } 1974 } 1975 /* loop over all elements of off process part of matrix zeroing removed columns*/ 1976 for (i = 0; i < l->B->rmap->N; ++i) { 1977 row = i/bs; 1978 for (j = baij->i[row]; j < baij->i[row+1]; ++j) { 1979 for (k = 0; k < bs; ++k) { 1980 col = bs*baij->j[j] + k; 1981 if (PetscAbsScalar(mask[col])) { 1982 aa = ((MatScalar*)(baij->a)) + j*bs2 + (i%bs) + bs*k; 1983 if (b) bb[i] -= aa[0]*xx[col]; 1984 aa[0] = 0.0; 1985 } 1986 } 1987 } 1988 } 1989 if (x) { 1990 ierr = VecRestoreArray(b,&bb);CHKERRQ(ierr); 1991 ierr = VecRestoreArrayRead(l->lvec,&xx);CHKERRQ(ierr); 1992 } 1993 ierr = VecRestoreArray(lmask,&mask);CHKERRQ(ierr); 1994 ierr = VecDestroy(&lmask);CHKERRQ(ierr); 1995 ierr = PetscFree(lrows);CHKERRQ(ierr); 1996 1997 /* only change matrix nonzero state if pattern was allowed to be changed */ 1998 if (!((Mat_SeqBAIJ*)(l->A->data))->keepnonzeropattern) { 1999 PetscObjectState state = l->A->nonzerostate + l->B->nonzerostate; 2000 ierr = MPI_Allreduce(&state,&A->nonzerostate,1,MPIU_INT64,MPI_SUM,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2001 } 2002 PetscFunctionReturn(0); 2003 } 2004 2005 #undef __FUNCT__ 2006 #define __FUNCT__ "MatSetUnfactored_MPIBAIJ" 2007 PetscErrorCode MatSetUnfactored_MPIBAIJ(Mat A) 2008 { 2009 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 2010 PetscErrorCode ierr; 2011 2012 PetscFunctionBegin; 2013 ierr = MatSetUnfactored(a->A);CHKERRQ(ierr); 2014 PetscFunctionReturn(0); 2015 } 2016 2017 static PetscErrorCode MatDuplicate_MPIBAIJ(Mat,MatDuplicateOption,Mat*); 2018 2019 #undef __FUNCT__ 2020 #define __FUNCT__ "MatEqual_MPIBAIJ" 2021 PetscErrorCode MatEqual_MPIBAIJ(Mat A,Mat B,PetscBool *flag) 2022 { 2023 Mat_MPIBAIJ *matB = (Mat_MPIBAIJ*)B->data,*matA = (Mat_MPIBAIJ*)A->data; 2024 Mat a,b,c,d; 2025 PetscBool flg; 2026 PetscErrorCode ierr; 2027 2028 PetscFunctionBegin; 2029 a = matA->A; b = matA->B; 2030 c = matB->A; d = matB->B; 2031 2032 ierr = MatEqual(a,c,&flg);CHKERRQ(ierr); 2033 if (flg) { 2034 ierr = MatEqual(b,d,&flg);CHKERRQ(ierr); 2035 } 2036 ierr = MPI_Allreduce(&flg,flag,1,MPIU_BOOL,MPI_LAND,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2037 PetscFunctionReturn(0); 2038 } 2039 2040 #undef __FUNCT__ 2041 #define __FUNCT__ "MatCopy_MPIBAIJ" 2042 PetscErrorCode MatCopy_MPIBAIJ(Mat A,Mat B,MatStructure str) 2043 { 2044 PetscErrorCode ierr; 2045 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 2046 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)B->data; 2047 2048 PetscFunctionBegin; 2049 /* If the two matrices don't have the same copy implementation, they aren't compatible for fast copy. */ 2050 if ((str != SAME_NONZERO_PATTERN) || (A->ops->copy != B->ops->copy)) { 2051 ierr = MatCopy_Basic(A,B,str);CHKERRQ(ierr); 2052 } else { 2053 ierr = MatCopy(a->A,b->A,str);CHKERRQ(ierr); 2054 ierr = MatCopy(a->B,b->B,str);CHKERRQ(ierr); 2055 } 2056 PetscFunctionReturn(0); 2057 } 2058 2059 #undef __FUNCT__ 2060 #define __FUNCT__ "MatSetUp_MPIBAIJ" 2061 PetscErrorCode MatSetUp_MPIBAIJ(Mat A) 2062 { 2063 PetscErrorCode ierr; 2064 2065 PetscFunctionBegin; 2066 ierr = MatMPIBAIJSetPreallocation(A,A->rmap->bs,PETSC_DEFAULT,0,PETSC_DEFAULT,0);CHKERRQ(ierr); 2067 PetscFunctionReturn(0); 2068 } 2069 2070 #undef __FUNCT__ 2071 #define __FUNCT__ "MatAXPYGetPreallocation_MPIBAIJ" 2072 PetscErrorCode MatAXPYGetPreallocation_MPIBAIJ(Mat Y,const PetscInt *yltog,Mat X,const PetscInt *xltog,PetscInt *nnz) 2073 { 2074 PetscErrorCode ierr; 2075 PetscInt bs = Y->rmap->bs,m = Y->rmap->N/bs; 2076 Mat_SeqBAIJ *x = (Mat_SeqBAIJ*)X->data; 2077 Mat_SeqBAIJ *y = (Mat_SeqBAIJ*)Y->data; 2078 2079 PetscFunctionBegin; 2080 ierr = MatAXPYGetPreallocation_MPIX_private(m,x->i,x->j,xltog,y->i,y->j,yltog,nnz);CHKERRQ(ierr); 2081 PetscFunctionReturn(0); 2082 } 2083 2084 #undef __FUNCT__ 2085 #define __FUNCT__ "MatAXPY_MPIBAIJ" 2086 PetscErrorCode MatAXPY_MPIBAIJ(Mat Y,PetscScalar a,Mat X,MatStructure str) 2087 { 2088 PetscErrorCode ierr; 2089 Mat_MPIBAIJ *xx=(Mat_MPIBAIJ*)X->data,*yy=(Mat_MPIBAIJ*)Y->data; 2090 PetscBLASInt bnz,one=1; 2091 Mat_SeqBAIJ *x,*y; 2092 2093 PetscFunctionBegin; 2094 if (str == SAME_NONZERO_PATTERN) { 2095 PetscScalar alpha = a; 2096 x = (Mat_SeqBAIJ*)xx->A->data; 2097 y = (Mat_SeqBAIJ*)yy->A->data; 2098 ierr = PetscBLASIntCast(x->nz,&bnz);CHKERRQ(ierr); 2099 PetscStackCallBLAS("BLASaxpy",BLASaxpy_(&bnz,&alpha,x->a,&one,y->a,&one)); 2100 x = (Mat_SeqBAIJ*)xx->B->data; 2101 y = (Mat_SeqBAIJ*)yy->B->data; 2102 ierr = PetscBLASIntCast(x->nz,&bnz);CHKERRQ(ierr); 2103 PetscStackCallBLAS("BLASaxpy",BLASaxpy_(&bnz,&alpha,x->a,&one,y->a,&one)); 2104 ierr = PetscObjectStateIncrease((PetscObject)Y);CHKERRQ(ierr); 2105 } else if (str == SUBSET_NONZERO_PATTERN) { /* nonzeros of X is a subset of Y's */ 2106 ierr = MatAXPY_Basic(Y,a,X,str);CHKERRQ(ierr); 2107 } else { 2108 Mat B; 2109 PetscInt *nnz_d,*nnz_o,bs=Y->rmap->bs; 2110 ierr = PetscMalloc1(yy->A->rmap->N,&nnz_d);CHKERRQ(ierr); 2111 ierr = PetscMalloc1(yy->B->rmap->N,&nnz_o);CHKERRQ(ierr); 2112 ierr = MatCreate(PetscObjectComm((PetscObject)Y),&B);CHKERRQ(ierr); 2113 ierr = PetscObjectSetName((PetscObject)B,((PetscObject)Y)->name);CHKERRQ(ierr); 2114 ierr = MatSetSizes(B,Y->rmap->n,Y->cmap->n,Y->rmap->N,Y->cmap->N);CHKERRQ(ierr); 2115 ierr = MatSetBlockSizesFromMats(B,Y,Y);CHKERRQ(ierr); 2116 ierr = MatSetType(B,MATMPIBAIJ);CHKERRQ(ierr); 2117 ierr = MatAXPYGetPreallocation_SeqBAIJ(yy->A,xx->A,nnz_d);CHKERRQ(ierr); 2118 ierr = MatAXPYGetPreallocation_MPIBAIJ(yy->B,yy->garray,xx->B,xx->garray,nnz_o);CHKERRQ(ierr); 2119 ierr = MatMPIBAIJSetPreallocation(B,bs,0,nnz_d,0,nnz_o);CHKERRQ(ierr); 2120 /* MatAXPY_BasicWithPreallocation() for BAIJ matrix is much slower than AIJ, even for bs=1 ! */ 2121 ierr = MatAXPY_BasicWithPreallocation(B,Y,a,X,str);CHKERRQ(ierr); 2122 ierr = MatHeaderReplace(Y,B);CHKERRQ(ierr); 2123 ierr = PetscFree(nnz_d);CHKERRQ(ierr); 2124 ierr = PetscFree(nnz_o);CHKERRQ(ierr); 2125 } 2126 PetscFunctionReturn(0); 2127 } 2128 2129 #undef __FUNCT__ 2130 #define __FUNCT__ "MatRealPart_MPIBAIJ" 2131 PetscErrorCode MatRealPart_MPIBAIJ(Mat A) 2132 { 2133 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 2134 PetscErrorCode ierr; 2135 2136 PetscFunctionBegin; 2137 ierr = MatRealPart(a->A);CHKERRQ(ierr); 2138 ierr = MatRealPart(a->B);CHKERRQ(ierr); 2139 PetscFunctionReturn(0); 2140 } 2141 2142 #undef __FUNCT__ 2143 #define __FUNCT__ "MatImaginaryPart_MPIBAIJ" 2144 PetscErrorCode MatImaginaryPart_MPIBAIJ(Mat A) 2145 { 2146 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 2147 PetscErrorCode ierr; 2148 2149 PetscFunctionBegin; 2150 ierr = MatImaginaryPart(a->A);CHKERRQ(ierr); 2151 ierr = MatImaginaryPart(a->B);CHKERRQ(ierr); 2152 PetscFunctionReturn(0); 2153 } 2154 2155 #undef __FUNCT__ 2156 #define __FUNCT__ "MatGetSubMatrix_MPIBAIJ" 2157 PetscErrorCode MatGetSubMatrix_MPIBAIJ(Mat mat,IS isrow,IS iscol,MatReuse call,Mat *newmat) 2158 { 2159 PetscErrorCode ierr; 2160 IS iscol_local; 2161 PetscInt csize; 2162 2163 PetscFunctionBegin; 2164 ierr = ISGetLocalSize(iscol,&csize);CHKERRQ(ierr); 2165 if (call == MAT_REUSE_MATRIX) { 2166 ierr = PetscObjectQuery((PetscObject)*newmat,"ISAllGather",(PetscObject*)&iscol_local);CHKERRQ(ierr); 2167 if (!iscol_local) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Submatrix passed in was not used before, cannot reuse"); 2168 } else { 2169 ierr = ISAllGather(iscol,&iscol_local);CHKERRQ(ierr); 2170 } 2171 ierr = MatGetSubMatrix_MPIBAIJ_Private(mat,isrow,iscol_local,csize,call,newmat);CHKERRQ(ierr); 2172 if (call == MAT_INITIAL_MATRIX) { 2173 ierr = PetscObjectCompose((PetscObject)*newmat,"ISAllGather",(PetscObject)iscol_local);CHKERRQ(ierr); 2174 ierr = ISDestroy(&iscol_local);CHKERRQ(ierr); 2175 } 2176 PetscFunctionReturn(0); 2177 } 2178 extern PetscErrorCode MatGetSubMatrices_MPIBAIJ_local(Mat,PetscInt,const IS[],const IS[],MatReuse,PetscBool*,PetscBool*,Mat*); 2179 #undef __FUNCT__ 2180 #define __FUNCT__ "MatGetSubMatrix_MPIBAIJ_Private" 2181 /* 2182 Not great since it makes two copies of the submatrix, first an SeqBAIJ 2183 in local and then by concatenating the local matrices the end result. 2184 Writing it directly would be much like MatGetSubMatrices_MPIBAIJ(). 2185 This routine is used for BAIJ and SBAIJ matrices (unfortunate dependency). 2186 */ 2187 PetscErrorCode MatGetSubMatrix_MPIBAIJ_Private(Mat mat,IS isrow,IS iscol,PetscInt csize,MatReuse call,Mat *newmat) 2188 { 2189 PetscErrorCode ierr; 2190 PetscMPIInt rank,size; 2191 PetscInt i,m,n,rstart,row,rend,nz,*cwork,j,bs; 2192 PetscInt *ii,*jj,nlocal,*dlens,*olens,dlen,olen,jend,mglobal,ncol,nrow; 2193 Mat M,Mreuse; 2194 MatScalar *vwork,*aa; 2195 MPI_Comm comm; 2196 IS isrow_new, iscol_new; 2197 PetscBool idflag,allrows, allcols; 2198 Mat_SeqBAIJ *aij; 2199 2200 PetscFunctionBegin; 2201 ierr = PetscObjectGetComm((PetscObject)mat,&comm);CHKERRQ(ierr); 2202 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 2203 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2204 /* The compression and expansion should be avoided. Doesn't point 2205 out errors, might change the indices, hence buggey */ 2206 ierr = ISCompressIndicesGeneral(mat->rmap->N,mat->rmap->n,mat->rmap->bs,1,&isrow,&isrow_new);CHKERRQ(ierr); 2207 ierr = ISCompressIndicesGeneral(mat->cmap->N,mat->cmap->n,mat->cmap->bs,1,&iscol,&iscol_new);CHKERRQ(ierr); 2208 2209 /* Check for special case: each processor gets entire matrix columns */ 2210 ierr = ISIdentity(iscol,&idflag);CHKERRQ(ierr); 2211 ierr = ISGetLocalSize(iscol,&ncol);CHKERRQ(ierr); 2212 if (idflag && ncol == mat->cmap->N) allcols = PETSC_TRUE; 2213 else allcols = PETSC_FALSE; 2214 2215 ierr = ISIdentity(isrow,&idflag);CHKERRQ(ierr); 2216 ierr = ISGetLocalSize(isrow,&nrow);CHKERRQ(ierr); 2217 if (idflag && nrow == mat->rmap->N) allrows = PETSC_TRUE; 2218 else allrows = PETSC_FALSE; 2219 2220 if (call == MAT_REUSE_MATRIX) { 2221 ierr = PetscObjectQuery((PetscObject)*newmat,"SubMatrix",(PetscObject*)&Mreuse);CHKERRQ(ierr); 2222 if (!Mreuse) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Submatrix passed in was not used before, cannot reuse"); 2223 ierr = MatGetSubMatrices_MPIBAIJ_local(mat,1,&isrow_new,&iscol_new,MAT_REUSE_MATRIX,&allrows,&allcols,&Mreuse);CHKERRQ(ierr); 2224 } else { 2225 ierr = MatGetSubMatrices_MPIBAIJ_local(mat,1,&isrow_new,&iscol_new,MAT_INITIAL_MATRIX,&allrows,&allcols,&Mreuse);CHKERRQ(ierr); 2226 } 2227 ierr = ISDestroy(&isrow_new);CHKERRQ(ierr); 2228 ierr = ISDestroy(&iscol_new);CHKERRQ(ierr); 2229 /* 2230 m - number of local rows 2231 n - number of columns (same on all processors) 2232 rstart - first row in new global matrix generated 2233 */ 2234 ierr = MatGetBlockSize(mat,&bs);CHKERRQ(ierr); 2235 ierr = MatGetSize(Mreuse,&m,&n);CHKERRQ(ierr); 2236 m = m/bs; 2237 n = n/bs; 2238 2239 if (call == MAT_INITIAL_MATRIX) { 2240 aij = (Mat_SeqBAIJ*)(Mreuse)->data; 2241 ii = aij->i; 2242 jj = aij->j; 2243 2244 /* 2245 Determine the number of non-zeros in the diagonal and off-diagonal 2246 portions of the matrix in order to do correct preallocation 2247 */ 2248 2249 /* first get start and end of "diagonal" columns */ 2250 if (csize == PETSC_DECIDE) { 2251 ierr = ISGetSize(isrow,&mglobal);CHKERRQ(ierr); 2252 if (mglobal == n*bs) { /* square matrix */ 2253 nlocal = m; 2254 } else { 2255 nlocal = n/size + ((n % size) > rank); 2256 } 2257 } else { 2258 nlocal = csize/bs; 2259 } 2260 ierr = MPI_Scan(&nlocal,&rend,1,MPIU_INT,MPI_SUM,comm);CHKERRQ(ierr); 2261 rstart = rend - nlocal; 2262 if (rank == size - 1 && rend != n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Local column sizes %D do not add up to total number of columns %D",rend,n); 2263 2264 /* next, compute all the lengths */ 2265 ierr = PetscMalloc2(m+1,&dlens,m+1,&olens);CHKERRQ(ierr); 2266 for (i=0; i<m; i++) { 2267 jend = ii[i+1] - ii[i]; 2268 olen = 0; 2269 dlen = 0; 2270 for (j=0; j<jend; j++) { 2271 if (*jj < rstart || *jj >= rend) olen++; 2272 else dlen++; 2273 jj++; 2274 } 2275 olens[i] = olen; 2276 dlens[i] = dlen; 2277 } 2278 ierr = MatCreate(comm,&M);CHKERRQ(ierr); 2279 ierr = MatSetSizes(M,bs*m,bs*nlocal,PETSC_DECIDE,bs*n);CHKERRQ(ierr); 2280 ierr = MatSetType(M,((PetscObject)mat)->type_name);CHKERRQ(ierr); 2281 ierr = MatMPIBAIJSetPreallocation(M,bs,0,dlens,0,olens);CHKERRQ(ierr); 2282 ierr = MatMPISBAIJSetPreallocation(M,bs,0,dlens,0,olens);CHKERRQ(ierr); 2283 ierr = PetscFree2(dlens,olens);CHKERRQ(ierr); 2284 } else { 2285 PetscInt ml,nl; 2286 2287 M = *newmat; 2288 ierr = MatGetLocalSize(M,&ml,&nl);CHKERRQ(ierr); 2289 if (ml != m) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Previous matrix must be same size/layout as request"); 2290 ierr = MatZeroEntries(M);CHKERRQ(ierr); 2291 /* 2292 The next two lines are needed so we may call MatSetValues_MPIAIJ() below directly, 2293 rather than the slower MatSetValues(). 2294 */ 2295 M->was_assembled = PETSC_TRUE; 2296 M->assembled = PETSC_FALSE; 2297 } 2298 ierr = MatSetOption(M,MAT_ROW_ORIENTED,PETSC_FALSE);CHKERRQ(ierr); 2299 ierr = MatGetOwnershipRange(M,&rstart,&rend);CHKERRQ(ierr); 2300 aij = (Mat_SeqBAIJ*)(Mreuse)->data; 2301 ii = aij->i; 2302 jj = aij->j; 2303 aa = aij->a; 2304 for (i=0; i<m; i++) { 2305 row = rstart/bs + i; 2306 nz = ii[i+1] - ii[i]; 2307 cwork = jj; jj += nz; 2308 vwork = aa; aa += nz*bs*bs; 2309 ierr = MatSetValuesBlocked_MPIBAIJ(M,1,&row,nz,cwork,vwork,INSERT_VALUES);CHKERRQ(ierr); 2310 } 2311 2312 ierr = MatAssemblyBegin(M,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2313 ierr = MatAssemblyEnd(M,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2314 *newmat = M; 2315 2316 /* save submatrix used in processor for next request */ 2317 if (call == MAT_INITIAL_MATRIX) { 2318 ierr = PetscObjectCompose((PetscObject)M,"SubMatrix",(PetscObject)Mreuse);CHKERRQ(ierr); 2319 ierr = PetscObjectDereference((PetscObject)Mreuse);CHKERRQ(ierr); 2320 } 2321 PetscFunctionReturn(0); 2322 } 2323 2324 #undef __FUNCT__ 2325 #define __FUNCT__ "MatPermute_MPIBAIJ" 2326 PetscErrorCode MatPermute_MPIBAIJ(Mat A,IS rowp,IS colp,Mat *B) 2327 { 2328 MPI_Comm comm,pcomm; 2329 PetscInt clocal_size,nrows; 2330 const PetscInt *rows; 2331 PetscMPIInt size; 2332 IS crowp,lcolp; 2333 PetscErrorCode ierr; 2334 2335 PetscFunctionBegin; 2336 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 2337 /* make a collective version of 'rowp' */ 2338 ierr = PetscObjectGetComm((PetscObject)rowp,&pcomm);CHKERRQ(ierr); 2339 if (pcomm==comm) { 2340 crowp = rowp; 2341 } else { 2342 ierr = ISGetSize(rowp,&nrows);CHKERRQ(ierr); 2343 ierr = ISGetIndices(rowp,&rows);CHKERRQ(ierr); 2344 ierr = ISCreateGeneral(comm,nrows,rows,PETSC_COPY_VALUES,&crowp);CHKERRQ(ierr); 2345 ierr = ISRestoreIndices(rowp,&rows);CHKERRQ(ierr); 2346 } 2347 ierr = ISSetPermutation(crowp);CHKERRQ(ierr); 2348 /* make a local version of 'colp' */ 2349 ierr = PetscObjectGetComm((PetscObject)colp,&pcomm);CHKERRQ(ierr); 2350 ierr = MPI_Comm_size(pcomm,&size);CHKERRQ(ierr); 2351 if (size==1) { 2352 lcolp = colp; 2353 } else { 2354 ierr = ISAllGather(colp,&lcolp);CHKERRQ(ierr); 2355 } 2356 ierr = ISSetPermutation(lcolp);CHKERRQ(ierr); 2357 /* now we just get the submatrix */ 2358 ierr = MatGetLocalSize(A,NULL,&clocal_size);CHKERRQ(ierr); 2359 ierr = MatGetSubMatrix_MPIBAIJ_Private(A,crowp,lcolp,clocal_size,MAT_INITIAL_MATRIX,B);CHKERRQ(ierr); 2360 /* clean up */ 2361 if (pcomm!=comm) { 2362 ierr = ISDestroy(&crowp);CHKERRQ(ierr); 2363 } 2364 if (size>1) { 2365 ierr = ISDestroy(&lcolp);CHKERRQ(ierr); 2366 } 2367 PetscFunctionReturn(0); 2368 } 2369 2370 #undef __FUNCT__ 2371 #define __FUNCT__ "MatGetGhosts_MPIBAIJ" 2372 PetscErrorCode MatGetGhosts_MPIBAIJ(Mat mat,PetscInt *nghosts,const PetscInt *ghosts[]) 2373 { 2374 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*) mat->data; 2375 Mat_SeqBAIJ *B = (Mat_SeqBAIJ*)baij->B->data; 2376 2377 PetscFunctionBegin; 2378 if (nghosts) *nghosts = B->nbs; 2379 if (ghosts) *ghosts = baij->garray; 2380 PetscFunctionReturn(0); 2381 } 2382 2383 #undef __FUNCT__ 2384 #define __FUNCT__ "MatGetSeqNonzeroStructure_MPIBAIJ" 2385 PetscErrorCode MatGetSeqNonzeroStructure_MPIBAIJ(Mat A,Mat *newmat) 2386 { 2387 Mat B; 2388 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 2389 Mat_SeqBAIJ *ad = (Mat_SeqBAIJ*)a->A->data,*bd = (Mat_SeqBAIJ*)a->B->data; 2390 Mat_SeqAIJ *b; 2391 PetscErrorCode ierr; 2392 PetscMPIInt size,rank,*recvcounts = 0,*displs = 0; 2393 PetscInt sendcount,i,*rstarts = A->rmap->range,n,cnt,j,bs = A->rmap->bs; 2394 PetscInt m,*garray = a->garray,*lens,*jsendbuf,*a_jsendbuf,*b_jsendbuf; 2395 2396 PetscFunctionBegin; 2397 ierr = MPI_Comm_size(PetscObjectComm((PetscObject)A),&size);CHKERRQ(ierr); 2398 ierr = MPI_Comm_rank(PetscObjectComm((PetscObject)A),&rank);CHKERRQ(ierr); 2399 2400 /* ---------------------------------------------------------------- 2401 Tell every processor the number of nonzeros per row 2402 */ 2403 ierr = PetscMalloc1(A->rmap->N/bs,&lens);CHKERRQ(ierr); 2404 for (i=A->rmap->rstart/bs; i<A->rmap->rend/bs; i++) { 2405 lens[i] = ad->i[i-A->rmap->rstart/bs+1] - ad->i[i-A->rmap->rstart/bs] + bd->i[i-A->rmap->rstart/bs+1] - bd->i[i-A->rmap->rstart/bs]; 2406 } 2407 sendcount = A->rmap->rend/bs - A->rmap->rstart/bs; 2408 ierr = PetscMalloc1(2*size,&recvcounts);CHKERRQ(ierr); 2409 displs = recvcounts + size; 2410 for (i=0; i<size; i++) { 2411 recvcounts[i] = A->rmap->range[i+1]/bs - A->rmap->range[i]/bs; 2412 displs[i] = A->rmap->range[i]/bs; 2413 } 2414 #if defined(PETSC_HAVE_MPI_IN_PLACE) 2415 ierr = MPI_Allgatherv(MPI_IN_PLACE,0,MPI_DATATYPE_NULL,lens,recvcounts,displs,MPIU_INT,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2416 #else 2417 ierr = MPI_Allgatherv(lens+A->rmap->rstart/bs,sendcount,MPIU_INT,lens,recvcounts,displs,MPIU_INT,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2418 #endif 2419 /* --------------------------------------------------------------- 2420 Create the sequential matrix of the same type as the local block diagonal 2421 */ 2422 ierr = MatCreate(PETSC_COMM_SELF,&B);CHKERRQ(ierr); 2423 ierr = MatSetSizes(B,A->rmap->N/bs,A->cmap->N/bs,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 2424 ierr = MatSetType(B,MATSEQAIJ);CHKERRQ(ierr); 2425 ierr = MatSeqAIJSetPreallocation(B,0,lens);CHKERRQ(ierr); 2426 b = (Mat_SeqAIJ*)B->data; 2427 2428 /*-------------------------------------------------------------------- 2429 Copy my part of matrix column indices over 2430 */ 2431 sendcount = ad->nz + bd->nz; 2432 jsendbuf = b->j + b->i[rstarts[rank]/bs]; 2433 a_jsendbuf = ad->j; 2434 b_jsendbuf = bd->j; 2435 n = A->rmap->rend/bs - A->rmap->rstart/bs; 2436 cnt = 0; 2437 for (i=0; i<n; i++) { 2438 2439 /* put in lower diagonal portion */ 2440 m = bd->i[i+1] - bd->i[i]; 2441 while (m > 0) { 2442 /* is it above diagonal (in bd (compressed) numbering) */ 2443 if (garray[*b_jsendbuf] > A->rmap->rstart/bs + i) break; 2444 jsendbuf[cnt++] = garray[*b_jsendbuf++]; 2445 m--; 2446 } 2447 2448 /* put in diagonal portion */ 2449 for (j=ad->i[i]; j<ad->i[i+1]; j++) { 2450 jsendbuf[cnt++] = A->rmap->rstart/bs + *a_jsendbuf++; 2451 } 2452 2453 /* put in upper diagonal portion */ 2454 while (m-- > 0) { 2455 jsendbuf[cnt++] = garray[*b_jsendbuf++]; 2456 } 2457 } 2458 if (cnt != sendcount) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Corrupted PETSc matrix: nz given %D actual nz %D",sendcount,cnt); 2459 2460 /*-------------------------------------------------------------------- 2461 Gather all column indices to all processors 2462 */ 2463 for (i=0; i<size; i++) { 2464 recvcounts[i] = 0; 2465 for (j=A->rmap->range[i]/bs; j<A->rmap->range[i+1]/bs; j++) { 2466 recvcounts[i] += lens[j]; 2467 } 2468 } 2469 displs[0] = 0; 2470 for (i=1; i<size; i++) { 2471 displs[i] = displs[i-1] + recvcounts[i-1]; 2472 } 2473 #if defined(PETSC_HAVE_MPI_IN_PLACE) 2474 ierr = MPI_Allgatherv(MPI_IN_PLACE,0,MPI_DATATYPE_NULL,b->j,recvcounts,displs,MPIU_INT,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2475 #else 2476 ierr = MPI_Allgatherv(jsendbuf,sendcount,MPIU_INT,b->j,recvcounts,displs,MPIU_INT,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2477 #endif 2478 /*-------------------------------------------------------------------- 2479 Assemble the matrix into useable form (note numerical values not yet set) 2480 */ 2481 /* set the b->ilen (length of each row) values */ 2482 ierr = PetscMemcpy(b->ilen,lens,(A->rmap->N/bs)*sizeof(PetscInt));CHKERRQ(ierr); 2483 /* set the b->i indices */ 2484 b->i[0] = 0; 2485 for (i=1; i<=A->rmap->N/bs; i++) { 2486 b->i[i] = b->i[i-1] + lens[i-1]; 2487 } 2488 ierr = PetscFree(lens);CHKERRQ(ierr); 2489 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2490 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2491 ierr = PetscFree(recvcounts);CHKERRQ(ierr); 2492 2493 if (A->symmetric) { 2494 ierr = MatSetOption(B,MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 2495 } else if (A->hermitian) { 2496 ierr = MatSetOption(B,MAT_HERMITIAN,PETSC_TRUE);CHKERRQ(ierr); 2497 } else if (A->structurally_symmetric) { 2498 ierr = MatSetOption(B,MAT_STRUCTURALLY_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 2499 } 2500 *newmat = B; 2501 PetscFunctionReturn(0); 2502 } 2503 2504 #undef __FUNCT__ 2505 #define __FUNCT__ "MatSOR_MPIBAIJ" 2506 PetscErrorCode MatSOR_MPIBAIJ(Mat matin,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx) 2507 { 2508 Mat_MPIBAIJ *mat = (Mat_MPIBAIJ*)matin->data; 2509 PetscErrorCode ierr; 2510 Vec bb1 = 0; 2511 2512 PetscFunctionBegin; 2513 if (flag == SOR_APPLY_UPPER) { 2514 ierr = (*mat->A->ops->sor)(mat->A,bb,omega,flag,fshift,lits,1,xx);CHKERRQ(ierr); 2515 PetscFunctionReturn(0); 2516 } 2517 2518 if (its > 1 || ~flag & SOR_ZERO_INITIAL_GUESS) { 2519 ierr = VecDuplicate(bb,&bb1);CHKERRQ(ierr); 2520 } 2521 2522 if ((flag & SOR_LOCAL_SYMMETRIC_SWEEP) == SOR_LOCAL_SYMMETRIC_SWEEP) { 2523 if (flag & SOR_ZERO_INITIAL_GUESS) { 2524 ierr = (*mat->A->ops->sor)(mat->A,bb,omega,flag,fshift,lits,1,xx);CHKERRQ(ierr); 2525 its--; 2526 } 2527 2528 while (its--) { 2529 ierr = VecScatterBegin(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2530 ierr = VecScatterEnd(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2531 2532 /* update rhs: bb1 = bb - B*x */ 2533 ierr = VecScale(mat->lvec,-1.0);CHKERRQ(ierr); 2534 ierr = (*mat->B->ops->multadd)(mat->B,mat->lvec,bb,bb1);CHKERRQ(ierr); 2535 2536 /* local sweep */ 2537 ierr = (*mat->A->ops->sor)(mat->A,bb1,omega,SOR_SYMMETRIC_SWEEP,fshift,lits,1,xx);CHKERRQ(ierr); 2538 } 2539 } else if (flag & SOR_LOCAL_FORWARD_SWEEP) { 2540 if (flag & SOR_ZERO_INITIAL_GUESS) { 2541 ierr = (*mat->A->ops->sor)(mat->A,bb,omega,flag,fshift,lits,1,xx);CHKERRQ(ierr); 2542 its--; 2543 } 2544 while (its--) { 2545 ierr = VecScatterBegin(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2546 ierr = VecScatterEnd(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2547 2548 /* update rhs: bb1 = bb - B*x */ 2549 ierr = VecScale(mat->lvec,-1.0);CHKERRQ(ierr); 2550 ierr = (*mat->B->ops->multadd)(mat->B,mat->lvec,bb,bb1);CHKERRQ(ierr); 2551 2552 /* local sweep */ 2553 ierr = (*mat->A->ops->sor)(mat->A,bb1,omega,SOR_FORWARD_SWEEP,fshift,lits,1,xx);CHKERRQ(ierr); 2554 } 2555 } else if (flag & SOR_LOCAL_BACKWARD_SWEEP) { 2556 if (flag & SOR_ZERO_INITIAL_GUESS) { 2557 ierr = (*mat->A->ops->sor)(mat->A,bb,omega,flag,fshift,lits,1,xx);CHKERRQ(ierr); 2558 its--; 2559 } 2560 while (its--) { 2561 ierr = VecScatterBegin(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2562 ierr = VecScatterEnd(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2563 2564 /* update rhs: bb1 = bb - B*x */ 2565 ierr = VecScale(mat->lvec,-1.0);CHKERRQ(ierr); 2566 ierr = (*mat->B->ops->multadd)(mat->B,mat->lvec,bb,bb1);CHKERRQ(ierr); 2567 2568 /* local sweep */ 2569 ierr = (*mat->A->ops->sor)(mat->A,bb1,omega,SOR_BACKWARD_SWEEP,fshift,lits,1,xx);CHKERRQ(ierr); 2570 } 2571 } else SETERRQ(PetscObjectComm((PetscObject)matin),PETSC_ERR_SUP,"Parallel version of SOR requested not supported"); 2572 2573 ierr = VecDestroy(&bb1);CHKERRQ(ierr); 2574 PetscFunctionReturn(0); 2575 } 2576 2577 #undef __FUNCT__ 2578 #define __FUNCT__ "MatGetColumnNorms_MPIBAIJ" 2579 PetscErrorCode MatGetColumnNorms_MPIBAIJ(Mat A,NormType type,PetscReal *norms) 2580 { 2581 PetscErrorCode ierr; 2582 Mat_MPIBAIJ *aij = (Mat_MPIBAIJ*)A->data; 2583 PetscInt N,i,*garray = aij->garray; 2584 PetscInt ib,jb,bs = A->rmap->bs; 2585 Mat_SeqBAIJ *a_aij = (Mat_SeqBAIJ*) aij->A->data; 2586 MatScalar *a_val = a_aij->a; 2587 Mat_SeqBAIJ *b_aij = (Mat_SeqBAIJ*) aij->B->data; 2588 MatScalar *b_val = b_aij->a; 2589 PetscReal *work; 2590 2591 PetscFunctionBegin; 2592 ierr = MatGetSize(A,NULL,&N);CHKERRQ(ierr); 2593 ierr = PetscCalloc1(N,&work);CHKERRQ(ierr); 2594 if (type == NORM_2) { 2595 for (i=a_aij->i[0]; i<a_aij->i[aij->A->rmap->n/bs]; i++) { 2596 for (jb=0; jb<bs; jb++) { 2597 for (ib=0; ib<bs; ib++) { 2598 work[A->cmap->rstart + a_aij->j[i] * bs + jb] += PetscAbsScalar(*a_val * *a_val); 2599 a_val++; 2600 } 2601 } 2602 } 2603 for (i=b_aij->i[0]; i<b_aij->i[aij->B->rmap->n/bs]; i++) { 2604 for (jb=0; jb<bs; jb++) { 2605 for (ib=0; ib<bs; ib++) { 2606 work[garray[b_aij->j[i]] * bs + jb] += PetscAbsScalar(*b_val * *b_val); 2607 b_val++; 2608 } 2609 } 2610 } 2611 } else if (type == NORM_1) { 2612 for (i=a_aij->i[0]; i<a_aij->i[aij->A->rmap->n/bs]; i++) { 2613 for (jb=0; jb<bs; jb++) { 2614 for (ib=0; ib<bs; ib++) { 2615 work[A->cmap->rstart + a_aij->j[i] * bs + jb] += PetscAbsScalar(*a_val); 2616 a_val++; 2617 } 2618 } 2619 } 2620 for (i=b_aij->i[0]; i<b_aij->i[aij->B->rmap->n/bs]; i++) { 2621 for (jb=0; jb<bs; jb++) { 2622 for (ib=0; ib<bs; ib++) { 2623 work[garray[b_aij->j[i]] * bs + jb] += PetscAbsScalar(*b_val); 2624 b_val++; 2625 } 2626 } 2627 } 2628 } else if (type == NORM_INFINITY) { 2629 for (i=a_aij->i[0]; i<a_aij->i[aij->A->rmap->n/bs]; i++) { 2630 for (jb=0; jb<bs; jb++) { 2631 for (ib=0; ib<bs; ib++) { 2632 int col = A->cmap->rstart + a_aij->j[i] * bs + jb; 2633 work[col] = PetscMax(PetscAbsScalar(*a_val), work[col]); 2634 a_val++; 2635 } 2636 } 2637 } 2638 for (i=b_aij->i[0]; i<b_aij->i[aij->B->rmap->n/bs]; i++) { 2639 for (jb=0; jb<bs; jb++) { 2640 for (ib=0; ib<bs; ib++) { 2641 int col = garray[b_aij->j[i]] * bs + jb; 2642 work[col] = PetscMax(PetscAbsScalar(*b_val), work[col]); 2643 b_val++; 2644 } 2645 } 2646 } 2647 } else SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_WRONG,"Unknown NormType"); 2648 if (type == NORM_INFINITY) { 2649 ierr = MPI_Allreduce(work,norms,N,MPIU_REAL,MPIU_MAX,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2650 } else { 2651 ierr = MPI_Allreduce(work,norms,N,MPIU_REAL,MPIU_SUM,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2652 } 2653 ierr = PetscFree(work);CHKERRQ(ierr); 2654 if (type == NORM_2) { 2655 for (i=0; i<N; i++) norms[i] = PetscSqrtReal(norms[i]); 2656 } 2657 PetscFunctionReturn(0); 2658 } 2659 2660 #undef __FUNCT__ 2661 #define __FUNCT__ "MatInvertBlockDiagonal_MPIBAIJ" 2662 PetscErrorCode MatInvertBlockDiagonal_MPIBAIJ(Mat A,const PetscScalar **values) 2663 { 2664 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*) A->data; 2665 PetscErrorCode ierr; 2666 2667 PetscFunctionBegin; 2668 ierr = MatInvertBlockDiagonal(a->A,values);CHKERRQ(ierr); 2669 PetscFunctionReturn(0); 2670 } 2671 2672 #undef __FUNCT__ 2673 #define __FUNCT__ "MatShift_MPIBAIJ" 2674 PetscErrorCode MatShift_MPIBAIJ(Mat Y,PetscScalar a) 2675 { 2676 PetscErrorCode ierr; 2677 Mat_MPIBAIJ *maij = (Mat_MPIBAIJ*)Y->data; 2678 Mat_SeqBAIJ *aij = (Mat_SeqBAIJ*)maij->A->data,*bij = (Mat_SeqBAIJ*)maij->B->data; 2679 2680 PetscFunctionBegin; 2681 if (!aij->nz && !bij->nz) { 2682 ierr = MatMPIBAIJSetPreallocation(Y,Y->rmap->bs,1,NULL,0,NULL);CHKERRQ(ierr); 2683 } 2684 ierr = MatShift_Basic(Y,a);CHKERRQ(ierr); 2685 PetscFunctionReturn(0); 2686 } 2687 2688 /* -------------------------------------------------------------------*/ 2689 static struct _MatOps MatOps_Values = {MatSetValues_MPIBAIJ, 2690 MatGetRow_MPIBAIJ, 2691 MatRestoreRow_MPIBAIJ, 2692 MatMult_MPIBAIJ, 2693 /* 4*/ MatMultAdd_MPIBAIJ, 2694 MatMultTranspose_MPIBAIJ, 2695 MatMultTransposeAdd_MPIBAIJ, 2696 0, 2697 0, 2698 0, 2699 /*10*/ 0, 2700 0, 2701 0, 2702 MatSOR_MPIBAIJ, 2703 MatTranspose_MPIBAIJ, 2704 /*15*/ MatGetInfo_MPIBAIJ, 2705 MatEqual_MPIBAIJ, 2706 MatGetDiagonal_MPIBAIJ, 2707 MatDiagonalScale_MPIBAIJ, 2708 MatNorm_MPIBAIJ, 2709 /*20*/ MatAssemblyBegin_MPIBAIJ, 2710 MatAssemblyEnd_MPIBAIJ, 2711 MatSetOption_MPIBAIJ, 2712 MatZeroEntries_MPIBAIJ, 2713 /*24*/ MatZeroRows_MPIBAIJ, 2714 0, 2715 0, 2716 0, 2717 0, 2718 /*29*/ MatSetUp_MPIBAIJ, 2719 0, 2720 0, 2721 0, 2722 0, 2723 /*34*/ MatDuplicate_MPIBAIJ, 2724 0, 2725 0, 2726 0, 2727 0, 2728 /*39*/ MatAXPY_MPIBAIJ, 2729 MatGetSubMatrices_MPIBAIJ, 2730 MatIncreaseOverlap_MPIBAIJ, 2731 MatGetValues_MPIBAIJ, 2732 MatCopy_MPIBAIJ, 2733 /*44*/ 0, 2734 MatScale_MPIBAIJ, 2735 MatShift_MPIBAIJ, 2736 0, 2737 MatZeroRowsColumns_MPIBAIJ, 2738 /*49*/ 0, 2739 0, 2740 0, 2741 0, 2742 0, 2743 /*54*/ MatFDColoringCreate_MPIXAIJ, 2744 0, 2745 MatSetUnfactored_MPIBAIJ, 2746 MatPermute_MPIBAIJ, 2747 MatSetValuesBlocked_MPIBAIJ, 2748 /*59*/ MatGetSubMatrix_MPIBAIJ, 2749 MatDestroy_MPIBAIJ, 2750 MatView_MPIBAIJ, 2751 0, 2752 0, 2753 /*64*/ 0, 2754 0, 2755 0, 2756 0, 2757 0, 2758 /*69*/ MatGetRowMaxAbs_MPIBAIJ, 2759 0, 2760 0, 2761 0, 2762 0, 2763 /*74*/ 0, 2764 MatFDColoringApply_BAIJ, 2765 0, 2766 0, 2767 0, 2768 /*79*/ 0, 2769 0, 2770 0, 2771 0, 2772 MatLoad_MPIBAIJ, 2773 /*84*/ 0, 2774 0, 2775 0, 2776 0, 2777 0, 2778 /*89*/ 0, 2779 0, 2780 0, 2781 0, 2782 0, 2783 /*94*/ 0, 2784 0, 2785 0, 2786 0, 2787 0, 2788 /*99*/ 0, 2789 0, 2790 0, 2791 0, 2792 0, 2793 /*104*/0, 2794 MatRealPart_MPIBAIJ, 2795 MatImaginaryPart_MPIBAIJ, 2796 0, 2797 0, 2798 /*109*/0, 2799 0, 2800 0, 2801 0, 2802 0, 2803 /*114*/MatGetSeqNonzeroStructure_MPIBAIJ, 2804 0, 2805 MatGetGhosts_MPIBAIJ, 2806 0, 2807 0, 2808 /*119*/0, 2809 0, 2810 0, 2811 0, 2812 MatGetMultiProcBlock_MPIBAIJ, 2813 /*124*/0, 2814 MatGetColumnNorms_MPIBAIJ, 2815 MatInvertBlockDiagonal_MPIBAIJ, 2816 0, 2817 0, 2818 /*129*/ 0, 2819 0, 2820 0, 2821 0, 2822 0, 2823 /*134*/ 0, 2824 0, 2825 0, 2826 0, 2827 0, 2828 /*139*/ 0, 2829 0, 2830 0, 2831 MatFDColoringSetUp_MPIXAIJ, 2832 0, 2833 /*144*/MatCreateMPIMatConcatenateSeqMat_MPIBAIJ 2834 }; 2835 2836 #undef __FUNCT__ 2837 #define __FUNCT__ "MatGetDiagonalBlock_MPIBAIJ" 2838 PetscErrorCode MatGetDiagonalBlock_MPIBAIJ(Mat A,Mat *a) 2839 { 2840 PetscFunctionBegin; 2841 *a = ((Mat_MPIBAIJ*)A->data)->A; 2842 PetscFunctionReturn(0); 2843 } 2844 2845 PETSC_EXTERN PetscErrorCode MatConvert_MPIBAIJ_MPISBAIJ(Mat, MatType,MatReuse,Mat*); 2846 2847 #undef __FUNCT__ 2848 #define __FUNCT__ "MatMPIBAIJSetPreallocationCSR_MPIBAIJ" 2849 PetscErrorCode MatMPIBAIJSetPreallocationCSR_MPIBAIJ(Mat B,PetscInt bs,const PetscInt ii[],const PetscInt jj[],const PetscScalar V[]) 2850 { 2851 PetscInt m,rstart,cstart,cend; 2852 PetscInt i,j,d,nz,nz_max=0,*d_nnz=0,*o_nnz=0; 2853 const PetscInt *JJ =0; 2854 PetscScalar *values=0; 2855 PetscBool roworiented = ((Mat_MPIBAIJ*)B->data)->roworiented; 2856 PetscErrorCode ierr; 2857 2858 PetscFunctionBegin; 2859 ierr = PetscLayoutSetBlockSize(B->rmap,bs);CHKERRQ(ierr); 2860 ierr = PetscLayoutSetBlockSize(B->cmap,bs);CHKERRQ(ierr); 2861 ierr = PetscLayoutSetUp(B->rmap);CHKERRQ(ierr); 2862 ierr = PetscLayoutSetUp(B->cmap);CHKERRQ(ierr); 2863 ierr = PetscLayoutGetBlockSize(B->rmap,&bs);CHKERRQ(ierr); 2864 m = B->rmap->n/bs; 2865 rstart = B->rmap->rstart/bs; 2866 cstart = B->cmap->rstart/bs; 2867 cend = B->cmap->rend/bs; 2868 2869 if (ii[0]) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"ii[0] must be 0 but it is %D",ii[0]); 2870 ierr = PetscMalloc2(m,&d_nnz,m,&o_nnz);CHKERRQ(ierr); 2871 for (i=0; i<m; i++) { 2872 nz = ii[i+1] - ii[i]; 2873 if (nz < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Local row %D has a negative number of columns %D",i,nz); 2874 nz_max = PetscMax(nz_max,nz); 2875 JJ = jj + ii[i]; 2876 for (j=0; j<nz; j++) { 2877 if (*JJ >= cstart) break; 2878 JJ++; 2879 } 2880 d = 0; 2881 for (; j<nz; j++) { 2882 if (*JJ++ >= cend) break; 2883 d++; 2884 } 2885 d_nnz[i] = d; 2886 o_nnz[i] = nz - d; 2887 } 2888 ierr = MatMPIBAIJSetPreallocation(B,bs,0,d_nnz,0,o_nnz);CHKERRQ(ierr); 2889 ierr = PetscFree2(d_nnz,o_nnz);CHKERRQ(ierr); 2890 2891 values = (PetscScalar*)V; 2892 if (!values) { 2893 ierr = PetscMalloc1(bs*bs*nz_max,&values);CHKERRQ(ierr); 2894 ierr = PetscMemzero(values,bs*bs*nz_max*sizeof(PetscScalar));CHKERRQ(ierr); 2895 } 2896 for (i=0; i<m; i++) { 2897 PetscInt row = i + rstart; 2898 PetscInt ncols = ii[i+1] - ii[i]; 2899 const PetscInt *icols = jj + ii[i]; 2900 if (!roworiented) { /* block ordering matches the non-nested layout of MatSetValues so we can insert entire rows */ 2901 const PetscScalar *svals = values + (V ? (bs*bs*ii[i]) : 0); 2902 ierr = MatSetValuesBlocked_MPIBAIJ(B,1,&row,ncols,icols,svals,INSERT_VALUES);CHKERRQ(ierr); 2903 } else { /* block ordering does not match so we can only insert one block at a time. */ 2904 PetscInt j; 2905 for (j=0; j<ncols; j++) { 2906 const PetscScalar *svals = values + (V ? (bs*bs*(ii[i]+j)) : 0); 2907 ierr = MatSetValuesBlocked_MPIBAIJ(B,1,&row,1,&icols[j],svals,INSERT_VALUES);CHKERRQ(ierr); 2908 } 2909 } 2910 } 2911 2912 if (!V) { ierr = PetscFree(values);CHKERRQ(ierr); } 2913 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2914 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2915 ierr = MatSetOption(B,MAT_NEW_NONZERO_LOCATION_ERR,PETSC_TRUE);CHKERRQ(ierr); 2916 PetscFunctionReturn(0); 2917 } 2918 2919 #undef __FUNCT__ 2920 #define __FUNCT__ "MatMPIBAIJSetPreallocationCSR" 2921 /*@C 2922 MatMPIBAIJSetPreallocationCSR - Allocates memory for a sparse parallel matrix in BAIJ format 2923 (the default parallel PETSc format). 2924 2925 Collective on MPI_Comm 2926 2927 Input Parameters: 2928 + B - the matrix 2929 . bs - the block size 2930 . i - the indices into j for the start of each local row (starts with zero) 2931 . j - the column indices for each local row (starts with zero) these must be sorted for each row 2932 - v - optional values in the matrix 2933 2934 Level: developer 2935 2936 Notes: The order of the entries in values is specified by the MatOption MAT_ROW_ORIENTED. For example, C programs 2937 may want to use the default MAT_ROW_ORIENTED=PETSC_TRUE and use an array v[nnz][bs][bs] where the second index is 2938 over rows within a block and the last index is over columns within a block row. Fortran programs will likely set 2939 MAT_ROW_ORIENTED=PETSC_FALSE and use a Fortran array v(bs,bs,nnz) in which the first index is over rows within a 2940 block column and the second index is over columns within a block. 2941 2942 .keywords: matrix, aij, compressed row, sparse, parallel 2943 2944 .seealso: MatCreate(), MatCreateSeqAIJ(), MatSetValues(), MatMPIBAIJSetPreallocation(), MatCreateAIJ(), MPIAIJ, MatCreateMPIBAIJWithArrays(), MPIBAIJ 2945 @*/ 2946 PetscErrorCode MatMPIBAIJSetPreallocationCSR(Mat B,PetscInt bs,const PetscInt i[],const PetscInt j[], const PetscScalar v[]) 2947 { 2948 PetscErrorCode ierr; 2949 2950 PetscFunctionBegin; 2951 PetscValidHeaderSpecific(B,MAT_CLASSID,1); 2952 PetscValidType(B,1); 2953 PetscValidLogicalCollectiveInt(B,bs,2); 2954 ierr = PetscTryMethod(B,"MatMPIBAIJSetPreallocationCSR_C",(Mat,PetscInt,const PetscInt[],const PetscInt[],const PetscScalar[]),(B,bs,i,j,v));CHKERRQ(ierr); 2955 PetscFunctionReturn(0); 2956 } 2957 2958 #undef __FUNCT__ 2959 #define __FUNCT__ "MatMPIBAIJSetPreallocation_MPIBAIJ" 2960 PetscErrorCode MatMPIBAIJSetPreallocation_MPIBAIJ(Mat B,PetscInt bs,PetscInt d_nz,const PetscInt *d_nnz,PetscInt o_nz,const PetscInt *o_nnz) 2961 { 2962 Mat_MPIBAIJ *b; 2963 PetscErrorCode ierr; 2964 PetscInt i; 2965 2966 PetscFunctionBegin; 2967 ierr = MatSetBlockSize(B,PetscAbs(bs));CHKERRQ(ierr); 2968 ierr = PetscLayoutSetUp(B->rmap);CHKERRQ(ierr); 2969 ierr = PetscLayoutSetUp(B->cmap);CHKERRQ(ierr); 2970 ierr = PetscLayoutGetBlockSize(B->rmap,&bs);CHKERRQ(ierr); 2971 2972 if (d_nnz) { 2973 for (i=0; i<B->rmap->n/bs; i++) { 2974 if (d_nnz[i] < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"d_nnz cannot be less than -1: local row %D value %D",i,d_nnz[i]); 2975 } 2976 } 2977 if (o_nnz) { 2978 for (i=0; i<B->rmap->n/bs; i++) { 2979 if (o_nnz[i] < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"o_nnz cannot be less than -1: local row %D value %D",i,o_nnz[i]); 2980 } 2981 } 2982 2983 b = (Mat_MPIBAIJ*)B->data; 2984 b->bs2 = bs*bs; 2985 b->mbs = B->rmap->n/bs; 2986 b->nbs = B->cmap->n/bs; 2987 b->Mbs = B->rmap->N/bs; 2988 b->Nbs = B->cmap->N/bs; 2989 2990 for (i=0; i<=b->size; i++) { 2991 b->rangebs[i] = B->rmap->range[i]/bs; 2992 } 2993 b->rstartbs = B->rmap->rstart/bs; 2994 b->rendbs = B->rmap->rend/bs; 2995 b->cstartbs = B->cmap->rstart/bs; 2996 b->cendbs = B->cmap->rend/bs; 2997 2998 if (!B->preallocated) { 2999 ierr = MatCreate(PETSC_COMM_SELF,&b->A);CHKERRQ(ierr); 3000 ierr = MatSetSizes(b->A,B->rmap->n,B->cmap->n,B->rmap->n,B->cmap->n);CHKERRQ(ierr); 3001 ierr = MatSetType(b->A,MATSEQBAIJ);CHKERRQ(ierr); 3002 ierr = PetscLogObjectParent((PetscObject)B,(PetscObject)b->A);CHKERRQ(ierr); 3003 ierr = MatCreate(PETSC_COMM_SELF,&b->B);CHKERRQ(ierr); 3004 ierr = MatSetSizes(b->B,B->rmap->n,B->cmap->N,B->rmap->n,B->cmap->N);CHKERRQ(ierr); 3005 ierr = MatSetType(b->B,MATSEQBAIJ);CHKERRQ(ierr); 3006 ierr = PetscLogObjectParent((PetscObject)B,(PetscObject)b->B);CHKERRQ(ierr); 3007 ierr = MatStashCreate_Private(PetscObjectComm((PetscObject)B),bs,&B->bstash);CHKERRQ(ierr); 3008 } 3009 3010 ierr = MatSeqBAIJSetPreallocation(b->A,bs,d_nz,d_nnz);CHKERRQ(ierr); 3011 ierr = MatSeqBAIJSetPreallocation(b->B,bs,o_nz,o_nnz);CHKERRQ(ierr); 3012 B->preallocated = PETSC_TRUE; 3013 PetscFunctionReturn(0); 3014 } 3015 3016 extern PetscErrorCode MatDiagonalScaleLocal_MPIBAIJ(Mat,Vec); 3017 extern PetscErrorCode MatSetHashTableFactor_MPIBAIJ(Mat,PetscReal); 3018 3019 #undef __FUNCT__ 3020 #define __FUNCT__ "MatConvert_MPIBAIJ_MPIAdj" 3021 PETSC_EXTERN PetscErrorCode MatConvert_MPIBAIJ_MPIAdj(Mat B, MatType newtype,MatReuse reuse,Mat *adj) 3022 { 3023 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)B->data; 3024 PetscErrorCode ierr; 3025 Mat_SeqBAIJ *d = (Mat_SeqBAIJ*) b->A->data,*o = (Mat_SeqBAIJ*) b->B->data; 3026 PetscInt M = B->rmap->n/B->rmap->bs,i,*ii,*jj,cnt,j,k,rstart = B->rmap->rstart/B->rmap->bs; 3027 const PetscInt *id = d->i, *jd = d->j, *io = o->i, *jo = o->j, *garray = b->garray; 3028 3029 PetscFunctionBegin; 3030 ierr = PetscMalloc1(M+1,&ii);CHKERRQ(ierr); 3031 ii[0] = 0; 3032 for (i=0; i<M; i++) { 3033 if ((id[i+1] - id[i]) < 0) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Indices wrong %D %D %D",i,id[i],id[i+1]); 3034 if ((io[i+1] - io[i]) < 0) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Indices wrong %D %D %D",i,io[i],io[i+1]); 3035 ii[i+1] = ii[i] + id[i+1] - id[i] + io[i+1] - io[i]; 3036 /* remove one from count of matrix has diagonal */ 3037 for (j=id[i]; j<id[i+1]; j++) { 3038 if (jd[j] == i) {ii[i+1]--;break;} 3039 } 3040 } 3041 ierr = PetscMalloc1(ii[M],&jj);CHKERRQ(ierr); 3042 cnt = 0; 3043 for (i=0; i<M; i++) { 3044 for (j=io[i]; j<io[i+1]; j++) { 3045 if (garray[jo[j]] > rstart) break; 3046 jj[cnt++] = garray[jo[j]]; 3047 } 3048 for (k=id[i]; k<id[i+1]; k++) { 3049 if (jd[k] != i) { 3050 jj[cnt++] = rstart + jd[k]; 3051 } 3052 } 3053 for (; j<io[i+1]; j++) { 3054 jj[cnt++] = garray[jo[j]]; 3055 } 3056 } 3057 ierr = MatCreateMPIAdj(PetscObjectComm((PetscObject)B),M,B->cmap->N/B->rmap->bs,ii,jj,NULL,adj);CHKERRQ(ierr); 3058 PetscFunctionReturn(0); 3059 } 3060 3061 #include <../src/mat/impls/aij/mpi/mpiaij.h> 3062 3063 PETSC_EXTERN PetscErrorCode MatConvert_SeqBAIJ_SeqAIJ(Mat,MatType,MatReuse,Mat*); 3064 3065 #undef __FUNCT__ 3066 #define __FUNCT__ "MatConvert_MPIBAIJ_MPIAIJ" 3067 PETSC_EXTERN PetscErrorCode MatConvert_MPIBAIJ_MPIAIJ(Mat A,MatType newtype,MatReuse reuse,Mat *newmat) 3068 { 3069 PetscErrorCode ierr; 3070 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 3071 Mat B; 3072 Mat_MPIAIJ *b; 3073 3074 PetscFunctionBegin; 3075 if (!A->assembled) SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_SUP,"Matrix must be assembled"); 3076 3077 ierr = MatCreate(PetscObjectComm((PetscObject)A),&B);CHKERRQ(ierr); 3078 ierr = MatSetType(B,MATMPIAIJ);CHKERRQ(ierr); 3079 ierr = MatSetSizes(B,A->rmap->n,A->cmap->n,A->rmap->N,A->cmap->N);CHKERRQ(ierr); 3080 ierr = MatSetBlockSizes(B,A->rmap->bs,A->cmap->bs);CHKERRQ(ierr); 3081 ierr = MatSeqAIJSetPreallocation(B,0,NULL);CHKERRQ(ierr); 3082 ierr = MatMPIAIJSetPreallocation(B,0,NULL,0,NULL);CHKERRQ(ierr); 3083 b = (Mat_MPIAIJ*) B->data; 3084 3085 ierr = MatDestroy(&b->A);CHKERRQ(ierr); 3086 ierr = MatDestroy(&b->B);CHKERRQ(ierr); 3087 ierr = MatDisAssemble_MPIBAIJ(A);CHKERRQ(ierr); 3088 ierr = MatConvert_SeqBAIJ_SeqAIJ(a->A, MATSEQAIJ, MAT_INITIAL_MATRIX, &b->A);CHKERRQ(ierr); 3089 ierr = MatConvert_SeqBAIJ_SeqAIJ(a->B, MATSEQAIJ, MAT_INITIAL_MATRIX, &b->B);CHKERRQ(ierr); 3090 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3091 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3092 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3093 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3094 if (reuse == MAT_REUSE_MATRIX) { 3095 ierr = MatHeaderReplace(A,B);CHKERRQ(ierr); 3096 } else { 3097 *newmat = B; 3098 } 3099 PetscFunctionReturn(0); 3100 } 3101 3102 /*MC 3103 MATMPIBAIJ - MATMPIBAIJ = "mpibaij" - A matrix type to be used for distributed block sparse matrices. 3104 3105 Options Database Keys: 3106 + -mat_type mpibaij - sets the matrix type to "mpibaij" during a call to MatSetFromOptions() 3107 . -mat_block_size <bs> - set the blocksize used to store the matrix 3108 - -mat_use_hash_table <fact> 3109 3110 Level: beginner 3111 3112 .seealso: MatCreateMPIBAIJ 3113 M*/ 3114 3115 PETSC_EXTERN PetscErrorCode MatConvert_MPIBAIJ_MPIBSTRM(Mat,MatType,MatReuse,Mat*); 3116 3117 #undef __FUNCT__ 3118 #define __FUNCT__ "MatCreate_MPIBAIJ" 3119 PETSC_EXTERN PetscErrorCode MatCreate_MPIBAIJ(Mat B) 3120 { 3121 Mat_MPIBAIJ *b; 3122 PetscErrorCode ierr; 3123 PetscBool flg = PETSC_FALSE; 3124 3125 PetscFunctionBegin; 3126 ierr = PetscNewLog(B,&b);CHKERRQ(ierr); 3127 B->data = (void*)b; 3128 3129 ierr = PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));CHKERRQ(ierr); 3130 B->assembled = PETSC_FALSE; 3131 3132 B->insertmode = NOT_SET_VALUES; 3133 ierr = MPI_Comm_rank(PetscObjectComm((PetscObject)B),&b->rank);CHKERRQ(ierr); 3134 ierr = MPI_Comm_size(PetscObjectComm((PetscObject)B),&b->size);CHKERRQ(ierr); 3135 3136 /* build local table of row and column ownerships */ 3137 ierr = PetscMalloc1(b->size+1,&b->rangebs);CHKERRQ(ierr); 3138 3139 /* build cache for off array entries formed */ 3140 ierr = MatStashCreate_Private(PetscObjectComm((PetscObject)B),1,&B->stash);CHKERRQ(ierr); 3141 3142 b->donotstash = PETSC_FALSE; 3143 b->colmap = NULL; 3144 b->garray = NULL; 3145 b->roworiented = PETSC_TRUE; 3146 3147 /* stuff used in block assembly */ 3148 b->barray = 0; 3149 3150 /* stuff used for matrix vector multiply */ 3151 b->lvec = 0; 3152 b->Mvctx = 0; 3153 3154 /* stuff for MatGetRow() */ 3155 b->rowindices = 0; 3156 b->rowvalues = 0; 3157 b->getrowactive = PETSC_FALSE; 3158 3159 /* hash table stuff */ 3160 b->ht = 0; 3161 b->hd = 0; 3162 b->ht_size = 0; 3163 b->ht_flag = PETSC_FALSE; 3164 b->ht_fact = 0; 3165 b->ht_total_ct = 0; 3166 b->ht_insert_ct = 0; 3167 3168 /* stuff for MatGetSubMatrices_MPIBAIJ_local() */ 3169 b->ijonly = PETSC_FALSE; 3170 3171 3172 ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_mpibaij_mpiadj_C",MatConvert_MPIBAIJ_MPIAdj);CHKERRQ(ierr); 3173 ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_mpibaij_mpiaij_C",MatConvert_MPIBAIJ_MPIAIJ);CHKERRQ(ierr); 3174 ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_mpibaij_mpisbaij_C",MatConvert_MPIBAIJ_MPISBAIJ);CHKERRQ(ierr); 3175 ierr = PetscObjectComposeFunction((PetscObject)B,"MatStoreValues_C",MatStoreValues_MPIBAIJ);CHKERRQ(ierr); 3176 ierr = PetscObjectComposeFunction((PetscObject)B,"MatRetrieveValues_C",MatRetrieveValues_MPIBAIJ);CHKERRQ(ierr); 3177 ierr = PetscObjectComposeFunction((PetscObject)B,"MatGetDiagonalBlock_C",MatGetDiagonalBlock_MPIBAIJ);CHKERRQ(ierr); 3178 ierr = PetscObjectComposeFunction((PetscObject)B,"MatMPIBAIJSetPreallocation_C",MatMPIBAIJSetPreallocation_MPIBAIJ);CHKERRQ(ierr); 3179 ierr = PetscObjectComposeFunction((PetscObject)B,"MatMPIBAIJSetPreallocationCSR_C",MatMPIBAIJSetPreallocationCSR_MPIBAIJ);CHKERRQ(ierr); 3180 ierr = PetscObjectComposeFunction((PetscObject)B,"MatDiagonalScaleLocal_C",MatDiagonalScaleLocal_MPIBAIJ);CHKERRQ(ierr); 3181 ierr = PetscObjectComposeFunction((PetscObject)B,"MatSetHashTableFactor_C",MatSetHashTableFactor_MPIBAIJ);CHKERRQ(ierr); 3182 ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_mpibaij_mpibstrm_C",MatConvert_MPIBAIJ_MPIBSTRM);CHKERRQ(ierr); 3183 ierr = PetscObjectChangeTypeName((PetscObject)B,MATMPIBAIJ);CHKERRQ(ierr); 3184 3185 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)B),NULL,"Options for loading MPIBAIJ matrix 1","Mat");CHKERRQ(ierr); 3186 ierr = PetscOptionsBool("-mat_use_hash_table","Use hash table to save memory in constructing matrix","MatSetOption",flg,&flg,NULL);CHKERRQ(ierr); 3187 if (flg) { 3188 PetscReal fact = 1.39; 3189 ierr = MatSetOption(B,MAT_USE_HASH_TABLE,PETSC_TRUE);CHKERRQ(ierr); 3190 ierr = PetscOptionsReal("-mat_use_hash_table","Use hash table factor","MatMPIBAIJSetHashTableFactor",fact,&fact,NULL);CHKERRQ(ierr); 3191 if (fact <= 1.0) fact = 1.39; 3192 ierr = MatMPIBAIJSetHashTableFactor(B,fact);CHKERRQ(ierr); 3193 ierr = PetscInfo1(B,"Hash table Factor used %5.2f\n",fact);CHKERRQ(ierr); 3194 } 3195 ierr = PetscOptionsEnd();CHKERRQ(ierr); 3196 PetscFunctionReturn(0); 3197 } 3198 3199 /*MC 3200 MATBAIJ - MATBAIJ = "baij" - A matrix type to be used for block sparse matrices. 3201 3202 This matrix type is identical to MATSEQBAIJ when constructed with a single process communicator, 3203 and MATMPIBAIJ otherwise. 3204 3205 Options Database Keys: 3206 . -mat_type baij - sets the matrix type to "baij" during a call to MatSetFromOptions() 3207 3208 Level: beginner 3209 3210 .seealso: MatCreateBAIJ(),MATSEQBAIJ,MATMPIBAIJ, MatMPIBAIJSetPreallocation(), MatMPIBAIJSetPreallocationCSR() 3211 M*/ 3212 3213 #undef __FUNCT__ 3214 #define __FUNCT__ "MatMPIBAIJSetPreallocation" 3215 /*@C 3216 MatMPIBAIJSetPreallocation - Allocates memory for a sparse parallel matrix in block AIJ format 3217 (block compressed row). For good matrix assembly performance 3218 the user should preallocate the matrix storage by setting the parameters 3219 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 3220 performance can be increased by more than a factor of 50. 3221 3222 Collective on Mat 3223 3224 Input Parameters: 3225 + B - the matrix 3226 . bs - size of block, the blocks are ALWAYS square. One can use MatSetBlockSizes() to set a different row and column blocksize but the row 3227 blocksize always defines the size of the blocks. The column blocksize sets the blocksize of the vectors obtained with MatCreateVecs() 3228 . d_nz - number of block nonzeros per block row in diagonal portion of local 3229 submatrix (same for all local rows) 3230 . d_nnz - array containing the number of block nonzeros in the various block rows 3231 of the in diagonal portion of the local (possibly different for each block 3232 row) or NULL. If you plan to factor the matrix you must leave room for the diagonal entry and 3233 set it even if it is zero. 3234 . o_nz - number of block nonzeros per block row in the off-diagonal portion of local 3235 submatrix (same for all local rows). 3236 - o_nnz - array containing the number of nonzeros in the various block rows of the 3237 off-diagonal portion of the local submatrix (possibly different for 3238 each block row) or NULL. 3239 3240 If the *_nnz parameter is given then the *_nz parameter is ignored 3241 3242 Options Database Keys: 3243 + -mat_block_size - size of the blocks to use 3244 - -mat_use_hash_table <fact> 3245 3246 Notes: 3247 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 3248 than it must be used on all processors that share the object for that argument. 3249 3250 Storage Information: 3251 For a square global matrix we define each processor's diagonal portion 3252 to be its local rows and the corresponding columns (a square submatrix); 3253 each processor's off-diagonal portion encompasses the remainder of the 3254 local matrix (a rectangular submatrix). 3255 3256 The user can specify preallocated storage for the diagonal part of 3257 the local submatrix with either d_nz or d_nnz (not both). Set 3258 d_nz=PETSC_DEFAULT and d_nnz=NULL for PETSc to control dynamic 3259 memory allocation. Likewise, specify preallocated storage for the 3260 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 3261 3262 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 3263 the figure below we depict these three local rows and all columns (0-11). 3264 3265 .vb 3266 0 1 2 3 4 5 6 7 8 9 10 11 3267 -------------------------- 3268 row 3 |o o o d d d o o o o o o 3269 row 4 |o o o d d d o o o o o o 3270 row 5 |o o o d d d o o o o o o 3271 -------------------------- 3272 .ve 3273 3274 Thus, any entries in the d locations are stored in the d (diagonal) 3275 submatrix, and any entries in the o locations are stored in the 3276 o (off-diagonal) submatrix. Note that the d and the o submatrices are 3277 stored simply in the MATSEQBAIJ format for compressed row storage. 3278 3279 Now d_nz should indicate the number of block nonzeros per row in the d matrix, 3280 and o_nz should indicate the number of block nonzeros per row in the o matrix. 3281 In general, for PDE problems in which most nonzeros are near the diagonal, 3282 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 3283 or you will get TERRIBLE performance; see the users' manual chapter on 3284 matrices. 3285 3286 You can call MatGetInfo() to get information on how effective the preallocation was; 3287 for example the fields mallocs,nz_allocated,nz_used,nz_unneeded; 3288 You can also run with the option -info and look for messages with the string 3289 malloc in them to see if additional memory allocation was needed. 3290 3291 Level: intermediate 3292 3293 .keywords: matrix, block, aij, compressed row, sparse, parallel 3294 3295 .seealso: MatCreate(), MatCreateSeqBAIJ(), MatSetValues(), MatCreateBAIJ(), MatMPIBAIJSetPreallocationCSR(), PetscSplitOwnership() 3296 @*/ 3297 PetscErrorCode MatMPIBAIJSetPreallocation(Mat B,PetscInt bs,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[]) 3298 { 3299 PetscErrorCode ierr; 3300 3301 PetscFunctionBegin; 3302 PetscValidHeaderSpecific(B,MAT_CLASSID,1); 3303 PetscValidType(B,1); 3304 PetscValidLogicalCollectiveInt(B,bs,2); 3305 ierr = PetscTryMethod(B,"MatMPIBAIJSetPreallocation_C",(Mat,PetscInt,PetscInt,const PetscInt[],PetscInt,const PetscInt[]),(B,bs,d_nz,d_nnz,o_nz,o_nnz));CHKERRQ(ierr); 3306 PetscFunctionReturn(0); 3307 } 3308 3309 #undef __FUNCT__ 3310 #define __FUNCT__ "MatCreateBAIJ" 3311 /*@C 3312 MatCreateBAIJ - Creates a sparse parallel matrix in block AIJ format 3313 (block compressed row). For good matrix assembly performance 3314 the user should preallocate the matrix storage by setting the parameters 3315 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 3316 performance can be increased by more than a factor of 50. 3317 3318 Collective on MPI_Comm 3319 3320 Input Parameters: 3321 + comm - MPI communicator 3322 . bs - size of block, the blocks are ALWAYS square. One can use MatSetBlockSizes() to set a different row and column blocksize but the row 3323 blocksize always defines the size of the blocks. The column blocksize sets the blocksize of the vectors obtained with MatCreateVecs() 3324 . m - number of local rows (or PETSC_DECIDE to have calculated if M is given) 3325 This value should be the same as the local size used in creating the 3326 y vector for the matrix-vector product y = Ax. 3327 . n - number of local columns (or PETSC_DECIDE to have calculated if N is given) 3328 This value should be the same as the local size used in creating the 3329 x vector for the matrix-vector product y = Ax. 3330 . M - number of global rows (or PETSC_DETERMINE to have calculated if m is given) 3331 . N - number of global columns (or PETSC_DETERMINE to have calculated if n is given) 3332 . d_nz - number of nonzero blocks per block row in diagonal portion of local 3333 submatrix (same for all local rows) 3334 . d_nnz - array containing the number of nonzero blocks in the various block rows 3335 of the in diagonal portion of the local (possibly different for each block 3336 row) or NULL. If you plan to factor the matrix you must leave room for the diagonal entry 3337 and set it even if it is zero. 3338 . o_nz - number of nonzero blocks per block row in the off-diagonal portion of local 3339 submatrix (same for all local rows). 3340 - o_nnz - array containing the number of nonzero blocks in the various block rows of the 3341 off-diagonal portion of the local submatrix (possibly different for 3342 each block row) or NULL. 3343 3344 Output Parameter: 3345 . A - the matrix 3346 3347 Options Database Keys: 3348 + -mat_block_size - size of the blocks to use 3349 - -mat_use_hash_table <fact> 3350 3351 It is recommended that one use the MatCreate(), MatSetType() and/or MatSetFromOptions(), 3352 MatXXXXSetPreallocation() paradgm instead of this routine directly. 3353 [MatXXXXSetPreallocation() is, for example, MatSeqAIJSetPreallocation] 3354 3355 Notes: 3356 If the *_nnz parameter is given then the *_nz parameter is ignored 3357 3358 A nonzero block is any block that as 1 or more nonzeros in it 3359 3360 The user MUST specify either the local or global matrix dimensions 3361 (possibly both). 3362 3363 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 3364 than it must be used on all processors that share the object for that argument. 3365 3366 Storage Information: 3367 For a square global matrix we define each processor's diagonal portion 3368 to be its local rows and the corresponding columns (a square submatrix); 3369 each processor's off-diagonal portion encompasses the remainder of the 3370 local matrix (a rectangular submatrix). 3371 3372 The user can specify preallocated storage for the diagonal part of 3373 the local submatrix with either d_nz or d_nnz (not both). Set 3374 d_nz=PETSC_DEFAULT and d_nnz=NULL for PETSc to control dynamic 3375 memory allocation. Likewise, specify preallocated storage for the 3376 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 3377 3378 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 3379 the figure below we depict these three local rows and all columns (0-11). 3380 3381 .vb 3382 0 1 2 3 4 5 6 7 8 9 10 11 3383 -------------------------- 3384 row 3 |o o o d d d o o o o o o 3385 row 4 |o o o d d d o o o o o o 3386 row 5 |o o o d d d o o o o o o 3387 -------------------------- 3388 .ve 3389 3390 Thus, any entries in the d locations are stored in the d (diagonal) 3391 submatrix, and any entries in the o locations are stored in the 3392 o (off-diagonal) submatrix. Note that the d and the o submatrices are 3393 stored simply in the MATSEQBAIJ format for compressed row storage. 3394 3395 Now d_nz should indicate the number of block nonzeros per row in the d matrix, 3396 and o_nz should indicate the number of block nonzeros per row in the o matrix. 3397 In general, for PDE problems in which most nonzeros are near the diagonal, 3398 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 3399 or you will get TERRIBLE performance; see the users' manual chapter on 3400 matrices. 3401 3402 Level: intermediate 3403 3404 .keywords: matrix, block, aij, compressed row, sparse, parallel 3405 3406 .seealso: MatCreate(), MatCreateSeqBAIJ(), MatSetValues(), MatCreateBAIJ(), MatMPIBAIJSetPreallocation(), MatMPIBAIJSetPreallocationCSR() 3407 @*/ 3408 PetscErrorCode MatCreateBAIJ(MPI_Comm comm,PetscInt bs,PetscInt m,PetscInt n,PetscInt M,PetscInt N,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[],Mat *A) 3409 { 3410 PetscErrorCode ierr; 3411 PetscMPIInt size; 3412 3413 PetscFunctionBegin; 3414 ierr = MatCreate(comm,A);CHKERRQ(ierr); 3415 ierr = MatSetSizes(*A,m,n,M,N);CHKERRQ(ierr); 3416 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 3417 if (size > 1) { 3418 ierr = MatSetType(*A,MATMPIBAIJ);CHKERRQ(ierr); 3419 ierr = MatMPIBAIJSetPreallocation(*A,bs,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 3420 } else { 3421 ierr = MatSetType(*A,MATSEQBAIJ);CHKERRQ(ierr); 3422 ierr = MatSeqBAIJSetPreallocation(*A,bs,d_nz,d_nnz);CHKERRQ(ierr); 3423 } 3424 PetscFunctionReturn(0); 3425 } 3426 3427 #undef __FUNCT__ 3428 #define __FUNCT__ "MatDuplicate_MPIBAIJ" 3429 static PetscErrorCode MatDuplicate_MPIBAIJ(Mat matin,MatDuplicateOption cpvalues,Mat *newmat) 3430 { 3431 Mat mat; 3432 Mat_MPIBAIJ *a,*oldmat = (Mat_MPIBAIJ*)matin->data; 3433 PetscErrorCode ierr; 3434 PetscInt len=0; 3435 3436 PetscFunctionBegin; 3437 *newmat = 0; 3438 ierr = MatCreate(PetscObjectComm((PetscObject)matin),&mat);CHKERRQ(ierr); 3439 ierr = MatSetSizes(mat,matin->rmap->n,matin->cmap->n,matin->rmap->N,matin->cmap->N);CHKERRQ(ierr); 3440 ierr = MatSetType(mat,((PetscObject)matin)->type_name);CHKERRQ(ierr); 3441 ierr = PetscMemcpy(mat->ops,matin->ops,sizeof(struct _MatOps));CHKERRQ(ierr); 3442 3443 mat->factortype = matin->factortype; 3444 mat->preallocated = PETSC_TRUE; 3445 mat->assembled = PETSC_TRUE; 3446 mat->insertmode = NOT_SET_VALUES; 3447 3448 a = (Mat_MPIBAIJ*)mat->data; 3449 mat->rmap->bs = matin->rmap->bs; 3450 a->bs2 = oldmat->bs2; 3451 a->mbs = oldmat->mbs; 3452 a->nbs = oldmat->nbs; 3453 a->Mbs = oldmat->Mbs; 3454 a->Nbs = oldmat->Nbs; 3455 3456 ierr = PetscLayoutReference(matin->rmap,&mat->rmap);CHKERRQ(ierr); 3457 ierr = PetscLayoutReference(matin->cmap,&mat->cmap);CHKERRQ(ierr); 3458 3459 a->size = oldmat->size; 3460 a->rank = oldmat->rank; 3461 a->donotstash = oldmat->donotstash; 3462 a->roworiented = oldmat->roworiented; 3463 a->rowindices = 0; 3464 a->rowvalues = 0; 3465 a->getrowactive = PETSC_FALSE; 3466 a->barray = 0; 3467 a->rstartbs = oldmat->rstartbs; 3468 a->rendbs = oldmat->rendbs; 3469 a->cstartbs = oldmat->cstartbs; 3470 a->cendbs = oldmat->cendbs; 3471 3472 /* hash table stuff */ 3473 a->ht = 0; 3474 a->hd = 0; 3475 a->ht_size = 0; 3476 a->ht_flag = oldmat->ht_flag; 3477 a->ht_fact = oldmat->ht_fact; 3478 a->ht_total_ct = 0; 3479 a->ht_insert_ct = 0; 3480 3481 ierr = PetscMemcpy(a->rangebs,oldmat->rangebs,(a->size+1)*sizeof(PetscInt));CHKERRQ(ierr); 3482 if (oldmat->colmap) { 3483 #if defined(PETSC_USE_CTABLE) 3484 ierr = PetscTableCreateCopy(oldmat->colmap,&a->colmap);CHKERRQ(ierr); 3485 #else 3486 ierr = PetscMalloc1(a->Nbs,&a->colmap);CHKERRQ(ierr); 3487 ierr = PetscLogObjectMemory((PetscObject)mat,(a->Nbs)*sizeof(PetscInt));CHKERRQ(ierr); 3488 ierr = PetscMemcpy(a->colmap,oldmat->colmap,(a->Nbs)*sizeof(PetscInt));CHKERRQ(ierr); 3489 #endif 3490 } else a->colmap = 0; 3491 3492 if (oldmat->garray && (len = ((Mat_SeqBAIJ*)(oldmat->B->data))->nbs)) { 3493 ierr = PetscMalloc1(len,&a->garray);CHKERRQ(ierr); 3494 ierr = PetscLogObjectMemory((PetscObject)mat,len*sizeof(PetscInt));CHKERRQ(ierr); 3495 ierr = PetscMemcpy(a->garray,oldmat->garray,len*sizeof(PetscInt));CHKERRQ(ierr); 3496 } else a->garray = 0; 3497 3498 ierr = MatStashCreate_Private(PetscObjectComm((PetscObject)matin),matin->rmap->bs,&mat->bstash);CHKERRQ(ierr); 3499 ierr = VecDuplicate(oldmat->lvec,&a->lvec);CHKERRQ(ierr); 3500 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)a->lvec);CHKERRQ(ierr); 3501 ierr = VecScatterCopy(oldmat->Mvctx,&a->Mvctx);CHKERRQ(ierr); 3502 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)a->Mvctx);CHKERRQ(ierr); 3503 3504 ierr = MatDuplicate(oldmat->A,cpvalues,&a->A);CHKERRQ(ierr); 3505 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)a->A);CHKERRQ(ierr); 3506 ierr = MatDuplicate(oldmat->B,cpvalues,&a->B);CHKERRQ(ierr); 3507 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)a->B);CHKERRQ(ierr); 3508 ierr = PetscFunctionListDuplicate(((PetscObject)matin)->qlist,&((PetscObject)mat)->qlist);CHKERRQ(ierr); 3509 *newmat = mat; 3510 PetscFunctionReturn(0); 3511 } 3512 3513 #undef __FUNCT__ 3514 #define __FUNCT__ "MatLoad_MPIBAIJ" 3515 PetscErrorCode MatLoad_MPIBAIJ(Mat newmat,PetscViewer viewer) 3516 { 3517 PetscErrorCode ierr; 3518 int fd; 3519 PetscInt i,nz,j,rstart,rend; 3520 PetscScalar *vals,*buf; 3521 MPI_Comm comm; 3522 MPI_Status status; 3523 PetscMPIInt rank,size,maxnz; 3524 PetscInt header[4],*rowlengths = 0,M,N,m,*rowners,*cols; 3525 PetscInt *locrowlens = NULL,*procsnz = NULL,*browners = NULL; 3526 PetscInt jj,*mycols,*ibuf,bs = newmat->rmap->bs,Mbs,mbs,extra_rows,mmax; 3527 PetscMPIInt tag = ((PetscObject)viewer)->tag; 3528 PetscInt *dlens = NULL,*odlens = NULL,*mask = NULL,*masked1 = NULL,*masked2 = NULL,rowcount,odcount; 3529 PetscInt dcount,kmax,k,nzcount,tmp,mend; 3530 3531 PetscFunctionBegin; 3532 /* force binary viewer to load .info file if it has not yet done so */ 3533 ierr = PetscViewerSetUp(viewer);CHKERRQ(ierr); 3534 ierr = PetscObjectGetComm((PetscObject)viewer,&comm);CHKERRQ(ierr); 3535 ierr = PetscOptionsBegin(comm,NULL,"Options for loading MPIBAIJ matrix 2","Mat");CHKERRQ(ierr); 3536 ierr = PetscOptionsInt("-matload_block_size","Set the blocksize used to store the matrix","MatLoad",bs,&bs,NULL);CHKERRQ(ierr); 3537 ierr = PetscOptionsEnd();CHKERRQ(ierr); 3538 if (bs < 0) bs = 1; 3539 3540 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 3541 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 3542 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 3543 if (!rank) { 3544 ierr = PetscBinaryRead(fd,(char*)header,4,PETSC_INT);CHKERRQ(ierr); 3545 if (header[0] != MAT_FILE_CLASSID) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED,"not matrix object"); 3546 } 3547 ierr = MPI_Bcast(header+1,3,MPIU_INT,0,comm);CHKERRQ(ierr); 3548 M = header[1]; N = header[2]; 3549 3550 /* If global sizes are set, check if they are consistent with that given in the file */ 3551 if (newmat->rmap->N >= 0 && newmat->rmap->N != M) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED,"Inconsistent # of rows:Matrix in file has (%D) and input matrix has (%D)",newmat->rmap->N,M); 3552 if (newmat->cmap->N >= 0 && newmat->cmap->N != N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED,"Inconsistent # of cols:Matrix in file has (%D) and input matrix has (%D)",newmat->cmap->N,N); 3553 3554 if (M != N) SETERRQ(PetscObjectComm((PetscObject)viewer),PETSC_ERR_SUP,"Can only do square matrices"); 3555 3556 /* 3557 This code adds extra rows to make sure the number of rows is 3558 divisible by the blocksize 3559 */ 3560 Mbs = M/bs; 3561 extra_rows = bs - M + bs*Mbs; 3562 if (extra_rows == bs) extra_rows = 0; 3563 else Mbs++; 3564 if (extra_rows && !rank) { 3565 ierr = PetscInfo(viewer,"Padding loaded matrix to match blocksize\n");CHKERRQ(ierr); 3566 } 3567 3568 /* determine ownership of all rows */ 3569 if (newmat->rmap->n < 0) { /* PETSC_DECIDE */ 3570 mbs = Mbs/size + ((Mbs % size) > rank); 3571 m = mbs*bs; 3572 } else { /* User set */ 3573 m = newmat->rmap->n; 3574 mbs = m/bs; 3575 } 3576 ierr = PetscMalloc2(size+1,&rowners,size+1,&browners);CHKERRQ(ierr); 3577 ierr = MPI_Allgather(&mbs,1,MPIU_INT,rowners+1,1,MPIU_INT,comm);CHKERRQ(ierr); 3578 3579 /* process 0 needs enough room for process with most rows */ 3580 if (!rank) { 3581 mmax = rowners[1]; 3582 for (i=2; i<=size; i++) { 3583 mmax = PetscMax(mmax,rowners[i]); 3584 } 3585 mmax*=bs; 3586 } else mmax = -1; /* unused, but compiler warns anyway */ 3587 3588 rowners[0] = 0; 3589 for (i=2; i<=size; i++) rowners[i] += rowners[i-1]; 3590 for (i=0; i<=size; i++) browners[i] = rowners[i]*bs; 3591 rstart = rowners[rank]; 3592 rend = rowners[rank+1]; 3593 3594 /* distribute row lengths to all processors */ 3595 ierr = PetscMalloc1(m,&locrowlens);CHKERRQ(ierr); 3596 if (!rank) { 3597 mend = m; 3598 if (size == 1) mend = mend - extra_rows; 3599 ierr = PetscBinaryRead(fd,locrowlens,mend,PETSC_INT);CHKERRQ(ierr); 3600 for (j=mend; j<m; j++) locrowlens[j] = 1; 3601 ierr = PetscMalloc1(mmax,&rowlengths);CHKERRQ(ierr); 3602 ierr = PetscCalloc1(size,&procsnz);CHKERRQ(ierr); 3603 for (j=0; j<m; j++) { 3604 procsnz[0] += locrowlens[j]; 3605 } 3606 for (i=1; i<size; i++) { 3607 mend = browners[i+1] - browners[i]; 3608 if (i == size-1) mend = mend - extra_rows; 3609 ierr = PetscBinaryRead(fd,rowlengths,mend,PETSC_INT);CHKERRQ(ierr); 3610 for (j=mend; j<browners[i+1] - browners[i]; j++) rowlengths[j] = 1; 3611 /* calculate the number of nonzeros on each processor */ 3612 for (j=0; j<browners[i+1]-browners[i]; j++) { 3613 procsnz[i] += rowlengths[j]; 3614 } 3615 ierr = MPI_Send(rowlengths,browners[i+1]-browners[i],MPIU_INT,i,tag,comm);CHKERRQ(ierr); 3616 } 3617 ierr = PetscFree(rowlengths);CHKERRQ(ierr); 3618 } else { 3619 ierr = MPI_Recv(locrowlens,m,MPIU_INT,0,tag,comm,&status);CHKERRQ(ierr); 3620 } 3621 3622 if (!rank) { 3623 /* determine max buffer needed and allocate it */ 3624 maxnz = procsnz[0]; 3625 for (i=1; i<size; i++) { 3626 maxnz = PetscMax(maxnz,procsnz[i]); 3627 } 3628 ierr = PetscMalloc1(maxnz,&cols);CHKERRQ(ierr); 3629 3630 /* read in my part of the matrix column indices */ 3631 nz = procsnz[0]; 3632 ierr = PetscMalloc1(nz+1,&ibuf);CHKERRQ(ierr); 3633 mycols = ibuf; 3634 if (size == 1) nz -= extra_rows; 3635 ierr = PetscBinaryRead(fd,mycols,nz,PETSC_INT);CHKERRQ(ierr); 3636 if (size == 1) { 3637 for (i=0; i< extra_rows; i++) mycols[nz+i] = M+i; 3638 } 3639 3640 /* read in every ones (except the last) and ship off */ 3641 for (i=1; i<size-1; i++) { 3642 nz = procsnz[i]; 3643 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 3644 ierr = MPI_Send(cols,nz,MPIU_INT,i,tag,comm);CHKERRQ(ierr); 3645 } 3646 /* read in the stuff for the last proc */ 3647 if (size != 1) { 3648 nz = procsnz[size-1] - extra_rows; /* the extra rows are not on the disk */ 3649 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 3650 for (i=0; i<extra_rows; i++) cols[nz+i] = M+i; 3651 ierr = MPI_Send(cols,nz+extra_rows,MPIU_INT,size-1,tag,comm);CHKERRQ(ierr); 3652 } 3653 ierr = PetscFree(cols);CHKERRQ(ierr); 3654 } else { 3655 /* determine buffer space needed for message */ 3656 nz = 0; 3657 for (i=0; i<m; i++) { 3658 nz += locrowlens[i]; 3659 } 3660 ierr = PetscMalloc1(nz+1,&ibuf);CHKERRQ(ierr); 3661 mycols = ibuf; 3662 /* receive message of column indices*/ 3663 ierr = MPI_Recv(mycols,nz,MPIU_INT,0,tag,comm,&status);CHKERRQ(ierr); 3664 ierr = MPI_Get_count(&status,MPIU_INT,&maxnz);CHKERRQ(ierr); 3665 if (maxnz != nz) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file"); 3666 } 3667 3668 /* loop over local rows, determining number of off diagonal entries */ 3669 ierr = PetscMalloc2(rend-rstart,&dlens,rend-rstart,&odlens);CHKERRQ(ierr); 3670 ierr = PetscCalloc3(Mbs,&mask,Mbs,&masked1,Mbs,&masked2);CHKERRQ(ierr); 3671 rowcount = 0; nzcount = 0; 3672 for (i=0; i<mbs; i++) { 3673 dcount = 0; 3674 odcount = 0; 3675 for (j=0; j<bs; j++) { 3676 kmax = locrowlens[rowcount]; 3677 for (k=0; k<kmax; k++) { 3678 tmp = mycols[nzcount++]/bs; 3679 if (!mask[tmp]) { 3680 mask[tmp] = 1; 3681 if (tmp < rstart || tmp >= rend) masked2[odcount++] = tmp; 3682 else masked1[dcount++] = tmp; 3683 } 3684 } 3685 rowcount++; 3686 } 3687 3688 dlens[i] = dcount; 3689 odlens[i] = odcount; 3690 3691 /* zero out the mask elements we set */ 3692 for (j=0; j<dcount; j++) mask[masked1[j]] = 0; 3693 for (j=0; j<odcount; j++) mask[masked2[j]] = 0; 3694 } 3695 3696 ierr = MatSetSizes(newmat,m,m,M+extra_rows,N+extra_rows);CHKERRQ(ierr); 3697 ierr = MatMPIBAIJSetPreallocation(newmat,bs,0,dlens,0,odlens);CHKERRQ(ierr); 3698 3699 if (!rank) { 3700 ierr = PetscMalloc1(maxnz+1,&buf);CHKERRQ(ierr); 3701 /* read in my part of the matrix numerical values */ 3702 nz = procsnz[0]; 3703 vals = buf; 3704 mycols = ibuf; 3705 if (size == 1) nz -= extra_rows; 3706 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 3707 if (size == 1) { 3708 for (i=0; i< extra_rows; i++) vals[nz+i] = 1.0; 3709 } 3710 3711 /* insert into matrix */ 3712 jj = rstart*bs; 3713 for (i=0; i<m; i++) { 3714 ierr = MatSetValues_MPIBAIJ(newmat,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 3715 mycols += locrowlens[i]; 3716 vals += locrowlens[i]; 3717 jj++; 3718 } 3719 /* read in other processors (except the last one) and ship out */ 3720 for (i=1; i<size-1; i++) { 3721 nz = procsnz[i]; 3722 vals = buf; 3723 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 3724 ierr = MPIULong_Send(vals,nz,MPIU_SCALAR,i,((PetscObject)newmat)->tag,comm);CHKERRQ(ierr); 3725 } 3726 /* the last proc */ 3727 if (size != 1) { 3728 nz = procsnz[i] - extra_rows; 3729 vals = buf; 3730 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 3731 for (i=0; i<extra_rows; i++) vals[nz+i] = 1.0; 3732 ierr = MPIULong_Send(vals,nz+extra_rows,MPIU_SCALAR,size-1,((PetscObject)newmat)->tag,comm);CHKERRQ(ierr); 3733 } 3734 ierr = PetscFree(procsnz);CHKERRQ(ierr); 3735 } else { 3736 /* receive numeric values */ 3737 ierr = PetscMalloc1(nz+1,&buf);CHKERRQ(ierr); 3738 3739 /* receive message of values*/ 3740 vals = buf; 3741 mycols = ibuf; 3742 ierr = MPIULong_Recv(vals,nz,MPIU_SCALAR,0,((PetscObject)newmat)->tag,comm);CHKERRQ(ierr); 3743 3744 /* insert into matrix */ 3745 jj = rstart*bs; 3746 for (i=0; i<m; i++) { 3747 ierr = MatSetValues_MPIBAIJ(newmat,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 3748 mycols += locrowlens[i]; 3749 vals += locrowlens[i]; 3750 jj++; 3751 } 3752 } 3753 ierr = PetscFree(locrowlens);CHKERRQ(ierr); 3754 ierr = PetscFree(buf);CHKERRQ(ierr); 3755 ierr = PetscFree(ibuf);CHKERRQ(ierr); 3756 ierr = PetscFree2(rowners,browners);CHKERRQ(ierr); 3757 ierr = PetscFree2(dlens,odlens);CHKERRQ(ierr); 3758 ierr = PetscFree3(mask,masked1,masked2);CHKERRQ(ierr); 3759 ierr = MatAssemblyBegin(newmat,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3760 ierr = MatAssemblyEnd(newmat,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3761 PetscFunctionReturn(0); 3762 } 3763 3764 #undef __FUNCT__ 3765 #define __FUNCT__ "MatMPIBAIJSetHashTableFactor" 3766 /*@ 3767 MatMPIBAIJSetHashTableFactor - Sets the factor required to compute the size of the HashTable. 3768 3769 Input Parameters: 3770 . mat - the matrix 3771 . fact - factor 3772 3773 Not Collective, each process can use a different factor 3774 3775 Level: advanced 3776 3777 Notes: 3778 This can also be set by the command line option: -mat_use_hash_table <fact> 3779 3780 .keywords: matrix, hashtable, factor, HT 3781 3782 .seealso: MatSetOption() 3783 @*/ 3784 PetscErrorCode MatMPIBAIJSetHashTableFactor(Mat mat,PetscReal fact) 3785 { 3786 PetscErrorCode ierr; 3787 3788 PetscFunctionBegin; 3789 ierr = PetscTryMethod(mat,"MatSetHashTableFactor_C",(Mat,PetscReal),(mat,fact));CHKERRQ(ierr); 3790 PetscFunctionReturn(0); 3791 } 3792 3793 #undef __FUNCT__ 3794 #define __FUNCT__ "MatSetHashTableFactor_MPIBAIJ" 3795 PetscErrorCode MatSetHashTableFactor_MPIBAIJ(Mat mat,PetscReal fact) 3796 { 3797 Mat_MPIBAIJ *baij; 3798 3799 PetscFunctionBegin; 3800 baij = (Mat_MPIBAIJ*)mat->data; 3801 baij->ht_fact = fact; 3802 PetscFunctionReturn(0); 3803 } 3804 3805 #undef __FUNCT__ 3806 #define __FUNCT__ "MatMPIBAIJGetSeqBAIJ" 3807 PetscErrorCode MatMPIBAIJGetSeqBAIJ(Mat A,Mat *Ad,Mat *Ao,const PetscInt *colmap[]) 3808 { 3809 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 3810 3811 PetscFunctionBegin; 3812 if (Ad) *Ad = a->A; 3813 if (Ao) *Ao = a->B; 3814 if (colmap) *colmap = a->garray; 3815 PetscFunctionReturn(0); 3816 } 3817 3818 /* 3819 Special version for direct calls from Fortran (to eliminate two function call overheads 3820 */ 3821 #if defined(PETSC_HAVE_FORTRAN_CAPS) 3822 #define matmpibaijsetvaluesblocked_ MATMPIBAIJSETVALUESBLOCKED 3823 #elif !defined(PETSC_HAVE_FORTRAN_UNDERSCORE) 3824 #define matmpibaijsetvaluesblocked_ matmpibaijsetvaluesblocked 3825 #endif 3826 3827 #undef __FUNCT__ 3828 #define __FUNCT__ "matmpibiajsetvaluesblocked" 3829 /*@C 3830 MatMPIBAIJSetValuesBlocked - Direct Fortran call to replace call to MatSetValuesBlocked() 3831 3832 Collective on Mat 3833 3834 Input Parameters: 3835 + mat - the matrix 3836 . min - number of input rows 3837 . im - input rows 3838 . nin - number of input columns 3839 . in - input columns 3840 . v - numerical values input 3841 - addvin - INSERT_VALUES or ADD_VALUES 3842 3843 Notes: This has a complete copy of MatSetValuesBlocked_MPIBAIJ() which is terrible code un-reuse. 3844 3845 Level: advanced 3846 3847 .seealso: MatSetValuesBlocked() 3848 @*/ 3849 PetscErrorCode matmpibaijsetvaluesblocked_(Mat *matin,PetscInt *min,const PetscInt im[],PetscInt *nin,const PetscInt in[],const MatScalar v[],InsertMode *addvin) 3850 { 3851 /* convert input arguments to C version */ 3852 Mat mat = *matin; 3853 PetscInt m = *min, n = *nin; 3854 InsertMode addv = *addvin; 3855 3856 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 3857 const MatScalar *value; 3858 MatScalar *barray = baij->barray; 3859 PetscBool roworiented = baij->roworiented; 3860 PetscErrorCode ierr; 3861 PetscInt i,j,ii,jj,row,col,rstart=baij->rstartbs; 3862 PetscInt rend=baij->rendbs,cstart=baij->cstartbs,stepval; 3863 PetscInt cend=baij->cendbs,bs=mat->rmap->bs,bs2=baij->bs2; 3864 3865 PetscFunctionBegin; 3866 /* tasks normally handled by MatSetValuesBlocked() */ 3867 if (mat->insertmode == NOT_SET_VALUES) mat->insertmode = addv; 3868 #if defined(PETSC_USE_DEBUG) 3869 else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values"); 3870 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 3871 #endif 3872 if (mat->assembled) { 3873 mat->was_assembled = PETSC_TRUE; 3874 mat->assembled = PETSC_FALSE; 3875 } 3876 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 3877 3878 3879 if (!barray) { 3880 ierr = PetscMalloc1(bs2,&barray);CHKERRQ(ierr); 3881 baij->barray = barray; 3882 } 3883 3884 if (roworiented) stepval = (n-1)*bs; 3885 else stepval = (m-1)*bs; 3886 3887 for (i=0; i<m; i++) { 3888 if (im[i] < 0) continue; 3889 #if defined(PETSC_USE_DEBUG) 3890 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large, row %D max %D",im[i],baij->Mbs-1); 3891 #endif 3892 if (im[i] >= rstart && im[i] < rend) { 3893 row = im[i] - rstart; 3894 for (j=0; j<n; j++) { 3895 /* If NumCol = 1 then a copy is not required */ 3896 if ((roworiented) && (n == 1)) { 3897 barray = (MatScalar*)v + i*bs2; 3898 } else if ((!roworiented) && (m == 1)) { 3899 barray = (MatScalar*)v + j*bs2; 3900 } else { /* Here a copy is required */ 3901 if (roworiented) { 3902 value = v + i*(stepval+bs)*bs + j*bs; 3903 } else { 3904 value = v + j*(stepval+bs)*bs + i*bs; 3905 } 3906 for (ii=0; ii<bs; ii++,value+=stepval) { 3907 for (jj=0; jj<bs; jj++) { 3908 *barray++ = *value++; 3909 } 3910 } 3911 barray -=bs2; 3912 } 3913 3914 if (in[j] >= cstart && in[j] < cend) { 3915 col = in[j] - cstart; 3916 ierr = MatSetValuesBlocked_SeqBAIJ_Inlined(baij->A,row,col,barray,addv,im[i],in[j]);CHKERRQ(ierr); 3917 } else if (in[j] < 0) continue; 3918 #if defined(PETSC_USE_DEBUG) 3919 else if (in[j] >= baij->Nbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column too large, col %D max %D",in[j],baij->Nbs-1); 3920 #endif 3921 else { 3922 if (mat->was_assembled) { 3923 if (!baij->colmap) { 3924 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 3925 } 3926 3927 #if defined(PETSC_USE_DEBUG) 3928 #if defined(PETSC_USE_CTABLE) 3929 { PetscInt data; 3930 ierr = PetscTableFind(baij->colmap,in[j]+1,&data);CHKERRQ(ierr); 3931 if ((data - 1) % bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Incorrect colmap"); 3932 } 3933 #else 3934 if ((baij->colmap[in[j]] - 1) % bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Incorrect colmap"); 3935 #endif 3936 #endif 3937 #if defined(PETSC_USE_CTABLE) 3938 ierr = PetscTableFind(baij->colmap,in[j]+1,&col);CHKERRQ(ierr); 3939 col = (col - 1)/bs; 3940 #else 3941 col = (baij->colmap[in[j]] - 1)/bs; 3942 #endif 3943 if (col < 0 && !((Mat_SeqBAIJ*)(baij->A->data))->nonew) { 3944 ierr = MatDisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 3945 col = in[j]; 3946 } 3947 } else col = in[j]; 3948 ierr = MatSetValuesBlocked_SeqBAIJ_Inlined(baij->B,row,col,barray,addv,im[i],in[j]);CHKERRQ(ierr); 3949 } 3950 } 3951 } else { 3952 if (!baij->donotstash) { 3953 if (roworiented) { 3954 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 3955 } else { 3956 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 3957 } 3958 } 3959 } 3960 } 3961 3962 /* task normally handled by MatSetValuesBlocked() */ 3963 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 3964 PetscFunctionReturn(0); 3965 } 3966 3967 #undef __FUNCT__ 3968 #define __FUNCT__ "MatCreateMPIBAIJWithArrays" 3969 /*@ 3970 MatCreateMPIBAIJWithArrays - creates a MPI BAIJ matrix using arrays that contain in standard 3971 CSR format the local rows. 3972 3973 Collective on MPI_Comm 3974 3975 Input Parameters: 3976 + comm - MPI communicator 3977 . bs - the block size, only a block size of 1 is supported 3978 . m - number of local rows (Cannot be PETSC_DECIDE) 3979 . n - This value should be the same as the local size used in creating the 3980 x vector for the matrix-vector product y = Ax. (or PETSC_DECIDE to have 3981 calculated if N is given) For square matrices n is almost always m. 3982 . M - number of global rows (or PETSC_DETERMINE to have calculated if m is given) 3983 . N - number of global columns (or PETSC_DETERMINE to have calculated if n is given) 3984 . i - row indices 3985 . j - column indices 3986 - a - matrix values 3987 3988 Output Parameter: 3989 . mat - the matrix 3990 3991 Level: intermediate 3992 3993 Notes: 3994 The i, j, and a arrays ARE copied by this routine into the internal format used by PETSc; 3995 thus you CANNOT change the matrix entries by changing the values of a[] after you have 3996 called this routine. Use MatCreateMPIAIJWithSplitArrays() to avoid needing to copy the arrays. 3997 3998 The order of the entries in values is the same as the block compressed sparse row storage format; that is, it is 3999 the same as a three dimensional array in Fortran values(bs,bs,nnz) that contains the first column of the first 4000 block, followed by the second column of the first block etc etc. That is, the blocks are contiguous in memory 4001 with column-major ordering within blocks. 4002 4003 The i and j indices are 0 based, and i indices are indices corresponding to the local j array. 4004 4005 .keywords: matrix, aij, compressed row, sparse, parallel 4006 4007 .seealso: MatCreate(), MatCreateSeqAIJ(), MatSetValues(), MatMPIAIJSetPreallocation(), MatMPIAIJSetPreallocationCSR(), 4008 MPIAIJ, MatCreateAIJ(), MatCreateMPIAIJWithSplitArrays() 4009 @*/ 4010 PetscErrorCode MatCreateMPIBAIJWithArrays(MPI_Comm comm,PetscInt bs,PetscInt m,PetscInt n,PetscInt M,PetscInt N,const PetscInt i[],const PetscInt j[],const PetscScalar a[],Mat *mat) 4011 { 4012 PetscErrorCode ierr; 4013 4014 PetscFunctionBegin; 4015 if (i[0]) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"i (row indices) must start with 0"); 4016 if (m < 0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"local number of rows (m) cannot be PETSC_DECIDE, or negative"); 4017 ierr = MatCreate(comm,mat);CHKERRQ(ierr); 4018 ierr = MatSetSizes(*mat,m,n,M,N);CHKERRQ(ierr); 4019 ierr = MatSetType(*mat,MATMPISBAIJ);CHKERRQ(ierr); 4020 ierr = MatSetOption(*mat,MAT_ROW_ORIENTED,PETSC_FALSE);CHKERRQ(ierr); 4021 ierr = MatMPIBAIJSetPreallocationCSR(*mat,bs,i,j,a);CHKERRQ(ierr); 4022 ierr = MatSetOption(*mat,MAT_ROW_ORIENTED,PETSC_TRUE);CHKERRQ(ierr); 4023 PetscFunctionReturn(0); 4024 } 4025 4026 #undef __FUNCT__ 4027 #define __FUNCT__ "MatCreateMPIMatConcatenateSeqMat_MPIBAIJ" 4028 PetscErrorCode MatCreateMPIMatConcatenateSeqMat_MPIBAIJ(MPI_Comm comm,Mat inmat,PetscInt n,MatReuse scall,Mat *outmat) 4029 { 4030 PetscErrorCode ierr; 4031 PetscInt m,N,i,rstart,nnz,Ii,bs,cbs; 4032 PetscInt *indx; 4033 PetscScalar *values; 4034 4035 PetscFunctionBegin; 4036 ierr = MatGetSize(inmat,&m,&N);CHKERRQ(ierr); 4037 if (scall == MAT_INITIAL_MATRIX) { /* symbolic phase */ 4038 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)inmat->data; 4039 PetscInt *dnz,*onz,sum,mbs,Nbs; 4040 PetscInt *bindx,rmax=a->rmax,j; 4041 4042 ierr = MatGetBlockSizes(inmat,&bs,&cbs);CHKERRQ(ierr); 4043 mbs = m/bs; Nbs = N/cbs; 4044 if (n == PETSC_DECIDE) { 4045 ierr = PetscSplitOwnership(comm,&n,&Nbs);CHKERRQ(ierr); 4046 } 4047 /* Check sum(n) = Nbs */ 4048 ierr = MPI_Allreduce(&n,&sum,1,MPIU_INT,MPI_SUM,comm);CHKERRQ(ierr); 4049 if (sum != Nbs) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_INCOMP,"Sum of local columns != global columns %d",Nbs); 4050 4051 ierr = MPI_Scan(&mbs, &rstart,1,MPIU_INT,MPI_SUM,comm);CHKERRQ(ierr); 4052 rstart -= mbs; 4053 4054 ierr = PetscMalloc1(rmax,&bindx);CHKERRQ(ierr); 4055 ierr = MatPreallocateInitialize(comm,mbs,n,dnz,onz);CHKERRQ(ierr); 4056 for (i=0; i<mbs; i++) { 4057 ierr = MatGetRow_SeqBAIJ(inmat,i*bs,&nnz,&indx,NULL);CHKERRQ(ierr); /* non-blocked nnz and indx */ 4058 nnz = nnz/bs; 4059 for (j=0; j<nnz; j++) bindx[j] = indx[j*bs]/bs; 4060 ierr = MatPreallocateSet(i+rstart,nnz,bindx,dnz,onz);CHKERRQ(ierr); 4061 ierr = MatRestoreRow_SeqBAIJ(inmat,i*bs,&nnz,&indx,NULL);CHKERRQ(ierr); 4062 } 4063 ierr = PetscFree(bindx);CHKERRQ(ierr); 4064 4065 ierr = MatCreate(comm,outmat);CHKERRQ(ierr); 4066 ierr = MatSetSizes(*outmat,m,n*bs,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 4067 ierr = MatSetBlockSizes(*outmat,bs,cbs);CHKERRQ(ierr); 4068 ierr = MatSetType(*outmat,MATMPIBAIJ);CHKERRQ(ierr); 4069 ierr = MatMPIBAIJSetPreallocation(*outmat,bs,0,dnz,0,onz);CHKERRQ(ierr); 4070 ierr = MatPreallocateFinalize(dnz,onz);CHKERRQ(ierr); 4071 } 4072 4073 /* numeric phase */ 4074 ierr = MatGetBlockSizes(inmat,&bs,&cbs);CHKERRQ(ierr); 4075 ierr = MatGetOwnershipRange(*outmat,&rstart,NULL);CHKERRQ(ierr); 4076 4077 for (i=0; i<m; i++) { 4078 ierr = MatGetRow_SeqBAIJ(inmat,i,&nnz,&indx,&values);CHKERRQ(ierr); 4079 Ii = i + rstart; 4080 ierr = MatSetValues(*outmat,1,&Ii,nnz,indx,values,INSERT_VALUES);CHKERRQ(ierr); 4081 ierr = MatRestoreRow_SeqBAIJ(inmat,i,&nnz,&indx,&values);CHKERRQ(ierr); 4082 } 4083 ierr = MatAssemblyBegin(*outmat,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 4084 ierr = MatAssemblyEnd(*outmat,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 4085 PetscFunctionReturn(0); 4086 } 4087