1 /*$Id: mpibaij.c,v 1.234 2001/09/25 22:56:49 balay Exp $*/ 2 3 #include "src/mat/impls/baij/mpi/mpibaij.h" /*I "petscmat.h" I*/ 4 #include "src/vec/vecimpl.h" 5 6 EXTERN int MatSetUpMultiply_MPIBAIJ(Mat); 7 EXTERN int DisAssemble_MPIBAIJ(Mat); 8 EXTERN int MatIncreaseOverlap_MPIBAIJ(Mat,int,IS[],int); 9 EXTERN int MatGetSubMatrices_MPIBAIJ(Mat,int,const IS[],const IS[],MatReuse,Mat *[]); 10 EXTERN int MatGetValues_SeqBAIJ(Mat,int,const int[],int,const int [],PetscScalar []); 11 EXTERN int MatSetValues_SeqBAIJ(Mat,int,const int[],int,const int [],const PetscScalar [],InsertMode); 12 EXTERN int MatSetValuesBlocked_SeqBAIJ(Mat,int,const int[],int,const int[],const PetscScalar[],InsertMode); 13 EXTERN int MatGetRow_SeqBAIJ(Mat,int,int*,int*[],PetscScalar*[]); 14 EXTERN int MatRestoreRow_SeqBAIJ(Mat,int,int*,int*[],PetscScalar*[]); 15 EXTERN int MatPrintHelp_SeqBAIJ(Mat); 16 EXTERN int MatZeroRows_SeqBAIJ(Mat,IS,const PetscScalar*); 17 18 /* UGLY, ugly, ugly 19 When MatScalar == PetscScalar the function MatSetValuesBlocked_MPIBAIJ_MatScalar() does 20 not exist. Otherwise ..._MatScalar() takes matrix elements in single precision and 21 inserts them into the single precision data structure. The function MatSetValuesBlocked_MPIBAIJ() 22 converts the entries into single precision and then calls ..._MatScalar() to put them 23 into the single precision data structures. 24 */ 25 #if defined(PETSC_USE_MAT_SINGLE) 26 EXTERN int MatSetValuesBlocked_SeqBAIJ_MatScalar(Mat,int,const int*,int,const int*,const MatScalar*,InsertMode); 27 EXTERN int MatSetValues_MPIBAIJ_MatScalar(Mat,int,const int*,int,const int*,const MatScalar*,InsertMode); 28 EXTERN int MatSetValuesBlocked_MPIBAIJ_MatScalar(Mat,int,const int*,int,const int*,const MatScalar*,InsertMode); 29 EXTERN int MatSetValues_MPIBAIJ_HT_MatScalar(Mat,int,const int*,int,const int*,const MatScalar*,InsertMode); 30 EXTERN int MatSetValuesBlocked_MPIBAIJ_HT_MatScalar(Mat,int,const int*,int,const int*,const MatScalar*,InsertMode); 31 #else 32 #define MatSetValuesBlocked_SeqBAIJ_MatScalar MatSetValuesBlocked_SeqBAIJ 33 #define MatSetValues_MPIBAIJ_MatScalar MatSetValues_MPIBAIJ 34 #define MatSetValuesBlocked_MPIBAIJ_MatScalar MatSetValuesBlocked_MPIBAIJ 35 #define MatSetValues_MPIBAIJ_HT_MatScalar MatSetValues_MPIBAIJ_HT 36 #define MatSetValuesBlocked_MPIBAIJ_HT_MatScalar MatSetValuesBlocked_MPIBAIJ_HT 37 #endif 38 39 #undef __FUNCT__ 40 #define __FUNCT__ "MatGetRowMax_MPIBAIJ" 41 int MatGetRowMax_MPIBAIJ(Mat A,Vec v) 42 { 43 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 44 int ierr,i; 45 PetscScalar *va,*vb; 46 Vec vtmp; 47 48 PetscFunctionBegin; 49 50 ierr = MatGetRowMax(a->A,v);CHKERRQ(ierr); 51 ierr = VecGetArray(v,&va);CHKERRQ(ierr); 52 53 ierr = VecCreateSeq(PETSC_COMM_SELF,A->m,&vtmp);CHKERRQ(ierr); 54 ierr = MatGetRowMax(a->B,vtmp);CHKERRQ(ierr); 55 ierr = VecGetArray(vtmp,&vb);CHKERRQ(ierr); 56 57 for (i=0; i<A->m; i++){ 58 if (PetscAbsScalar(va[i]) < PetscAbsScalar(vb[i])) va[i] = vb[i]; 59 } 60 61 ierr = VecRestoreArray(v,&va);CHKERRQ(ierr); 62 ierr = VecRestoreArray(vtmp,&vb);CHKERRQ(ierr); 63 ierr = VecDestroy(vtmp);CHKERRQ(ierr); 64 65 PetscFunctionReturn(0); 66 } 67 68 EXTERN_C_BEGIN 69 #undef __FUNCT__ 70 #define __FUNCT__ "MatStoreValues_MPIBAIJ" 71 int MatStoreValues_MPIBAIJ(Mat mat) 72 { 73 Mat_MPIBAIJ *aij = (Mat_MPIBAIJ *)mat->data; 74 int ierr; 75 76 PetscFunctionBegin; 77 ierr = MatStoreValues(aij->A);CHKERRQ(ierr); 78 ierr = MatStoreValues(aij->B);CHKERRQ(ierr); 79 PetscFunctionReturn(0); 80 } 81 EXTERN_C_END 82 83 EXTERN_C_BEGIN 84 #undef __FUNCT__ 85 #define __FUNCT__ "MatRetrieveValues_MPIBAIJ" 86 int MatRetrieveValues_MPIBAIJ(Mat mat) 87 { 88 Mat_MPIBAIJ *aij = (Mat_MPIBAIJ *)mat->data; 89 int ierr; 90 91 PetscFunctionBegin; 92 ierr = MatRetrieveValues(aij->A);CHKERRQ(ierr); 93 ierr = MatRetrieveValues(aij->B);CHKERRQ(ierr); 94 PetscFunctionReturn(0); 95 } 96 EXTERN_C_END 97 98 /* 99 Local utility routine that creates a mapping from the global column 100 number to the local number in the off-diagonal part of the local 101 storage of the matrix. This is done in a non scable way since the 102 length of colmap equals the global matrix length. 103 */ 104 #undef __FUNCT__ 105 #define __FUNCT__ "CreateColmap_MPIBAIJ_Private" 106 static int CreateColmap_MPIBAIJ_Private(Mat mat) 107 { 108 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 109 Mat_SeqBAIJ *B = (Mat_SeqBAIJ*)baij->B->data; 110 int nbs = B->nbs,i,bs=B->bs,ierr; 111 112 PetscFunctionBegin; 113 #if defined (PETSC_USE_CTABLE) 114 ierr = PetscTableCreate(baij->nbs,&baij->colmap);CHKERRQ(ierr); 115 for (i=0; i<nbs; i++){ 116 ierr = PetscTableAdd(baij->colmap,baij->garray[i]+1,i*bs+1);CHKERRQ(ierr); 117 } 118 #else 119 ierr = PetscMalloc((baij->Nbs+1)*sizeof(int),&baij->colmap);CHKERRQ(ierr); 120 PetscLogObjectMemory(mat,baij->Nbs*sizeof(int)); 121 ierr = PetscMemzero(baij->colmap,baij->Nbs*sizeof(int));CHKERRQ(ierr); 122 for (i=0; i<nbs; i++) baij->colmap[baij->garray[i]] = i*bs+1; 123 #endif 124 PetscFunctionReturn(0); 125 } 126 127 #define CHUNKSIZE 10 128 129 #define MatSetValues_SeqBAIJ_A_Private(row,col,value,addv) \ 130 { \ 131 \ 132 brow = row/bs; \ 133 rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; \ 134 rmax = aimax[brow]; nrow = ailen[brow]; \ 135 bcol = col/bs; \ 136 ridx = row % bs; cidx = col % bs; \ 137 low = 0; high = nrow; \ 138 while (high-low > 3) { \ 139 t = (low+high)/2; \ 140 if (rp[t] > bcol) high = t; \ 141 else low = t; \ 142 } \ 143 for (_i=low; _i<high; _i++) { \ 144 if (rp[_i] > bcol) break; \ 145 if (rp[_i] == bcol) { \ 146 bap = ap + bs2*_i + bs*cidx + ridx; \ 147 if (addv == ADD_VALUES) *bap += value; \ 148 else *bap = value; \ 149 goto a_noinsert; \ 150 } \ 151 } \ 152 if (a->nonew == 1) goto a_noinsert; \ 153 else if (a->nonew == -1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero into matrix"); \ 154 if (nrow >= rmax) { \ 155 /* there is no extra room in row, therefore enlarge */ \ 156 int new_nz = ai[a->mbs] + CHUNKSIZE,len,*new_i,*new_j; \ 157 MatScalar *new_a; \ 158 \ 159 if (a->nonew == -2) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero in the matrix"); \ 160 \ 161 /* malloc new storage space */ \ 162 len = new_nz*(sizeof(int)+bs2*sizeof(MatScalar))+(a->mbs+1)*sizeof(int); \ 163 ierr = PetscMalloc(len,&new_a);CHKERRQ(ierr); \ 164 new_j = (int*)(new_a + bs2*new_nz); \ 165 new_i = new_j + new_nz; \ 166 \ 167 /* copy over old data into new slots */ \ 168 for (ii=0; ii<brow+1; ii++) {new_i[ii] = ai[ii];} \ 169 for (ii=brow+1; ii<a->mbs+1; ii++) {new_i[ii] = ai[ii]+CHUNKSIZE;} \ 170 ierr = PetscMemcpy(new_j,aj,(ai[brow]+nrow)*sizeof(int));CHKERRQ(ierr); \ 171 len = (new_nz - CHUNKSIZE - ai[brow] - nrow); \ 172 ierr = PetscMemcpy(new_j+ai[brow]+nrow+CHUNKSIZE,aj+ai[brow]+nrow,len*sizeof(int));CHKERRQ(ierr); \ 173 ierr = PetscMemcpy(new_a,aa,(ai[brow]+nrow)*bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 174 ierr = PetscMemzero(new_a+bs2*(ai[brow]+nrow),bs2*CHUNKSIZE*sizeof(PetscScalar));CHKERRQ(ierr); \ 175 ierr = PetscMemcpy(new_a+bs2*(ai[brow]+nrow+CHUNKSIZE), \ 176 aa+bs2*(ai[brow]+nrow),bs2*len*sizeof(MatScalar));CHKERRQ(ierr); \ 177 /* free up old matrix storage */ \ 178 ierr = PetscFree(a->a);CHKERRQ(ierr); \ 179 if (!a->singlemalloc) { \ 180 ierr = PetscFree(a->i);CHKERRQ(ierr); \ 181 ierr = PetscFree(a->j);CHKERRQ(ierr);\ 182 } \ 183 aa = a->a = new_a; ai = a->i = new_i; aj = a->j = new_j; \ 184 a->singlemalloc = PETSC_TRUE; \ 185 \ 186 rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; \ 187 rmax = aimax[brow] = aimax[brow] + CHUNKSIZE; \ 188 PetscLogObjectMemory(A,CHUNKSIZE*(sizeof(int) + bs2*sizeof(MatScalar))); \ 189 a->maxnz += bs2*CHUNKSIZE; \ 190 a->reallocs++; \ 191 a->nz++; \ 192 } \ 193 N = nrow++ - 1; \ 194 /* shift up all the later entries in this row */ \ 195 for (ii=N; ii>=_i; ii--) { \ 196 rp[ii+1] = rp[ii]; \ 197 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 198 } \ 199 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr); } \ 200 rp[_i] = bcol; \ 201 ap[bs2*_i + bs*cidx + ridx] = value; \ 202 a_noinsert:; \ 203 ailen[brow] = nrow; \ 204 } 205 206 #define MatSetValues_SeqBAIJ_B_Private(row,col,value,addv) \ 207 { \ 208 brow = row/bs; \ 209 rp = bj + bi[brow]; ap = ba + bs2*bi[brow]; \ 210 rmax = bimax[brow]; nrow = bilen[brow]; \ 211 bcol = col/bs; \ 212 ridx = row % bs; cidx = col % bs; \ 213 low = 0; high = nrow; \ 214 while (high-low > 3) { \ 215 t = (low+high)/2; \ 216 if (rp[t] > bcol) high = t; \ 217 else low = t; \ 218 } \ 219 for (_i=low; _i<high; _i++) { \ 220 if (rp[_i] > bcol) break; \ 221 if (rp[_i] == bcol) { \ 222 bap = ap + bs2*_i + bs*cidx + ridx; \ 223 if (addv == ADD_VALUES) *bap += value; \ 224 else *bap = value; \ 225 goto b_noinsert; \ 226 } \ 227 } \ 228 if (b->nonew == 1) goto b_noinsert; \ 229 else if (b->nonew == -1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero into matrix"); \ 230 if (nrow >= rmax) { \ 231 /* there is no extra room in row, therefore enlarge */ \ 232 int new_nz = bi[b->mbs] + CHUNKSIZE,len,*new_i,*new_j; \ 233 MatScalar *new_a; \ 234 \ 235 if (b->nonew == -2) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero in the matrix"); \ 236 \ 237 /* malloc new storage space */ \ 238 len = new_nz*(sizeof(int)+bs2*sizeof(MatScalar))+(b->mbs+1)*sizeof(int); \ 239 ierr = PetscMalloc(len,&new_a);CHKERRQ(ierr); \ 240 new_j = (int*)(new_a + bs2*new_nz); \ 241 new_i = new_j + new_nz; \ 242 \ 243 /* copy over old data into new slots */ \ 244 for (ii=0; ii<brow+1; ii++) {new_i[ii] = bi[ii];} \ 245 for (ii=brow+1; ii<b->mbs+1; ii++) {new_i[ii] = bi[ii]+CHUNKSIZE;} \ 246 ierr = PetscMemcpy(new_j,bj,(bi[brow]+nrow)*sizeof(int));CHKERRQ(ierr); \ 247 len = (new_nz - CHUNKSIZE - bi[brow] - nrow); \ 248 ierr = PetscMemcpy(new_j+bi[brow]+nrow+CHUNKSIZE,bj+bi[brow]+nrow,len*sizeof(int));CHKERRQ(ierr); \ 249 ierr = PetscMemcpy(new_a,ba,(bi[brow]+nrow)*bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 250 ierr = PetscMemzero(new_a+bs2*(bi[brow]+nrow),bs2*CHUNKSIZE*sizeof(MatScalar));CHKERRQ(ierr); \ 251 ierr = PetscMemcpy(new_a+bs2*(bi[brow]+nrow+CHUNKSIZE), \ 252 ba+bs2*(bi[brow]+nrow),bs2*len*sizeof(MatScalar));CHKERRQ(ierr); \ 253 /* free up old matrix storage */ \ 254 ierr = PetscFree(b->a);CHKERRQ(ierr); \ 255 if (!b->singlemalloc) { \ 256 ierr = PetscFree(b->i);CHKERRQ(ierr); \ 257 ierr = PetscFree(b->j);CHKERRQ(ierr); \ 258 } \ 259 ba = b->a = new_a; bi = b->i = new_i; bj = b->j = new_j; \ 260 b->singlemalloc = PETSC_TRUE; \ 261 \ 262 rp = bj + bi[brow]; ap = ba + bs2*bi[brow]; \ 263 rmax = bimax[brow] = bimax[brow] + CHUNKSIZE; \ 264 PetscLogObjectMemory(B,CHUNKSIZE*(sizeof(int) + bs2*sizeof(MatScalar))); \ 265 b->maxnz += bs2*CHUNKSIZE; \ 266 b->reallocs++; \ 267 b->nz++; \ 268 } \ 269 N = nrow++ - 1; \ 270 /* shift up all the later entries in this row */ \ 271 for (ii=N; ii>=_i; ii--) { \ 272 rp[ii+1] = rp[ii]; \ 273 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 274 } \ 275 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr);} \ 276 rp[_i] = bcol; \ 277 ap[bs2*_i + bs*cidx + ridx] = value; \ 278 b_noinsert:; \ 279 bilen[brow] = nrow; \ 280 } 281 282 #if defined(PETSC_USE_MAT_SINGLE) 283 #undef __FUNCT__ 284 #define __FUNCT__ "MatSetValues_MPIBAIJ" 285 int MatSetValues_MPIBAIJ(Mat mat,int m,const int im[],int n,const int in[],const PetscScalar v[],InsertMode addv) 286 { 287 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data; 288 int ierr,i,N = m*n; 289 MatScalar *vsingle; 290 291 PetscFunctionBegin; 292 if (N > b->setvalueslen) { 293 if (b->setvaluescopy) {ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr);} 294 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 295 b->setvalueslen = N; 296 } 297 vsingle = b->setvaluescopy; 298 299 for (i=0; i<N; i++) { 300 vsingle[i] = v[i]; 301 } 302 ierr = MatSetValues_MPIBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 303 PetscFunctionReturn(0); 304 } 305 306 #undef __FUNCT__ 307 #define __FUNCT__ "MatSetValuesBlocked_MPIBAIJ" 308 int MatSetValuesBlocked_MPIBAIJ(Mat mat,int m,const int im[],int n,const int in[],const PetscScalar v[],InsertMode addv) 309 { 310 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data; 311 int ierr,i,N = m*n*b->bs2; 312 MatScalar *vsingle; 313 314 PetscFunctionBegin; 315 if (N > b->setvalueslen) { 316 if (b->setvaluescopy) {ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr);} 317 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 318 b->setvalueslen = N; 319 } 320 vsingle = b->setvaluescopy; 321 for (i=0; i<N; i++) { 322 vsingle[i] = v[i]; 323 } 324 ierr = MatSetValuesBlocked_MPIBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 325 PetscFunctionReturn(0); 326 } 327 328 #undef __FUNCT__ 329 #define __FUNCT__ "MatSetValues_MPIBAIJ_HT" 330 int MatSetValues_MPIBAIJ_HT(Mat mat,int m,const int im[],int n,const int in[],const PetscScalar v[],InsertMode addv) 331 { 332 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data; 333 int ierr,i,N = m*n; 334 MatScalar *vsingle; 335 336 PetscFunctionBegin; 337 if (N > b->setvalueslen) { 338 if (b->setvaluescopy) {ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr);} 339 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 340 b->setvalueslen = N; 341 } 342 vsingle = b->setvaluescopy; 343 for (i=0; i<N; i++) { 344 vsingle[i] = v[i]; 345 } 346 ierr = MatSetValues_MPIBAIJ_HT_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 347 PetscFunctionReturn(0); 348 } 349 350 #undef __FUNCT__ 351 #define __FUNCT__ "MatSetValuesBlocked_MPIBAIJ_HT" 352 int MatSetValuesBlocked_MPIBAIJ_HT(Mat mat,int m,const int im[],int n,const int in[],const PetscScalar v[],InsertMode addv) 353 { 354 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data; 355 int ierr,i,N = m*n*b->bs2; 356 MatScalar *vsingle; 357 358 PetscFunctionBegin; 359 if (N > b->setvalueslen) { 360 if (b->setvaluescopy) {ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr);} 361 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 362 b->setvalueslen = N; 363 } 364 vsingle = b->setvaluescopy; 365 for (i=0; i<N; i++) { 366 vsingle[i] = v[i]; 367 } 368 ierr = MatSetValuesBlocked_MPIBAIJ_HT_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 369 PetscFunctionReturn(0); 370 } 371 #endif 372 373 #undef __FUNCT__ 374 #define __FUNCT__ "MatSetValues_MPIBAIJ_MatScalar" 375 int MatSetValues_MPIBAIJ_MatScalar(Mat mat,int m,const int im[],int n,const int in[],const MatScalar v[],InsertMode addv) 376 { 377 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 378 MatScalar value; 379 PetscTruth roworiented = baij->roworiented; 380 int ierr,i,j,row,col; 381 int rstart_orig=baij->rstart_bs; 382 int rend_orig=baij->rend_bs,cstart_orig=baij->cstart_bs; 383 int cend_orig=baij->cend_bs,bs=baij->bs; 384 385 /* Some Variables required in the macro */ 386 Mat A = baij->A; 387 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)(A)->data; 388 int *aimax=a->imax,*ai=a->i,*ailen=a->ilen,*aj=a->j; 389 MatScalar *aa=a->a; 390 391 Mat B = baij->B; 392 Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(B)->data; 393 int *bimax=b->imax,*bi=b->i,*bilen=b->ilen,*bj=b->j; 394 MatScalar *ba=b->a; 395 396 int *rp,ii,nrow,_i,rmax,N,brow,bcol; 397 int low,high,t,ridx,cidx,bs2=a->bs2; 398 MatScalar *ap,*bap; 399 400 PetscFunctionBegin; 401 for (i=0; i<m; i++) { 402 if (im[i] < 0) continue; 403 #if defined(PETSC_USE_BOPT_g) 404 if (im[i] >= mat->M) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %d max %d",im[i],mat->M-1); 405 #endif 406 if (im[i] >= rstart_orig && im[i] < rend_orig) { 407 row = im[i] - rstart_orig; 408 for (j=0; j<n; j++) { 409 if (in[j] >= cstart_orig && in[j] < cend_orig){ 410 col = in[j] - cstart_orig; 411 if (roworiented) value = v[i*n+j]; else value = v[i+j*m]; 412 MatSetValues_SeqBAIJ_A_Private(row,col,value,addv); 413 /* ierr = MatSetValues_SeqBAIJ(baij->A,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 414 } else if (in[j] < 0) continue; 415 #if defined(PETSC_USE_BOPT_g) 416 else if (in[j] >= mat->N) {SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %d max %d",in[i],mat->N-1);} 417 #endif 418 else { 419 if (mat->was_assembled) { 420 if (!baij->colmap) { 421 ierr = CreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 422 } 423 #if defined (PETSC_USE_CTABLE) 424 ierr = PetscTableFind(baij->colmap,in[j]/bs + 1,&col);CHKERRQ(ierr); 425 col = col - 1; 426 #else 427 col = baij->colmap[in[j]/bs] - 1; 428 #endif 429 if (col < 0 && !((Mat_SeqBAIJ*)(baij->A->data))->nonew) { 430 ierr = DisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 431 col = in[j]; 432 /* Reinitialize the variables required by MatSetValues_SeqBAIJ_B_Private() */ 433 B = baij->B; 434 b = (Mat_SeqBAIJ*)(B)->data; 435 bimax=b->imax;bi=b->i;bilen=b->ilen;bj=b->j; 436 ba=b->a; 437 } else col += in[j]%bs; 438 } else col = in[j]; 439 if (roworiented) value = v[i*n+j]; else value = v[i+j*m]; 440 MatSetValues_SeqBAIJ_B_Private(row,col,value,addv); 441 /* ierr = MatSetValues_SeqBAIJ(baij->B,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 442 } 443 } 444 } else { 445 if (!baij->donotstash) { 446 if (roworiented) { 447 ierr = MatStashValuesRow_Private(&mat->stash,im[i],n,in,v+i*n);CHKERRQ(ierr); 448 } else { 449 ierr = MatStashValuesCol_Private(&mat->stash,im[i],n,in,v+i,m);CHKERRQ(ierr); 450 } 451 } 452 } 453 } 454 PetscFunctionReturn(0); 455 } 456 457 #undef __FUNCT__ 458 #define __FUNCT__ "MatSetValuesBlocked_MPIBAIJ_MatScalar" 459 int MatSetValuesBlocked_MPIBAIJ_MatScalar(Mat mat,int m,const int im[],int n,const int in[],const MatScalar v[],InsertMode addv) 460 { 461 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 462 const MatScalar *value; 463 MatScalar *barray=baij->barray; 464 PetscTruth roworiented = baij->roworiented; 465 int ierr,i,j,ii,jj,row,col,rstart=baij->rstart; 466 int rend=baij->rend,cstart=baij->cstart,stepval; 467 int cend=baij->cend,bs=baij->bs,bs2=baij->bs2; 468 469 PetscFunctionBegin; 470 if(!barray) { 471 ierr = PetscMalloc(bs2*sizeof(MatScalar),&barray);CHKERRQ(ierr); 472 baij->barray = barray; 473 } 474 475 if (roworiented) { 476 stepval = (n-1)*bs; 477 } else { 478 stepval = (m-1)*bs; 479 } 480 for (i=0; i<m; i++) { 481 if (im[i] < 0) continue; 482 #if defined(PETSC_USE_BOPT_g) 483 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large, row %d max %d",im[i],baij->Mbs-1); 484 #endif 485 if (im[i] >= rstart && im[i] < rend) { 486 row = im[i] - rstart; 487 for (j=0; j<n; j++) { 488 /* If NumCol = 1 then a copy is not required */ 489 if ((roworiented) && (n == 1)) { 490 barray = (MatScalar*)v + i*bs2; 491 } else if((!roworiented) && (m == 1)) { 492 barray = (MatScalar*)v + j*bs2; 493 } else { /* Here a copy is required */ 494 if (roworiented) { 495 value = v + i*(stepval+bs)*bs + j*bs; 496 } else { 497 value = v + j*(stepval+bs)*bs + i*bs; 498 } 499 for (ii=0; ii<bs; ii++,value+=stepval) { 500 for (jj=0; jj<bs; jj++) { 501 *barray++ = *value++; 502 } 503 } 504 barray -=bs2; 505 } 506 507 if (in[j] >= cstart && in[j] < cend){ 508 col = in[j] - cstart; 509 ierr = MatSetValuesBlocked_SeqBAIJ_MatScalar(baij->A,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 510 } 511 else if (in[j] < 0) continue; 512 #if defined(PETSC_USE_BOPT_g) 513 else if (in[j] >= baij->Nbs) {SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large, col %d max %d",in[j],baij->Nbs-1);} 514 #endif 515 else { 516 if (mat->was_assembled) { 517 if (!baij->colmap) { 518 ierr = CreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 519 } 520 521 #if defined(PETSC_USE_BOPT_g) 522 #if defined (PETSC_USE_CTABLE) 523 { int data; 524 ierr = PetscTableFind(baij->colmap,in[j]+1,&data);CHKERRQ(ierr); 525 if ((data - 1) % bs) SETERRQ(PETSC_ERR_PLIB,"Incorrect colmap"); 526 } 527 #else 528 if ((baij->colmap[in[j]] - 1) % bs) SETERRQ(PETSC_ERR_PLIB,"Incorrect colmap"); 529 #endif 530 #endif 531 #if defined (PETSC_USE_CTABLE) 532 ierr = PetscTableFind(baij->colmap,in[j]+1,&col);CHKERRQ(ierr); 533 col = (col - 1)/bs; 534 #else 535 col = (baij->colmap[in[j]] - 1)/bs; 536 #endif 537 if (col < 0 && !((Mat_SeqBAIJ*)(baij->A->data))->nonew) { 538 ierr = DisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 539 col = in[j]; 540 } 541 } 542 else col = in[j]; 543 ierr = MatSetValuesBlocked_SeqBAIJ_MatScalar(baij->B,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 544 } 545 } 546 } else { 547 if (!baij->donotstash) { 548 if (roworiented) { 549 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 550 } else { 551 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 552 } 553 } 554 } 555 } 556 PetscFunctionReturn(0); 557 } 558 559 #define HASH_KEY 0.6180339887 560 #define HASH(size,key,tmp) (tmp = (key)*HASH_KEY,(int)((size)*(tmp-(int)tmp))) 561 /* #define HASH(size,key) ((int)((size)*fmod(((key)*HASH_KEY),1))) */ 562 /* #define HASH(size,key,tmp) ((int)((size)*fmod(((key)*HASH_KEY),1))) */ 563 #undef __FUNCT__ 564 #define __FUNCT__ "MatSetValues_MPIBAIJ_HT_MatScalar" 565 int MatSetValues_MPIBAIJ_HT_MatScalar(Mat mat,int m,const int im[],int n,const int in[],const MatScalar v[],InsertMode addv) 566 { 567 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 568 PetscTruth roworiented = baij->roworiented; 569 int ierr,i,j,row,col; 570 int rstart_orig=baij->rstart_bs; 571 int rend_orig=baij->rend_bs,Nbs=baij->Nbs; 572 int h1,key,size=baij->ht_size,bs=baij->bs,*HT=baij->ht,idx; 573 PetscReal tmp; 574 MatScalar **HD = baij->hd,value; 575 #if defined(PETSC_USE_BOPT_g) 576 int total_ct=baij->ht_total_ct,insert_ct=baij->ht_insert_ct; 577 #endif 578 579 PetscFunctionBegin; 580 581 for (i=0; i<m; i++) { 582 #if defined(PETSC_USE_BOPT_g) 583 if (im[i] < 0) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Negative row"); 584 if (im[i] >= mat->M) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %d max %d",im[i],mat->M-1); 585 #endif 586 row = im[i]; 587 if (row >= rstart_orig && row < rend_orig) { 588 for (j=0; j<n; j++) { 589 col = in[j]; 590 if (roworiented) value = v[i*n+j]; else value = v[i+j*m]; 591 /* Look up into the Hash Table */ 592 key = (row/bs)*Nbs+(col/bs)+1; 593 h1 = HASH(size,key,tmp); 594 595 596 idx = h1; 597 #if defined(PETSC_USE_BOPT_g) 598 insert_ct++; 599 total_ct++; 600 if (HT[idx] != key) { 601 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++,total_ct++); 602 if (idx == size) { 603 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++,total_ct++); 604 if (idx == h1) { 605 SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"(row,col) has no entry in the hash table"); 606 } 607 } 608 } 609 #else 610 if (HT[idx] != key) { 611 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++); 612 if (idx == size) { 613 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++); 614 if (idx == h1) { 615 SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"(row,col) has no entry in the hash table"); 616 } 617 } 618 } 619 #endif 620 /* A HASH table entry is found, so insert the values at the correct address */ 621 if (addv == ADD_VALUES) *(HD[idx]+ (col % bs)*bs + (row % bs)) += value; 622 else *(HD[idx]+ (col % bs)*bs + (row % bs)) = value; 623 } 624 } else { 625 if (!baij->donotstash) { 626 if (roworiented) { 627 ierr = MatStashValuesRow_Private(&mat->stash,im[i],n,in,v+i*n);CHKERRQ(ierr); 628 } else { 629 ierr = MatStashValuesCol_Private(&mat->stash,im[i],n,in,v+i,m);CHKERRQ(ierr); 630 } 631 } 632 } 633 } 634 #if defined(PETSC_USE_BOPT_g) 635 baij->ht_total_ct = total_ct; 636 baij->ht_insert_ct = insert_ct; 637 #endif 638 PetscFunctionReturn(0); 639 } 640 641 #undef __FUNCT__ 642 #define __FUNCT__ "MatSetValuesBlocked_MPIBAIJ_HT_MatScalar" 643 int MatSetValuesBlocked_MPIBAIJ_HT_MatScalar(Mat mat,int m,const int im[],int n,const int in[],const MatScalar v[],InsertMode addv) 644 { 645 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 646 PetscTruth roworiented = baij->roworiented; 647 int ierr,i,j,ii,jj,row,col; 648 int rstart=baij->rstart ; 649 int rend=baij->rend,stepval,bs=baij->bs,bs2=baij->bs2; 650 int h1,key,size=baij->ht_size,idx,*HT=baij->ht,Nbs=baij->Nbs; 651 PetscReal tmp; 652 MatScalar **HD = baij->hd,*baij_a; 653 const MatScalar *v_t,*value; 654 #if defined(PETSC_USE_BOPT_g) 655 int total_ct=baij->ht_total_ct,insert_ct=baij->ht_insert_ct; 656 #endif 657 658 PetscFunctionBegin; 659 660 if (roworiented) { 661 stepval = (n-1)*bs; 662 } else { 663 stepval = (m-1)*bs; 664 } 665 for (i=0; i<m; i++) { 666 #if defined(PETSC_USE_BOPT_g) 667 if (im[i] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %d",im[i]); 668 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %d max %d",im[i],baij->Mbs-1); 669 #endif 670 row = im[i]; 671 v_t = v + i*bs2; 672 if (row >= rstart && row < rend) { 673 for (j=0; j<n; j++) { 674 col = in[j]; 675 676 /* Look up into the Hash Table */ 677 key = row*Nbs+col+1; 678 h1 = HASH(size,key,tmp); 679 680 idx = h1; 681 #if defined(PETSC_USE_BOPT_g) 682 total_ct++; 683 insert_ct++; 684 if (HT[idx] != key) { 685 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++,total_ct++); 686 if (idx == size) { 687 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++,total_ct++); 688 if (idx == h1) { 689 SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"(row,col) has no entry in the hash table"); 690 } 691 } 692 } 693 #else 694 if (HT[idx] != key) { 695 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++); 696 if (idx == size) { 697 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++); 698 if (idx == h1) { 699 SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"(row,col) has no entry in the hash table"); 700 } 701 } 702 } 703 #endif 704 baij_a = HD[idx]; 705 if (roworiented) { 706 /*value = v + i*(stepval+bs)*bs + j*bs;*/ 707 /* value = v + (i*(stepval+bs)+j)*bs; */ 708 value = v_t; 709 v_t += bs; 710 if (addv == ADD_VALUES) { 711 for (ii=0; ii<bs; ii++,value+=stepval) { 712 for (jj=ii; jj<bs2; jj+=bs) { 713 baij_a[jj] += *value++; 714 } 715 } 716 } else { 717 for (ii=0; ii<bs; ii++,value+=stepval) { 718 for (jj=ii; jj<bs2; jj+=bs) { 719 baij_a[jj] = *value++; 720 } 721 } 722 } 723 } else { 724 value = v + j*(stepval+bs)*bs + i*bs; 725 if (addv == ADD_VALUES) { 726 for (ii=0; ii<bs; ii++,value+=stepval,baij_a+=bs) { 727 for (jj=0; jj<bs; jj++) { 728 baij_a[jj] += *value++; 729 } 730 } 731 } else { 732 for (ii=0; ii<bs; ii++,value+=stepval,baij_a+=bs) { 733 for (jj=0; jj<bs; jj++) { 734 baij_a[jj] = *value++; 735 } 736 } 737 } 738 } 739 } 740 } else { 741 if (!baij->donotstash) { 742 if (roworiented) { 743 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 744 } else { 745 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 746 } 747 } 748 } 749 } 750 #if defined(PETSC_USE_BOPT_g) 751 baij->ht_total_ct = total_ct; 752 baij->ht_insert_ct = insert_ct; 753 #endif 754 PetscFunctionReturn(0); 755 } 756 757 #undef __FUNCT__ 758 #define __FUNCT__ "MatGetValues_MPIBAIJ" 759 int MatGetValues_MPIBAIJ(Mat mat,int m,const int idxm[],int n,const int idxn[],PetscScalar v[]) 760 { 761 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 762 int bs=baij->bs,ierr,i,j,bsrstart = baij->rstart*bs,bsrend = baij->rend*bs; 763 int bscstart = baij->cstart*bs,bscend = baij->cend*bs,row,col,data; 764 765 PetscFunctionBegin; 766 for (i=0; i<m; i++) { 767 if (idxm[i] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %d",idxm[i]); 768 if (idxm[i] >= mat->M) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %d max %d",idxm[i],mat->M-1); 769 if (idxm[i] >= bsrstart && idxm[i] < bsrend) { 770 row = idxm[i] - bsrstart; 771 for (j=0; j<n; j++) { 772 if (idxn[j] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative column: %d",idxn[j]); 773 if (idxn[j] >= mat->N) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %d max %d",idxn[j],mat->N-1); 774 if (idxn[j] >= bscstart && idxn[j] < bscend){ 775 col = idxn[j] - bscstart; 776 ierr = MatGetValues_SeqBAIJ(baij->A,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 777 } else { 778 if (!baij->colmap) { 779 ierr = CreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 780 } 781 #if defined (PETSC_USE_CTABLE) 782 ierr = PetscTableFind(baij->colmap,idxn[j]/bs+1,&data);CHKERRQ(ierr); 783 data --; 784 #else 785 data = baij->colmap[idxn[j]/bs]-1; 786 #endif 787 if((data < 0) || (baij->garray[data/bs] != idxn[j]/bs)) *(v+i*n+j) = 0.0; 788 else { 789 col = data + idxn[j]%bs; 790 ierr = MatGetValues_SeqBAIJ(baij->B,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 791 } 792 } 793 } 794 } else { 795 SETERRQ(PETSC_ERR_SUP,"Only local values currently supported"); 796 } 797 } 798 PetscFunctionReturn(0); 799 } 800 801 #undef __FUNCT__ 802 #define __FUNCT__ "MatNorm_MPIBAIJ" 803 int MatNorm_MPIBAIJ(Mat mat,NormType type,PetscReal *nrm) 804 { 805 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 806 Mat_SeqBAIJ *amat = (Mat_SeqBAIJ*)baij->A->data,*bmat = (Mat_SeqBAIJ*)baij->B->data; 807 int ierr,i,bs2=baij->bs2; 808 PetscReal sum = 0.0; 809 MatScalar *v; 810 811 PetscFunctionBegin; 812 if (baij->size == 1) { 813 ierr = MatNorm(baij->A,type,nrm);CHKERRQ(ierr); 814 } else { 815 if (type == NORM_FROBENIUS) { 816 v = amat->a; 817 for (i=0; i<amat->nz*bs2; i++) { 818 #if defined(PETSC_USE_COMPLEX) 819 sum += PetscRealPart(PetscConj(*v)*(*v)); v++; 820 #else 821 sum += (*v)*(*v); v++; 822 #endif 823 } 824 v = bmat->a; 825 for (i=0; i<bmat->nz*bs2; i++) { 826 #if defined(PETSC_USE_COMPLEX) 827 sum += PetscRealPart(PetscConj(*v)*(*v)); v++; 828 #else 829 sum += (*v)*(*v); v++; 830 #endif 831 } 832 ierr = MPI_Allreduce(&sum,nrm,1,MPIU_REAL,MPI_SUM,mat->comm);CHKERRQ(ierr); 833 *nrm = sqrt(*nrm); 834 } else { 835 SETERRQ(PETSC_ERR_SUP,"No support for this norm yet"); 836 } 837 } 838 PetscFunctionReturn(0); 839 } 840 841 842 /* 843 Creates the hash table, and sets the table 844 This table is created only once. 845 If new entried need to be added to the matrix 846 then the hash table has to be destroyed and 847 recreated. 848 */ 849 #undef __FUNCT__ 850 #define __FUNCT__ "MatCreateHashTable_MPIBAIJ_Private" 851 int MatCreateHashTable_MPIBAIJ_Private(Mat mat,PetscReal factor) 852 { 853 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 854 Mat A = baij->A,B=baij->B; 855 Mat_SeqBAIJ *a=(Mat_SeqBAIJ *)A->data,*b=(Mat_SeqBAIJ *)B->data; 856 int i,j,k,nz=a->nz+b->nz,h1,*ai=a->i,*aj=a->j,*bi=b->i,*bj=b->j; 857 int size,bs2=baij->bs2,rstart=baij->rstart,ierr; 858 int cstart=baij->cstart,*garray=baij->garray,row,col,Nbs=baij->Nbs; 859 int *HT,key; 860 MatScalar **HD; 861 PetscReal tmp; 862 #if defined(PETSC_USE_BOPT_g) 863 int ct=0,max=0; 864 #endif 865 866 PetscFunctionBegin; 867 baij->ht_size=(int)(factor*nz); 868 size = baij->ht_size; 869 870 if (baij->ht) { 871 PetscFunctionReturn(0); 872 } 873 874 /* Allocate Memory for Hash Table */ 875 ierr = PetscMalloc((size)*(sizeof(int)+sizeof(MatScalar*))+1,&baij->hd);CHKERRQ(ierr); 876 baij->ht = (int*)(baij->hd + size); 877 HD = baij->hd; 878 HT = baij->ht; 879 880 881 ierr = PetscMemzero(HD,size*(sizeof(int)+sizeof(PetscScalar*)));CHKERRQ(ierr); 882 883 884 /* Loop Over A */ 885 for (i=0; i<a->mbs; i++) { 886 for (j=ai[i]; j<ai[i+1]; j++) { 887 row = i+rstart; 888 col = aj[j]+cstart; 889 890 key = row*Nbs + col + 1; 891 h1 = HASH(size,key,tmp); 892 for (k=0; k<size; k++){ 893 if (HT[(h1+k)%size] == 0.0) { 894 HT[(h1+k)%size] = key; 895 HD[(h1+k)%size] = a->a + j*bs2; 896 break; 897 #if defined(PETSC_USE_BOPT_g) 898 } else { 899 ct++; 900 #endif 901 } 902 } 903 #if defined(PETSC_USE_BOPT_g) 904 if (k> max) max = k; 905 #endif 906 } 907 } 908 /* Loop Over B */ 909 for (i=0; i<b->mbs; i++) { 910 for (j=bi[i]; j<bi[i+1]; j++) { 911 row = i+rstart; 912 col = garray[bj[j]]; 913 key = row*Nbs + col + 1; 914 h1 = HASH(size,key,tmp); 915 for (k=0; k<size; k++){ 916 if (HT[(h1+k)%size] == 0.0) { 917 HT[(h1+k)%size] = key; 918 HD[(h1+k)%size] = b->a + j*bs2; 919 break; 920 #if defined(PETSC_USE_BOPT_g) 921 } else { 922 ct++; 923 #endif 924 } 925 } 926 #if defined(PETSC_USE_BOPT_g) 927 if (k> max) max = k; 928 #endif 929 } 930 } 931 932 /* Print Summary */ 933 #if defined(PETSC_USE_BOPT_g) 934 for (i=0,j=0; i<size; i++) { 935 if (HT[i]) {j++;} 936 } 937 PetscLogInfo(0,"MatCreateHashTable_MPIBAIJ_Private: Average Search = %5.2f,max search = %d\n",(j== 0)? 0.0:((PetscReal)(ct+j))/j,max); 938 #endif 939 PetscFunctionReturn(0); 940 } 941 942 #undef __FUNCT__ 943 #define __FUNCT__ "MatAssemblyBegin_MPIBAIJ" 944 int MatAssemblyBegin_MPIBAIJ(Mat mat,MatAssemblyType mode) 945 { 946 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 947 int ierr,nstash,reallocs; 948 InsertMode addv; 949 950 PetscFunctionBegin; 951 if (baij->donotstash) { 952 PetscFunctionReturn(0); 953 } 954 955 /* make sure all processors are either in INSERTMODE or ADDMODE */ 956 ierr = MPI_Allreduce(&mat->insertmode,&addv,1,MPI_INT,MPI_BOR,mat->comm);CHKERRQ(ierr); 957 if (addv == (ADD_VALUES|INSERT_VALUES)) { 958 SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Some processors inserted others added"); 959 } 960 mat->insertmode = addv; /* in case this processor had no cache */ 961 962 ierr = MatStashScatterBegin_Private(&mat->stash,baij->rowners_bs);CHKERRQ(ierr); 963 ierr = MatStashScatterBegin_Private(&mat->bstash,baij->rowners);CHKERRQ(ierr); 964 ierr = MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);CHKERRQ(ierr); 965 PetscLogInfo(0,"MatAssemblyBegin_MPIBAIJ:Stash has %d entries,uses %d mallocs.\n",nstash,reallocs); 966 ierr = MatStashGetInfo_Private(&mat->bstash,&nstash,&reallocs);CHKERRQ(ierr); 967 PetscLogInfo(0,"MatAssemblyBegin_MPIBAIJ:Block-Stash has %d entries, uses %d mallocs.\n",nstash,reallocs); 968 PetscFunctionReturn(0); 969 } 970 971 #undef __FUNCT__ 972 #define __FUNCT__ "MatAssemblyEnd_MPIBAIJ" 973 int MatAssemblyEnd_MPIBAIJ(Mat mat,MatAssemblyType mode) 974 { 975 Mat_MPIBAIJ *baij=(Mat_MPIBAIJ*)mat->data; 976 Mat_SeqBAIJ *a=(Mat_SeqBAIJ*)baij->A->data,*b=(Mat_SeqBAIJ*)baij->B->data; 977 int i,j,rstart,ncols,n,ierr,flg,bs2=baij->bs2; 978 int *row,*col,other_disassembled; 979 PetscTruth r1,r2,r3; 980 MatScalar *val; 981 InsertMode addv = mat->insertmode; 982 983 PetscFunctionBegin; 984 if (!baij->donotstash) { 985 while (1) { 986 ierr = MatStashScatterGetMesg_Private(&mat->stash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 987 if (!flg) break; 988 989 for (i=0; i<n;) { 990 /* Now identify the consecutive vals belonging to the same row */ 991 for (j=i,rstart=row[j]; j<n; j++) { if (row[j] != rstart) break; } 992 if (j < n) ncols = j-i; 993 else ncols = n-i; 994 /* Now assemble all these values with a single function call */ 995 ierr = MatSetValues_MPIBAIJ_MatScalar(mat,1,row+i,ncols,col+i,val+i,addv);CHKERRQ(ierr); 996 i = j; 997 } 998 } 999 ierr = MatStashScatterEnd_Private(&mat->stash);CHKERRQ(ierr); 1000 /* Now process the block-stash. Since the values are stashed column-oriented, 1001 set the roworiented flag to column oriented, and after MatSetValues() 1002 restore the original flags */ 1003 r1 = baij->roworiented; 1004 r2 = a->roworiented; 1005 r3 = b->roworiented; 1006 baij->roworiented = PETSC_FALSE; 1007 a->roworiented = PETSC_FALSE; 1008 b->roworiented = PETSC_FALSE; 1009 while (1) { 1010 ierr = MatStashScatterGetMesg_Private(&mat->bstash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 1011 if (!flg) break; 1012 1013 for (i=0; i<n;) { 1014 /* Now identify the consecutive vals belonging to the same row */ 1015 for (j=i,rstart=row[j]; j<n; j++) { if (row[j] != rstart) break; } 1016 if (j < n) ncols = j-i; 1017 else ncols = n-i; 1018 ierr = MatSetValuesBlocked_MPIBAIJ_MatScalar(mat,1,row+i,ncols,col+i,val+i*bs2,addv);CHKERRQ(ierr); 1019 i = j; 1020 } 1021 } 1022 ierr = MatStashScatterEnd_Private(&mat->bstash);CHKERRQ(ierr); 1023 baij->roworiented = r1; 1024 a->roworiented = r2; 1025 b->roworiented = r3; 1026 } 1027 1028 ierr = MatAssemblyBegin(baij->A,mode);CHKERRQ(ierr); 1029 ierr = MatAssemblyEnd(baij->A,mode);CHKERRQ(ierr); 1030 1031 /* determine if any processor has disassembled, if so we must 1032 also disassemble ourselfs, in order that we may reassemble. */ 1033 /* 1034 if nonzero structure of submatrix B cannot change then we know that 1035 no processor disassembled thus we can skip this stuff 1036 */ 1037 if (!((Mat_SeqBAIJ*)baij->B->data)->nonew) { 1038 ierr = MPI_Allreduce(&mat->was_assembled,&other_disassembled,1,MPI_INT,MPI_PROD,mat->comm);CHKERRQ(ierr); 1039 if (mat->was_assembled && !other_disassembled) { 1040 ierr = DisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 1041 } 1042 } 1043 1044 if (!mat->was_assembled && mode == MAT_FINAL_ASSEMBLY) { 1045 ierr = MatSetUpMultiply_MPIBAIJ(mat);CHKERRQ(ierr); 1046 } 1047 ierr = MatAssemblyBegin(baij->B,mode);CHKERRQ(ierr); 1048 ierr = MatAssemblyEnd(baij->B,mode);CHKERRQ(ierr); 1049 1050 #if defined(PETSC_USE_BOPT_g) 1051 if (baij->ht && mode== MAT_FINAL_ASSEMBLY) { 1052 PetscLogInfo(0,"MatAssemblyEnd_MPIBAIJ:Average Hash Table Search in MatSetValues = %5.2f\n",((PetscReal)baij->ht_total_ct)/baij->ht_insert_ct); 1053 baij->ht_total_ct = 0; 1054 baij->ht_insert_ct = 0; 1055 } 1056 #endif 1057 if (baij->ht_flag && !baij->ht && mode == MAT_FINAL_ASSEMBLY) { 1058 ierr = MatCreateHashTable_MPIBAIJ_Private(mat,baij->ht_fact);CHKERRQ(ierr); 1059 mat->ops->setvalues = MatSetValues_MPIBAIJ_HT; 1060 mat->ops->setvaluesblocked = MatSetValuesBlocked_MPIBAIJ_HT; 1061 } 1062 1063 if (baij->rowvalues) { 1064 ierr = PetscFree(baij->rowvalues);CHKERRQ(ierr); 1065 baij->rowvalues = 0; 1066 } 1067 PetscFunctionReturn(0); 1068 } 1069 1070 #undef __FUNCT__ 1071 #define __FUNCT__ "MatView_MPIBAIJ_ASCIIorDraworSocket" 1072 static int MatView_MPIBAIJ_ASCIIorDraworSocket(Mat mat,PetscViewer viewer) 1073 { 1074 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1075 int ierr,bs = baij->bs,size = baij->size,rank = baij->rank; 1076 PetscTruth isascii,isdraw; 1077 PetscViewer sviewer; 1078 PetscViewerFormat format; 1079 1080 PetscFunctionBegin; 1081 /* printf(" MatView_MPIBAIJ_ASCIIorDraworSocket is called ...\n"); */ 1082 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&isascii);CHKERRQ(ierr); 1083 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);CHKERRQ(ierr); 1084 if (isascii) { 1085 ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr); 1086 if (format == PETSC_VIEWER_ASCII_INFO_DETAIL) { 1087 MatInfo info; 1088 ierr = MPI_Comm_rank(mat->comm,&rank);CHKERRQ(ierr); 1089 ierr = MatGetInfo(mat,MAT_LOCAL,&info);CHKERRQ(ierr); 1090 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] Local rows %d nz %d nz alloced %d bs %d mem %d\n", 1091 rank,mat->m,(int)info.nz_used*bs,(int)info.nz_allocated*bs, 1092 baij->bs,(int)info.memory);CHKERRQ(ierr); 1093 ierr = MatGetInfo(baij->A,MAT_LOCAL,&info);CHKERRQ(ierr); 1094 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] on-diagonal part: nz %d \n",rank,(int)info.nz_used*bs);CHKERRQ(ierr); 1095 ierr = MatGetInfo(baij->B,MAT_LOCAL,&info);CHKERRQ(ierr); 1096 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] off-diagonal part: nz %d \n",rank,(int)info.nz_used*bs);CHKERRQ(ierr); 1097 ierr = PetscViewerFlush(viewer);CHKERRQ(ierr); 1098 ierr = VecScatterView(baij->Mvctx,viewer);CHKERRQ(ierr); 1099 PetscFunctionReturn(0); 1100 } else if (format == PETSC_VIEWER_ASCII_INFO) { 1101 ierr = PetscViewerASCIIPrintf(viewer," block size is %d\n",bs);CHKERRQ(ierr); 1102 PetscFunctionReturn(0); 1103 } else if (format == PETSC_VIEWER_ASCII_FACTOR_INFO) { 1104 PetscFunctionReturn(0); 1105 } 1106 } 1107 1108 if (isdraw) { 1109 PetscDraw draw; 1110 PetscTruth isnull; 1111 ierr = PetscViewerDrawGetDraw(viewer,0,&draw);CHKERRQ(ierr); 1112 ierr = PetscDrawIsNull(draw,&isnull);CHKERRQ(ierr); if (isnull) PetscFunctionReturn(0); 1113 } 1114 1115 if (size == 1) { 1116 ierr = PetscObjectSetName((PetscObject)baij->A,mat->name);CHKERRQ(ierr); 1117 ierr = MatView(baij->A,viewer);CHKERRQ(ierr); 1118 } else { 1119 /* assemble the entire matrix onto first processor. */ 1120 Mat A; 1121 Mat_SeqBAIJ *Aloc; 1122 int M = mat->M,N = mat->N,*ai,*aj,col,i,j,k,*rvals,mbs = baij->mbs; 1123 MatScalar *a; 1124 1125 if (!rank) { 1126 ierr = MatCreateMPIBAIJ(mat->comm,baij->bs,M,N,M,N,0,PETSC_NULL,0,PETSC_NULL,&A);CHKERRQ(ierr); 1127 } else { 1128 ierr = MatCreateMPIBAIJ(mat->comm,baij->bs,0,0,M,N,0,PETSC_NULL,0,PETSC_NULL,&A);CHKERRQ(ierr); 1129 } 1130 PetscLogObjectParent(mat,A); 1131 1132 /* copy over the A part */ 1133 Aloc = (Mat_SeqBAIJ*)baij->A->data; 1134 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1135 ierr = PetscMalloc(bs*sizeof(int),&rvals);CHKERRQ(ierr); 1136 1137 for (i=0; i<mbs; i++) { 1138 rvals[0] = bs*(baij->rstart + i); 1139 for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; } 1140 for (j=ai[i]; j<ai[i+1]; j++) { 1141 col = (baij->cstart+aj[j])*bs; 1142 for (k=0; k<bs; k++) { 1143 ierr = MatSetValues_MPIBAIJ_MatScalar(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 1144 col++; a += bs; 1145 } 1146 } 1147 } 1148 /* copy over the B part */ 1149 Aloc = (Mat_SeqBAIJ*)baij->B->data; 1150 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1151 for (i=0; i<mbs; i++) { 1152 rvals[0] = bs*(baij->rstart + i); 1153 for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; } 1154 for (j=ai[i]; j<ai[i+1]; j++) { 1155 col = baij->garray[aj[j]]*bs; 1156 for (k=0; k<bs; k++) { 1157 ierr = MatSetValues_MPIBAIJ_MatScalar(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 1158 col++; a += bs; 1159 } 1160 } 1161 } 1162 ierr = PetscFree(rvals);CHKERRQ(ierr); 1163 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1164 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1165 /* 1166 Everyone has to call to draw the matrix since the graphics waits are 1167 synchronized across all processors that share the PetscDraw object 1168 */ 1169 ierr = PetscViewerGetSingleton(viewer,&sviewer);CHKERRQ(ierr); 1170 if (!rank) { 1171 ierr = PetscObjectSetName((PetscObject)((Mat_MPIBAIJ*)(A->data))->A,mat->name);CHKERRQ(ierr); 1172 ierr = MatView(((Mat_MPIBAIJ*)(A->data))->A,sviewer);CHKERRQ(ierr); 1173 } 1174 ierr = PetscViewerRestoreSingleton(viewer,&sviewer);CHKERRQ(ierr); 1175 ierr = MatDestroy(A);CHKERRQ(ierr); 1176 } 1177 PetscFunctionReturn(0); 1178 } 1179 1180 #undef __FUNCT__ 1181 #define __FUNCT__ "MatView_MPIBAIJ" 1182 int MatView_MPIBAIJ(Mat mat,PetscViewer viewer) 1183 { 1184 int ierr; 1185 PetscTruth isascii,isdraw,issocket,isbinary; 1186 1187 PetscFunctionBegin; 1188 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&isascii);CHKERRQ(ierr); 1189 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);CHKERRQ(ierr); 1190 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_SOCKET,&issocket);CHKERRQ(ierr); 1191 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_BINARY,&isbinary);CHKERRQ(ierr); 1192 if (isascii || isdraw || issocket || isbinary) { 1193 ierr = MatView_MPIBAIJ_ASCIIorDraworSocket(mat,viewer);CHKERRQ(ierr); 1194 } else { 1195 SETERRQ1(1,"Viewer type %s not supported by MPIBAIJ matrices",((PetscObject)viewer)->type_name); 1196 } 1197 PetscFunctionReturn(0); 1198 } 1199 1200 #undef __FUNCT__ 1201 #define __FUNCT__ "MatDestroy_MPIBAIJ" 1202 int MatDestroy_MPIBAIJ(Mat mat) 1203 { 1204 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1205 int ierr; 1206 1207 PetscFunctionBegin; 1208 #if defined(PETSC_USE_LOG) 1209 PetscLogObjectState((PetscObject)mat,"Rows=%d,Cols=%d",mat->M,mat->N); 1210 #endif 1211 ierr = MatStashDestroy_Private(&mat->stash);CHKERRQ(ierr); 1212 ierr = MatStashDestroy_Private(&mat->bstash);CHKERRQ(ierr); 1213 ierr = PetscFree(baij->rowners);CHKERRQ(ierr); 1214 ierr = MatDestroy(baij->A);CHKERRQ(ierr); 1215 ierr = MatDestroy(baij->B);CHKERRQ(ierr); 1216 #if defined (PETSC_USE_CTABLE) 1217 if (baij->colmap) {ierr = PetscTableDelete(baij->colmap);CHKERRQ(ierr);} 1218 #else 1219 if (baij->colmap) {ierr = PetscFree(baij->colmap);CHKERRQ(ierr);} 1220 #endif 1221 if (baij->garray) {ierr = PetscFree(baij->garray);CHKERRQ(ierr);} 1222 if (baij->lvec) {ierr = VecDestroy(baij->lvec);CHKERRQ(ierr);} 1223 if (baij->Mvctx) {ierr = VecScatterDestroy(baij->Mvctx);CHKERRQ(ierr);} 1224 if (baij->rowvalues) {ierr = PetscFree(baij->rowvalues);CHKERRQ(ierr);} 1225 if (baij->barray) {ierr = PetscFree(baij->barray);CHKERRQ(ierr);} 1226 if (baij->hd) {ierr = PetscFree(baij->hd);CHKERRQ(ierr);} 1227 #if defined(PETSC_USE_MAT_SINGLE) 1228 if (baij->setvaluescopy) {ierr = PetscFree(baij->setvaluescopy);CHKERRQ(ierr);} 1229 #endif 1230 ierr = PetscFree(baij);CHKERRQ(ierr); 1231 PetscFunctionReturn(0); 1232 } 1233 1234 #undef __FUNCT__ 1235 #define __FUNCT__ "MatMult_MPIBAIJ" 1236 int MatMult_MPIBAIJ(Mat A,Vec xx,Vec yy) 1237 { 1238 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1239 int ierr,nt; 1240 1241 PetscFunctionBegin; 1242 ierr = VecGetLocalSize(xx,&nt);CHKERRQ(ierr); 1243 if (nt != A->n) { 1244 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx"); 1245 } 1246 ierr = VecGetLocalSize(yy,&nt);CHKERRQ(ierr); 1247 if (nt != A->m) { 1248 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible parition of A and yy"); 1249 } 1250 ierr = VecScatterBegin(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1251 ierr = (*a->A->ops->mult)(a->A,xx,yy);CHKERRQ(ierr); 1252 ierr = VecScatterEnd(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1253 ierr = (*a->B->ops->multadd)(a->B,a->lvec,yy,yy);CHKERRQ(ierr); 1254 ierr = VecScatterPostRecvs(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1255 PetscFunctionReturn(0); 1256 } 1257 1258 #undef __FUNCT__ 1259 #define __FUNCT__ "MatMultAdd_MPIBAIJ" 1260 int MatMultAdd_MPIBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 1261 { 1262 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1263 int ierr; 1264 1265 PetscFunctionBegin; 1266 ierr = VecScatterBegin(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1267 ierr = (*a->A->ops->multadd)(a->A,xx,yy,zz);CHKERRQ(ierr); 1268 ierr = VecScatterEnd(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1269 ierr = (*a->B->ops->multadd)(a->B,a->lvec,zz,zz);CHKERRQ(ierr); 1270 PetscFunctionReturn(0); 1271 } 1272 1273 #undef __FUNCT__ 1274 #define __FUNCT__ "MatMultTranspose_MPIBAIJ" 1275 int MatMultTranspose_MPIBAIJ(Mat A,Vec xx,Vec yy) 1276 { 1277 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1278 int ierr; 1279 1280 PetscFunctionBegin; 1281 /* do nondiagonal part */ 1282 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 1283 /* send it on its way */ 1284 ierr = VecScatterBegin(a->lvec,yy,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 1285 /* do local part */ 1286 ierr = (*a->A->ops->multtranspose)(a->A,xx,yy);CHKERRQ(ierr); 1287 /* receive remote parts: note this assumes the values are not actually */ 1288 /* inserted in yy until the next line, which is true for my implementation*/ 1289 /* but is not perhaps always true. */ 1290 ierr = VecScatterEnd(a->lvec,yy,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 1291 PetscFunctionReturn(0); 1292 } 1293 1294 #undef __FUNCT__ 1295 #define __FUNCT__ "MatMultTransposeAdd_MPIBAIJ" 1296 int MatMultTransposeAdd_MPIBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 1297 { 1298 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1299 int ierr; 1300 1301 PetscFunctionBegin; 1302 /* do nondiagonal part */ 1303 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 1304 /* send it on its way */ 1305 ierr = VecScatterBegin(a->lvec,zz,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 1306 /* do local part */ 1307 ierr = (*a->A->ops->multtransposeadd)(a->A,xx,yy,zz);CHKERRQ(ierr); 1308 /* receive remote parts: note this assumes the values are not actually */ 1309 /* inserted in yy until the next line, which is true for my implementation*/ 1310 /* but is not perhaps always true. */ 1311 ierr = VecScatterEnd(a->lvec,zz,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 1312 PetscFunctionReturn(0); 1313 } 1314 1315 /* 1316 This only works correctly for square matrices where the subblock A->A is the 1317 diagonal block 1318 */ 1319 #undef __FUNCT__ 1320 #define __FUNCT__ "MatGetDiagonal_MPIBAIJ" 1321 int MatGetDiagonal_MPIBAIJ(Mat A,Vec v) 1322 { 1323 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1324 int ierr; 1325 1326 PetscFunctionBegin; 1327 if (A->M != A->N) SETERRQ(PETSC_ERR_SUP,"Supports only square matrix where A->A is diag block"); 1328 ierr = MatGetDiagonal(a->A,v);CHKERRQ(ierr); 1329 PetscFunctionReturn(0); 1330 } 1331 1332 #undef __FUNCT__ 1333 #define __FUNCT__ "MatScale_MPIBAIJ" 1334 int MatScale_MPIBAIJ(const PetscScalar *aa,Mat A) 1335 { 1336 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1337 int ierr; 1338 1339 PetscFunctionBegin; 1340 ierr = MatScale(aa,a->A);CHKERRQ(ierr); 1341 ierr = MatScale(aa,a->B);CHKERRQ(ierr); 1342 PetscFunctionReturn(0); 1343 } 1344 1345 #undef __FUNCT__ 1346 #define __FUNCT__ "MatGetRow_MPIBAIJ" 1347 int MatGetRow_MPIBAIJ(Mat matin,int row,int *nz,int **idx,PetscScalar **v) 1348 { 1349 Mat_MPIBAIJ *mat = (Mat_MPIBAIJ*)matin->data; 1350 PetscScalar *vworkA,*vworkB,**pvA,**pvB,*v_p; 1351 int bs = mat->bs,bs2 = mat->bs2,i,ierr,*cworkA,*cworkB,**pcA,**pcB; 1352 int nztot,nzA,nzB,lrow,brstart = mat->rstart*bs,brend = mat->rend*bs; 1353 int *cmap,*idx_p,cstart = mat->cstart; 1354 1355 PetscFunctionBegin; 1356 if (mat->getrowactive == PETSC_TRUE) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Already active"); 1357 mat->getrowactive = PETSC_TRUE; 1358 1359 if (!mat->rowvalues && (idx || v)) { 1360 /* 1361 allocate enough space to hold information from the longest row. 1362 */ 1363 Mat_SeqBAIJ *Aa = (Mat_SeqBAIJ*)mat->A->data,*Ba = (Mat_SeqBAIJ*)mat->B->data; 1364 int max = 1,mbs = mat->mbs,tmp; 1365 for (i=0; i<mbs; i++) { 1366 tmp = Aa->i[i+1] - Aa->i[i] + Ba->i[i+1] - Ba->i[i]; 1367 if (max < tmp) { max = tmp; } 1368 } 1369 ierr = PetscMalloc(max*bs2*(sizeof(int)+sizeof(PetscScalar)),&mat->rowvalues);CHKERRQ(ierr); 1370 mat->rowindices = (int*)(mat->rowvalues + max*bs2); 1371 } 1372 1373 if (row < brstart || row >= brend) SETERRQ(PETSC_ERR_SUP,"Only local rows") 1374 lrow = row - brstart; 1375 1376 pvA = &vworkA; pcA = &cworkA; pvB = &vworkB; pcB = &cworkB; 1377 if (!v) {pvA = 0; pvB = 0;} 1378 if (!idx) {pcA = 0; if (!v) pcB = 0;} 1379 ierr = (*mat->A->ops->getrow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1380 ierr = (*mat->B->ops->getrow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1381 nztot = nzA + nzB; 1382 1383 cmap = mat->garray; 1384 if (v || idx) { 1385 if (nztot) { 1386 /* Sort by increasing column numbers, assuming A and B already sorted */ 1387 int imark = -1; 1388 if (v) { 1389 *v = v_p = mat->rowvalues; 1390 for (i=0; i<nzB; i++) { 1391 if (cmap[cworkB[i]/bs] < cstart) v_p[i] = vworkB[i]; 1392 else break; 1393 } 1394 imark = i; 1395 for (i=0; i<nzA; i++) v_p[imark+i] = vworkA[i]; 1396 for (i=imark; i<nzB; i++) v_p[nzA+i] = vworkB[i]; 1397 } 1398 if (idx) { 1399 *idx = idx_p = mat->rowindices; 1400 if (imark > -1) { 1401 for (i=0; i<imark; i++) { 1402 idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs; 1403 } 1404 } else { 1405 for (i=0; i<nzB; i++) { 1406 if (cmap[cworkB[i]/bs] < cstart) 1407 idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ; 1408 else break; 1409 } 1410 imark = i; 1411 } 1412 for (i=0; i<nzA; i++) idx_p[imark+i] = cstart*bs + cworkA[i]; 1413 for (i=imark; i<nzB; i++) idx_p[nzA+i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ; 1414 } 1415 } else { 1416 if (idx) *idx = 0; 1417 if (v) *v = 0; 1418 } 1419 } 1420 *nz = nztot; 1421 ierr = (*mat->A->ops->restorerow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1422 ierr = (*mat->B->ops->restorerow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1423 PetscFunctionReturn(0); 1424 } 1425 1426 #undef __FUNCT__ 1427 #define __FUNCT__ "MatRestoreRow_MPIBAIJ" 1428 int MatRestoreRow_MPIBAIJ(Mat mat,int row,int *nz,int **idx,PetscScalar **v) 1429 { 1430 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1431 1432 PetscFunctionBegin; 1433 if (baij->getrowactive == PETSC_FALSE) { 1434 SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"MatGetRow not called"); 1435 } 1436 baij->getrowactive = PETSC_FALSE; 1437 PetscFunctionReturn(0); 1438 } 1439 1440 #undef __FUNCT__ 1441 #define __FUNCT__ "MatGetBlockSize_MPIBAIJ" 1442 int MatGetBlockSize_MPIBAIJ(Mat mat,int *bs) 1443 { 1444 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1445 1446 PetscFunctionBegin; 1447 *bs = baij->bs; 1448 PetscFunctionReturn(0); 1449 } 1450 1451 #undef __FUNCT__ 1452 #define __FUNCT__ "MatZeroEntries_MPIBAIJ" 1453 int MatZeroEntries_MPIBAIJ(Mat A) 1454 { 1455 Mat_MPIBAIJ *l = (Mat_MPIBAIJ*)A->data; 1456 int ierr; 1457 1458 PetscFunctionBegin; 1459 ierr = MatZeroEntries(l->A);CHKERRQ(ierr); 1460 ierr = MatZeroEntries(l->B);CHKERRQ(ierr); 1461 PetscFunctionReturn(0); 1462 } 1463 1464 #undef __FUNCT__ 1465 #define __FUNCT__ "MatGetInfo_MPIBAIJ" 1466 int MatGetInfo_MPIBAIJ(Mat matin,MatInfoType flag,MatInfo *info) 1467 { 1468 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)matin->data; 1469 Mat A = a->A,B = a->B; 1470 int ierr; 1471 PetscReal isend[5],irecv[5]; 1472 1473 PetscFunctionBegin; 1474 info->block_size = (PetscReal)a->bs; 1475 ierr = MatGetInfo(A,MAT_LOCAL,info);CHKERRQ(ierr); 1476 isend[0] = info->nz_used; isend[1] = info->nz_allocated; isend[2] = info->nz_unneeded; 1477 isend[3] = info->memory; isend[4] = info->mallocs; 1478 ierr = MatGetInfo(B,MAT_LOCAL,info);CHKERRQ(ierr); 1479 isend[0] += info->nz_used; isend[1] += info->nz_allocated; isend[2] += info->nz_unneeded; 1480 isend[3] += info->memory; isend[4] += info->mallocs; 1481 if (flag == MAT_LOCAL) { 1482 info->nz_used = isend[0]; 1483 info->nz_allocated = isend[1]; 1484 info->nz_unneeded = isend[2]; 1485 info->memory = isend[3]; 1486 info->mallocs = isend[4]; 1487 } else if (flag == MAT_GLOBAL_MAX) { 1488 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPI_MAX,matin->comm);CHKERRQ(ierr); 1489 info->nz_used = irecv[0]; 1490 info->nz_allocated = irecv[1]; 1491 info->nz_unneeded = irecv[2]; 1492 info->memory = irecv[3]; 1493 info->mallocs = irecv[4]; 1494 } else if (flag == MAT_GLOBAL_SUM) { 1495 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPI_SUM,matin->comm);CHKERRQ(ierr); 1496 info->nz_used = irecv[0]; 1497 info->nz_allocated = irecv[1]; 1498 info->nz_unneeded = irecv[2]; 1499 info->memory = irecv[3]; 1500 info->mallocs = irecv[4]; 1501 } else { 1502 SETERRQ1(1,"Unknown MatInfoType argument %d",flag); 1503 } 1504 info->rows_global = (PetscReal)A->M; 1505 info->columns_global = (PetscReal)A->N; 1506 info->rows_local = (PetscReal)A->m; 1507 info->columns_local = (PetscReal)A->N; 1508 info->fill_ratio_given = 0; /* no parallel LU/ILU/Cholesky */ 1509 info->fill_ratio_needed = 0; 1510 info->factor_mallocs = 0; 1511 PetscFunctionReturn(0); 1512 } 1513 1514 #undef __FUNCT__ 1515 #define __FUNCT__ "MatSetOption_MPIBAIJ" 1516 int MatSetOption_MPIBAIJ(Mat A,MatOption op) 1517 { 1518 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1519 int ierr; 1520 1521 PetscFunctionBegin; 1522 switch (op) { 1523 case MAT_NO_NEW_NONZERO_LOCATIONS: 1524 case MAT_YES_NEW_NONZERO_LOCATIONS: 1525 case MAT_COLUMNS_UNSORTED: 1526 case MAT_COLUMNS_SORTED: 1527 case MAT_NEW_NONZERO_ALLOCATION_ERR: 1528 case MAT_KEEP_ZEROED_ROWS: 1529 case MAT_NEW_NONZERO_LOCATION_ERR: 1530 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1531 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1532 break; 1533 case MAT_ROW_ORIENTED: 1534 a->roworiented = PETSC_TRUE; 1535 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1536 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1537 break; 1538 case MAT_ROWS_SORTED: 1539 case MAT_ROWS_UNSORTED: 1540 case MAT_YES_NEW_DIAGONALS: 1541 PetscLogInfo(A,"Info:MatSetOption_MPIBAIJ:Option ignored\n"); 1542 break; 1543 case MAT_COLUMN_ORIENTED: 1544 a->roworiented = PETSC_FALSE; 1545 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1546 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1547 break; 1548 case MAT_IGNORE_OFF_PROC_ENTRIES: 1549 a->donotstash = PETSC_TRUE; 1550 break; 1551 case MAT_NO_NEW_DIAGONALS: 1552 SETERRQ(PETSC_ERR_SUP,"MAT_NO_NEW_DIAGONALS"); 1553 case MAT_USE_HASH_TABLE: 1554 a->ht_flag = PETSC_TRUE; 1555 break; 1556 default: 1557 SETERRQ(PETSC_ERR_SUP,"unknown option"); 1558 } 1559 PetscFunctionReturn(0); 1560 } 1561 1562 #undef __FUNCT__ 1563 #define __FUNCT__ "MatTranspose_MPIBAIJ(" 1564 int MatTranspose_MPIBAIJ(Mat A,Mat *matout) 1565 { 1566 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)A->data; 1567 Mat_SeqBAIJ *Aloc; 1568 Mat B; 1569 int ierr,M=A->M,N=A->N,*ai,*aj,i,*rvals,j,k,col; 1570 int bs=baij->bs,mbs=baij->mbs; 1571 MatScalar *a; 1572 1573 PetscFunctionBegin; 1574 if (!matout && M != N) SETERRQ(PETSC_ERR_ARG_SIZ,"Square matrix only for in-place"); 1575 ierr = MatCreateMPIBAIJ(A->comm,baij->bs,A->n,A->m,N,M,0,PETSC_NULL,0,PETSC_NULL,&B);CHKERRQ(ierr); 1576 1577 /* copy over the A part */ 1578 Aloc = (Mat_SeqBAIJ*)baij->A->data; 1579 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1580 ierr = PetscMalloc(bs*sizeof(int),&rvals);CHKERRQ(ierr); 1581 1582 for (i=0; i<mbs; i++) { 1583 rvals[0] = bs*(baij->rstart + i); 1584 for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; } 1585 for (j=ai[i]; j<ai[i+1]; j++) { 1586 col = (baij->cstart+aj[j])*bs; 1587 for (k=0; k<bs; k++) { 1588 ierr = MatSetValues_MPIBAIJ_MatScalar(B,1,&col,bs,rvals,a,INSERT_VALUES);CHKERRQ(ierr); 1589 col++; a += bs; 1590 } 1591 } 1592 } 1593 /* copy over the B part */ 1594 Aloc = (Mat_SeqBAIJ*)baij->B->data; 1595 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1596 for (i=0; i<mbs; i++) { 1597 rvals[0] = bs*(baij->rstart + i); 1598 for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; } 1599 for (j=ai[i]; j<ai[i+1]; j++) { 1600 col = baij->garray[aj[j]]*bs; 1601 for (k=0; k<bs; k++) { 1602 ierr = MatSetValues_MPIBAIJ_MatScalar(B,1,&col,bs,rvals,a,INSERT_VALUES);CHKERRQ(ierr); 1603 col++; a += bs; 1604 } 1605 } 1606 } 1607 ierr = PetscFree(rvals);CHKERRQ(ierr); 1608 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1609 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1610 1611 if (matout) { 1612 *matout = B; 1613 } else { 1614 ierr = MatHeaderCopy(A,B);CHKERRQ(ierr); 1615 } 1616 PetscFunctionReturn(0); 1617 } 1618 1619 #undef __FUNCT__ 1620 #define __FUNCT__ "MatDiagonalScale_MPIBAIJ" 1621 int MatDiagonalScale_MPIBAIJ(Mat mat,Vec ll,Vec rr) 1622 { 1623 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1624 Mat a = baij->A,b = baij->B; 1625 int ierr,s1,s2,s3; 1626 1627 PetscFunctionBegin; 1628 ierr = MatGetLocalSize(mat,&s2,&s3);CHKERRQ(ierr); 1629 if (rr) { 1630 ierr = VecGetLocalSize(rr,&s1);CHKERRQ(ierr); 1631 if (s1!=s3) SETERRQ(PETSC_ERR_ARG_SIZ,"right vector non-conforming local size"); 1632 /* Overlap communication with computation. */ 1633 ierr = VecScatterBegin(rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD,baij->Mvctx);CHKERRQ(ierr); 1634 } 1635 if (ll) { 1636 ierr = VecGetLocalSize(ll,&s1);CHKERRQ(ierr); 1637 if (s1!=s2) SETERRQ(PETSC_ERR_ARG_SIZ,"left vector non-conforming local size"); 1638 ierr = (*b->ops->diagonalscale)(b,ll,PETSC_NULL);CHKERRQ(ierr); 1639 } 1640 /* scale the diagonal block */ 1641 ierr = (*a->ops->diagonalscale)(a,ll,rr);CHKERRQ(ierr); 1642 1643 if (rr) { 1644 /* Do a scatter end and then right scale the off-diagonal block */ 1645 ierr = VecScatterEnd(rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD,baij->Mvctx);CHKERRQ(ierr); 1646 ierr = (*b->ops->diagonalscale)(b,PETSC_NULL,baij->lvec);CHKERRQ(ierr); 1647 } 1648 1649 PetscFunctionReturn(0); 1650 } 1651 1652 #undef __FUNCT__ 1653 #define __FUNCT__ "MatZeroRows_MPIBAIJ" 1654 int MatZeroRows_MPIBAIJ(Mat A,IS is,const PetscScalar *diag) 1655 { 1656 Mat_MPIBAIJ *l = (Mat_MPIBAIJ*)A->data; 1657 int i,ierr,N,*rows,*owners = l->rowners,size = l->size; 1658 int *nprocs,j,idx,nsends,row; 1659 int nmax,*svalues,*starts,*owner,nrecvs,rank = l->rank; 1660 int *rvalues,tag = A->tag,count,base,slen,n,*source; 1661 int *lens,imdex,*lrows,*values,bs=l->bs,rstart_bs=l->rstart_bs; 1662 MPI_Comm comm = A->comm; 1663 MPI_Request *send_waits,*recv_waits; 1664 MPI_Status recv_status,*send_status; 1665 IS istmp; 1666 PetscTruth found; 1667 1668 PetscFunctionBegin; 1669 ierr = ISGetLocalSize(is,&N);CHKERRQ(ierr); 1670 ierr = ISGetIndices(is,&rows);CHKERRQ(ierr); 1671 1672 /* first count number of contributors to each processor */ 1673 ierr = PetscMalloc(2*size*sizeof(int),&nprocs);CHKERRQ(ierr); 1674 ierr = PetscMemzero(nprocs,2*size*sizeof(int));CHKERRQ(ierr); 1675 ierr = PetscMalloc((N+1)*sizeof(int),&owner);CHKERRQ(ierr); /* see note*/ 1676 for (i=0; i<N; i++) { 1677 idx = rows[i]; 1678 found = PETSC_FALSE; 1679 for (j=0; j<size; j++) { 1680 if (idx >= owners[j]*bs && idx < owners[j+1]*bs) { 1681 nprocs[2*j]++; nprocs[2*j+1] = 1; owner[i] = j; found = PETSC_TRUE; break; 1682 } 1683 } 1684 if (!found) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Index out of range"); 1685 } 1686 nsends = 0; for (i=0; i<size; i++) { nsends += nprocs[2*i+1];} 1687 1688 /* inform other processors of number of messages and max length*/ 1689 ierr = PetscMaxSum(comm,nprocs,&nmax,&nrecvs);CHKERRQ(ierr); 1690 1691 /* post receives: */ 1692 ierr = PetscMalloc((nrecvs+1)*(nmax+1)*sizeof(int),&rvalues);CHKERRQ(ierr); 1693 ierr = PetscMalloc((nrecvs+1)*sizeof(MPI_Request),&recv_waits);CHKERRQ(ierr); 1694 for (i=0; i<nrecvs; i++) { 1695 ierr = MPI_Irecv(rvalues+nmax*i,nmax,MPI_INT,MPI_ANY_SOURCE,tag,comm,recv_waits+i);CHKERRQ(ierr); 1696 } 1697 1698 /* do sends: 1699 1) starts[i] gives the starting index in svalues for stuff going to 1700 the ith processor 1701 */ 1702 ierr = PetscMalloc((N+1)*sizeof(int),&svalues);CHKERRQ(ierr); 1703 ierr = PetscMalloc((nsends+1)*sizeof(MPI_Request),&send_waits);CHKERRQ(ierr); 1704 ierr = PetscMalloc((size+1)*sizeof(int),&starts);CHKERRQ(ierr); 1705 starts[0] = 0; 1706 for (i=1; i<size; i++) { starts[i] = starts[i-1] + nprocs[2*i-2];} 1707 for (i=0; i<N; i++) { 1708 svalues[starts[owner[i]]++] = rows[i]; 1709 } 1710 ierr = ISRestoreIndices(is,&rows);CHKERRQ(ierr); 1711 1712 starts[0] = 0; 1713 for (i=1; i<size+1; i++) { starts[i] = starts[i-1] + nprocs[2*i-2];} 1714 count = 0; 1715 for (i=0; i<size; i++) { 1716 if (nprocs[2*i+1]) { 1717 ierr = MPI_Isend(svalues+starts[i],nprocs[2*i],MPI_INT,i,tag,comm,send_waits+count++);CHKERRQ(ierr); 1718 } 1719 } 1720 ierr = PetscFree(starts);CHKERRQ(ierr); 1721 1722 base = owners[rank]*bs; 1723 1724 /* wait on receives */ 1725 ierr = PetscMalloc(2*(nrecvs+1)*sizeof(int),&lens);CHKERRQ(ierr); 1726 source = lens + nrecvs; 1727 count = nrecvs; slen = 0; 1728 while (count) { 1729 ierr = MPI_Waitany(nrecvs,recv_waits,&imdex,&recv_status);CHKERRQ(ierr); 1730 /* unpack receives into our local space */ 1731 ierr = MPI_Get_count(&recv_status,MPI_INT,&n);CHKERRQ(ierr); 1732 source[imdex] = recv_status.MPI_SOURCE; 1733 lens[imdex] = n; 1734 slen += n; 1735 count--; 1736 } 1737 ierr = PetscFree(recv_waits);CHKERRQ(ierr); 1738 1739 /* move the data into the send scatter */ 1740 ierr = PetscMalloc((slen+1)*sizeof(int),&lrows);CHKERRQ(ierr); 1741 count = 0; 1742 for (i=0; i<nrecvs; i++) { 1743 values = rvalues + i*nmax; 1744 for (j=0; j<lens[i]; j++) { 1745 lrows[count++] = values[j] - base; 1746 } 1747 } 1748 ierr = PetscFree(rvalues);CHKERRQ(ierr); 1749 ierr = PetscFree(lens);CHKERRQ(ierr); 1750 ierr = PetscFree(owner);CHKERRQ(ierr); 1751 ierr = PetscFree(nprocs);CHKERRQ(ierr); 1752 1753 /* actually zap the local rows */ 1754 ierr = ISCreateGeneral(PETSC_COMM_SELF,slen,lrows,&istmp);CHKERRQ(ierr); 1755 PetscLogObjectParent(A,istmp); 1756 1757 /* 1758 Zero the required rows. If the "diagonal block" of the matrix 1759 is square and the user wishes to set the diagonal we use seperate 1760 code so that MatSetValues() is not called for each diagonal allocating 1761 new memory, thus calling lots of mallocs and slowing things down. 1762 1763 Contributed by: Mathew Knepley 1764 */ 1765 /* must zero l->B before l->A because the (diag) case below may put values into l->B*/ 1766 ierr = MatZeroRows_SeqBAIJ(l->B,istmp,0);CHKERRQ(ierr); 1767 if (diag && (l->A->M == l->A->N)) { 1768 ierr = MatZeroRows_SeqBAIJ(l->A,istmp,diag);CHKERRQ(ierr); 1769 } else if (diag) { 1770 ierr = MatZeroRows_SeqBAIJ(l->A,istmp,0);CHKERRQ(ierr); 1771 if (((Mat_SeqBAIJ*)l->A->data)->nonew) { 1772 SETERRQ(PETSC_ERR_SUP,"MatZeroRows() on rectangular matrices cannot be used with the Mat options \n\ 1773 MAT_NO_NEW_NONZERO_LOCATIONS,MAT_NEW_NONZERO_LOCATION_ERR,MAT_NEW_NONZERO_ALLOCATION_ERR"); 1774 } 1775 for (i=0; i<slen; i++) { 1776 row = lrows[i] + rstart_bs; 1777 ierr = MatSetValues(A,1,&row,1,&row,diag,INSERT_VALUES);CHKERRQ(ierr); 1778 } 1779 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1780 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1781 } else { 1782 ierr = MatZeroRows_SeqBAIJ(l->A,istmp,0);CHKERRQ(ierr); 1783 } 1784 1785 ierr = ISDestroy(istmp);CHKERRQ(ierr); 1786 ierr = PetscFree(lrows);CHKERRQ(ierr); 1787 1788 /* wait on sends */ 1789 if (nsends) { 1790 ierr = PetscMalloc(nsends*sizeof(MPI_Status),&send_status);CHKERRQ(ierr); 1791 ierr = MPI_Waitall(nsends,send_waits,send_status);CHKERRQ(ierr); 1792 ierr = PetscFree(send_status);CHKERRQ(ierr); 1793 } 1794 ierr = PetscFree(send_waits);CHKERRQ(ierr); 1795 ierr = PetscFree(svalues);CHKERRQ(ierr); 1796 1797 PetscFunctionReturn(0); 1798 } 1799 1800 #undef __FUNCT__ 1801 #define __FUNCT__ "MatPrintHelp_MPIBAIJ" 1802 int MatPrintHelp_MPIBAIJ(Mat A) 1803 { 1804 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1805 MPI_Comm comm = A->comm; 1806 static int called = 0; 1807 int ierr; 1808 1809 PetscFunctionBegin; 1810 if (!a->rank) { 1811 ierr = MatPrintHelp_SeqBAIJ(a->A);CHKERRQ(ierr); 1812 } 1813 if (called) {PetscFunctionReturn(0);} else called = 1; 1814 ierr = (*PetscHelpPrintf)(comm," Options for MATMPIBAIJ matrix format (the defaults):\n");CHKERRQ(ierr); 1815 ierr = (*PetscHelpPrintf)(comm," -mat_use_hash_table <factor>: Use hashtable for efficient matrix assembly\n");CHKERRQ(ierr); 1816 PetscFunctionReturn(0); 1817 } 1818 1819 #undef __FUNCT__ 1820 #define __FUNCT__ "MatSetUnfactored_MPIBAIJ" 1821 int MatSetUnfactored_MPIBAIJ(Mat A) 1822 { 1823 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1824 int ierr; 1825 1826 PetscFunctionBegin; 1827 ierr = MatSetUnfactored(a->A);CHKERRQ(ierr); 1828 PetscFunctionReturn(0); 1829 } 1830 1831 static int MatDuplicate_MPIBAIJ(Mat,MatDuplicateOption,Mat *); 1832 1833 #undef __FUNCT__ 1834 #define __FUNCT__ "MatEqual_MPIBAIJ" 1835 int MatEqual_MPIBAIJ(Mat A,Mat B,PetscTruth *flag) 1836 { 1837 Mat_MPIBAIJ *matB = (Mat_MPIBAIJ*)B->data,*matA = (Mat_MPIBAIJ*)A->data; 1838 Mat a,b,c,d; 1839 PetscTruth flg; 1840 int ierr; 1841 1842 PetscFunctionBegin; 1843 a = matA->A; b = matA->B; 1844 c = matB->A; d = matB->B; 1845 1846 ierr = MatEqual(a,c,&flg);CHKERRQ(ierr); 1847 if (flg == PETSC_TRUE) { 1848 ierr = MatEqual(b,d,&flg);CHKERRQ(ierr); 1849 } 1850 ierr = MPI_Allreduce(&flg,flag,1,MPI_INT,MPI_LAND,A->comm);CHKERRQ(ierr); 1851 PetscFunctionReturn(0); 1852 } 1853 1854 1855 #undef __FUNCT__ 1856 #define __FUNCT__ "MatSetUpPreallocation_MPIBAIJ" 1857 int MatSetUpPreallocation_MPIBAIJ(Mat A) 1858 { 1859 int ierr; 1860 1861 PetscFunctionBegin; 1862 ierr = MatMPIBAIJSetPreallocation(A,1,PETSC_DEFAULT,0,PETSC_DEFAULT,0);CHKERRQ(ierr); 1863 PetscFunctionReturn(0); 1864 } 1865 1866 /* -------------------------------------------------------------------*/ 1867 static struct _MatOps MatOps_Values = { 1868 MatSetValues_MPIBAIJ, 1869 MatGetRow_MPIBAIJ, 1870 MatRestoreRow_MPIBAIJ, 1871 MatMult_MPIBAIJ, 1872 /* 4*/ MatMultAdd_MPIBAIJ, 1873 MatMultTranspose_MPIBAIJ, 1874 MatMultTransposeAdd_MPIBAIJ, 1875 0, 1876 0, 1877 0, 1878 /*10*/ 0, 1879 0, 1880 0, 1881 0, 1882 MatTranspose_MPIBAIJ, 1883 /*15*/ MatGetInfo_MPIBAIJ, 1884 MatEqual_MPIBAIJ, 1885 MatGetDiagonal_MPIBAIJ, 1886 MatDiagonalScale_MPIBAIJ, 1887 MatNorm_MPIBAIJ, 1888 /*20*/ MatAssemblyBegin_MPIBAIJ, 1889 MatAssemblyEnd_MPIBAIJ, 1890 0, 1891 MatSetOption_MPIBAIJ, 1892 MatZeroEntries_MPIBAIJ, 1893 /*25*/ MatZeroRows_MPIBAIJ, 1894 0, 1895 0, 1896 0, 1897 0, 1898 /*30*/ MatSetUpPreallocation_MPIBAIJ, 1899 0, 1900 0, 1901 0, 1902 0, 1903 /*35*/ MatDuplicate_MPIBAIJ, 1904 0, 1905 0, 1906 0, 1907 0, 1908 /*40*/ 0, 1909 MatGetSubMatrices_MPIBAIJ, 1910 MatIncreaseOverlap_MPIBAIJ, 1911 MatGetValues_MPIBAIJ, 1912 0, 1913 /*45*/ MatPrintHelp_MPIBAIJ, 1914 MatScale_MPIBAIJ, 1915 0, 1916 0, 1917 0, 1918 /*50*/ MatGetBlockSize_MPIBAIJ, 1919 0, 1920 0, 1921 0, 1922 0, 1923 /*55*/ 0, 1924 0, 1925 MatSetUnfactored_MPIBAIJ, 1926 0, 1927 MatSetValuesBlocked_MPIBAIJ, 1928 /*60*/ 0, 1929 MatDestroy_MPIBAIJ, 1930 MatView_MPIBAIJ, 1931 MatGetPetscMaps_Petsc, 1932 0, 1933 /*65*/ 0, 1934 0, 1935 0, 1936 0, 1937 0, 1938 /*70*/ MatGetRowMax_MPIBAIJ, 1939 0, 1940 0, 1941 0, 1942 0, 1943 /*75*/ 0, 1944 0, 1945 0, 1946 0, 1947 0, 1948 /*80*/ 0, 1949 0, 1950 0, 1951 0, 1952 0, 1953 /*85*/ MatLoad_MPIBAIJ 1954 }; 1955 1956 1957 EXTERN_C_BEGIN 1958 #undef __FUNCT__ 1959 #define __FUNCT__ "MatGetDiagonalBlock_MPIBAIJ" 1960 int MatGetDiagonalBlock_MPIBAIJ(Mat A,PetscTruth *iscopy,MatReuse reuse,Mat *a) 1961 { 1962 PetscFunctionBegin; 1963 *a = ((Mat_MPIBAIJ *)A->data)->A; 1964 *iscopy = PETSC_FALSE; 1965 PetscFunctionReturn(0); 1966 } 1967 EXTERN_C_END 1968 1969 EXTERN_C_BEGIN 1970 #undef __FUNCT__ 1971 #define __FUNCT__ "MatMPIBAIJSetPreallocation_MPIBAIJ" 1972 int MatMPIBAIJSetPreallocation_MPIBAIJ(Mat B,int bs,int d_nz,int *d_nnz,int o_nz,int *o_nnz) 1973 { 1974 Mat_MPIBAIJ *b; 1975 int ierr,i; 1976 1977 PetscFunctionBegin; 1978 B->preallocated = PETSC_TRUE; 1979 ierr = PetscOptionsGetInt(PETSC_NULL,"-mat_block_size",&bs,PETSC_NULL);CHKERRQ(ierr); 1980 1981 if (bs < 1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Invalid block size specified, must be positive"); 1982 if (d_nz == PETSC_DEFAULT || d_nz == PETSC_DECIDE) d_nz = 5; 1983 if (o_nz == PETSC_DEFAULT || o_nz == PETSC_DECIDE) o_nz = 2; 1984 if (d_nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"d_nz cannot be less than 0: value %d",d_nz); 1985 if (o_nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"o_nz cannot be less than 0: value %d",o_nz); 1986 if (d_nnz) { 1987 for (i=0; i<B->m/bs; i++) { 1988 if (d_nnz[i] < 0) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"d_nnz cannot be less than -1: local row %d value %d",i,d_nnz[i]); 1989 } 1990 } 1991 if (o_nnz) { 1992 for (i=0; i<B->m/bs; i++) { 1993 if (o_nnz[i] < 0) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"o_nnz cannot be less than -1: local row %d value %d",i,o_nnz[i]); 1994 } 1995 } 1996 1997 ierr = PetscSplitOwnershipBlock(B->comm,bs,&B->m,&B->M);CHKERRQ(ierr); 1998 ierr = PetscSplitOwnershipBlock(B->comm,bs,&B->n,&B->N);CHKERRQ(ierr); 1999 ierr = PetscMapCreateMPI(B->comm,B->m,B->M,&B->rmap);CHKERRQ(ierr); 2000 ierr = PetscMapCreateMPI(B->comm,B->n,B->N,&B->cmap);CHKERRQ(ierr); 2001 2002 b = (Mat_MPIBAIJ*)B->data; 2003 b->bs = bs; 2004 b->bs2 = bs*bs; 2005 b->mbs = B->m/bs; 2006 b->nbs = B->n/bs; 2007 b->Mbs = B->M/bs; 2008 b->Nbs = B->N/bs; 2009 2010 ierr = MPI_Allgather(&b->mbs,1,MPI_INT,b->rowners+1,1,MPI_INT,B->comm);CHKERRQ(ierr); 2011 b->rowners[0] = 0; 2012 for (i=2; i<=b->size; i++) { 2013 b->rowners[i] += b->rowners[i-1]; 2014 } 2015 b->rstart = b->rowners[b->rank]; 2016 b->rend = b->rowners[b->rank+1]; 2017 2018 ierr = MPI_Allgather(&b->nbs,1,MPI_INT,b->cowners+1,1,MPI_INT,B->comm);CHKERRQ(ierr); 2019 b->cowners[0] = 0; 2020 for (i=2; i<=b->size; i++) { 2021 b->cowners[i] += b->cowners[i-1]; 2022 } 2023 b->cstart = b->cowners[b->rank]; 2024 b->cend = b->cowners[b->rank+1]; 2025 2026 for (i=0; i<=b->size; i++) { 2027 b->rowners_bs[i] = b->rowners[i]*bs; 2028 } 2029 b->rstart_bs = b->rstart*bs; 2030 b->rend_bs = b->rend*bs; 2031 b->cstart_bs = b->cstart*bs; 2032 b->cend_bs = b->cend*bs; 2033 2034 ierr = MatCreateSeqBAIJ(PETSC_COMM_SELF,bs,B->m,B->n,d_nz,d_nnz,&b->A);CHKERRQ(ierr); 2035 PetscLogObjectParent(B,b->A); 2036 ierr = MatCreateSeqBAIJ(PETSC_COMM_SELF,bs,B->m,B->N,o_nz,o_nnz,&b->B);CHKERRQ(ierr); 2037 PetscLogObjectParent(B,b->B); 2038 ierr = MatStashCreate_Private(B->comm,bs,&B->bstash);CHKERRQ(ierr); 2039 2040 PetscFunctionReturn(0); 2041 } 2042 EXTERN_C_END 2043 2044 EXTERN_C_BEGIN 2045 EXTERN int MatDiagonalScaleLocal_MPIBAIJ(Mat,Vec); 2046 EXTERN int MatSetHashTableFactor_MPIBAIJ(Mat,PetscReal); 2047 EXTERN_C_END 2048 2049 /*MC 2050 MATMPIBAIJ = "mpibaij" - A matrix type to be used for distributed block sparse matrices. 2051 2052 Options Database Keys: 2053 . -mat_type mpibaij - sets the matrix type to "mpibaij" during a call to MatSetFromOptions() 2054 2055 Level: beginner 2056 2057 .seealso: MatCreateMPIBAIJ 2058 M*/ 2059 2060 EXTERN_C_BEGIN 2061 #undef __FUNCT__ 2062 #define __FUNCT__ "MatCreate_MPIBAIJ" 2063 int MatCreate_MPIBAIJ(Mat B) 2064 { 2065 Mat_MPIBAIJ *b; 2066 int ierr; 2067 PetscTruth flg; 2068 2069 PetscFunctionBegin; 2070 2071 ierr = PetscNew(Mat_MPIBAIJ,&b);CHKERRQ(ierr); 2072 B->data = (void*)b; 2073 2074 ierr = PetscMemzero(b,sizeof(Mat_MPIBAIJ));CHKERRQ(ierr); 2075 ierr = PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));CHKERRQ(ierr); 2076 B->mapping = 0; 2077 B->factor = 0; 2078 B->assembled = PETSC_FALSE; 2079 2080 B->insertmode = NOT_SET_VALUES; 2081 ierr = MPI_Comm_rank(B->comm,&b->rank);CHKERRQ(ierr); 2082 ierr = MPI_Comm_size(B->comm,&b->size);CHKERRQ(ierr); 2083 2084 /* build local table of row and column ownerships */ 2085 ierr = PetscMalloc(3*(b->size+2)*sizeof(int),&b->rowners);CHKERRQ(ierr); 2086 PetscLogObjectMemory(B,3*(b->size+2)*sizeof(int)+sizeof(struct _p_Mat)+sizeof(Mat_MPIBAIJ)); 2087 b->cowners = b->rowners + b->size + 2; 2088 b->rowners_bs = b->cowners + b->size + 2; 2089 2090 /* build cache for off array entries formed */ 2091 ierr = MatStashCreate_Private(B->comm,1,&B->stash);CHKERRQ(ierr); 2092 b->donotstash = PETSC_FALSE; 2093 b->colmap = PETSC_NULL; 2094 b->garray = PETSC_NULL; 2095 b->roworiented = PETSC_TRUE; 2096 2097 #if defined(PETSC_USE_MAT_SINGLE) 2098 /* stuff for MatSetValues_XXX in single precision */ 2099 b->setvalueslen = 0; 2100 b->setvaluescopy = PETSC_NULL; 2101 #endif 2102 2103 /* stuff used in block assembly */ 2104 b->barray = 0; 2105 2106 /* stuff used for matrix vector multiply */ 2107 b->lvec = 0; 2108 b->Mvctx = 0; 2109 2110 /* stuff for MatGetRow() */ 2111 b->rowindices = 0; 2112 b->rowvalues = 0; 2113 b->getrowactive = PETSC_FALSE; 2114 2115 /* hash table stuff */ 2116 b->ht = 0; 2117 b->hd = 0; 2118 b->ht_size = 0; 2119 b->ht_flag = PETSC_FALSE; 2120 b->ht_fact = 0; 2121 b->ht_total_ct = 0; 2122 b->ht_insert_ct = 0; 2123 2124 ierr = PetscOptionsHasName(PETSC_NULL,"-mat_use_hash_table",&flg);CHKERRQ(ierr); 2125 if (flg) { 2126 PetscReal fact = 1.39; 2127 ierr = MatSetOption(B,MAT_USE_HASH_TABLE);CHKERRQ(ierr); 2128 ierr = PetscOptionsGetReal(PETSC_NULL,"-mat_use_hash_table",&fact,PETSC_NULL);CHKERRQ(ierr); 2129 if (fact <= 1.0) fact = 1.39; 2130 ierr = MatMPIBAIJSetHashTableFactor(B,fact);CHKERRQ(ierr); 2131 PetscLogInfo(0,"MatCreateMPIBAIJ:Hash table Factor used %5.2f\n",fact); 2132 } 2133 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatStoreValues_C", 2134 "MatStoreValues_MPIBAIJ", 2135 MatStoreValues_MPIBAIJ);CHKERRQ(ierr); 2136 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatRetrieveValues_C", 2137 "MatRetrieveValues_MPIBAIJ", 2138 MatRetrieveValues_MPIBAIJ);CHKERRQ(ierr); 2139 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatGetDiagonalBlock_C", 2140 "MatGetDiagonalBlock_MPIBAIJ", 2141 MatGetDiagonalBlock_MPIBAIJ);CHKERRQ(ierr); 2142 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatMPIBAIJSetPreallocation_C", 2143 "MatMPIBAIJSetPreallocation_MPIBAIJ", 2144 MatMPIBAIJSetPreallocation_MPIBAIJ);CHKERRQ(ierr); 2145 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatDiagonalScaleLocal_C", 2146 "MatDiagonalScaleLocal_MPIBAIJ", 2147 MatDiagonalScaleLocal_MPIBAIJ);CHKERRQ(ierr); 2148 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatSetHashTableFactor_C", 2149 "MatSetHashTableFactor_MPIBAIJ", 2150 MatSetHashTableFactor_MPIBAIJ);CHKERRQ(ierr); 2151 PetscFunctionReturn(0); 2152 } 2153 EXTERN_C_END 2154 2155 #undef __FUNCT__ 2156 #define __FUNCT__ "MatMPIBAIJSetPreallocation" 2157 /*@C 2158 MatMPIBAIJSetPreallocation - Creates a sparse parallel matrix in block AIJ format 2159 (block compressed row). For good matrix assembly performance 2160 the user should preallocate the matrix storage by setting the parameters 2161 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 2162 performance can be increased by more than a factor of 50. 2163 2164 Collective on Mat 2165 2166 Input Parameters: 2167 + A - the matrix 2168 . bs - size of blockk 2169 . d_nz - number of block nonzeros per block row in diagonal portion of local 2170 submatrix (same for all local rows) 2171 . d_nnz - array containing the number of block nonzeros in the various block rows 2172 of the in diagonal portion of the local (possibly different for each block 2173 row) or PETSC_NULL. You must leave room for the diagonal entry even if it is zero. 2174 . o_nz - number of block nonzeros per block row in the off-diagonal portion of local 2175 submatrix (same for all local rows). 2176 - o_nnz - array containing the number of nonzeros in the various block rows of the 2177 off-diagonal portion of the local submatrix (possibly different for 2178 each block row) or PETSC_NULL. 2179 2180 Output Parameter: 2181 2182 2183 Options Database Keys: 2184 . -mat_no_unroll - uses code that does not unroll the loops in the 2185 block calculations (much slower) 2186 . -mat_block_size - size of the blocks to use 2187 2188 Notes: 2189 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 2190 than it must be used on all processors that share the object for that argument. 2191 2192 Storage Information: 2193 For a square global matrix we define each processor's diagonal portion 2194 to be its local rows and the corresponding columns (a square submatrix); 2195 each processor's off-diagonal portion encompasses the remainder of the 2196 local matrix (a rectangular submatrix). 2197 2198 The user can specify preallocated storage for the diagonal part of 2199 the local submatrix with either d_nz or d_nnz (not both). Set 2200 d_nz=PETSC_DEFAULT and d_nnz=PETSC_NULL for PETSc to control dynamic 2201 memory allocation. Likewise, specify preallocated storage for the 2202 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 2203 2204 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 2205 the figure below we depict these three local rows and all columns (0-11). 2206 2207 .vb 2208 0 1 2 3 4 5 6 7 8 9 10 11 2209 ------------------- 2210 row 3 | o o o d d d o o o o o o 2211 row 4 | o o o d d d o o o o o o 2212 row 5 | o o o d d d o o o o o o 2213 ------------------- 2214 .ve 2215 2216 Thus, any entries in the d locations are stored in the d (diagonal) 2217 submatrix, and any entries in the o locations are stored in the 2218 o (off-diagonal) submatrix. Note that the d and the o submatrices are 2219 stored simply in the MATSEQBAIJ format for compressed row storage. 2220 2221 Now d_nz should indicate the number of block nonzeros per row in the d matrix, 2222 and o_nz should indicate the number of block nonzeros per row in the o matrix. 2223 In general, for PDE problems in which most nonzeros are near the diagonal, 2224 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 2225 or you will get TERRIBLE performance; see the users' manual chapter on 2226 matrices. 2227 2228 Level: intermediate 2229 2230 .keywords: matrix, block, aij, compressed row, sparse, parallel 2231 2232 .seealso: MatCreate(), MatCreateSeqBAIJ(), MatSetValues(), MatCreateMPIBAIJ() 2233 @*/ 2234 int MatMPIBAIJSetPreallocation(Mat B,int bs,int d_nz,const int d_nnz[],int o_nz,const int o_nnz[]) 2235 { 2236 int ierr,(*f)(Mat,int,int,const int[],int,const int[]); 2237 2238 PetscFunctionBegin; 2239 ierr = PetscObjectQueryFunction((PetscObject)B,"MatMPIBAIJSetPreallocation_C",(void (**)(void))&f);CHKERRQ(ierr); 2240 if (f) { 2241 ierr = (*f)(B,bs,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 2242 } 2243 PetscFunctionReturn(0); 2244 } 2245 2246 #undef __FUNCT__ 2247 #define __FUNCT__ "MatCreateMPIBAIJ" 2248 /*@C 2249 MatCreateMPIBAIJ - Creates a sparse parallel matrix in block AIJ format 2250 (block compressed row). For good matrix assembly performance 2251 the user should preallocate the matrix storage by setting the parameters 2252 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 2253 performance can be increased by more than a factor of 50. 2254 2255 Collective on MPI_Comm 2256 2257 Input Parameters: 2258 + comm - MPI communicator 2259 . bs - size of blockk 2260 . m - number of local rows (or PETSC_DECIDE to have calculated if M is given) 2261 This value should be the same as the local size used in creating the 2262 y vector for the matrix-vector product y = Ax. 2263 . n - number of local columns (or PETSC_DECIDE to have calculated if N is given) 2264 This value should be the same as the local size used in creating the 2265 x vector for the matrix-vector product y = Ax. 2266 . M - number of global rows (or PETSC_DETERMINE to have calculated if m is given) 2267 . N - number of global columns (or PETSC_DETERMINE to have calculated if n is given) 2268 . d_nz - number of nonzero blocks per block row in diagonal portion of local 2269 submatrix (same for all local rows) 2270 . d_nnz - array containing the number of nonzero blocks in the various block rows 2271 of the in diagonal portion of the local (possibly different for each block 2272 row) or PETSC_NULL. You must leave room for the diagonal entry even if it is zero. 2273 . o_nz - number of nonzero blocks per block row in the off-diagonal portion of local 2274 submatrix (same for all local rows). 2275 - o_nnz - array containing the number of nonzero blocks in the various block rows of the 2276 off-diagonal portion of the local submatrix (possibly different for 2277 each block row) or PETSC_NULL. 2278 2279 Output Parameter: 2280 . A - the matrix 2281 2282 Options Database Keys: 2283 . -mat_no_unroll - uses code that does not unroll the loops in the 2284 block calculations (much slower) 2285 . -mat_block_size - size of the blocks to use 2286 2287 Notes: 2288 A nonzero block is any block that as 1 or more nonzeros in it 2289 2290 The user MUST specify either the local or global matrix dimensions 2291 (possibly both). 2292 2293 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 2294 than it must be used on all processors that share the object for that argument. 2295 2296 Storage Information: 2297 For a square global matrix we define each processor's diagonal portion 2298 to be its local rows and the corresponding columns (a square submatrix); 2299 each processor's off-diagonal portion encompasses the remainder of the 2300 local matrix (a rectangular submatrix). 2301 2302 The user can specify preallocated storage for the diagonal part of 2303 the local submatrix with either d_nz or d_nnz (not both). Set 2304 d_nz=PETSC_DEFAULT and d_nnz=PETSC_NULL for PETSc to control dynamic 2305 memory allocation. Likewise, specify preallocated storage for the 2306 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 2307 2308 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 2309 the figure below we depict these three local rows and all columns (0-11). 2310 2311 .vb 2312 0 1 2 3 4 5 6 7 8 9 10 11 2313 ------------------- 2314 row 3 | o o o d d d o o o o o o 2315 row 4 | o o o d d d o o o o o o 2316 row 5 | o o o d d d o o o o o o 2317 ------------------- 2318 .ve 2319 2320 Thus, any entries in the d locations are stored in the d (diagonal) 2321 submatrix, and any entries in the o locations are stored in the 2322 o (off-diagonal) submatrix. Note that the d and the o submatrices are 2323 stored simply in the MATSEQBAIJ format for compressed row storage. 2324 2325 Now d_nz should indicate the number of block nonzeros per row in the d matrix, 2326 and o_nz should indicate the number of block nonzeros per row in the o matrix. 2327 In general, for PDE problems in which most nonzeros are near the diagonal, 2328 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 2329 or you will get TERRIBLE performance; see the users' manual chapter on 2330 matrices. 2331 2332 Level: intermediate 2333 2334 .keywords: matrix, block, aij, compressed row, sparse, parallel 2335 2336 .seealso: MatCreate(), MatCreateSeqBAIJ(), MatSetValues(), MatCreateMPIBAIJ() 2337 @*/ 2338 int MatCreateMPIBAIJ(MPI_Comm comm,int bs,int m,int n,int M,int N,int d_nz,const int d_nnz[],int o_nz,const int o_nnz[],Mat *A) 2339 { 2340 int ierr,size; 2341 2342 PetscFunctionBegin; 2343 ierr = MatCreate(comm,m,n,M,N,A);CHKERRQ(ierr); 2344 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2345 if (size > 1) { 2346 ierr = MatSetType(*A,MATMPIBAIJ);CHKERRQ(ierr); 2347 ierr = MatMPIBAIJSetPreallocation(*A,bs,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 2348 } else { 2349 ierr = MatSetType(*A,MATSEQBAIJ);CHKERRQ(ierr); 2350 ierr = MatSeqBAIJSetPreallocation(*A,bs,d_nz,d_nnz);CHKERRQ(ierr); 2351 } 2352 PetscFunctionReturn(0); 2353 } 2354 2355 #undef __FUNCT__ 2356 #define __FUNCT__ "MatDuplicate_MPIBAIJ" 2357 static int MatDuplicate_MPIBAIJ(Mat matin,MatDuplicateOption cpvalues,Mat *newmat) 2358 { 2359 Mat mat; 2360 Mat_MPIBAIJ *a,*oldmat = (Mat_MPIBAIJ*)matin->data; 2361 int ierr,len=0; 2362 2363 PetscFunctionBegin; 2364 *newmat = 0; 2365 ierr = MatCreate(matin->comm,matin->m,matin->n,matin->M,matin->N,&mat);CHKERRQ(ierr); 2366 ierr = MatSetType(mat,MATMPIBAIJ);CHKERRQ(ierr); 2367 mat->preallocated = PETSC_TRUE; 2368 mat->assembled = PETSC_TRUE; 2369 a = (Mat_MPIBAIJ*)mat->data; 2370 a->bs = oldmat->bs; 2371 a->bs2 = oldmat->bs2; 2372 a->mbs = oldmat->mbs; 2373 a->nbs = oldmat->nbs; 2374 a->Mbs = oldmat->Mbs; 2375 a->Nbs = oldmat->Nbs; 2376 2377 a->rstart = oldmat->rstart; 2378 a->rend = oldmat->rend; 2379 a->cstart = oldmat->cstart; 2380 a->cend = oldmat->cend; 2381 a->size = oldmat->size; 2382 a->rank = oldmat->rank; 2383 a->donotstash = oldmat->donotstash; 2384 a->roworiented = oldmat->roworiented; 2385 a->rowindices = 0; 2386 a->rowvalues = 0; 2387 a->getrowactive = PETSC_FALSE; 2388 a->barray = 0; 2389 a->rstart_bs = oldmat->rstart_bs; 2390 a->rend_bs = oldmat->rend_bs; 2391 a->cstart_bs = oldmat->cstart_bs; 2392 a->cend_bs = oldmat->cend_bs; 2393 2394 /* hash table stuff */ 2395 a->ht = 0; 2396 a->hd = 0; 2397 a->ht_size = 0; 2398 a->ht_flag = oldmat->ht_flag; 2399 a->ht_fact = oldmat->ht_fact; 2400 a->ht_total_ct = 0; 2401 a->ht_insert_ct = 0; 2402 2403 ierr = PetscMemcpy(a->rowners,oldmat->rowners,3*(a->size+2)*sizeof(int));CHKERRQ(ierr); 2404 ierr = MatStashCreate_Private(matin->comm,1,&mat->stash);CHKERRQ(ierr); 2405 ierr = MatStashCreate_Private(matin->comm,oldmat->bs,&mat->bstash);CHKERRQ(ierr); 2406 if (oldmat->colmap) { 2407 #if defined (PETSC_USE_CTABLE) 2408 ierr = PetscTableCreateCopy(oldmat->colmap,&a->colmap);CHKERRQ(ierr); 2409 #else 2410 ierr = PetscMalloc((a->Nbs)*sizeof(int),&a->colmap);CHKERRQ(ierr); 2411 PetscLogObjectMemory(mat,(a->Nbs)*sizeof(int)); 2412 ierr = PetscMemcpy(a->colmap,oldmat->colmap,(a->Nbs)*sizeof(int));CHKERRQ(ierr); 2413 #endif 2414 } else a->colmap = 0; 2415 if (oldmat->garray && (len = ((Mat_SeqBAIJ*)(oldmat->B->data))->nbs)) { 2416 ierr = PetscMalloc(len*sizeof(int),&a->garray);CHKERRQ(ierr); 2417 PetscLogObjectMemory(mat,len*sizeof(int)); 2418 ierr = PetscMemcpy(a->garray,oldmat->garray,len*sizeof(int));CHKERRQ(ierr); 2419 } else a->garray = 0; 2420 2421 ierr = VecDuplicate(oldmat->lvec,&a->lvec);CHKERRQ(ierr); 2422 PetscLogObjectParent(mat,a->lvec); 2423 ierr = VecScatterCopy(oldmat->Mvctx,&a->Mvctx);CHKERRQ(ierr); 2424 2425 PetscLogObjectParent(mat,a->Mvctx); 2426 ierr = MatDuplicate(oldmat->A,cpvalues,&a->A);CHKERRQ(ierr); 2427 PetscLogObjectParent(mat,a->A); 2428 ierr = MatDuplicate(oldmat->B,cpvalues,&a->B);CHKERRQ(ierr); 2429 PetscLogObjectParent(mat,a->B); 2430 ierr = PetscFListDuplicate(matin->qlist,&mat->qlist);CHKERRQ(ierr); 2431 *newmat = mat; 2432 PetscFunctionReturn(0); 2433 } 2434 2435 #include "petscsys.h" 2436 2437 #undef __FUNCT__ 2438 #define __FUNCT__ "MatLoad_MPIBAIJ" 2439 int MatLoad_MPIBAIJ(PetscViewer viewer,MatType type,Mat *newmat) 2440 { 2441 Mat A; 2442 int i,nz,ierr,j,rstart,rend,fd; 2443 PetscScalar *vals,*buf; 2444 MPI_Comm comm = ((PetscObject)viewer)->comm; 2445 MPI_Status status; 2446 int header[4],rank,size,*rowlengths = 0,M,N,m,*rowners,*browners,maxnz,*cols; 2447 int *locrowlens,*sndcounts = 0,*procsnz = 0,jj,*mycols,*ibuf; 2448 int tag = ((PetscObject)viewer)->tag,bs=1,Mbs,mbs,extra_rows; 2449 int *dlens,*odlens,*mask,*masked1,*masked2,rowcount,odcount; 2450 int dcount,kmax,k,nzcount,tmp; 2451 2452 PetscFunctionBegin; 2453 ierr = PetscOptionsGetInt(PETSC_NULL,"-matload_block_size",&bs,PETSC_NULL);CHKERRQ(ierr); 2454 2455 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2456 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 2457 if (!rank) { 2458 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 2459 ierr = PetscBinaryRead(fd,(char *)header,4,PETSC_INT);CHKERRQ(ierr); 2460 if (header[0] != MAT_FILE_COOKIE) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"not matrix object"); 2461 if (header[3] < 0) { 2462 SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"Matrix stored in special format, cannot load as MPIBAIJ"); 2463 } 2464 } 2465 2466 ierr = MPI_Bcast(header+1,3,MPI_INT,0,comm);CHKERRQ(ierr); 2467 M = header[1]; N = header[2]; 2468 2469 if (M != N) SETERRQ(PETSC_ERR_SUP,"Can only do square matrices"); 2470 2471 /* 2472 This code adds extra rows to make sure the number of rows is 2473 divisible by the blocksize 2474 */ 2475 Mbs = M/bs; 2476 extra_rows = bs - M + bs*(Mbs); 2477 if (extra_rows == bs) extra_rows = 0; 2478 else Mbs++; 2479 if (extra_rows &&!rank) { 2480 PetscLogInfo(0,"MatLoad_MPIBAIJ:Padding loaded matrix to match blocksize\n"); 2481 } 2482 2483 /* determine ownership of all rows */ 2484 mbs = Mbs/size + ((Mbs % size) > rank); 2485 m = mbs*bs; 2486 ierr = PetscMalloc(2*(size+2)*sizeof(int),&rowners);CHKERRQ(ierr); 2487 browners = rowners + size + 1; 2488 ierr = MPI_Allgather(&mbs,1,MPI_INT,rowners+1,1,MPI_INT,comm);CHKERRQ(ierr); 2489 rowners[0] = 0; 2490 for (i=2; i<=size; i++) rowners[i] += rowners[i-1]; 2491 for (i=0; i<=size; i++) browners[i] = rowners[i]*bs; 2492 rstart = rowners[rank]; 2493 rend = rowners[rank+1]; 2494 2495 /* distribute row lengths to all processors */ 2496 ierr = PetscMalloc((rend-rstart)*bs*sizeof(int),&locrowlens);CHKERRQ(ierr); 2497 if (!rank) { 2498 ierr = PetscMalloc((M+extra_rows)*sizeof(int),&rowlengths);CHKERRQ(ierr); 2499 ierr = PetscBinaryRead(fd,rowlengths,M,PETSC_INT);CHKERRQ(ierr); 2500 for (i=0; i<extra_rows; i++) rowlengths[M+i] = 1; 2501 ierr = PetscMalloc(size*sizeof(int),&sndcounts);CHKERRQ(ierr); 2502 for (i=0; i<size; i++) sndcounts[i] = browners[i+1] - browners[i]; 2503 ierr = MPI_Scatterv(rowlengths,sndcounts,browners,MPI_INT,locrowlens,(rend-rstart)*bs,MPI_INT,0,comm);CHKERRQ(ierr); 2504 ierr = PetscFree(sndcounts);CHKERRQ(ierr); 2505 } else { 2506 ierr = MPI_Scatterv(0,0,0,MPI_INT,locrowlens,(rend-rstart)*bs,MPI_INT,0,comm);CHKERRQ(ierr); 2507 } 2508 2509 if (!rank) { 2510 /* calculate the number of nonzeros on each processor */ 2511 ierr = PetscMalloc(size*sizeof(int),&procsnz);CHKERRQ(ierr); 2512 ierr = PetscMemzero(procsnz,size*sizeof(int));CHKERRQ(ierr); 2513 for (i=0; i<size; i++) { 2514 for (j=rowners[i]*bs; j< rowners[i+1]*bs; j++) { 2515 procsnz[i] += rowlengths[j]; 2516 } 2517 } 2518 ierr = PetscFree(rowlengths);CHKERRQ(ierr); 2519 2520 /* determine max buffer needed and allocate it */ 2521 maxnz = 0; 2522 for (i=0; i<size; i++) { 2523 maxnz = PetscMax(maxnz,procsnz[i]); 2524 } 2525 ierr = PetscMalloc(maxnz*sizeof(int),&cols);CHKERRQ(ierr); 2526 2527 /* read in my part of the matrix column indices */ 2528 nz = procsnz[0]; 2529 ierr = PetscMalloc(nz*sizeof(int),&ibuf);CHKERRQ(ierr); 2530 mycols = ibuf; 2531 if (size == 1) nz -= extra_rows; 2532 ierr = PetscBinaryRead(fd,mycols,nz,PETSC_INT);CHKERRQ(ierr); 2533 if (size == 1) for (i=0; i< extra_rows; i++) { mycols[nz+i] = M+i; } 2534 2535 /* read in every ones (except the last) and ship off */ 2536 for (i=1; i<size-1; i++) { 2537 nz = procsnz[i]; 2538 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 2539 ierr = MPI_Send(cols,nz,MPI_INT,i,tag,comm);CHKERRQ(ierr); 2540 } 2541 /* read in the stuff for the last proc */ 2542 if (size != 1) { 2543 nz = procsnz[size-1] - extra_rows; /* the extra rows are not on the disk */ 2544 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 2545 for (i=0; i<extra_rows; i++) cols[nz+i] = M+i; 2546 ierr = MPI_Send(cols,nz+extra_rows,MPI_INT,size-1,tag,comm);CHKERRQ(ierr); 2547 } 2548 ierr = PetscFree(cols);CHKERRQ(ierr); 2549 } else { 2550 /* determine buffer space needed for message */ 2551 nz = 0; 2552 for (i=0; i<m; i++) { 2553 nz += locrowlens[i]; 2554 } 2555 ierr = PetscMalloc(nz*sizeof(int),&ibuf);CHKERRQ(ierr); 2556 mycols = ibuf; 2557 /* receive message of column indices*/ 2558 ierr = MPI_Recv(mycols,nz,MPI_INT,0,tag,comm,&status);CHKERRQ(ierr); 2559 ierr = MPI_Get_count(&status,MPI_INT,&maxnz);CHKERRQ(ierr); 2560 if (maxnz != nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file"); 2561 } 2562 2563 /* loop over local rows, determining number of off diagonal entries */ 2564 ierr = PetscMalloc(2*(rend-rstart+1)*sizeof(int),&dlens);CHKERRQ(ierr); 2565 odlens = dlens + (rend-rstart); 2566 ierr = PetscMalloc(3*Mbs*sizeof(int),&mask);CHKERRQ(ierr); 2567 ierr = PetscMemzero(mask,3*Mbs*sizeof(int));CHKERRQ(ierr); 2568 masked1 = mask + Mbs; 2569 masked2 = masked1 + Mbs; 2570 rowcount = 0; nzcount = 0; 2571 for (i=0; i<mbs; i++) { 2572 dcount = 0; 2573 odcount = 0; 2574 for (j=0; j<bs; j++) { 2575 kmax = locrowlens[rowcount]; 2576 for (k=0; k<kmax; k++) { 2577 tmp = mycols[nzcount++]/bs; 2578 if (!mask[tmp]) { 2579 mask[tmp] = 1; 2580 if (tmp < rstart || tmp >= rend) masked2[odcount++] = tmp; 2581 else masked1[dcount++] = tmp; 2582 } 2583 } 2584 rowcount++; 2585 } 2586 2587 dlens[i] = dcount; 2588 odlens[i] = odcount; 2589 2590 /* zero out the mask elements we set */ 2591 for (j=0; j<dcount; j++) mask[masked1[j]] = 0; 2592 for (j=0; j<odcount; j++) mask[masked2[j]] = 0; 2593 } 2594 2595 /* create our matrix */ 2596 ierr = MatCreate(comm,m,m,M+extra_rows,N+extra_rows,&A);CHKERRQ(ierr); 2597 ierr = MatSetType(A,type);CHKERRQ(ierr) 2598 ierr = MatMPIBAIJSetPreallocation(A,bs,0,dlens,0,odlens);CHKERRQ(ierr); 2599 2600 /* Why doesn't this called using ierr = MatSetOption(A,MAT_COLUMNS_SORTED);CHKERRQ(ierr); */ 2601 MatSetOption(A,MAT_COLUMNS_SORTED); 2602 2603 if (!rank) { 2604 ierr = PetscMalloc(maxnz*sizeof(PetscScalar),&buf);CHKERRQ(ierr); 2605 /* read in my part of the matrix numerical values */ 2606 nz = procsnz[0]; 2607 vals = buf; 2608 mycols = ibuf; 2609 if (size == 1) nz -= extra_rows; 2610 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2611 if (size == 1) for (i=0; i< extra_rows; i++) { vals[nz+i] = 1.0; } 2612 2613 /* insert into matrix */ 2614 jj = rstart*bs; 2615 for (i=0; i<m; i++) { 2616 ierr = MatSetValues(A,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 2617 mycols += locrowlens[i]; 2618 vals += locrowlens[i]; 2619 jj++; 2620 } 2621 /* read in other processors (except the last one) and ship out */ 2622 for (i=1; i<size-1; i++) { 2623 nz = procsnz[i]; 2624 vals = buf; 2625 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2626 ierr = MPI_Send(vals,nz,MPIU_SCALAR,i,A->tag,comm);CHKERRQ(ierr); 2627 } 2628 /* the last proc */ 2629 if (size != 1){ 2630 nz = procsnz[i] - extra_rows; 2631 vals = buf; 2632 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2633 for (i=0; i<extra_rows; i++) vals[nz+i] = 1.0; 2634 ierr = MPI_Send(vals,nz+extra_rows,MPIU_SCALAR,size-1,A->tag,comm);CHKERRQ(ierr); 2635 } 2636 ierr = PetscFree(procsnz);CHKERRQ(ierr); 2637 } else { 2638 /* receive numeric values */ 2639 ierr = PetscMalloc(nz*sizeof(PetscScalar),&buf);CHKERRQ(ierr); 2640 2641 /* receive message of values*/ 2642 vals = buf; 2643 mycols = ibuf; 2644 ierr = MPI_Recv(vals,nz,MPIU_SCALAR,0,A->tag,comm,&status);CHKERRQ(ierr); 2645 ierr = MPI_Get_count(&status,MPIU_SCALAR,&maxnz);CHKERRQ(ierr); 2646 if (maxnz != nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file"); 2647 2648 /* insert into matrix */ 2649 jj = rstart*bs; 2650 for (i=0; i<m; i++) { 2651 ierr = MatSetValues(A,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 2652 mycols += locrowlens[i]; 2653 vals += locrowlens[i]; 2654 jj++; 2655 } 2656 } 2657 ierr = PetscFree(locrowlens);CHKERRQ(ierr); 2658 ierr = PetscFree(buf);CHKERRQ(ierr); 2659 ierr = PetscFree(ibuf);CHKERRQ(ierr); 2660 ierr = PetscFree(rowners);CHKERRQ(ierr); 2661 ierr = PetscFree(dlens);CHKERRQ(ierr); 2662 ierr = PetscFree(mask);CHKERRQ(ierr); 2663 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2664 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2665 2666 *newmat = A; 2667 PetscFunctionReturn(0); 2668 } 2669 2670 #undef __FUNCT__ 2671 #define __FUNCT__ "MatMPIBAIJSetHashTableFactor" 2672 /*@ 2673 MatMPIBAIJSetHashTableFactor - Sets the factor required to compute the size of the HashTable. 2674 2675 Input Parameters: 2676 . mat - the matrix 2677 . fact - factor 2678 2679 Collective on Mat 2680 2681 Level: advanced 2682 2683 Notes: 2684 This can also be set by the command line option: -mat_use_hash_table fact 2685 2686 .keywords: matrix, hashtable, factor, HT 2687 2688 .seealso: MatSetOption() 2689 @*/ 2690 int MatMPIBAIJSetHashTableFactor(Mat mat,PetscReal fact) 2691 { 2692 int ierr,(*f)(Mat,PetscReal); 2693 2694 PetscFunctionBegin; 2695 ierr = PetscObjectQueryFunction((PetscObject)mat,"MatSetHashTableFactor_C",(void (**)(void))&f);CHKERRQ(ierr); 2696 if (f) { 2697 ierr = (*f)(mat,fact);CHKERRQ(ierr); 2698 } 2699 PetscFunctionReturn(0); 2700 } 2701 2702 #undef __FUNCT__ 2703 #define __FUNCT__ "MatSetHashTableFactor_MPIBAIJ" 2704 int MatSetHashTableFactor_MPIBAIJ(Mat mat,PetscReal fact) 2705 { 2706 Mat_MPIBAIJ *baij; 2707 2708 PetscFunctionBegin; 2709 PetscValidHeaderSpecific(mat,MAT_COOKIE); 2710 baij = (Mat_MPIBAIJ*)mat->data; 2711 baij->ht_fact = fact; 2712 PetscFunctionReturn(0); 2713 } 2714 2715 #undef __FUNCT__ 2716 #define __FUNCT__ "MatMPIBAIJGetSeqBAIJ" 2717 int MatMPIBAIJGetSeqBAIJ(Mat A,Mat *Ad,Mat *Ao,int *colmap[]) 2718 { 2719 Mat_MPIBAIJ *a = (Mat_MPIBAIJ *)A->data; 2720 PetscFunctionBegin; 2721 *Ad = a->A; 2722 *Ao = a->B; 2723 *colmap = a->garray; 2724 PetscFunctionReturn(0); 2725 } 2726