1 #define PETSCMAT_DLL 2 3 #include "src/mat/impls/baij/mpi/mpibaij.h" /*I "petscmat.h" I*/ 4 #include "mpisbaij.h" 5 #include "src/mat/impls/sbaij/seq/sbaij.h" 6 7 EXTERN PetscErrorCode MatSetUpMultiply_MPISBAIJ(Mat); 8 EXTERN PetscErrorCode MatSetUpMultiply_MPISBAIJ_2comm(Mat); 9 EXTERN PetscErrorCode DisAssemble_MPISBAIJ(Mat); 10 EXTERN PetscErrorCode MatIncreaseOverlap_MPISBAIJ(Mat,PetscInt,IS[],PetscInt); 11 EXTERN PetscErrorCode MatGetValues_SeqSBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],PetscScalar []); 12 EXTERN PetscErrorCode MatGetValues_SeqBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],PetscScalar []); 13 EXTERN PetscErrorCode MatSetValues_SeqSBAIJ(Mat,PetscInt,const PetscInt [],PetscInt,const PetscInt [],const PetscScalar [],InsertMode); 14 EXTERN PetscErrorCode MatSetValuesBlocked_SeqSBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode); 15 EXTERN PetscErrorCode MatSetValuesBlocked_SeqBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode); 16 EXTERN PetscErrorCode MatGetRow_SeqSBAIJ(Mat,PetscInt,PetscInt*,PetscInt**,PetscScalar**); 17 EXTERN PetscErrorCode MatRestoreRow_SeqSBAIJ(Mat,PetscInt,PetscInt*,PetscInt**,PetscScalar**); 18 EXTERN PetscErrorCode MatZeroRows_SeqSBAIJ(Mat,IS,PetscScalar*); 19 EXTERN PetscErrorCode MatZeroRows_SeqBAIJ(Mat,IS,PetscScalar *); 20 EXTERN PetscErrorCode MatGetRowMax_MPISBAIJ(Mat,Vec); 21 EXTERN PetscErrorCode MatRelax_MPISBAIJ(Mat,Vec,PetscReal,MatSORType,PetscReal,PetscInt,PetscInt,Vec); 22 23 /* UGLY, ugly, ugly 24 When MatScalar == PetscScalar the function MatSetValuesBlocked_MPIBAIJ_MatScalar() does 25 not exist. Otherwise ..._MatScalar() takes matrix elements in single precision and 26 inserts them into the single precision data structure. The function MatSetValuesBlocked_MPIBAIJ() 27 converts the entries into single precision and then calls ..._MatScalar() to put them 28 into the single precision data structures. 29 */ 30 #if defined(PETSC_USE_MAT_SINGLE) 31 EXTERN PetscErrorCode MatSetValuesBlocked_SeqSBAIJ_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode); 32 EXTERN PetscErrorCode MatSetValues_MPISBAIJ_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode); 33 EXTERN PetscErrorCode MatSetValuesBlocked_MPISBAIJ_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode); 34 EXTERN PetscErrorCode MatSetValues_MPISBAIJ_HT_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode); 35 EXTERN PetscErrorCode MatSetValuesBlocked_MPISBAIJ_HT_MatScalar(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode); 36 #else 37 #define MatSetValuesBlocked_SeqSBAIJ_MatScalar MatSetValuesBlocked_SeqSBAIJ 38 #define MatSetValues_MPISBAIJ_MatScalar MatSetValues_MPISBAIJ 39 #define MatSetValuesBlocked_MPISBAIJ_MatScalar MatSetValuesBlocked_MPISBAIJ 40 #define MatSetValues_MPISBAIJ_HT_MatScalar MatSetValues_MPISBAIJ_HT 41 #define MatSetValuesBlocked_MPISBAIJ_HT_MatScalar MatSetValuesBlocked_MPISBAIJ_HT 42 #endif 43 44 EXTERN_C_BEGIN 45 #undef __FUNCT__ 46 #define __FUNCT__ "MatStoreValues_MPISBAIJ" 47 PetscErrorCode PETSCMAT_DLLEXPORT MatStoreValues_MPISBAIJ(Mat mat) 48 { 49 Mat_MPISBAIJ *aij = (Mat_MPISBAIJ *)mat->data; 50 PetscErrorCode ierr; 51 52 PetscFunctionBegin; 53 ierr = MatStoreValues(aij->A);CHKERRQ(ierr); 54 ierr = MatStoreValues(aij->B);CHKERRQ(ierr); 55 PetscFunctionReturn(0); 56 } 57 EXTERN_C_END 58 59 EXTERN_C_BEGIN 60 #undef __FUNCT__ 61 #define __FUNCT__ "MatRetrieveValues_MPISBAIJ" 62 PetscErrorCode PETSCMAT_DLLEXPORT MatRetrieveValues_MPISBAIJ(Mat mat) 63 { 64 Mat_MPISBAIJ *aij = (Mat_MPISBAIJ *)mat->data; 65 PetscErrorCode ierr; 66 67 PetscFunctionBegin; 68 ierr = MatRetrieveValues(aij->A);CHKERRQ(ierr); 69 ierr = MatRetrieveValues(aij->B);CHKERRQ(ierr); 70 PetscFunctionReturn(0); 71 } 72 EXTERN_C_END 73 74 75 #define CHUNKSIZE 10 76 77 #define MatSetValues_SeqSBAIJ_A_Private(row,col,value,addv) \ 78 { \ 79 \ 80 brow = row/bs; \ 81 rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; \ 82 rmax = aimax[brow]; nrow = ailen[brow]; \ 83 bcol = col/bs; \ 84 ridx = row % bs; cidx = col % bs; \ 85 low = 0; high = nrow; \ 86 while (high-low > 3) { \ 87 t = (low+high)/2; \ 88 if (rp[t] > bcol) high = t; \ 89 else low = t; \ 90 } \ 91 for (_i=low; _i<high; _i++) { \ 92 if (rp[_i] > bcol) break; \ 93 if (rp[_i] == bcol) { \ 94 bap = ap + bs2*_i + bs*cidx + ridx; \ 95 if (addv == ADD_VALUES) *bap += value; \ 96 else *bap = value; \ 97 goto a_noinsert; \ 98 } \ 99 } \ 100 if (a->nonew == 1) goto a_noinsert; \ 101 if (a->nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", row, col); \ 102 MatSeqXAIJReallocateAIJ(A,a->mbs,bs2,nrow,brow,bcol,rmax,aa,ai,aj,rp,ap,aimax,a->nonew,MatScalar); \ 103 N = nrow++ - 1; \ 104 /* shift up all the later entries in this row */ \ 105 for (ii=N; ii>=_i; ii--) { \ 106 rp[ii+1] = rp[ii]; \ 107 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 108 } \ 109 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr); } \ 110 rp[_i] = bcol; \ 111 ap[bs2*_i + bs*cidx + ridx] = value; \ 112 a_noinsert:; \ 113 ailen[brow] = nrow; \ 114 } 115 #ifndef MatSetValues_SeqBAIJ_B_Private 116 #define MatSetValues_SeqSBAIJ_B_Private(row,col,value,addv) \ 117 { \ 118 brow = row/bs; \ 119 rp = bj + bi[brow]; ap = ba + bs2*bi[brow]; \ 120 rmax = bimax[brow]; nrow = bilen[brow]; \ 121 bcol = col/bs; \ 122 ridx = row % bs; cidx = col % bs; \ 123 low = 0; high = nrow; \ 124 while (high-low > 3) { \ 125 t = (low+high)/2; \ 126 if (rp[t] > bcol) high = t; \ 127 else low = t; \ 128 } \ 129 for (_i=low; _i<high; _i++) { \ 130 if (rp[_i] > bcol) break; \ 131 if (rp[_i] == bcol) { \ 132 bap = ap + bs2*_i + bs*cidx + ridx; \ 133 if (addv == ADD_VALUES) *bap += value; \ 134 else *bap = value; \ 135 goto b_noinsert; \ 136 } \ 137 } \ 138 if (b->nonew == 1) goto b_noinsert; \ 139 if (b->nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", row, col); \ 140 MatSeqXAIJReallocateAIJ(B,b->mbs,bs2,nrow,brow,bcol,rmax,ba,bi,bj,rp,ap,bimax,b->nonew,MatScalar); \ 141 N = nrow++ - 1; \ 142 /* shift up all the later entries in this row */ \ 143 for (ii=N; ii>=_i; ii--) { \ 144 rp[ii+1] = rp[ii]; \ 145 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 146 } \ 147 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr);} \ 148 rp[_i] = bcol; \ 149 ap[bs2*_i + bs*cidx + ridx] = value; \ 150 b_noinsert:; \ 151 bilen[brow] = nrow; \ 152 } 153 #endif 154 155 #if defined(PETSC_USE_MAT_SINGLE) 156 #undef __FUNCT__ 157 #define __FUNCT__ "MatSetValues_MPISBAIJ" 158 PetscErrorCode MatSetValues_MPISBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 159 { 160 Mat_MPISBAIJ *b = (Mat_MPISBAIJ*)mat->data; 161 PetscErrorCode ierr; 162 PetscInt i,N = m*n; 163 MatScalar *vsingle; 164 165 PetscFunctionBegin; 166 if (N > b->setvalueslen) { 167 ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr); 168 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 169 b->setvalueslen = N; 170 } 171 vsingle = b->setvaluescopy; 172 173 for (i=0; i<N; i++) { 174 vsingle[i] = v[i]; 175 } 176 ierr = MatSetValues_MPISBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 177 PetscFunctionReturn(0); 178 } 179 180 #undef __FUNCT__ 181 #define __FUNCT__ "MatSetValuesBlocked_MPISBAIJ" 182 PetscErrorCode MatSetValuesBlocked_MPISBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 183 { 184 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data; 185 PetscErrorCode ierr; 186 PetscInt i,N = m*n*b->bs2; 187 MatScalar *vsingle; 188 189 PetscFunctionBegin; 190 if (N > b->setvalueslen) { 191 ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr); 192 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 193 b->setvalueslen = N; 194 } 195 vsingle = b->setvaluescopy; 196 for (i=0; i<N; i++) { 197 vsingle[i] = v[i]; 198 } 199 ierr = MatSetValuesBlocked_MPISBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 200 PetscFunctionReturn(0); 201 } 202 #endif 203 204 /* Only add/insert a(i,j) with i<=j (blocks). 205 Any a(i,j) with i>j input by user is ingored. 206 */ 207 #undef __FUNCT__ 208 #define __FUNCT__ "MatSetValues_MPISBAIJ_MatScalar" 209 PetscErrorCode MatSetValues_MPISBAIJ_MatScalar(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const MatScalar v[],InsertMode addv) 210 { 211 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 212 MatScalar value; 213 PetscTruth roworiented = baij->roworiented; 214 PetscErrorCode ierr; 215 PetscInt i,j,row,col; 216 PetscInt rstart_orig=mat->rmap.rstart; 217 PetscInt rend_orig=mat->rmap.rend,cstart_orig=mat->cmap.rstart; 218 PetscInt cend_orig=mat->cmap.rend,bs=mat->rmap.bs; 219 220 /* Some Variables required in the macro */ 221 Mat A = baij->A; 222 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)(A)->data; 223 PetscInt *aimax=a->imax,*ai=a->i,*ailen=a->ilen,*aj=a->j; 224 MatScalar *aa=a->a; 225 226 Mat B = baij->B; 227 Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(B)->data; 228 PetscInt *bimax=b->imax,*bi=b->i,*bilen=b->ilen,*bj=b->j; 229 MatScalar *ba=b->a; 230 231 PetscInt *rp,ii,nrow,_i,rmax,N,brow,bcol; 232 PetscInt low,high,t,ridx,cidx,bs2=a->bs2; 233 MatScalar *ap,*bap; 234 235 /* for stash */ 236 PetscInt n_loc, *in_loc = PETSC_NULL; 237 MatScalar *v_loc = PETSC_NULL; 238 239 PetscFunctionBegin; 240 241 if (!baij->donotstash){ 242 if (n > baij->n_loc) { 243 ierr = PetscFree(baij->in_loc);CHKERRQ(ierr); 244 ierr = PetscFree(baij->v_loc);CHKERRQ(ierr); 245 ierr = PetscMalloc(n*sizeof(PetscInt),&baij->in_loc);CHKERRQ(ierr); 246 ierr = PetscMalloc(n*sizeof(MatScalar),&baij->v_loc);CHKERRQ(ierr); 247 baij->n_loc = n; 248 } 249 in_loc = baij->in_loc; 250 v_loc = baij->v_loc; 251 } 252 253 for (i=0; i<m; i++) { 254 if (im[i] < 0) continue; 255 #if defined(PETSC_USE_DEBUG) 256 if (im[i] >= mat->rmap.N) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",im[i],mat->rmap.N-1); 257 #endif 258 if (im[i] >= rstart_orig && im[i] < rend_orig) { /* this processor entry */ 259 row = im[i] - rstart_orig; /* local row index */ 260 for (j=0; j<n; j++) { 261 if (im[i]/bs > in[j]/bs){ 262 if (a->ignore_ltriangular){ 263 continue; /* ignore lower triangular blocks */ 264 } else { 265 SETERRQ(PETSC_ERR_USER,"Lower triangular value cannot be set for sbaij format. Ignoring these values, run with -mat_ignore_lower_triangular or call MatSetOption(mat,MAT_IGNORE_LOWER_TRIANGULAR)"); 266 } 267 } 268 if (in[j] >= cstart_orig && in[j] < cend_orig){ /* diag entry (A) */ 269 col = in[j] - cstart_orig; /* local col index */ 270 brow = row/bs; bcol = col/bs; 271 if (brow > bcol) continue; /* ignore lower triangular blocks of A */ 272 if (roworiented) value = v[i*n+j]; else value = v[i+j*m]; 273 MatSetValues_SeqSBAIJ_A_Private(row,col,value,addv); 274 /* ierr = MatSetValues_SeqBAIJ(baij->A,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 275 } else if (in[j] < 0) continue; 276 #if defined(PETSC_USE_DEBUG) 277 else if (in[j] >= mat->cmap.N) {SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[j],mat->cmap.N-1);} 278 #endif 279 else { /* off-diag entry (B) */ 280 if (mat->was_assembled) { 281 if (!baij->colmap) { 282 ierr = CreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 283 } 284 #if defined (PETSC_USE_CTABLE) 285 ierr = PetscTableFind(baij->colmap,in[j]/bs + 1,&col);CHKERRQ(ierr); 286 col = col - 1; 287 #else 288 col = baij->colmap[in[j]/bs] - 1; 289 #endif 290 if (col < 0 && !((Mat_SeqSBAIJ*)(baij->A->data))->nonew) { 291 ierr = DisAssemble_MPISBAIJ(mat);CHKERRQ(ierr); 292 col = in[j]; 293 /* Reinitialize the variables required by MatSetValues_SeqBAIJ_B_Private() */ 294 B = baij->B; 295 b = (Mat_SeqBAIJ*)(B)->data; 296 bimax=b->imax;bi=b->i;bilen=b->ilen;bj=b->j; 297 ba=b->a; 298 } else col += in[j]%bs; 299 } else col = in[j]; 300 if (roworiented) value = v[i*n+j]; else value = v[i+j*m]; 301 MatSetValues_SeqSBAIJ_B_Private(row,col,value,addv); 302 /* ierr = MatSetValues_SeqBAIJ(baij->B,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 303 } 304 } 305 } else { /* off processor entry */ 306 if (!baij->donotstash) { 307 n_loc = 0; 308 for (j=0; j<n; j++){ 309 if (im[i]/bs > in[j]/bs) continue; /* ignore lower triangular blocks */ 310 in_loc[n_loc] = in[j]; 311 if (roworiented) { 312 v_loc[n_loc] = v[i*n+j]; 313 } else { 314 v_loc[n_loc] = v[j*m+i]; 315 } 316 n_loc++; 317 } 318 ierr = MatStashValuesRow_Private(&mat->stash,im[i],n_loc,in_loc,v_loc);CHKERRQ(ierr); 319 } 320 } 321 } 322 PetscFunctionReturn(0); 323 } 324 325 #undef __FUNCT__ 326 #define __FUNCT__ "MatSetValuesBlocked_MPISBAIJ_MatScalar" 327 PetscErrorCode MatSetValuesBlocked_MPISBAIJ_MatScalar(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const MatScalar v[],InsertMode addv) 328 { 329 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 330 const MatScalar *value; 331 MatScalar *barray=baij->barray; 332 PetscTruth roworiented = baij->roworiented; 333 PetscErrorCode ierr; 334 PetscInt i,j,ii,jj,row,col,rstart=baij->rstartbs; 335 PetscInt rend=baij->rendbs,cstart=baij->rstartbs,stepval; 336 PetscInt cend=baij->rendbs,bs=mat->rmap.bs,bs2=baij->bs2; 337 338 PetscFunctionBegin; 339 if(!barray) { 340 ierr = PetscMalloc(bs2*sizeof(MatScalar),&barray);CHKERRQ(ierr); 341 baij->barray = barray; 342 } 343 344 if (roworiented) { 345 stepval = (n-1)*bs; 346 } else { 347 stepval = (m-1)*bs; 348 } 349 for (i=0; i<m; i++) { 350 if (im[i] < 0) continue; 351 #if defined(PETSC_USE_DEBUG) 352 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large, row %D max %D",im[i],baij->Mbs-1); 353 #endif 354 if (im[i] >= rstart && im[i] < rend) { 355 row = im[i] - rstart; 356 for (j=0; j<n; j++) { 357 /* If NumCol = 1 then a copy is not required */ 358 if ((roworiented) && (n == 1)) { 359 barray = (MatScalar*) v + i*bs2; 360 } else if((!roworiented) && (m == 1)) { 361 barray = (MatScalar*) v + j*bs2; 362 } else { /* Here a copy is required */ 363 if (roworiented) { 364 value = v + i*(stepval+bs)*bs + j*bs; 365 } else { 366 value = v + j*(stepval+bs)*bs + i*bs; 367 } 368 for (ii=0; ii<bs; ii++,value+=stepval) { 369 for (jj=0; jj<bs; jj++) { 370 *barray++ = *value++; 371 } 372 } 373 barray -=bs2; 374 } 375 376 if (in[j] >= cstart && in[j] < cend){ 377 col = in[j] - cstart; 378 ierr = MatSetValuesBlocked_SeqSBAIJ(baij->A,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 379 } 380 else if (in[j] < 0) continue; 381 #if defined(PETSC_USE_DEBUG) 382 else if (in[j] >= baij->Nbs) {SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large, col %D max %D",in[j],baij->Nbs-1);} 383 #endif 384 else { 385 if (mat->was_assembled) { 386 if (!baij->colmap) { 387 ierr = CreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 388 } 389 390 #if defined(PETSC_USE_DEBUG) 391 #if defined (PETSC_USE_CTABLE) 392 { PetscInt data; 393 ierr = PetscTableFind(baij->colmap,in[j]+1,&data);CHKERRQ(ierr); 394 if ((data - 1) % bs) SETERRQ(PETSC_ERR_PLIB,"Incorrect colmap"); 395 } 396 #else 397 if ((baij->colmap[in[j]] - 1) % bs) SETERRQ(PETSC_ERR_PLIB,"Incorrect colmap"); 398 #endif 399 #endif 400 #if defined (PETSC_USE_CTABLE) 401 ierr = PetscTableFind(baij->colmap,in[j]+1,&col);CHKERRQ(ierr); 402 col = (col - 1)/bs; 403 #else 404 col = (baij->colmap[in[j]] - 1)/bs; 405 #endif 406 if (col < 0 && !((Mat_SeqBAIJ*)(baij->A->data))->nonew) { 407 ierr = DisAssemble_MPISBAIJ(mat);CHKERRQ(ierr); 408 col = in[j]; 409 } 410 } 411 else col = in[j]; 412 ierr = MatSetValuesBlocked_SeqBAIJ(baij->B,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 413 } 414 } 415 } else { 416 if (!baij->donotstash) { 417 if (roworiented) { 418 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 419 } else { 420 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 421 } 422 } 423 } 424 } 425 PetscFunctionReturn(0); 426 } 427 428 #undef __FUNCT__ 429 #define __FUNCT__ "MatGetValues_MPISBAIJ" 430 PetscErrorCode MatGetValues_MPISBAIJ(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],PetscScalar v[]) 431 { 432 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 433 PetscErrorCode ierr; 434 PetscInt bs=mat->rmap.bs,i,j,bsrstart = mat->rmap.rstart,bsrend = mat->rmap.rend; 435 PetscInt bscstart = mat->cmap.rstart,bscend = mat->cmap.rend,row,col,data; 436 437 PetscFunctionBegin; 438 for (i=0; i<m; i++) { 439 if (idxm[i] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %D",idxm[i]); 440 if (idxm[i] >= mat->rmap.N) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",idxm[i],mat->rmap.N-1); 441 if (idxm[i] >= bsrstart && idxm[i] < bsrend) { 442 row = idxm[i] - bsrstart; 443 for (j=0; j<n; j++) { 444 if (idxn[j] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative column %D",idxn[j]); 445 if (idxn[j] >= mat->cmap.N) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",idxn[j],mat->cmap.N-1); 446 if (idxn[j] >= bscstart && idxn[j] < bscend){ 447 col = idxn[j] - bscstart; 448 ierr = MatGetValues_SeqSBAIJ(baij->A,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 449 } else { 450 if (!baij->colmap) { 451 ierr = CreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 452 } 453 #if defined (PETSC_USE_CTABLE) 454 ierr = PetscTableFind(baij->colmap,idxn[j]/bs+1,&data);CHKERRQ(ierr); 455 data --; 456 #else 457 data = baij->colmap[idxn[j]/bs]-1; 458 #endif 459 if((data < 0) || (baij->garray[data/bs] != idxn[j]/bs)) *(v+i*n+j) = 0.0; 460 else { 461 col = data + idxn[j]%bs; 462 ierr = MatGetValues_SeqBAIJ(baij->B,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 463 } 464 } 465 } 466 } else { 467 SETERRQ(PETSC_ERR_SUP,"Only local values currently supported"); 468 } 469 } 470 PetscFunctionReturn(0); 471 } 472 473 #undef __FUNCT__ 474 #define __FUNCT__ "MatNorm_MPISBAIJ" 475 PetscErrorCode MatNorm_MPISBAIJ(Mat mat,NormType type,PetscReal *norm) 476 { 477 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 478 PetscErrorCode ierr; 479 PetscReal sum[2],*lnorm2; 480 481 PetscFunctionBegin; 482 if (baij->size == 1) { 483 ierr = MatNorm(baij->A,type,norm);CHKERRQ(ierr); 484 } else { 485 if (type == NORM_FROBENIUS) { 486 ierr = PetscMalloc(2*sizeof(PetscReal),&lnorm2);CHKERRQ(ierr); 487 ierr = MatNorm(baij->A,type,lnorm2);CHKERRQ(ierr); 488 *lnorm2 = (*lnorm2)*(*lnorm2); lnorm2++; /* squar power of norm(A) */ 489 ierr = MatNorm(baij->B,type,lnorm2);CHKERRQ(ierr); 490 *lnorm2 = (*lnorm2)*(*lnorm2); lnorm2--; /* squar power of norm(B) */ 491 ierr = MPI_Allreduce(lnorm2,&sum,2,MPIU_REAL,MPI_SUM,mat->comm);CHKERRQ(ierr); 492 *norm = sqrt(sum[0] + 2*sum[1]); 493 ierr = PetscFree(lnorm2);CHKERRQ(ierr); 494 } else if (type == NORM_INFINITY || type == NORM_1) { /* max row/column sum */ 495 Mat_SeqSBAIJ *amat=(Mat_SeqSBAIJ*)baij->A->data; 496 Mat_SeqBAIJ *bmat=(Mat_SeqBAIJ*)baij->B->data; 497 PetscReal *rsum,*rsum2,vabs; 498 PetscInt *jj,*garray=baij->garray,rstart=baij->rstartbs,nz; 499 PetscInt brow,bcol,col,bs=baij->A->rmap.bs,row,grow,gcol,mbs=amat->mbs; 500 MatScalar *v; 501 502 ierr = PetscMalloc((2*mat->cmap.N+1)*sizeof(PetscReal),&rsum);CHKERRQ(ierr); 503 rsum2 = rsum + mat->cmap.N; 504 ierr = PetscMemzero(rsum,mat->cmap.N*sizeof(PetscReal));CHKERRQ(ierr); 505 /* Amat */ 506 v = amat->a; jj = amat->j; 507 for (brow=0; brow<mbs; brow++) { 508 grow = bs*(rstart + brow); 509 nz = amat->i[brow+1] - amat->i[brow]; 510 for (bcol=0; bcol<nz; bcol++){ 511 gcol = bs*(rstart + *jj); jj++; 512 for (col=0; col<bs; col++){ 513 for (row=0; row<bs; row++){ 514 vabs = PetscAbsScalar(*v); v++; 515 rsum[gcol+col] += vabs; 516 /* non-diagonal block */ 517 if (bcol > 0 && vabs > 0.0) rsum[grow+row] += vabs; 518 } 519 } 520 } 521 } 522 /* Bmat */ 523 v = bmat->a; jj = bmat->j; 524 for (brow=0; brow<mbs; brow++) { 525 grow = bs*(rstart + brow); 526 nz = bmat->i[brow+1] - bmat->i[brow]; 527 for (bcol=0; bcol<nz; bcol++){ 528 gcol = bs*garray[*jj]; jj++; 529 for (col=0; col<bs; col++){ 530 for (row=0; row<bs; row++){ 531 vabs = PetscAbsScalar(*v); v++; 532 rsum[gcol+col] += vabs; 533 rsum[grow+row] += vabs; 534 } 535 } 536 } 537 } 538 ierr = MPI_Allreduce(rsum,rsum2,mat->cmap.N,MPIU_REAL,MPI_SUM,mat->comm);CHKERRQ(ierr); 539 *norm = 0.0; 540 for (col=0; col<mat->cmap.N; col++) { 541 if (rsum2[col] > *norm) *norm = rsum2[col]; 542 } 543 ierr = PetscFree(rsum);CHKERRQ(ierr); 544 } else { 545 SETERRQ(PETSC_ERR_SUP,"No support for this norm yet"); 546 } 547 } 548 PetscFunctionReturn(0); 549 } 550 551 #undef __FUNCT__ 552 #define __FUNCT__ "MatAssemblyBegin_MPISBAIJ" 553 PetscErrorCode MatAssemblyBegin_MPISBAIJ(Mat mat,MatAssemblyType mode) 554 { 555 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 556 PetscErrorCode ierr; 557 PetscInt nstash,reallocs; 558 InsertMode addv; 559 560 PetscFunctionBegin; 561 if (baij->donotstash) { 562 PetscFunctionReturn(0); 563 } 564 565 /* make sure all processors are either in INSERTMODE or ADDMODE */ 566 ierr = MPI_Allreduce(&mat->insertmode,&addv,1,MPI_INT,MPI_BOR,mat->comm);CHKERRQ(ierr); 567 if (addv == (ADD_VALUES|INSERT_VALUES)) { 568 SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Some processors inserted others added"); 569 } 570 mat->insertmode = addv; /* in case this processor had no cache */ 571 572 ierr = MatStashScatterBegin_Private(&mat->stash,mat->rmap.range);CHKERRQ(ierr); 573 ierr = MatStashScatterBegin_Private(&mat->bstash,baij->rangebs);CHKERRQ(ierr); 574 ierr = MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);CHKERRQ(ierr); 575 ierr = PetscInfo2(0,"Stash has %D entries,uses %D mallocs.\n",nstash,reallocs);CHKERRQ(ierr); 576 ierr = MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);CHKERRQ(ierr); 577 ierr = PetscInfo2(0,"Block-Stash has %D entries, uses %D mallocs.\n",nstash,reallocs);CHKERRQ(ierr); 578 PetscFunctionReturn(0); 579 } 580 581 #undef __FUNCT__ 582 #define __FUNCT__ "MatAssemblyEnd_MPISBAIJ" 583 PetscErrorCode MatAssemblyEnd_MPISBAIJ(Mat mat,MatAssemblyType mode) 584 { 585 Mat_MPISBAIJ *baij=(Mat_MPISBAIJ*)mat->data; 586 Mat_SeqSBAIJ *a=(Mat_SeqSBAIJ*)baij->A->data; 587 PetscErrorCode ierr; 588 PetscInt i,j,rstart,ncols,flg,bs2=baij->bs2; 589 PetscInt *row,*col,other_disassembled; 590 PetscMPIInt n; 591 PetscTruth r1,r2,r3; 592 MatScalar *val; 593 InsertMode addv = mat->insertmode; 594 595 /* do not use 'b=(Mat_SeqBAIJ*)baij->B->data' as B can be reset in disassembly */ 596 PetscFunctionBegin; 597 598 if (!baij->donotstash) { 599 while (1) { 600 ierr = MatStashScatterGetMesg_Private(&mat->stash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 601 if (!flg) break; 602 603 for (i=0; i<n;) { 604 /* Now identify the consecutive vals belonging to the same row */ 605 for (j=i,rstart=row[j]; j<n; j++) { if (row[j] != rstart) break; } 606 if (j < n) ncols = j-i; 607 else ncols = n-i; 608 /* Now assemble all these values with a single function call */ 609 ierr = MatSetValues_MPISBAIJ_MatScalar(mat,1,row+i,ncols,col+i,val+i,addv);CHKERRQ(ierr); 610 i = j; 611 } 612 } 613 ierr = MatStashScatterEnd_Private(&mat->stash);CHKERRQ(ierr); 614 /* Now process the block-stash. Since the values are stashed column-oriented, 615 set the roworiented flag to column oriented, and after MatSetValues() 616 restore the original flags */ 617 r1 = baij->roworiented; 618 r2 = a->roworiented; 619 r3 = ((Mat_SeqBAIJ*)baij->B->data)->roworiented; 620 baij->roworiented = PETSC_FALSE; 621 a->roworiented = PETSC_FALSE; 622 ((Mat_SeqBAIJ*)baij->B->data)->roworiented = PETSC_FALSE; /* b->roworinted */ 623 while (1) { 624 ierr = MatStashScatterGetMesg_Private(&mat->bstash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 625 if (!flg) break; 626 627 for (i=0; i<n;) { 628 /* Now identify the consecutive vals belonging to the same row */ 629 for (j=i,rstart=row[j]; j<n; j++) { if (row[j] != rstart) break; } 630 if (j < n) ncols = j-i; 631 else ncols = n-i; 632 ierr = MatSetValuesBlocked_MPISBAIJ_MatScalar(mat,1,row+i,ncols,col+i,val+i*bs2,addv);CHKERRQ(ierr); 633 i = j; 634 } 635 } 636 ierr = MatStashScatterEnd_Private(&mat->bstash);CHKERRQ(ierr); 637 baij->roworiented = r1; 638 a->roworiented = r2; 639 ((Mat_SeqBAIJ*)baij->B->data)->roworiented = r3; /* b->roworinted */ 640 } 641 642 ierr = MatAssemblyBegin(baij->A,mode);CHKERRQ(ierr); 643 ierr = MatAssemblyEnd(baij->A,mode);CHKERRQ(ierr); 644 645 /* determine if any processor has disassembled, if so we must 646 also disassemble ourselfs, in order that we may reassemble. */ 647 /* 648 if nonzero structure of submatrix B cannot change then we know that 649 no processor disassembled thus we can skip this stuff 650 */ 651 if (!((Mat_SeqBAIJ*)baij->B->data)->nonew) { 652 ierr = MPI_Allreduce(&mat->was_assembled,&other_disassembled,1,MPI_INT,MPI_PROD,mat->comm);CHKERRQ(ierr); 653 if (mat->was_assembled && !other_disassembled) { 654 ierr = DisAssemble_MPISBAIJ(mat);CHKERRQ(ierr); 655 } 656 } 657 658 if (!mat->was_assembled && mode == MAT_FINAL_ASSEMBLY) { 659 ierr = MatSetUpMultiply_MPISBAIJ(mat);CHKERRQ(ierr); /* setup Mvctx and sMvctx */ 660 } 661 ((Mat_SeqBAIJ*)baij->B->data)->compressedrow.use = PETSC_TRUE; /* b->compressedrow.use */ 662 ierr = MatAssemblyBegin(baij->B,mode);CHKERRQ(ierr); 663 ierr = MatAssemblyEnd(baij->B,mode);CHKERRQ(ierr); 664 665 ierr = PetscFree(baij->rowvalues);CHKERRQ(ierr); 666 baij->rowvalues = 0; 667 668 PetscFunctionReturn(0); 669 } 670 671 extern PetscErrorCode MatSetValues_MPIBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const MatScalar[],InsertMode); 672 #undef __FUNCT__ 673 #define __FUNCT__ "MatView_MPISBAIJ_ASCIIorDraworSocket" 674 static PetscErrorCode MatView_MPISBAIJ_ASCIIorDraworSocket(Mat mat,PetscViewer viewer) 675 { 676 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 677 PetscErrorCode ierr; 678 PetscInt bs = mat->rmap.bs; 679 PetscMPIInt size = baij->size,rank = baij->rank; 680 PetscTruth iascii,isdraw; 681 PetscViewer sviewer; 682 PetscViewerFormat format; 683 684 PetscFunctionBegin; 685 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&iascii);CHKERRQ(ierr); 686 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);CHKERRQ(ierr); 687 if (iascii) { 688 ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr); 689 if (format == PETSC_VIEWER_ASCII_INFO_DETAIL) { 690 MatInfo info; 691 ierr = MPI_Comm_rank(mat->comm,&rank);CHKERRQ(ierr); 692 ierr = MatGetInfo(mat,MAT_LOCAL,&info);CHKERRQ(ierr); 693 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] Local rows %D nz %D nz alloced %D bs %D mem %D\n", 694 rank,mat->rmap.N,(PetscInt)info.nz_used*bs,(PetscInt)info.nz_allocated*bs, 695 mat->rmap.bs,(PetscInt)info.memory);CHKERRQ(ierr); 696 ierr = MatGetInfo(baij->A,MAT_LOCAL,&info);CHKERRQ(ierr); 697 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] on-diagonal part: nz %D \n",rank,(PetscInt)info.nz_used*bs);CHKERRQ(ierr); 698 ierr = MatGetInfo(baij->B,MAT_LOCAL,&info);CHKERRQ(ierr); 699 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] off-diagonal part: nz %D \n",rank,(PetscInt)info.nz_used*bs);CHKERRQ(ierr); 700 ierr = PetscViewerFlush(viewer);CHKERRQ(ierr); 701 ierr = VecScatterView(baij->Mvctx,viewer);CHKERRQ(ierr); 702 PetscFunctionReturn(0); 703 } else if (format == PETSC_VIEWER_ASCII_INFO) { 704 ierr = PetscViewerASCIIPrintf(viewer," block size is %D\n",bs);CHKERRQ(ierr); 705 PetscFunctionReturn(0); 706 } else if (format == PETSC_VIEWER_ASCII_FACTOR_INFO) { 707 PetscFunctionReturn(0); 708 } 709 } 710 711 if (isdraw) { 712 PetscDraw draw; 713 PetscTruth isnull; 714 ierr = PetscViewerDrawGetDraw(viewer,0,&draw);CHKERRQ(ierr); 715 ierr = PetscDrawIsNull(draw,&isnull);CHKERRQ(ierr); if (isnull) PetscFunctionReturn(0); 716 } 717 718 if (size == 1) { 719 ierr = PetscObjectSetName((PetscObject)baij->A,mat->name);CHKERRQ(ierr); 720 ierr = MatView(baij->A,viewer);CHKERRQ(ierr); 721 } else { 722 /* assemble the entire matrix onto first processor. */ 723 Mat A; 724 Mat_SeqSBAIJ *Aloc; 725 Mat_SeqBAIJ *Bloc; 726 PetscInt M = mat->rmap.N,N = mat->cmap.N,*ai,*aj,col,i,j,k,*rvals,mbs = baij->mbs; 727 MatScalar *a; 728 729 /* Should this be the same type as mat? */ 730 ierr = MatCreate(mat->comm,&A);CHKERRQ(ierr); 731 if (!rank) { 732 ierr = MatSetSizes(A,M,N,M,N);CHKERRQ(ierr); 733 } else { 734 ierr = MatSetSizes(A,0,0,M,N);CHKERRQ(ierr); 735 } 736 ierr = MatSetType(A,MATMPISBAIJ);CHKERRQ(ierr); 737 ierr = MatMPISBAIJSetPreallocation(A,mat->rmap.bs,0,PETSC_NULL,0,PETSC_NULL);CHKERRQ(ierr); 738 ierr = PetscLogObjectParent(mat,A);CHKERRQ(ierr); 739 740 /* copy over the A part */ 741 Aloc = (Mat_SeqSBAIJ*)baij->A->data; 742 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 743 ierr = PetscMalloc(bs*sizeof(PetscInt),&rvals);CHKERRQ(ierr); 744 745 for (i=0; i<mbs; i++) { 746 rvals[0] = bs*(baij->rstartbs + i); 747 for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; } 748 for (j=ai[i]; j<ai[i+1]; j++) { 749 col = (baij->cstartbs+aj[j])*bs; 750 for (k=0; k<bs; k++) { 751 ierr = MatSetValues_MPISBAIJ_MatScalar(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 752 col++; a += bs; 753 } 754 } 755 } 756 /* copy over the B part */ 757 Bloc = (Mat_SeqBAIJ*)baij->B->data; 758 ai = Bloc->i; aj = Bloc->j; a = Bloc->a; 759 for (i=0; i<mbs; i++) { 760 761 rvals[0] = bs*(baij->rstartbs + i); 762 for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; } 763 for (j=ai[i]; j<ai[i+1]; j++) { 764 col = baij->garray[aj[j]]*bs; 765 for (k=0; k<bs; k++) { 766 ierr = MatSetValues_MPIBAIJ(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 767 col++; a += bs; 768 } 769 } 770 } 771 ierr = PetscFree(rvals);CHKERRQ(ierr); 772 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 773 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 774 /* 775 Everyone has to call to draw the matrix since the graphics waits are 776 synchronized across all processors that share the PetscDraw object 777 */ 778 ierr = PetscViewerGetSingleton(viewer,&sviewer);CHKERRQ(ierr); 779 if (!rank) { 780 ierr = PetscObjectSetName((PetscObject)((Mat_MPISBAIJ*)(A->data))->A,mat->name);CHKERRQ(ierr); 781 ierr = MatView(((Mat_MPISBAIJ*)(A->data))->A,sviewer);CHKERRQ(ierr); 782 } 783 ierr = PetscViewerRestoreSingleton(viewer,&sviewer);CHKERRQ(ierr); 784 ierr = MatDestroy(A);CHKERRQ(ierr); 785 } 786 PetscFunctionReturn(0); 787 } 788 789 #undef __FUNCT__ 790 #define __FUNCT__ "MatView_MPISBAIJ" 791 PetscErrorCode MatView_MPISBAIJ(Mat mat,PetscViewer viewer) 792 { 793 PetscErrorCode ierr; 794 PetscTruth iascii,isdraw,issocket,isbinary; 795 796 PetscFunctionBegin; 797 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&iascii);CHKERRQ(ierr); 798 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);CHKERRQ(ierr); 799 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_SOCKET,&issocket);CHKERRQ(ierr); 800 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_BINARY,&isbinary);CHKERRQ(ierr); 801 if (iascii || isdraw || issocket || isbinary) { 802 ierr = MatView_MPISBAIJ_ASCIIorDraworSocket(mat,viewer);CHKERRQ(ierr); 803 } else { 804 SETERRQ1(PETSC_ERR_SUP,"Viewer type %s not supported by MPISBAIJ matrices",((PetscObject)viewer)->type_name); 805 } 806 PetscFunctionReturn(0); 807 } 808 809 #undef __FUNCT__ 810 #define __FUNCT__ "MatDestroy_MPISBAIJ" 811 PetscErrorCode MatDestroy_MPISBAIJ(Mat mat) 812 { 813 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 814 PetscErrorCode ierr; 815 816 PetscFunctionBegin; 817 #if defined(PETSC_USE_LOG) 818 PetscLogObjectState((PetscObject)mat,"Rows=%D,Cols=%D",mat->rmap.N,mat->cmap.N); 819 #endif 820 ierr = MatStashDestroy_Private(&mat->stash);CHKERRQ(ierr); 821 ierr = MatStashDestroy_Private(&mat->bstash);CHKERRQ(ierr); 822 ierr = MatDestroy(baij->A);CHKERRQ(ierr); 823 ierr = MatDestroy(baij->B);CHKERRQ(ierr); 824 #if defined (PETSC_USE_CTABLE) 825 if (baij->colmap) {ierr = PetscTableDelete(baij->colmap);CHKERRQ(ierr);} 826 #else 827 ierr = PetscFree(baij->colmap);CHKERRQ(ierr); 828 #endif 829 ierr = PetscFree(baij->garray);CHKERRQ(ierr); 830 if (baij->lvec) {ierr = VecDestroy(baij->lvec);CHKERRQ(ierr);} 831 if (baij->Mvctx) {ierr = VecScatterDestroy(baij->Mvctx);CHKERRQ(ierr);} 832 if (baij->slvec0) { 833 ierr = VecDestroy(baij->slvec0);CHKERRQ(ierr); 834 ierr = VecDestroy(baij->slvec0b);CHKERRQ(ierr); 835 } 836 if (baij->slvec1) { 837 ierr = VecDestroy(baij->slvec1);CHKERRQ(ierr); 838 ierr = VecDestroy(baij->slvec1a);CHKERRQ(ierr); 839 ierr = VecDestroy(baij->slvec1b);CHKERRQ(ierr); 840 } 841 if (baij->sMvctx) {ierr = VecScatterDestroy(baij->sMvctx);CHKERRQ(ierr);} 842 ierr = PetscFree(baij->rowvalues);CHKERRQ(ierr); 843 ierr = PetscFree(baij->barray);CHKERRQ(ierr); 844 ierr = PetscFree(baij->hd);CHKERRQ(ierr); 845 #if defined(PETSC_USE_MAT_SINGLE) 846 ierr = PetscFree(baij->setvaluescopy);CHKERRQ(ierr); 847 #endif 848 ierr = PetscFree(baij->in_loc);CHKERRQ(ierr); 849 ierr = PetscFree(baij->v_loc);CHKERRQ(ierr); 850 ierr = PetscFree(baij->rangebs);CHKERRQ(ierr); 851 ierr = PetscFree(baij);CHKERRQ(ierr); 852 853 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatStoreValues_C","",PETSC_NULL);CHKERRQ(ierr); 854 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatRetrieveValues_C","",PETSC_NULL);CHKERRQ(ierr); 855 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatGetDiagonalBlock_C","",PETSC_NULL);CHKERRQ(ierr); 856 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatMPISBAIJSetPreallocation_C","",PETSC_NULL);CHKERRQ(ierr); 857 PetscFunctionReturn(0); 858 } 859 860 #undef __FUNCT__ 861 #define __FUNCT__ "MatMult_MPISBAIJ" 862 PetscErrorCode MatMult_MPISBAIJ(Mat A,Vec xx,Vec yy) 863 { 864 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 865 PetscErrorCode ierr; 866 PetscInt nt,mbs=a->mbs,bs=A->rmap.bs; 867 PetscScalar *x,*from,zero=0.0; 868 869 PetscFunctionBegin; 870 ierr = VecGetLocalSize(xx,&nt);CHKERRQ(ierr); 871 if (nt != A->cmap.n) { 872 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx"); 873 } 874 875 /* diagonal part */ 876 ierr = (*a->A->ops->mult)(a->A,xx,a->slvec1a);CHKERRQ(ierr); 877 ierr = VecSet(a->slvec1b,zero);CHKERRQ(ierr); 878 879 /* subdiagonal part */ 880 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->slvec0b);CHKERRQ(ierr); 881 CHKMEMQ; 882 /* copy x into the vec slvec0 */ 883 ierr = VecGetArray(a->slvec0,&from);CHKERRQ(ierr); 884 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 885 CHKMEMQ; 886 ierr = PetscMemcpy(from,x,bs*mbs*sizeof(MatScalar));CHKERRQ(ierr); 887 CHKMEMQ; 888 ierr = VecRestoreArray(a->slvec0,&from);CHKERRQ(ierr); 889 890 CHKMEMQ; 891 ierr = VecScatterBegin(a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD,a->sMvctx);CHKERRQ(ierr); 892 CHKMEMQ; 893 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 894 CHKMEMQ; 895 ierr = VecScatterEnd(a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD,a->sMvctx);CHKERRQ(ierr); 896 CHKMEMQ; 897 /* supperdiagonal part */ 898 ierr = (*a->B->ops->multadd)(a->B,a->slvec1b,a->slvec1a,yy);CHKERRQ(ierr); 899 CHKMEMQ; 900 PetscFunctionReturn(0); 901 } 902 903 #undef __FUNCT__ 904 #define __FUNCT__ "MatMult_MPISBAIJ_2comm" 905 PetscErrorCode MatMult_MPISBAIJ_2comm(Mat A,Vec xx,Vec yy) 906 { 907 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 908 PetscErrorCode ierr; 909 PetscInt nt; 910 911 PetscFunctionBegin; 912 ierr = VecGetLocalSize(xx,&nt);CHKERRQ(ierr); 913 if (nt != A->cmap.n) { 914 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx"); 915 } 916 ierr = VecGetLocalSize(yy,&nt);CHKERRQ(ierr); 917 if (nt != A->rmap.N) { 918 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible parition of A and yy"); 919 } 920 921 ierr = VecScatterBegin(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 922 /* do diagonal part */ 923 ierr = (*a->A->ops->mult)(a->A,xx,yy);CHKERRQ(ierr); 924 /* do supperdiagonal part */ 925 ierr = VecScatterEnd(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 926 ierr = (*a->B->ops->multadd)(a->B,a->lvec,yy,yy);CHKERRQ(ierr); 927 /* do subdiagonal part */ 928 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 929 ierr = VecScatterBegin(a->lvec,yy,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 930 ierr = VecScatterEnd(a->lvec,yy,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 931 932 PetscFunctionReturn(0); 933 } 934 935 #undef __FUNCT__ 936 #define __FUNCT__ "MatMultAdd_MPISBAIJ" 937 PetscErrorCode MatMultAdd_MPISBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 938 { 939 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 940 PetscErrorCode ierr; 941 PetscInt mbs=a->mbs,bs=A->rmap.bs; 942 PetscScalar *x,*from,zero=0.0; 943 944 PetscFunctionBegin; 945 /* 946 PetscSynchronizedPrintf(A->comm," MatMultAdd is called ...\n"); 947 PetscSynchronizedFlush(A->comm); 948 */ 949 /* diagonal part */ 950 ierr = (*a->A->ops->multadd)(a->A,xx,yy,a->slvec1a);CHKERRQ(ierr); 951 ierr = VecSet(a->slvec1b,zero);CHKERRQ(ierr); 952 953 /* subdiagonal part */ 954 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->slvec0b);CHKERRQ(ierr); 955 956 /* copy x into the vec slvec0 */ 957 ierr = VecGetArray(a->slvec0,&from);CHKERRQ(ierr); 958 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 959 ierr = PetscMemcpy(from,x,bs*mbs*sizeof(MatScalar));CHKERRQ(ierr); 960 ierr = VecRestoreArray(a->slvec0,&from);CHKERRQ(ierr); 961 962 ierr = VecScatterBegin(a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD,a->sMvctx);CHKERRQ(ierr); 963 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 964 ierr = VecScatterEnd(a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD,a->sMvctx);CHKERRQ(ierr); 965 966 /* supperdiagonal part */ 967 ierr = (*a->B->ops->multadd)(a->B,a->slvec1b,a->slvec1a,zz);CHKERRQ(ierr); 968 969 PetscFunctionReturn(0); 970 } 971 972 #undef __FUNCT__ 973 #define __FUNCT__ "MatMultAdd_MPISBAIJ_2comm" 974 PetscErrorCode MatMultAdd_MPISBAIJ_2comm(Mat A,Vec xx,Vec yy,Vec zz) 975 { 976 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 977 PetscErrorCode ierr; 978 979 PetscFunctionBegin; 980 ierr = VecScatterBegin(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 981 /* do diagonal part */ 982 ierr = (*a->A->ops->multadd)(a->A,xx,yy,zz);CHKERRQ(ierr); 983 /* do supperdiagonal part */ 984 ierr = VecScatterEnd(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 985 ierr = (*a->B->ops->multadd)(a->B,a->lvec,zz,zz);CHKERRQ(ierr); 986 987 /* do subdiagonal part */ 988 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 989 ierr = VecScatterBegin(a->lvec,zz,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 990 ierr = VecScatterEnd(a->lvec,zz,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 991 992 PetscFunctionReturn(0); 993 } 994 995 /* 996 This only works correctly for square matrices where the subblock A->A is the 997 diagonal block 998 */ 999 #undef __FUNCT__ 1000 #define __FUNCT__ "MatGetDiagonal_MPISBAIJ" 1001 PetscErrorCode MatGetDiagonal_MPISBAIJ(Mat A,Vec v) 1002 { 1003 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1004 PetscErrorCode ierr; 1005 1006 PetscFunctionBegin; 1007 /* if (a->rmap.N != a->cmap.N) SETERRQ(PETSC_ERR_SUP,"Supports only square matrix where A->A is diag block"); */ 1008 ierr = MatGetDiagonal(a->A,v);CHKERRQ(ierr); 1009 PetscFunctionReturn(0); 1010 } 1011 1012 #undef __FUNCT__ 1013 #define __FUNCT__ "MatScale_MPISBAIJ" 1014 PetscErrorCode MatScale_MPISBAIJ(Mat A,PetscScalar aa) 1015 { 1016 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1017 PetscErrorCode ierr; 1018 1019 PetscFunctionBegin; 1020 ierr = MatScale(a->A,aa);CHKERRQ(ierr); 1021 ierr = MatScale(a->B,aa);CHKERRQ(ierr); 1022 PetscFunctionReturn(0); 1023 } 1024 1025 #undef __FUNCT__ 1026 #define __FUNCT__ "MatGetRow_MPISBAIJ" 1027 PetscErrorCode MatGetRow_MPISBAIJ(Mat matin,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v) 1028 { 1029 Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data; 1030 PetscScalar *vworkA,*vworkB,**pvA,**pvB,*v_p; 1031 PetscErrorCode ierr; 1032 PetscInt bs = matin->rmap.bs,bs2 = mat->bs2,i,*cworkA,*cworkB,**pcA,**pcB; 1033 PetscInt nztot,nzA,nzB,lrow,brstart = matin->rmap.rstart,brend = matin->rmap.rend; 1034 PetscInt *cmap,*idx_p,cstart = mat->rstartbs; 1035 1036 PetscFunctionBegin; 1037 if (mat->getrowactive == PETSC_TRUE) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Already active"); 1038 mat->getrowactive = PETSC_TRUE; 1039 1040 if (!mat->rowvalues && (idx || v)) { 1041 /* 1042 allocate enough space to hold information from the longest row. 1043 */ 1044 Mat_SeqSBAIJ *Aa = (Mat_SeqSBAIJ*)mat->A->data; 1045 Mat_SeqBAIJ *Ba = (Mat_SeqBAIJ*)mat->B->data; 1046 PetscInt max = 1,mbs = mat->mbs,tmp; 1047 for (i=0; i<mbs; i++) { 1048 tmp = Aa->i[i+1] - Aa->i[i] + Ba->i[i+1] - Ba->i[i]; /* row length */ 1049 if (max < tmp) { max = tmp; } 1050 } 1051 ierr = PetscMalloc(max*bs2*(sizeof(PetscInt)+sizeof(PetscScalar)),&mat->rowvalues);CHKERRQ(ierr); 1052 mat->rowindices = (PetscInt*)(mat->rowvalues + max*bs2); 1053 } 1054 1055 if (row < brstart || row >= brend) SETERRQ(PETSC_ERR_SUP,"Only local rows") 1056 lrow = row - brstart; /* local row index */ 1057 1058 pvA = &vworkA; pcA = &cworkA; pvB = &vworkB; pcB = &cworkB; 1059 if (!v) {pvA = 0; pvB = 0;} 1060 if (!idx) {pcA = 0; if (!v) pcB = 0;} 1061 ierr = (*mat->A->ops->getrow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1062 ierr = (*mat->B->ops->getrow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1063 nztot = nzA + nzB; 1064 1065 cmap = mat->garray; 1066 if (v || idx) { 1067 if (nztot) { 1068 /* Sort by increasing column numbers, assuming A and B already sorted */ 1069 PetscInt imark = -1; 1070 if (v) { 1071 *v = v_p = mat->rowvalues; 1072 for (i=0; i<nzB; i++) { 1073 if (cmap[cworkB[i]/bs] < cstart) v_p[i] = vworkB[i]; 1074 else break; 1075 } 1076 imark = i; 1077 for (i=0; i<nzA; i++) v_p[imark+i] = vworkA[i]; 1078 for (i=imark; i<nzB; i++) v_p[nzA+i] = vworkB[i]; 1079 } 1080 if (idx) { 1081 *idx = idx_p = mat->rowindices; 1082 if (imark > -1) { 1083 for (i=0; i<imark; i++) { 1084 idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs; 1085 } 1086 } else { 1087 for (i=0; i<nzB; i++) { 1088 if (cmap[cworkB[i]/bs] < cstart) 1089 idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ; 1090 else break; 1091 } 1092 imark = i; 1093 } 1094 for (i=0; i<nzA; i++) idx_p[imark+i] = cstart*bs + cworkA[i]; 1095 for (i=imark; i<nzB; i++) idx_p[nzA+i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ; 1096 } 1097 } else { 1098 if (idx) *idx = 0; 1099 if (v) *v = 0; 1100 } 1101 } 1102 *nz = nztot; 1103 ierr = (*mat->A->ops->restorerow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1104 ierr = (*mat->B->ops->restorerow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1105 PetscFunctionReturn(0); 1106 } 1107 1108 #undef __FUNCT__ 1109 #define __FUNCT__ "MatRestoreRow_MPISBAIJ" 1110 PetscErrorCode MatRestoreRow_MPISBAIJ(Mat mat,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v) 1111 { 1112 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 1113 1114 PetscFunctionBegin; 1115 if (!baij->getrowactive) { 1116 SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"MatGetRow() must be called first"); 1117 } 1118 baij->getrowactive = PETSC_FALSE; 1119 PetscFunctionReturn(0); 1120 } 1121 1122 #undef __FUNCT__ 1123 #define __FUNCT__ "MatGetRowUpperTriangular_MPISBAIJ" 1124 PetscErrorCode MatGetRowUpperTriangular_MPISBAIJ(Mat A) 1125 { 1126 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1127 Mat_SeqSBAIJ *aA = (Mat_SeqSBAIJ*)a->A->data; 1128 1129 PetscFunctionBegin; 1130 aA->getrow_utriangular = PETSC_TRUE; 1131 PetscFunctionReturn(0); 1132 } 1133 #undef __FUNCT__ 1134 #define __FUNCT__ "MatRestoreRowUpperTriangular_MPISBAIJ" 1135 PetscErrorCode MatRestoreRowUpperTriangular_MPISBAIJ(Mat A) 1136 { 1137 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1138 Mat_SeqSBAIJ *aA = (Mat_SeqSBAIJ*)a->A->data; 1139 1140 PetscFunctionBegin; 1141 aA->getrow_utriangular = PETSC_FALSE; 1142 PetscFunctionReturn(0); 1143 } 1144 1145 #undef __FUNCT__ 1146 #define __FUNCT__ "MatRealPart_MPISBAIJ" 1147 PetscErrorCode MatRealPart_MPISBAIJ(Mat A) 1148 { 1149 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1150 PetscErrorCode ierr; 1151 1152 PetscFunctionBegin; 1153 ierr = MatRealPart(a->A);CHKERRQ(ierr); 1154 ierr = MatRealPart(a->B);CHKERRQ(ierr); 1155 PetscFunctionReturn(0); 1156 } 1157 1158 #undef __FUNCT__ 1159 #define __FUNCT__ "MatImaginaryPart_MPISBAIJ" 1160 PetscErrorCode MatImaginaryPart_MPISBAIJ(Mat A) 1161 { 1162 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1163 PetscErrorCode ierr; 1164 1165 PetscFunctionBegin; 1166 ierr = MatImaginaryPart(a->A);CHKERRQ(ierr); 1167 ierr = MatImaginaryPart(a->B);CHKERRQ(ierr); 1168 PetscFunctionReturn(0); 1169 } 1170 1171 #undef __FUNCT__ 1172 #define __FUNCT__ "MatZeroEntries_MPISBAIJ" 1173 PetscErrorCode MatZeroEntries_MPISBAIJ(Mat A) 1174 { 1175 Mat_MPISBAIJ *l = (Mat_MPISBAIJ*)A->data; 1176 PetscErrorCode ierr; 1177 1178 PetscFunctionBegin; 1179 ierr = MatZeroEntries(l->A);CHKERRQ(ierr); 1180 ierr = MatZeroEntries(l->B);CHKERRQ(ierr); 1181 PetscFunctionReturn(0); 1182 } 1183 1184 #undef __FUNCT__ 1185 #define __FUNCT__ "MatGetInfo_MPISBAIJ" 1186 PetscErrorCode MatGetInfo_MPISBAIJ(Mat matin,MatInfoType flag,MatInfo *info) 1187 { 1188 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)matin->data; 1189 Mat A = a->A,B = a->B; 1190 PetscErrorCode ierr; 1191 PetscReal isend[5],irecv[5]; 1192 1193 PetscFunctionBegin; 1194 info->block_size = (PetscReal)matin->rmap.bs; 1195 ierr = MatGetInfo(A,MAT_LOCAL,info);CHKERRQ(ierr); 1196 isend[0] = info->nz_used; isend[1] = info->nz_allocated; isend[2] = info->nz_unneeded; 1197 isend[3] = info->memory; isend[4] = info->mallocs; 1198 ierr = MatGetInfo(B,MAT_LOCAL,info);CHKERRQ(ierr); 1199 isend[0] += info->nz_used; isend[1] += info->nz_allocated; isend[2] += info->nz_unneeded; 1200 isend[3] += info->memory; isend[4] += info->mallocs; 1201 if (flag == MAT_LOCAL) { 1202 info->nz_used = isend[0]; 1203 info->nz_allocated = isend[1]; 1204 info->nz_unneeded = isend[2]; 1205 info->memory = isend[3]; 1206 info->mallocs = isend[4]; 1207 } else if (flag == MAT_GLOBAL_MAX) { 1208 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPI_MAX,matin->comm);CHKERRQ(ierr); 1209 info->nz_used = irecv[0]; 1210 info->nz_allocated = irecv[1]; 1211 info->nz_unneeded = irecv[2]; 1212 info->memory = irecv[3]; 1213 info->mallocs = irecv[4]; 1214 } else if (flag == MAT_GLOBAL_SUM) { 1215 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPI_SUM,matin->comm);CHKERRQ(ierr); 1216 info->nz_used = irecv[0]; 1217 info->nz_allocated = irecv[1]; 1218 info->nz_unneeded = irecv[2]; 1219 info->memory = irecv[3]; 1220 info->mallocs = irecv[4]; 1221 } else { 1222 SETERRQ1(PETSC_ERR_ARG_WRONG,"Unknown MatInfoType argument %d",(int)flag); 1223 } 1224 info->rows_global = (PetscReal)A->rmap.N; 1225 info->columns_global = (PetscReal)A->cmap.N; 1226 info->rows_local = (PetscReal)A->rmap.N; 1227 info->columns_local = (PetscReal)A->cmap.N; 1228 info->fill_ratio_given = 0; /* no parallel LU/ILU/Cholesky */ 1229 info->fill_ratio_needed = 0; 1230 info->factor_mallocs = 0; 1231 PetscFunctionReturn(0); 1232 } 1233 1234 #undef __FUNCT__ 1235 #define __FUNCT__ "MatSetOption_MPISBAIJ" 1236 PetscErrorCode MatSetOption_MPISBAIJ(Mat A,MatOption op) 1237 { 1238 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1239 Mat_SeqSBAIJ *aA = (Mat_SeqSBAIJ*)a->A->data; 1240 PetscErrorCode ierr; 1241 1242 PetscFunctionBegin; 1243 switch (op) { 1244 case MAT_NO_NEW_NONZERO_LOCATIONS: 1245 case MAT_YES_NEW_NONZERO_LOCATIONS: 1246 case MAT_COLUMNS_UNSORTED: 1247 case MAT_COLUMNS_SORTED: 1248 case MAT_NEW_NONZERO_ALLOCATION_ERR: 1249 case MAT_KEEP_ZEROED_ROWS: 1250 case MAT_NEW_NONZERO_LOCATION_ERR: 1251 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1252 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1253 break; 1254 case MAT_ROW_ORIENTED: 1255 a->roworiented = PETSC_TRUE; 1256 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1257 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1258 break; 1259 case MAT_ROWS_SORTED: 1260 case MAT_ROWS_UNSORTED: 1261 case MAT_YES_NEW_DIAGONALS: 1262 ierr = PetscInfo1(A,"Option %s ignored\n",MatOptions[op]);CHKERRQ(ierr); 1263 break; 1264 case MAT_COLUMN_ORIENTED: 1265 a->roworiented = PETSC_FALSE; 1266 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1267 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1268 break; 1269 case MAT_IGNORE_OFF_PROC_ENTRIES: 1270 a->donotstash = PETSC_TRUE; 1271 break; 1272 case MAT_NO_NEW_DIAGONALS: 1273 SETERRQ(PETSC_ERR_SUP,"MAT_NO_NEW_DIAGONALS"); 1274 case MAT_USE_HASH_TABLE: 1275 a->ht_flag = PETSC_TRUE; 1276 break; 1277 case MAT_NOT_SYMMETRIC: 1278 case MAT_NOT_STRUCTURALLY_SYMMETRIC: 1279 case MAT_HERMITIAN: 1280 SETERRQ(PETSC_ERR_SUP,"Matrix must be symmetric"); 1281 case MAT_SYMMETRIC: 1282 case MAT_STRUCTURALLY_SYMMETRIC: 1283 case MAT_NOT_HERMITIAN: 1284 case MAT_SYMMETRY_ETERNAL: 1285 case MAT_NOT_SYMMETRY_ETERNAL: 1286 ierr = PetscInfo1(A,"Option %s ignored\n",MatOptions[op]);CHKERRQ(ierr); 1287 break; 1288 case MAT_IGNORE_LOWER_TRIANGULAR: 1289 aA->ignore_ltriangular = PETSC_TRUE; 1290 break; 1291 case MAT_ERROR_LOWER_TRIANGULAR: 1292 aA->ignore_ltriangular = PETSC_FALSE; 1293 break; 1294 case MAT_GETROW_UPPERTRIANGULAR: 1295 aA->getrow_utriangular = PETSC_TRUE; 1296 break; 1297 default: 1298 SETERRQ1(PETSC_ERR_SUP,"unknown option %d",op); 1299 } 1300 PetscFunctionReturn(0); 1301 } 1302 1303 #undef __FUNCT__ 1304 #define __FUNCT__ "MatTranspose_MPISBAIJ" 1305 PetscErrorCode MatTranspose_MPISBAIJ(Mat A,Mat *B) 1306 { 1307 PetscErrorCode ierr; 1308 PetscFunctionBegin; 1309 ierr = MatDuplicate(A,MAT_COPY_VALUES,B);CHKERRQ(ierr); 1310 PetscFunctionReturn(0); 1311 } 1312 1313 #undef __FUNCT__ 1314 #define __FUNCT__ "MatDiagonalScale_MPISBAIJ" 1315 PetscErrorCode MatDiagonalScale_MPISBAIJ(Mat mat,Vec ll,Vec rr) 1316 { 1317 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 1318 Mat a=baij->A, b=baij->B; 1319 PetscErrorCode ierr; 1320 PetscInt nv,m,n; 1321 PetscTruth flg; 1322 1323 PetscFunctionBegin; 1324 if (ll != rr){ 1325 ierr = VecEqual(ll,rr,&flg);CHKERRQ(ierr); 1326 if (!flg) 1327 SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"For symmetric format, left and right scaling vectors must be same\n"); 1328 } 1329 if (!ll) PetscFunctionReturn(0); 1330 1331 ierr = MatGetLocalSize(mat,&m,&n);CHKERRQ(ierr); 1332 if (m != n) SETERRQ2(PETSC_ERR_ARG_SIZ,"For symmetric format, local size %d %d must be same",m,n); 1333 1334 ierr = VecGetLocalSize(rr,&nv);CHKERRQ(ierr); 1335 if (nv!=n) SETERRQ(PETSC_ERR_ARG_SIZ,"Left and right vector non-conforming local size"); 1336 1337 ierr = VecScatterBegin(rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD,baij->Mvctx);CHKERRQ(ierr); 1338 1339 /* left diagonalscale the off-diagonal part */ 1340 ierr = (*b->ops->diagonalscale)(b,ll,PETSC_NULL);CHKERRQ(ierr); 1341 1342 /* scale the diagonal part */ 1343 ierr = (*a->ops->diagonalscale)(a,ll,rr);CHKERRQ(ierr); 1344 1345 /* right diagonalscale the off-diagonal part */ 1346 ierr = VecScatterEnd(rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD,baij->Mvctx);CHKERRQ(ierr); 1347 ierr = (*b->ops->diagonalscale)(b,PETSC_NULL,baij->lvec);CHKERRQ(ierr); 1348 PetscFunctionReturn(0); 1349 } 1350 1351 #undef __FUNCT__ 1352 #define __FUNCT__ "MatSetUnfactored_MPISBAIJ" 1353 PetscErrorCode MatSetUnfactored_MPISBAIJ(Mat A) 1354 { 1355 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1356 PetscErrorCode ierr; 1357 1358 PetscFunctionBegin; 1359 ierr = MatSetUnfactored(a->A);CHKERRQ(ierr); 1360 PetscFunctionReturn(0); 1361 } 1362 1363 static PetscErrorCode MatDuplicate_MPISBAIJ(Mat,MatDuplicateOption,Mat *); 1364 1365 #undef __FUNCT__ 1366 #define __FUNCT__ "MatEqual_MPISBAIJ" 1367 PetscErrorCode MatEqual_MPISBAIJ(Mat A,Mat B,PetscTruth *flag) 1368 { 1369 Mat_MPISBAIJ *matB = (Mat_MPISBAIJ*)B->data,*matA = (Mat_MPISBAIJ*)A->data; 1370 Mat a,b,c,d; 1371 PetscTruth flg; 1372 PetscErrorCode ierr; 1373 1374 PetscFunctionBegin; 1375 a = matA->A; b = matA->B; 1376 c = matB->A; d = matB->B; 1377 1378 ierr = MatEqual(a,c,&flg);CHKERRQ(ierr); 1379 if (flg) { 1380 ierr = MatEqual(b,d,&flg);CHKERRQ(ierr); 1381 } 1382 ierr = MPI_Allreduce(&flg,flag,1,MPI_INT,MPI_LAND,A->comm);CHKERRQ(ierr); 1383 PetscFunctionReturn(0); 1384 } 1385 1386 #undef __FUNCT__ 1387 #define __FUNCT__ "MatCopy_MPISBAIJ" 1388 PetscErrorCode MatCopy_MPISBAIJ(Mat A,Mat B,MatStructure str) 1389 { 1390 PetscErrorCode ierr; 1391 Mat_MPISBAIJ *a = (Mat_MPISBAIJ *)A->data; 1392 Mat_MPISBAIJ *b = (Mat_MPISBAIJ *)B->data; 1393 1394 PetscFunctionBegin; 1395 /* If the two matrices don't have the same copy implementation, they aren't compatible for fast copy. */ 1396 if ((str != SAME_NONZERO_PATTERN) || (A->ops->copy != B->ops->copy)) { 1397 ierr = MatGetRowUpperTriangular(A);CHKERRQ(ierr); 1398 ierr = MatCopy_Basic(A,B,str);CHKERRQ(ierr); 1399 ierr = MatRestoreRowUpperTriangular(A);CHKERRQ(ierr); 1400 } else { 1401 ierr = MatCopy(a->A,b->A,str);CHKERRQ(ierr); 1402 ierr = MatCopy(a->B,b->B,str);CHKERRQ(ierr); 1403 } 1404 PetscFunctionReturn(0); 1405 } 1406 1407 #undef __FUNCT__ 1408 #define __FUNCT__ "MatSetUpPreallocation_MPISBAIJ" 1409 PetscErrorCode MatSetUpPreallocation_MPISBAIJ(Mat A) 1410 { 1411 PetscErrorCode ierr; 1412 1413 PetscFunctionBegin; 1414 ierr = MatMPISBAIJSetPreallocation(A,PetscMax(A->rmap.bs,1),PETSC_DEFAULT,0,PETSC_DEFAULT,0);CHKERRQ(ierr); 1415 PetscFunctionReturn(0); 1416 } 1417 1418 #include "petscblaslapack.h" 1419 #undef __FUNCT__ 1420 #define __FUNCT__ "MatAXPY_MPISBAIJ" 1421 PetscErrorCode MatAXPY_MPISBAIJ(Mat Y,PetscScalar a,Mat X,MatStructure str) 1422 { 1423 PetscErrorCode ierr; 1424 Mat_MPISBAIJ *xx=(Mat_MPISBAIJ *)X->data,*yy=(Mat_MPISBAIJ *)Y->data; 1425 PetscBLASInt bnz,one=1; 1426 Mat_SeqSBAIJ *xa,*ya; 1427 Mat_SeqBAIJ *xb,*yb; 1428 1429 PetscFunctionBegin; 1430 if (str == SAME_NONZERO_PATTERN) { 1431 PetscScalar alpha = a; 1432 xa = (Mat_SeqSBAIJ *)xx->A->data; 1433 ya = (Mat_SeqSBAIJ *)yy->A->data; 1434 bnz = (PetscBLASInt)xa->nz; 1435 BLASaxpy_(&bnz,&alpha,xa->a,&one,ya->a,&one); 1436 xb = (Mat_SeqBAIJ *)xx->B->data; 1437 yb = (Mat_SeqBAIJ *)yy->B->data; 1438 bnz = (PetscBLASInt)xb->nz; 1439 BLASaxpy_(&bnz,&alpha,xb->a,&one,yb->a,&one); 1440 } else { 1441 ierr = MatGetRowUpperTriangular(X);CHKERRQ(ierr); 1442 ierr = MatAXPY_Basic(Y,a,X,str);CHKERRQ(ierr); 1443 ierr = MatRestoreRowUpperTriangular(X);CHKERRQ(ierr); 1444 } 1445 PetscFunctionReturn(0); 1446 } 1447 1448 #undef __FUNCT__ 1449 #define __FUNCT__ "MatGetSubMatrices_MPISBAIJ" 1450 PetscErrorCode MatGetSubMatrices_MPISBAIJ(Mat A,PetscInt n,const IS irow[],const IS icol[],MatReuse scall,Mat *B[]) 1451 { 1452 PetscErrorCode ierr; 1453 PetscInt i; 1454 PetscTruth flg; 1455 1456 PetscFunctionBegin; 1457 for (i=0; i<n; i++) { 1458 ierr = ISEqual(irow[i],icol[i],&flg);CHKERRQ(ierr); 1459 if (!flg) { 1460 SETERRQ(PETSC_ERR_SUP,"Can only get symmetric submatrix for MPISBAIJ matrices"); 1461 } 1462 } 1463 ierr = MatGetSubMatrices_MPIBAIJ(A,n,irow,icol,scall,B);CHKERRQ(ierr); 1464 PetscFunctionReturn(0); 1465 } 1466 1467 1468 /* -------------------------------------------------------------------*/ 1469 static struct _MatOps MatOps_Values = { 1470 MatSetValues_MPISBAIJ, 1471 MatGetRow_MPISBAIJ, 1472 MatRestoreRow_MPISBAIJ, 1473 MatMult_MPISBAIJ, 1474 /* 4*/ MatMultAdd_MPISBAIJ, 1475 MatMult_MPISBAIJ, /* transpose versions are same as non-transpose */ 1476 MatMultAdd_MPISBAIJ, 1477 0, 1478 0, 1479 0, 1480 /*10*/ 0, 1481 0, 1482 0, 1483 MatRelax_MPISBAIJ, 1484 MatTranspose_MPISBAIJ, 1485 /*15*/ MatGetInfo_MPISBAIJ, 1486 MatEqual_MPISBAIJ, 1487 MatGetDiagonal_MPISBAIJ, 1488 MatDiagonalScale_MPISBAIJ, 1489 MatNorm_MPISBAIJ, 1490 /*20*/ MatAssemblyBegin_MPISBAIJ, 1491 MatAssemblyEnd_MPISBAIJ, 1492 0, 1493 MatSetOption_MPISBAIJ, 1494 MatZeroEntries_MPISBAIJ, 1495 /*25*/ 0, 1496 0, 1497 0, 1498 0, 1499 0, 1500 /*30*/ MatSetUpPreallocation_MPISBAIJ, 1501 0, 1502 0, 1503 0, 1504 0, 1505 /*35*/ MatDuplicate_MPISBAIJ, 1506 0, 1507 0, 1508 0, 1509 0, 1510 /*40*/ MatAXPY_MPISBAIJ, 1511 MatGetSubMatrices_MPISBAIJ, 1512 MatIncreaseOverlap_MPISBAIJ, 1513 MatGetValues_MPISBAIJ, 1514 MatCopy_MPISBAIJ, 1515 /*45*/ 0, 1516 MatScale_MPISBAIJ, 1517 0, 1518 0, 1519 0, 1520 /*50*/ 0, 1521 0, 1522 0, 1523 0, 1524 0, 1525 /*55*/ 0, 1526 0, 1527 MatSetUnfactored_MPISBAIJ, 1528 0, 1529 MatSetValuesBlocked_MPISBAIJ, 1530 /*60*/ 0, 1531 0, 1532 0, 1533 0, 1534 0, 1535 /*65*/ 0, 1536 0, 1537 0, 1538 0, 1539 0, 1540 /*70*/ MatGetRowMax_MPISBAIJ, 1541 0, 1542 0, 1543 0, 1544 0, 1545 /*75*/ 0, 1546 0, 1547 0, 1548 0, 1549 0, 1550 /*80*/ 0, 1551 0, 1552 0, 1553 0, 1554 MatLoad_MPISBAIJ, 1555 /*85*/ 0, 1556 0, 1557 0, 1558 0, 1559 0, 1560 /*90*/ 0, 1561 0, 1562 0, 1563 0, 1564 0, 1565 /*95*/ 0, 1566 0, 1567 0, 1568 0, 1569 0, 1570 /*100*/0, 1571 0, 1572 0, 1573 0, 1574 0, 1575 /*105*/0, 1576 MatRealPart_MPISBAIJ, 1577 MatImaginaryPart_MPISBAIJ, 1578 MatGetRowUpperTriangular_MPISBAIJ, 1579 MatRestoreRowUpperTriangular_MPISBAIJ 1580 }; 1581 1582 1583 EXTERN_C_BEGIN 1584 #undef __FUNCT__ 1585 #define __FUNCT__ "MatGetDiagonalBlock_MPISBAIJ" 1586 PetscErrorCode PETSCMAT_DLLEXPORT MatGetDiagonalBlock_MPISBAIJ(Mat A,PetscTruth *iscopy,MatReuse reuse,Mat *a) 1587 { 1588 PetscFunctionBegin; 1589 *a = ((Mat_MPISBAIJ *)A->data)->A; 1590 *iscopy = PETSC_FALSE; 1591 PetscFunctionReturn(0); 1592 } 1593 EXTERN_C_END 1594 1595 EXTERN_C_BEGIN 1596 #undef __FUNCT__ 1597 #define __FUNCT__ "MatMPISBAIJSetPreallocation_MPISBAIJ" 1598 PetscErrorCode PETSCMAT_DLLEXPORT MatMPISBAIJSetPreallocation_MPISBAIJ(Mat B,PetscInt bs,PetscInt d_nz,PetscInt *d_nnz,PetscInt o_nz,PetscInt *o_nnz) 1599 { 1600 Mat_MPISBAIJ *b; 1601 PetscErrorCode ierr; 1602 PetscInt i,mbs,Mbs; 1603 1604 PetscFunctionBegin; 1605 ierr = PetscOptionsBegin(B->comm,B->prefix,"Options for MPIBAIJ matrix","Mat");CHKERRQ(ierr); 1606 ierr = PetscOptionsInt("-mat_block_size","Set the blocksize used to store the matrix","MatMPIBAIJSetPreallocation",bs,&bs,PETSC_NULL);CHKERRQ(ierr); 1607 ierr = PetscOptionsEnd();CHKERRQ(ierr); 1608 1609 if (bs < 1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Invalid block size specified, must be positive"); 1610 if (d_nz == PETSC_DECIDE || d_nz == PETSC_DEFAULT) d_nz = 3; 1611 if (o_nz == PETSC_DECIDE || o_nz == PETSC_DEFAULT) o_nz = 1; 1612 if (d_nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"d_nz cannot be less than 0: value %D",d_nz); 1613 if (o_nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"o_nz cannot be less than 0: value %D",o_nz); 1614 1615 B->rmap.bs = B->cmap.bs = bs; 1616 ierr = PetscMapInitialize(B->comm,&B->rmap);CHKERRQ(ierr); 1617 ierr = PetscMapInitialize(B->comm,&B->cmap);CHKERRQ(ierr); 1618 1619 if (d_nnz) { 1620 for (i=0; i<B->rmap.n/bs; i++) { 1621 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]); 1622 } 1623 } 1624 if (o_nnz) { 1625 for (i=0; i<B->rmap.n/bs; i++) { 1626 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]); 1627 } 1628 } 1629 B->preallocated = PETSC_TRUE; 1630 1631 b = (Mat_MPISBAIJ*)B->data; 1632 mbs = B->rmap.n/bs; 1633 Mbs = B->rmap.N/bs; 1634 if (mbs*bs != B->rmap.n) { 1635 SETERRQ2(PETSC_ERR_ARG_SIZ,"No of local rows %D must be divisible by blocksize %D",B->rmap.N,bs); 1636 } 1637 1638 B->rmap.bs = bs; 1639 b->bs2 = bs*bs; 1640 b->mbs = mbs; 1641 b->nbs = mbs; 1642 b->Mbs = Mbs; 1643 b->Nbs = Mbs; 1644 1645 for (i=0; i<=b->size; i++) { 1646 b->rangebs[i] = B->rmap.range[i]/bs; 1647 } 1648 b->rstartbs = B->rmap.rstart/bs; 1649 b->rendbs = B->rmap.rend/bs; 1650 1651 b->cstartbs = B->cmap.rstart/bs; 1652 b->cendbs = B->cmap.rend/bs; 1653 1654 ierr = MatCreate(PETSC_COMM_SELF,&b->A);CHKERRQ(ierr); 1655 ierr = MatSetSizes(b->A,B->rmap.n,B->cmap.n,B->rmap.n,B->cmap.n);CHKERRQ(ierr); 1656 ierr = MatSetType(b->A,MATSEQSBAIJ);CHKERRQ(ierr); 1657 ierr = MatSeqSBAIJSetPreallocation(b->A,bs,d_nz,d_nnz);CHKERRQ(ierr); 1658 ierr = PetscLogObjectParent(B,b->A);CHKERRQ(ierr); 1659 1660 ierr = MatCreate(PETSC_COMM_SELF,&b->B);CHKERRQ(ierr); 1661 ierr = MatSetSizes(b->B,B->rmap.n,B->cmap.N,B->rmap.n,B->cmap.N);CHKERRQ(ierr); 1662 ierr = MatSetType(b->B,MATSEQBAIJ);CHKERRQ(ierr); 1663 ierr = MatSeqBAIJSetPreallocation(b->B,bs,o_nz,o_nnz);CHKERRQ(ierr); 1664 ierr = PetscLogObjectParent(B,b->B);CHKERRQ(ierr); 1665 1666 /* build cache for off array entries formed */ 1667 ierr = MatStashCreate_Private(B->comm,bs,&B->bstash);CHKERRQ(ierr); 1668 1669 PetscFunctionReturn(0); 1670 } 1671 EXTERN_C_END 1672 1673 /*MC 1674 MATMPISBAIJ - MATMPISBAIJ = "mpisbaij" - A matrix type to be used for distributed symmetric sparse block matrices, 1675 based on block compressed sparse row format. Only the upper triangular portion of the matrix is stored. 1676 1677 Options Database Keys: 1678 . -mat_type mpisbaij - sets the matrix type to "mpisbaij" during a call to MatSetFromOptions() 1679 1680 Level: beginner 1681 1682 .seealso: MatCreateMPISBAIJ 1683 M*/ 1684 1685 EXTERN_C_BEGIN 1686 #undef __FUNCT__ 1687 #define __FUNCT__ "MatCreate_MPISBAIJ" 1688 PetscErrorCode PETSCMAT_DLLEXPORT MatCreate_MPISBAIJ(Mat B) 1689 { 1690 Mat_MPISBAIJ *b; 1691 PetscErrorCode ierr; 1692 PetscTruth flg; 1693 1694 PetscFunctionBegin; 1695 1696 ierr = PetscNew(Mat_MPISBAIJ,&b);CHKERRQ(ierr); 1697 B->data = (void*)b; 1698 ierr = PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));CHKERRQ(ierr); 1699 1700 B->ops->destroy = MatDestroy_MPISBAIJ; 1701 B->ops->view = MatView_MPISBAIJ; 1702 B->mapping = 0; 1703 B->factor = 0; 1704 B->assembled = PETSC_FALSE; 1705 1706 B->insertmode = NOT_SET_VALUES; 1707 ierr = MPI_Comm_rank(B->comm,&b->rank);CHKERRQ(ierr); 1708 ierr = MPI_Comm_size(B->comm,&b->size);CHKERRQ(ierr); 1709 1710 /* build local table of row and column ownerships */ 1711 ierr = PetscMalloc((b->size+2)*sizeof(PetscInt),&b->rangebs);CHKERRQ(ierr); 1712 1713 /* build cache for off array entries formed */ 1714 ierr = MatStashCreate_Private(B->comm,1,&B->stash);CHKERRQ(ierr); 1715 b->donotstash = PETSC_FALSE; 1716 b->colmap = PETSC_NULL; 1717 b->garray = PETSC_NULL; 1718 b->roworiented = PETSC_TRUE; 1719 1720 #if defined(PETSC_USE_MAT_SINGLE) 1721 /* stuff for MatSetValues_XXX in single precision */ 1722 b->setvalueslen = 0; 1723 b->setvaluescopy = PETSC_NULL; 1724 #endif 1725 1726 /* stuff used in block assembly */ 1727 b->barray = 0; 1728 1729 /* stuff used for matrix vector multiply */ 1730 b->lvec = 0; 1731 b->Mvctx = 0; 1732 b->slvec0 = 0; 1733 b->slvec0b = 0; 1734 b->slvec1 = 0; 1735 b->slvec1a = 0; 1736 b->slvec1b = 0; 1737 b->sMvctx = 0; 1738 1739 /* stuff for MatGetRow() */ 1740 b->rowindices = 0; 1741 b->rowvalues = 0; 1742 b->getrowactive = PETSC_FALSE; 1743 1744 /* hash table stuff */ 1745 b->ht = 0; 1746 b->hd = 0; 1747 b->ht_size = 0; 1748 b->ht_flag = PETSC_FALSE; 1749 b->ht_fact = 0; 1750 b->ht_total_ct = 0; 1751 b->ht_insert_ct = 0; 1752 1753 b->in_loc = 0; 1754 b->v_loc = 0; 1755 b->n_loc = 0; 1756 ierr = PetscOptionsBegin(B->comm,PETSC_NULL,"Options for loading MPIBAIJ matrix","Mat");CHKERRQ(ierr); 1757 ierr = PetscOptionsTruth("-mat_use_hash_table","Use hash table to save memory in constructing matrix","MatSetOption",PETSC_FALSE,&flg,PETSC_NULL);CHKERRQ(ierr); 1758 if (flg) { 1759 PetscReal fact = 1.39; 1760 ierr = MatSetOption(B,MAT_USE_HASH_TABLE);CHKERRQ(ierr); 1761 ierr = PetscOptionsReal("-mat_use_hash_table","Use hash table factor","MatMPIBAIJSetHashTableFactor",fact,&fact,PETSC_NULL);CHKERRQ(ierr); 1762 if (fact <= 1.0) fact = 1.39; 1763 ierr = MatMPIBAIJSetHashTableFactor(B,fact);CHKERRQ(ierr); 1764 ierr = PetscInfo1(0,"Hash table Factor used %5.2f\n",fact);CHKERRQ(ierr); 1765 } 1766 ierr = PetscOptionsEnd();CHKERRQ(ierr); 1767 1768 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatStoreValues_C", 1769 "MatStoreValues_MPISBAIJ", 1770 MatStoreValues_MPISBAIJ);CHKERRQ(ierr); 1771 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatRetrieveValues_C", 1772 "MatRetrieveValues_MPISBAIJ", 1773 MatRetrieveValues_MPISBAIJ);CHKERRQ(ierr); 1774 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatGetDiagonalBlock_C", 1775 "MatGetDiagonalBlock_MPISBAIJ", 1776 MatGetDiagonalBlock_MPISBAIJ);CHKERRQ(ierr); 1777 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatMPISBAIJSetPreallocation_C", 1778 "MatMPISBAIJSetPreallocation_MPISBAIJ", 1779 MatMPISBAIJSetPreallocation_MPISBAIJ);CHKERRQ(ierr); 1780 B->symmetric = PETSC_TRUE; 1781 B->structurally_symmetric = PETSC_TRUE; 1782 B->symmetric_set = PETSC_TRUE; 1783 B->structurally_symmetric_set = PETSC_TRUE; 1784 ierr = PetscObjectChangeTypeName((PetscObject)B,MATMPISBAIJ);CHKERRQ(ierr); 1785 PetscFunctionReturn(0); 1786 } 1787 EXTERN_C_END 1788 1789 /*MC 1790 MATSBAIJ - MATSBAIJ = "sbaij" - A matrix type to be used for symmetric block sparse matrices. 1791 1792 This matrix type is identical to MATSEQSBAIJ when constructed with a single process communicator, 1793 and MATMPISBAIJ otherwise. 1794 1795 Options Database Keys: 1796 . -mat_type sbaij - sets the matrix type to "sbaij" during a call to MatSetFromOptions() 1797 1798 Level: beginner 1799 1800 .seealso: MatCreateMPISBAIJ,MATSEQSBAIJ,MATMPISBAIJ 1801 M*/ 1802 1803 EXTERN_C_BEGIN 1804 #undef __FUNCT__ 1805 #define __FUNCT__ "MatCreate_SBAIJ" 1806 PetscErrorCode PETSCMAT_DLLEXPORT MatCreate_SBAIJ(Mat A) 1807 { 1808 PetscErrorCode ierr; 1809 PetscMPIInt size; 1810 1811 PetscFunctionBegin; 1812 ierr = MPI_Comm_size(A->comm,&size);CHKERRQ(ierr); 1813 if (size == 1) { 1814 ierr = MatSetType(A,MATSEQSBAIJ);CHKERRQ(ierr); 1815 } else { 1816 ierr = MatSetType(A,MATMPISBAIJ);CHKERRQ(ierr); 1817 } 1818 PetscFunctionReturn(0); 1819 } 1820 EXTERN_C_END 1821 1822 #undef __FUNCT__ 1823 #define __FUNCT__ "MatMPISBAIJSetPreallocation" 1824 /*@C 1825 MatMPISBAIJSetPreallocation - For good matrix assembly performance 1826 the user should preallocate the matrix storage by setting the parameters 1827 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 1828 performance can be increased by more than a factor of 50. 1829 1830 Collective on Mat 1831 1832 Input Parameters: 1833 + A - the matrix 1834 . bs - size of blockk 1835 . d_nz - number of block nonzeros per block row in diagonal portion of local 1836 submatrix (same for all local rows) 1837 . d_nnz - array containing the number of block nonzeros in the various block rows 1838 in the upper triangular and diagonal part of the in diagonal portion of the local 1839 (possibly different for each block row) or PETSC_NULL. You must leave room 1840 for the diagonal entry even if it is zero. 1841 . o_nz - number of block nonzeros per block row in the off-diagonal portion of local 1842 submatrix (same for all local rows). 1843 - o_nnz - array containing the number of nonzeros in the various block rows of the 1844 off-diagonal portion of the local submatrix (possibly different for 1845 each block row) or PETSC_NULL. 1846 1847 1848 Options Database Keys: 1849 . -mat_no_unroll - uses code that does not unroll the loops in the 1850 block calculations (much slower) 1851 . -mat_block_size - size of the blocks to use 1852 1853 Notes: 1854 1855 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 1856 than it must be used on all processors that share the object for that argument. 1857 1858 If the *_nnz parameter is given then the *_nz parameter is ignored 1859 1860 Storage Information: 1861 For a square global matrix we define each processor's diagonal portion 1862 to be its local rows and the corresponding columns (a square submatrix); 1863 each processor's off-diagonal portion encompasses the remainder of the 1864 local matrix (a rectangular submatrix). 1865 1866 The user can specify preallocated storage for the diagonal part of 1867 the local submatrix with either d_nz or d_nnz (not both). Set 1868 d_nz=PETSC_DEFAULT and d_nnz=PETSC_NULL for PETSc to control dynamic 1869 memory allocation. Likewise, specify preallocated storage for the 1870 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 1871 1872 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 1873 the figure below we depict these three local rows and all columns (0-11). 1874 1875 .vb 1876 0 1 2 3 4 5 6 7 8 9 10 11 1877 ------------------- 1878 row 3 | o o o d d d o o o o o o 1879 row 4 | o o o d d d o o o o o o 1880 row 5 | o o o d d d o o o o o o 1881 ------------------- 1882 .ve 1883 1884 Thus, any entries in the d locations are stored in the d (diagonal) 1885 submatrix, and any entries in the o locations are stored in the 1886 o (off-diagonal) submatrix. Note that the d matrix is stored in 1887 MatSeqSBAIJ format and the o submatrix in MATSEQBAIJ format. 1888 1889 Now d_nz should indicate the number of block nonzeros per row in the upper triangular 1890 plus the diagonal part of the d matrix, 1891 and o_nz should indicate the number of block nonzeros per row in the o matrix. 1892 In general, for PDE problems in which most nonzeros are near the diagonal, 1893 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 1894 or you will get TERRIBLE performance; see the users' manual chapter on 1895 matrices. 1896 1897 Level: intermediate 1898 1899 .keywords: matrix, block, aij, compressed row, sparse, parallel 1900 1901 .seealso: MatCreate(), MatCreateSeqSBAIJ(), MatSetValues(), MatCreateMPIBAIJ() 1902 @*/ 1903 PetscErrorCode PETSCMAT_DLLEXPORT MatMPISBAIJSetPreallocation(Mat B,PetscInt bs,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[]) 1904 { 1905 PetscErrorCode ierr,(*f)(Mat,PetscInt,PetscInt,const PetscInt[],PetscInt,const PetscInt[]); 1906 1907 PetscFunctionBegin; 1908 ierr = PetscObjectQueryFunction((PetscObject)B,"MatMPISBAIJSetPreallocation_C",(void (**)(void))&f);CHKERRQ(ierr); 1909 if (f) { 1910 ierr = (*f)(B,bs,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 1911 } 1912 PetscFunctionReturn(0); 1913 } 1914 1915 #undef __FUNCT__ 1916 #define __FUNCT__ "MatCreateMPISBAIJ" 1917 /*@C 1918 MatCreateMPISBAIJ - Creates a sparse parallel matrix in symmetric block AIJ format 1919 (block compressed row). For good matrix assembly performance 1920 the user should preallocate the matrix storage by setting the parameters 1921 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 1922 performance can be increased by more than a factor of 50. 1923 1924 Collective on MPI_Comm 1925 1926 Input Parameters: 1927 + comm - MPI communicator 1928 . bs - size of blockk 1929 . m - number of local rows (or PETSC_DECIDE to have calculated if M is given) 1930 This value should be the same as the local size used in creating the 1931 y vector for the matrix-vector product y = Ax. 1932 . n - number of local columns (or PETSC_DECIDE to have calculated if N is given) 1933 This value should be the same as the local size used in creating the 1934 x vector for the matrix-vector product y = Ax. 1935 . M - number of global rows (or PETSC_DETERMINE to have calculated if m is given) 1936 . N - number of global columns (or PETSC_DETERMINE to have calculated if n is given) 1937 . d_nz - number of block nonzeros per block row in diagonal portion of local 1938 submatrix (same for all local rows) 1939 . d_nnz - array containing the number of block nonzeros in the various block rows 1940 in the upper triangular portion of the in diagonal portion of the local 1941 (possibly different for each block block row) or PETSC_NULL. 1942 You must leave room for the diagonal entry even if it is zero. 1943 . o_nz - number of block nonzeros per block row in the off-diagonal portion of local 1944 submatrix (same for all local rows). 1945 - o_nnz - array containing the number of nonzeros in the various block rows of the 1946 off-diagonal portion of the local submatrix (possibly different for 1947 each block row) or PETSC_NULL. 1948 1949 Output Parameter: 1950 . A - the matrix 1951 1952 Options Database Keys: 1953 . -mat_no_unroll - uses code that does not unroll the loops in the 1954 block calculations (much slower) 1955 . -mat_block_size - size of the blocks to use 1956 . -mat_mpi - use the parallel matrix data structures even on one processor 1957 (defaults to using SeqBAIJ format on one processor) 1958 1959 Notes: 1960 The number of rows and columns must be divisible by blocksize. 1961 1962 The user MUST specify either the local or global matrix dimensions 1963 (possibly both). 1964 1965 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 1966 than it must be used on all processors that share the object for that argument. 1967 1968 If the *_nnz parameter is given then the *_nz parameter is ignored 1969 1970 Storage Information: 1971 For a square global matrix we define each processor's diagonal portion 1972 to be its local rows and the corresponding columns (a square submatrix); 1973 each processor's off-diagonal portion encompasses the remainder of the 1974 local matrix (a rectangular submatrix). 1975 1976 The user can specify preallocated storage for the diagonal part of 1977 the local submatrix with either d_nz or d_nnz (not both). Set 1978 d_nz=PETSC_DEFAULT and d_nnz=PETSC_NULL for PETSc to control dynamic 1979 memory allocation. Likewise, specify preallocated storage for the 1980 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 1981 1982 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 1983 the figure below we depict these three local rows and all columns (0-11). 1984 1985 .vb 1986 0 1 2 3 4 5 6 7 8 9 10 11 1987 ------------------- 1988 row 3 | o o o d d d o o o o o o 1989 row 4 | o o o d d d o o o o o o 1990 row 5 | o o o d d d o o o o o o 1991 ------------------- 1992 .ve 1993 1994 Thus, any entries in the d locations are stored in the d (diagonal) 1995 submatrix, and any entries in the o locations are stored in the 1996 o (off-diagonal) submatrix. Note that the d matrix is stored in 1997 MatSeqSBAIJ format and the o submatrix in MATSEQBAIJ format. 1998 1999 Now d_nz should indicate the number of block nonzeros per row in the upper triangular 2000 plus the diagonal part of the d matrix, 2001 and o_nz should indicate the number of block nonzeros per row in the o matrix. 2002 In general, for PDE problems in which most nonzeros are near the diagonal, 2003 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 2004 or you will get TERRIBLE performance; see the users' manual chapter on 2005 matrices. 2006 2007 Level: intermediate 2008 2009 .keywords: matrix, block, aij, compressed row, sparse, parallel 2010 2011 .seealso: MatCreate(), MatCreateSeqSBAIJ(), MatSetValues(), MatCreateMPIBAIJ() 2012 @*/ 2013 2014 PetscErrorCode PETSCMAT_DLLEXPORT MatCreateMPISBAIJ(MPI_Comm comm,PetscInt bs,PetscInt m,PetscInt n,PetscInt M,PetscInt N,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[],Mat *A) 2015 { 2016 PetscErrorCode ierr; 2017 PetscMPIInt size; 2018 2019 PetscFunctionBegin; 2020 ierr = MatCreate(comm,A);CHKERRQ(ierr); 2021 ierr = MatSetSizes(*A,m,n,M,N);CHKERRQ(ierr); 2022 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2023 if (size > 1) { 2024 ierr = MatSetType(*A,MATMPISBAIJ);CHKERRQ(ierr); 2025 ierr = MatMPISBAIJSetPreallocation(*A,bs,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 2026 } else { 2027 ierr = MatSetType(*A,MATSEQSBAIJ);CHKERRQ(ierr); 2028 ierr = MatSeqSBAIJSetPreallocation(*A,bs,d_nz,d_nnz);CHKERRQ(ierr); 2029 } 2030 PetscFunctionReturn(0); 2031 } 2032 2033 2034 #undef __FUNCT__ 2035 #define __FUNCT__ "MatDuplicate_MPISBAIJ" 2036 static PetscErrorCode MatDuplicate_MPISBAIJ(Mat matin,MatDuplicateOption cpvalues,Mat *newmat) 2037 { 2038 Mat mat; 2039 Mat_MPISBAIJ *a,*oldmat = (Mat_MPISBAIJ*)matin->data; 2040 PetscErrorCode ierr; 2041 PetscInt len=0,nt,bs=matin->rmap.bs,mbs=oldmat->mbs; 2042 PetscScalar *array; 2043 2044 PetscFunctionBegin; 2045 *newmat = 0; 2046 ierr = MatCreate(matin->comm,&mat);CHKERRQ(ierr); 2047 ierr = MatSetSizes(mat,matin->rmap.n,matin->cmap.n,matin->rmap.N,matin->cmap.N);CHKERRQ(ierr); 2048 ierr = MatSetType(mat,matin->type_name);CHKERRQ(ierr); 2049 ierr = PetscMemcpy(mat->ops,matin->ops,sizeof(struct _MatOps));CHKERRQ(ierr); 2050 ierr = PetscMapCopy(matin->comm,&matin->rmap,&mat->rmap);CHKERRQ(ierr); 2051 ierr = PetscMapCopy(matin->comm,&matin->cmap,&mat->cmap);CHKERRQ(ierr); 2052 2053 mat->factor = matin->factor; 2054 mat->preallocated = PETSC_TRUE; 2055 mat->assembled = PETSC_TRUE; 2056 mat->insertmode = NOT_SET_VALUES; 2057 2058 a = (Mat_MPISBAIJ*)mat->data; 2059 a->bs2 = oldmat->bs2; 2060 a->mbs = oldmat->mbs; 2061 a->nbs = oldmat->nbs; 2062 a->Mbs = oldmat->Mbs; 2063 a->Nbs = oldmat->Nbs; 2064 2065 2066 a->size = oldmat->size; 2067 a->rank = oldmat->rank; 2068 a->donotstash = oldmat->donotstash; 2069 a->roworiented = oldmat->roworiented; 2070 a->rowindices = 0; 2071 a->rowvalues = 0; 2072 a->getrowactive = PETSC_FALSE; 2073 a->barray = 0; 2074 a->rstartbs = oldmat->rstartbs; 2075 a->rendbs = oldmat->rendbs; 2076 a->cstartbs = oldmat->cstartbs; 2077 a->cendbs = oldmat->cendbs; 2078 2079 /* hash table stuff */ 2080 a->ht = 0; 2081 a->hd = 0; 2082 a->ht_size = 0; 2083 a->ht_flag = oldmat->ht_flag; 2084 a->ht_fact = oldmat->ht_fact; 2085 a->ht_total_ct = 0; 2086 a->ht_insert_ct = 0; 2087 2088 ierr = PetscMemcpy(a->rangebs,oldmat->rangebs,(a->size+2)*sizeof(PetscInt));CHKERRQ(ierr); 2089 ierr = MatStashCreate_Private(matin->comm,1,&mat->stash);CHKERRQ(ierr); 2090 ierr = MatStashCreate_Private(matin->comm,matin->rmap.bs,&mat->bstash);CHKERRQ(ierr); 2091 if (oldmat->colmap) { 2092 #if defined (PETSC_USE_CTABLE) 2093 ierr = PetscTableCreateCopy(oldmat->colmap,&a->colmap);CHKERRQ(ierr); 2094 #else 2095 ierr = PetscMalloc((a->Nbs)*sizeof(PetscInt),&a->colmap);CHKERRQ(ierr); 2096 ierr = PetscLogObjectMemory(mat,(a->Nbs)*sizeof(PetscInt));CHKERRQ(ierr); 2097 ierr = PetscMemcpy(a->colmap,oldmat->colmap,(a->Nbs)*sizeof(PetscInt));CHKERRQ(ierr); 2098 #endif 2099 } else a->colmap = 0; 2100 2101 if (oldmat->garray && (len = ((Mat_SeqBAIJ*)(oldmat->B->data))->nbs)) { 2102 ierr = PetscMalloc(len*sizeof(PetscInt),&a->garray);CHKERRQ(ierr); 2103 ierr = PetscLogObjectMemory(mat,len*sizeof(PetscInt));CHKERRQ(ierr); 2104 ierr = PetscMemcpy(a->garray,oldmat->garray,len*sizeof(PetscInt));CHKERRQ(ierr); 2105 } else a->garray = 0; 2106 2107 ierr = VecDuplicate(oldmat->lvec,&a->lvec);CHKERRQ(ierr); 2108 ierr = PetscLogObjectParent(mat,a->lvec);CHKERRQ(ierr); 2109 ierr = VecScatterCopy(oldmat->Mvctx,&a->Mvctx);CHKERRQ(ierr); 2110 ierr = PetscLogObjectParent(mat,a->Mvctx);CHKERRQ(ierr); 2111 2112 ierr = VecDuplicate(oldmat->slvec0,&a->slvec0);CHKERRQ(ierr); 2113 ierr = PetscLogObjectParent(mat,a->slvec0);CHKERRQ(ierr); 2114 ierr = VecDuplicate(oldmat->slvec1,&a->slvec1);CHKERRQ(ierr); 2115 ierr = PetscLogObjectParent(mat,a->slvec1);CHKERRQ(ierr); 2116 2117 ierr = VecGetLocalSize(a->slvec1,&nt);CHKERRQ(ierr); 2118 ierr = VecGetArray(a->slvec1,&array);CHKERRQ(ierr); 2119 ierr = VecCreateSeqWithArray(PETSC_COMM_SELF,bs*mbs,array,&a->slvec1a);CHKERRQ(ierr); 2120 ierr = VecCreateSeqWithArray(PETSC_COMM_SELF,nt-bs*mbs,array+bs*mbs,&a->slvec1b);CHKERRQ(ierr); 2121 ierr = VecRestoreArray(a->slvec1,&array);CHKERRQ(ierr); 2122 ierr = VecGetArray(a->slvec0,&array);CHKERRQ(ierr); 2123 ierr = VecCreateSeqWithArray(PETSC_COMM_SELF,nt-bs*mbs,array+bs*mbs,&a->slvec0b);CHKERRQ(ierr); 2124 ierr = VecRestoreArray(a->slvec0,&array);CHKERRQ(ierr); 2125 ierr = PetscLogObjectParent(mat,a->slvec0);CHKERRQ(ierr); 2126 ierr = PetscLogObjectParent(mat,a->slvec1);CHKERRQ(ierr); 2127 ierr = PetscLogObjectParent(mat,a->slvec0b);CHKERRQ(ierr); 2128 ierr = PetscLogObjectParent(mat,a->slvec1a);CHKERRQ(ierr); 2129 ierr = PetscLogObjectParent(mat,a->slvec1b);CHKERRQ(ierr); 2130 2131 /* ierr = VecScatterCopy(oldmat->sMvctx,&a->sMvctx); - not written yet, replaced by the lazy trick: */ 2132 ierr = PetscObjectReference((PetscObject)oldmat->sMvctx);CHKERRQ(ierr); 2133 a->sMvctx = oldmat->sMvctx; 2134 ierr = PetscLogObjectParent(mat,a->sMvctx);CHKERRQ(ierr); 2135 2136 ierr = MatDuplicate(oldmat->A,cpvalues,&a->A);CHKERRQ(ierr); 2137 ierr = PetscLogObjectParent(mat,a->A);CHKERRQ(ierr); 2138 ierr = MatDuplicate(oldmat->B,cpvalues,&a->B);CHKERRQ(ierr); 2139 ierr = PetscLogObjectParent(mat,a->B);CHKERRQ(ierr); 2140 ierr = PetscFListDuplicate(mat->qlist,&matin->qlist);CHKERRQ(ierr); 2141 *newmat = mat; 2142 PetscFunctionReturn(0); 2143 } 2144 2145 #include "petscsys.h" 2146 2147 #undef __FUNCT__ 2148 #define __FUNCT__ "MatLoad_MPISBAIJ" 2149 PetscErrorCode MatLoad_MPISBAIJ(PetscViewer viewer, MatType type,Mat *newmat) 2150 { 2151 Mat A; 2152 PetscErrorCode ierr; 2153 PetscInt i,nz,j,rstart,rend; 2154 PetscScalar *vals,*buf; 2155 MPI_Comm comm = ((PetscObject)viewer)->comm; 2156 MPI_Status status; 2157 PetscMPIInt rank,size,tag = ((PetscObject)viewer)->tag,*sndcounts = 0,*browners,maxnz,*rowners,*locrowlens; 2158 PetscInt header[4],*rowlengths = 0,M,N,m,*cols; 2159 PetscInt *procsnz = 0,jj,*mycols,*ibuf; 2160 PetscInt bs=1,Mbs,mbs,extra_rows; 2161 PetscInt *dlens,*odlens,*mask,*masked1,*masked2,rowcount,odcount; 2162 PetscInt dcount,kmax,k,nzcount,tmp; 2163 int fd; 2164 2165 PetscFunctionBegin; 2166 ierr = PetscOptionsBegin(comm,PETSC_NULL,"Options for loading MPIBAIJ matrix","Mat");CHKERRQ(ierr); 2167 ierr = PetscOptionsInt("-matload_block_size","Set the blocksize used to store the matrix","MatLoad",bs,&bs,PETSC_NULL);CHKERRQ(ierr); 2168 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2169 2170 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2171 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 2172 if (!rank) { 2173 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 2174 ierr = PetscBinaryRead(fd,(char *)header,4,PETSC_INT);CHKERRQ(ierr); 2175 if (header[0] != MAT_FILE_COOKIE) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"not matrix object"); 2176 if (header[3] < 0) { 2177 SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"Matrix stored in special format, cannot load as MPISBAIJ"); 2178 } 2179 } 2180 2181 ierr = MPI_Bcast(header+1,3,MPIU_INT,0,comm);CHKERRQ(ierr); 2182 M = header[1]; N = header[2]; 2183 2184 if (M != N) SETERRQ(PETSC_ERR_SUP,"Can only do square matrices"); 2185 2186 /* 2187 This code adds extra rows to make sure the number of rows is 2188 divisible by the blocksize 2189 */ 2190 Mbs = M/bs; 2191 extra_rows = bs - M + bs*(Mbs); 2192 if (extra_rows == bs) extra_rows = 0; 2193 else Mbs++; 2194 if (extra_rows &&!rank) { 2195 ierr = PetscInfo(0,"Padding loaded matrix to match blocksize\n");CHKERRQ(ierr); 2196 } 2197 2198 /* determine ownership of all rows */ 2199 mbs = Mbs/size + ((Mbs % size) > rank); 2200 m = mbs*bs; 2201 ierr = PetscMalloc(2*(size+2)*sizeof(PetscMPIInt),&rowners);CHKERRQ(ierr); 2202 browners = rowners + size + 1; 2203 ierr = MPI_Allgather(&mbs,1,MPI_INT,rowners+1,1,MPI_INT,comm);CHKERRQ(ierr); 2204 rowners[0] = 0; 2205 for (i=2; i<=size; i++) rowners[i] += rowners[i-1]; 2206 for (i=0; i<=size; i++) browners[i] = rowners[i]*bs; 2207 rstart = rowners[rank]; 2208 rend = rowners[rank+1]; 2209 2210 /* distribute row lengths to all processors */ 2211 ierr = PetscMalloc((rend-rstart)*bs*sizeof(PetscMPIInt),&locrowlens);CHKERRQ(ierr); 2212 if (!rank) { 2213 ierr = PetscMalloc((M+extra_rows)*sizeof(PetscInt),&rowlengths);CHKERRQ(ierr); 2214 ierr = PetscBinaryRead(fd,rowlengths,M,PETSC_INT);CHKERRQ(ierr); 2215 for (i=0; i<extra_rows; i++) rowlengths[M+i] = 1; 2216 ierr = PetscMalloc(size*sizeof(PetscMPIInt),&sndcounts);CHKERRQ(ierr); 2217 for (i=0; i<size; i++) sndcounts[i] = browners[i+1] - browners[i]; 2218 ierr = MPI_Scatterv(rowlengths,sndcounts,browners,MPIU_INT,locrowlens,(rend-rstart)*bs,MPIU_INT,0,comm);CHKERRQ(ierr); 2219 ierr = PetscFree(sndcounts);CHKERRQ(ierr); 2220 } else { 2221 ierr = MPI_Scatterv(0,0,0,MPIU_INT,locrowlens,(rend-rstart)*bs,MPIU_INT,0,comm);CHKERRQ(ierr); 2222 } 2223 2224 if (!rank) { /* procs[0] */ 2225 /* calculate the number of nonzeros on each processor */ 2226 ierr = PetscMalloc(size*sizeof(PetscInt),&procsnz);CHKERRQ(ierr); 2227 ierr = PetscMemzero(procsnz,size*sizeof(PetscInt));CHKERRQ(ierr); 2228 for (i=0; i<size; i++) { 2229 for (j=rowners[i]*bs; j< rowners[i+1]*bs; j++) { 2230 procsnz[i] += rowlengths[j]; 2231 } 2232 } 2233 ierr = PetscFree(rowlengths);CHKERRQ(ierr); 2234 2235 /* determine max buffer needed and allocate it */ 2236 maxnz = 0; 2237 for (i=0; i<size; i++) { 2238 maxnz = PetscMax(maxnz,procsnz[i]); 2239 } 2240 ierr = PetscMalloc(maxnz*sizeof(PetscInt),&cols);CHKERRQ(ierr); 2241 2242 /* read in my part of the matrix column indices */ 2243 nz = procsnz[0]; 2244 ierr = PetscMalloc(nz*sizeof(PetscInt),&ibuf);CHKERRQ(ierr); 2245 mycols = ibuf; 2246 if (size == 1) nz -= extra_rows; 2247 ierr = PetscBinaryRead(fd,mycols,nz,PETSC_INT);CHKERRQ(ierr); 2248 if (size == 1) for (i=0; i< extra_rows; i++) { mycols[nz+i] = M+i; } 2249 2250 /* read in every ones (except the last) and ship off */ 2251 for (i=1; i<size-1; i++) { 2252 nz = procsnz[i]; 2253 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 2254 ierr = MPI_Send(cols,nz,MPIU_INT,i,tag,comm);CHKERRQ(ierr); 2255 } 2256 /* read in the stuff for the last proc */ 2257 if (size != 1) { 2258 nz = procsnz[size-1] - extra_rows; /* the extra rows are not on the disk */ 2259 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 2260 for (i=0; i<extra_rows; i++) cols[nz+i] = M+i; 2261 ierr = MPI_Send(cols,nz+extra_rows,MPIU_INT,size-1,tag,comm);CHKERRQ(ierr); 2262 } 2263 ierr = PetscFree(cols);CHKERRQ(ierr); 2264 } else { /* procs[i], i>0 */ 2265 /* determine buffer space needed for message */ 2266 nz = 0; 2267 for (i=0; i<m; i++) { 2268 nz += locrowlens[i]; 2269 } 2270 ierr = PetscMalloc(nz*sizeof(PetscInt),&ibuf);CHKERRQ(ierr); 2271 mycols = ibuf; 2272 /* receive message of column indices*/ 2273 ierr = MPI_Recv(mycols,nz,MPIU_INT,0,tag,comm,&status);CHKERRQ(ierr); 2274 ierr = MPI_Get_count(&status,MPIU_INT,&maxnz);CHKERRQ(ierr); 2275 if (maxnz != nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file"); 2276 } 2277 2278 /* loop over local rows, determining number of off diagonal entries */ 2279 ierr = PetscMalloc(2*(rend-rstart+1)*sizeof(PetscInt),&dlens);CHKERRQ(ierr); 2280 odlens = dlens + (rend-rstart); 2281 ierr = PetscMalloc(3*Mbs*sizeof(PetscInt),&mask);CHKERRQ(ierr); 2282 ierr = PetscMemzero(mask,3*Mbs*sizeof(PetscInt));CHKERRQ(ierr); 2283 masked1 = mask + Mbs; 2284 masked2 = masked1 + Mbs; 2285 rowcount = 0; nzcount = 0; 2286 for (i=0; i<mbs; i++) { 2287 dcount = 0; 2288 odcount = 0; 2289 for (j=0; j<bs; j++) { 2290 kmax = locrowlens[rowcount]; 2291 for (k=0; k<kmax; k++) { 2292 tmp = mycols[nzcount++]/bs; /* block col. index */ 2293 if (!mask[tmp]) { 2294 mask[tmp] = 1; 2295 if (tmp < rstart || tmp >= rend) masked2[odcount++] = tmp; /* entry in off-diag portion */ 2296 else masked1[dcount++] = tmp; /* entry in diag portion */ 2297 } 2298 } 2299 rowcount++; 2300 } 2301 2302 dlens[i] = dcount; /* d_nzz[i] */ 2303 odlens[i] = odcount; /* o_nzz[i] */ 2304 2305 /* zero out the mask elements we set */ 2306 for (j=0; j<dcount; j++) mask[masked1[j]] = 0; 2307 for (j=0; j<odcount; j++) mask[masked2[j]] = 0; 2308 } 2309 2310 /* create our matrix */ 2311 ierr = MatCreate(comm,&A);CHKERRQ(ierr); 2312 ierr = MatSetSizes(A,m,m,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 2313 ierr = MatSetType(A,type);CHKERRQ(ierr); 2314 ierr = MatMPISBAIJSetPreallocation(A,bs,0,dlens,0,odlens);CHKERRQ(ierr); 2315 ierr = MatSetOption(A,MAT_COLUMNS_SORTED);CHKERRQ(ierr); 2316 2317 if (!rank) { 2318 ierr = PetscMalloc(maxnz*sizeof(PetscScalar),&buf);CHKERRQ(ierr); 2319 /* read in my part of the matrix numerical values */ 2320 nz = procsnz[0]; 2321 vals = buf; 2322 mycols = ibuf; 2323 if (size == 1) nz -= extra_rows; 2324 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2325 if (size == 1) for (i=0; i< extra_rows; i++) { vals[nz+i] = 1.0; } 2326 2327 /* insert into matrix */ 2328 jj = rstart*bs; 2329 for (i=0; i<m; i++) { 2330 ierr = MatSetValues(A,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 2331 mycols += locrowlens[i]; 2332 vals += locrowlens[i]; 2333 jj++; 2334 } 2335 2336 /* read in other processors (except the last one) and ship out */ 2337 for (i=1; i<size-1; i++) { 2338 nz = procsnz[i]; 2339 vals = buf; 2340 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2341 ierr = MPI_Send(vals,nz,MPIU_SCALAR,i,A->tag,comm);CHKERRQ(ierr); 2342 } 2343 /* the last proc */ 2344 if (size != 1){ 2345 nz = procsnz[i] - extra_rows; 2346 vals = buf; 2347 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2348 for (i=0; i<extra_rows; i++) vals[nz+i] = 1.0; 2349 ierr = MPI_Send(vals,nz+extra_rows,MPIU_SCALAR,size-1,A->tag,comm);CHKERRQ(ierr); 2350 } 2351 ierr = PetscFree(procsnz);CHKERRQ(ierr); 2352 2353 } else { 2354 /* receive numeric values */ 2355 ierr = PetscMalloc(nz*sizeof(PetscScalar),&buf);CHKERRQ(ierr); 2356 2357 /* receive message of values*/ 2358 vals = buf; 2359 mycols = ibuf; 2360 ierr = MPI_Recv(vals,nz,MPIU_SCALAR,0,A->tag,comm,&status);CHKERRQ(ierr); 2361 ierr = MPI_Get_count(&status,MPIU_SCALAR,&maxnz);CHKERRQ(ierr); 2362 if (maxnz != nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file"); 2363 2364 /* insert into matrix */ 2365 jj = rstart*bs; 2366 for (i=0; i<m; i++) { 2367 ierr = MatSetValues_MPISBAIJ(A,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 2368 mycols += locrowlens[i]; 2369 vals += locrowlens[i]; 2370 jj++; 2371 } 2372 } 2373 2374 ierr = PetscFree(locrowlens);CHKERRQ(ierr); 2375 ierr = PetscFree(buf);CHKERRQ(ierr); 2376 ierr = PetscFree(ibuf);CHKERRQ(ierr); 2377 ierr = PetscFree(rowners);CHKERRQ(ierr); 2378 ierr = PetscFree(dlens);CHKERRQ(ierr); 2379 ierr = PetscFree(mask);CHKERRQ(ierr); 2380 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2381 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2382 *newmat = A; 2383 PetscFunctionReturn(0); 2384 } 2385 2386 #undef __FUNCT__ 2387 #define __FUNCT__ "MatMPISBAIJSetHashTableFactor" 2388 /*XXXXX@ 2389 MatMPISBAIJSetHashTableFactor - Sets the factor required to compute the size of the HashTable. 2390 2391 Input Parameters: 2392 . mat - the matrix 2393 . fact - factor 2394 2395 Collective on Mat 2396 2397 Level: advanced 2398 2399 Notes: 2400 This can also be set by the command line option: -mat_use_hash_table fact 2401 2402 .keywords: matrix, hashtable, factor, HT 2403 2404 .seealso: MatSetOption() 2405 @XXXXX*/ 2406 2407 2408 #undef __FUNCT__ 2409 #define __FUNCT__ "MatGetRowMax_MPISBAIJ" 2410 PetscErrorCode MatGetRowMax_MPISBAIJ(Mat A,Vec v) 2411 { 2412 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 2413 Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(a->B)->data; 2414 PetscReal atmp; 2415 PetscReal *work,*svalues,*rvalues; 2416 PetscErrorCode ierr; 2417 PetscInt i,bs,mbs,*bi,*bj,brow,j,ncols,krow,kcol,col,row,Mbs,bcol; 2418 PetscMPIInt rank,size; 2419 PetscInt *rowners_bs,dest,count,source; 2420 PetscScalar *va; 2421 MatScalar *ba; 2422 MPI_Status stat; 2423 2424 PetscFunctionBegin; 2425 ierr = MatGetRowMax(a->A,v);CHKERRQ(ierr); 2426 ierr = VecGetArray(v,&va);CHKERRQ(ierr); 2427 2428 ierr = MPI_Comm_size(A->comm,&size);CHKERRQ(ierr); 2429 ierr = MPI_Comm_rank(A->comm,&rank);CHKERRQ(ierr); 2430 2431 bs = A->rmap.bs; 2432 mbs = a->mbs; 2433 Mbs = a->Mbs; 2434 ba = b->a; 2435 bi = b->i; 2436 bj = b->j; 2437 2438 /* find ownerships */ 2439 rowners_bs = A->rmap.range; 2440 2441 /* each proc creates an array to be distributed */ 2442 ierr = PetscMalloc(bs*Mbs*sizeof(PetscReal),&work);CHKERRQ(ierr); 2443 ierr = PetscMemzero(work,bs*Mbs*sizeof(PetscReal));CHKERRQ(ierr); 2444 2445 /* row_max for B */ 2446 if (rank != size-1){ 2447 for (i=0; i<mbs; i++) { 2448 ncols = bi[1] - bi[0]; bi++; 2449 brow = bs*i; 2450 for (j=0; j<ncols; j++){ 2451 bcol = bs*(*bj); 2452 for (kcol=0; kcol<bs; kcol++){ 2453 col = bcol + kcol; /* local col index */ 2454 col += rowners_bs[rank+1]; /* global col index */ 2455 for (krow=0; krow<bs; krow++){ 2456 atmp = PetscAbsScalar(*ba); ba++; 2457 row = brow + krow; /* local row index */ 2458 if (PetscRealPart(va[row]) < atmp) va[row] = atmp; 2459 if (work[col] < atmp) work[col] = atmp; 2460 } 2461 } 2462 bj++; 2463 } 2464 } 2465 2466 /* send values to its owners */ 2467 for (dest=rank+1; dest<size; dest++){ 2468 svalues = work + rowners_bs[dest]; 2469 count = rowners_bs[dest+1]-rowners_bs[dest]; 2470 ierr = MPI_Send(svalues,count,MPIU_REAL,dest,rank,A->comm);CHKERRQ(ierr); 2471 } 2472 } 2473 2474 /* receive values */ 2475 if (rank){ 2476 rvalues = work; 2477 count = rowners_bs[rank+1]-rowners_bs[rank]; 2478 for (source=0; source<rank; source++){ 2479 ierr = MPI_Recv(rvalues,count,MPIU_REAL,MPI_ANY_SOURCE,MPI_ANY_TAG,A->comm,&stat);CHKERRQ(ierr); 2480 /* process values */ 2481 for (i=0; i<count; i++){ 2482 if (PetscRealPart(va[i]) < rvalues[i]) va[i] = rvalues[i]; 2483 } 2484 } 2485 } 2486 2487 ierr = VecRestoreArray(v,&va);CHKERRQ(ierr); 2488 ierr = PetscFree(work);CHKERRQ(ierr); 2489 PetscFunctionReturn(0); 2490 } 2491 2492 #undef __FUNCT__ 2493 #define __FUNCT__ "MatRelax_MPISBAIJ" 2494 PetscErrorCode MatRelax_MPISBAIJ(Mat matin,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx) 2495 { 2496 Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data; 2497 PetscErrorCode ierr; 2498 PetscInt mbs=mat->mbs,bs=matin->rmap.bs; 2499 PetscScalar *x,*b,*ptr,zero=0.0; 2500 Vec bb1; 2501 2502 PetscFunctionBegin; 2503 if (its <= 0 || lits <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits); 2504 if (bs > 1) 2505 SETERRQ(PETSC_ERR_SUP,"SSOR for block size > 1 is not yet implemented"); 2506 2507 if ((flag & SOR_LOCAL_SYMMETRIC_SWEEP) == SOR_LOCAL_SYMMETRIC_SWEEP){ 2508 if ( flag & SOR_ZERO_INITIAL_GUESS ) { 2509 ierr = (*mat->A->ops->relax)(mat->A,bb,omega,flag,fshift,lits,lits,xx);CHKERRQ(ierr); 2510 its--; 2511 } 2512 2513 ierr = VecDuplicate(bb,&bb1);CHKERRQ(ierr); 2514 while (its--){ 2515 2516 /* lower triangular part: slvec0b = - B^T*xx */ 2517 ierr = (*mat->B->ops->multtranspose)(mat->B,xx,mat->slvec0b);CHKERRQ(ierr); 2518 2519 /* copy xx into slvec0a */ 2520 ierr = VecGetArray(mat->slvec0,&ptr);CHKERRQ(ierr); 2521 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 2522 ierr = PetscMemcpy(ptr,x,bs*mbs*sizeof(MatScalar));CHKERRQ(ierr); 2523 ierr = VecRestoreArray(mat->slvec0,&ptr);CHKERRQ(ierr); 2524 2525 ierr = VecScale(mat->slvec0,-1.0);CHKERRQ(ierr); 2526 2527 /* copy bb into slvec1a */ 2528 ierr = VecGetArray(mat->slvec1,&ptr);CHKERRQ(ierr); 2529 ierr = VecGetArray(bb,&b);CHKERRQ(ierr); 2530 ierr = PetscMemcpy(ptr,b,bs*mbs*sizeof(MatScalar));CHKERRQ(ierr); 2531 ierr = VecRestoreArray(mat->slvec1,&ptr);CHKERRQ(ierr); 2532 2533 /* set slvec1b = 0 */ 2534 ierr = VecSet(mat->slvec1b,zero);CHKERRQ(ierr); 2535 2536 ierr = VecScatterBegin(mat->slvec0,mat->slvec1,ADD_VALUES,SCATTER_FORWARD,mat->sMvctx);CHKERRQ(ierr); 2537 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 2538 ierr = VecRestoreArray(bb,&b);CHKERRQ(ierr); 2539 ierr = VecScatterEnd(mat->slvec0,mat->slvec1,ADD_VALUES,SCATTER_FORWARD,mat->sMvctx);CHKERRQ(ierr); 2540 2541 /* upper triangular part: bb1 = bb1 - B*x */ 2542 ierr = (*mat->B->ops->multadd)(mat->B,mat->slvec1b,mat->slvec1a,bb1);CHKERRQ(ierr); 2543 2544 /* local diagonal sweep */ 2545 ierr = (*mat->A->ops->relax)(mat->A,bb1,omega,SOR_SYMMETRIC_SWEEP,fshift,lits,lits,xx);CHKERRQ(ierr); 2546 } 2547 ierr = VecDestroy(bb1);CHKERRQ(ierr); 2548 } else { 2549 SETERRQ(PETSC_ERR_SUP,"MatSORType is not supported for SBAIJ matrix format"); 2550 } 2551 PetscFunctionReturn(0); 2552 } 2553 2554 #undef __FUNCT__ 2555 #define __FUNCT__ "MatRelax_MPISBAIJ_2comm" 2556 PetscErrorCode MatRelax_MPISBAIJ_2comm(Mat matin,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx) 2557 { 2558 Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data; 2559 PetscErrorCode ierr; 2560 Vec lvec1,bb1; 2561 2562 PetscFunctionBegin; 2563 if (its <= 0 || lits <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %D and local its %D both positive",its,lits); 2564 if (matin->rmap.bs > 1) 2565 SETERRQ(PETSC_ERR_SUP,"SSOR for block size > 1 is not yet implemented"); 2566 2567 if ((flag & SOR_LOCAL_SYMMETRIC_SWEEP) == SOR_LOCAL_SYMMETRIC_SWEEP){ 2568 if ( flag & SOR_ZERO_INITIAL_GUESS ) { 2569 ierr = (*mat->A->ops->relax)(mat->A,bb,omega,flag,fshift,lits,lits,xx);CHKERRQ(ierr); 2570 its--; 2571 } 2572 2573 ierr = VecDuplicate(mat->lvec,&lvec1);CHKERRQ(ierr); 2574 ierr = VecDuplicate(bb,&bb1);CHKERRQ(ierr); 2575 while (its--){ 2576 ierr = VecScatterBegin(xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD,mat->Mvctx);CHKERRQ(ierr); 2577 2578 /* lower diagonal part: bb1 = bb - B^T*xx */ 2579 ierr = (*mat->B->ops->multtranspose)(mat->B,xx,lvec1);CHKERRQ(ierr); 2580 ierr = VecScale(lvec1,-1.0);CHKERRQ(ierr); 2581 2582 ierr = VecScatterEnd(xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD,mat->Mvctx);CHKERRQ(ierr); 2583 ierr = VecCopy(bb,bb1);CHKERRQ(ierr); 2584 ierr = VecScatterBegin(lvec1,bb1,ADD_VALUES,SCATTER_REVERSE,mat->Mvctx);CHKERRQ(ierr); 2585 2586 /* upper diagonal part: bb1 = bb1 - B*x */ 2587 ierr = VecScale(mat->lvec,-1.0);CHKERRQ(ierr); 2588 ierr = (*mat->B->ops->multadd)(mat->B,mat->lvec,bb1,bb1);CHKERRQ(ierr); 2589 2590 ierr = VecScatterEnd(lvec1,bb1,ADD_VALUES,SCATTER_REVERSE,mat->Mvctx);CHKERRQ(ierr); 2591 2592 /* diagonal sweep */ 2593 ierr = (*mat->A->ops->relax)(mat->A,bb1,omega,SOR_SYMMETRIC_SWEEP,fshift,lits,lits,xx);CHKERRQ(ierr); 2594 } 2595 ierr = VecDestroy(lvec1);CHKERRQ(ierr); 2596 ierr = VecDestroy(bb1);CHKERRQ(ierr); 2597 } else { 2598 SETERRQ(PETSC_ERR_SUP,"MatSORType is not supported for SBAIJ matrix format"); 2599 } 2600 PetscFunctionReturn(0); 2601 } 2602 2603