1 /*$Id: mpisbaij.c,v 1.61 2001/08/10 03:31:37 bsmith Exp $*/ 2 3 #include "src/mat/impls/baij/mpi/mpibaij.h" /*I "petscmat.h" I*/ 4 #include "src/vec/vecimpl.h" 5 #include "mpisbaij.h" 6 #include "src/mat/impls/sbaij/seq/sbaij.h" 7 8 extern int MatSetUpMultiply_MPISBAIJ(Mat); 9 extern int MatSetUpMultiply_MPISBAIJ_2comm(Mat); 10 extern int DisAssemble_MPISBAIJ(Mat); 11 extern int MatGetValues_SeqSBAIJ(Mat,int,const int[],int,const int[],PetscScalar []); 12 extern int MatGetValues_SeqBAIJ(Mat,int,const int[],int,const int[],PetscScalar []); 13 extern int MatSetValues_SeqSBAIJ(Mat,int,const int [],int,const int [],const PetscScalar [],InsertMode); 14 extern int MatSetValuesBlocked_SeqSBAIJ(Mat,int,const int[],int,const int[],const PetscScalar[],InsertMode); 15 extern int MatSetValuesBlocked_SeqBAIJ(Mat,int,const int[],int,const int[],const PetscScalar[],InsertMode); 16 extern int MatGetRow_SeqSBAIJ(Mat,int,int*,int**,PetscScalar**); 17 extern int MatRestoreRow_SeqSBAIJ(Mat,int,int*,int**,PetscScalar**); 18 extern int MatPrintHelp_SeqSBAIJ(Mat); 19 extern int MatZeroRows_SeqSBAIJ(Mat,IS,PetscScalar*); 20 extern int MatZeroRows_SeqBAIJ(Mat,IS,PetscScalar *); 21 extern int MatGetRowMax_MPISBAIJ(Mat,Vec); 22 extern int MatRelax_MPISBAIJ(Mat,Vec,PetscReal,MatSORType,PetscReal,int,int,Vec); 23 24 /* UGLY, ugly, ugly 25 When MatScalar == PetscScalar the function MatSetValuesBlocked_MPIBAIJ_MatScalar() does 26 not exist. Otherwise ..._MatScalar() takes matrix elements in single precision and 27 inserts them into the single precision data structure. The function MatSetValuesBlocked_MPIBAIJ() 28 converts the entries into single precision and then calls ..._MatScalar() to put them 29 into the single precision data structures. 30 */ 31 #if defined(PETSC_USE_MAT_SINGLE) 32 extern int MatSetValuesBlocked_SeqSBAIJ_MatScalar(Mat,int,const int[],int,const int[],const MatScalar[],InsertMode); 33 extern int MatSetValues_MPISBAIJ_MatScalar(Mat,int,const int[],int,const int[],const MatScalar[],InsertMode); 34 extern int MatSetValuesBlocked_MPISBAIJ_MatScalar(Mat,int,const int[],int,const int[],const MatScalar[],InsertMode); 35 extern int MatSetValues_MPISBAIJ_HT_MatScalar(Mat,int,const int[],int,const int[],const MatScalar[],InsertMode); 36 extern int MatSetValuesBlocked_MPISBAIJ_HT_MatScalar(Mat,int,const int[],int,const int[],const MatScalar[],InsertMode); 37 #else 38 #define MatSetValuesBlocked_SeqSBAIJ_MatScalar MatSetValuesBlocked_SeqSBAIJ 39 #define MatSetValues_MPISBAIJ_MatScalar MatSetValues_MPISBAIJ 40 #define MatSetValuesBlocked_MPISBAIJ_MatScalar MatSetValuesBlocked_MPISBAIJ 41 #define MatSetValues_MPISBAIJ_HT_MatScalar MatSetValues_MPISBAIJ_HT 42 #define MatSetValuesBlocked_MPISBAIJ_HT_MatScalar MatSetValuesBlocked_MPISBAIJ_HT 43 #endif 44 45 EXTERN_C_BEGIN 46 #undef __FUNCT__ 47 #define __FUNCT__ "MatStoreValues_MPISBAIJ" 48 int MatStoreValues_MPISBAIJ(Mat mat) 49 { 50 Mat_MPISBAIJ *aij = (Mat_MPISBAIJ *)mat->data; 51 int ierr; 52 53 PetscFunctionBegin; 54 ierr = MatStoreValues(aij->A);CHKERRQ(ierr); 55 ierr = MatStoreValues(aij->B);CHKERRQ(ierr); 56 PetscFunctionReturn(0); 57 } 58 EXTERN_C_END 59 60 EXTERN_C_BEGIN 61 #undef __FUNCT__ 62 #define __FUNCT__ "MatRetrieveValues_MPISBAIJ" 63 int MatRetrieveValues_MPISBAIJ(Mat mat) 64 { 65 Mat_MPISBAIJ *aij = (Mat_MPISBAIJ *)mat->data; 66 int ierr; 67 68 PetscFunctionBegin; 69 ierr = MatRetrieveValues(aij->A);CHKERRQ(ierr); 70 ierr = MatRetrieveValues(aij->B);CHKERRQ(ierr); 71 PetscFunctionReturn(0); 72 } 73 EXTERN_C_END 74 75 /* 76 Local utility routine that creates a mapping from the global column 77 number to the local number in the off-diagonal part of the local 78 storage of the matrix. This is done in a non scable way since the 79 length of colmap equals the global matrix length. 80 */ 81 #undef __FUNCT__ 82 #define __FUNCT__ "CreateColmap_MPISBAIJ_Private" 83 static int CreateColmap_MPISBAIJ_Private(Mat mat) 84 { 85 PetscFunctionBegin; 86 SETERRQ(1,"Function not yet written for SBAIJ format"); 87 /* PetscFunctionReturn(0); */ 88 } 89 90 #define CHUNKSIZE 10 91 92 #define MatSetValues_SeqSBAIJ_A_Private(row,col,value,addv) \ 93 { \ 94 \ 95 brow = row/bs; \ 96 rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; \ 97 rmax = aimax[brow]; nrow = ailen[brow]; \ 98 bcol = col/bs; \ 99 ridx = row % bs; cidx = col % bs; \ 100 low = 0; high = nrow; \ 101 while (high-low > 3) { \ 102 t = (low+high)/2; \ 103 if (rp[t] > bcol) high = t; \ 104 else low = t; \ 105 } \ 106 for (_i=low; _i<high; _i++) { \ 107 if (rp[_i] > bcol) break; \ 108 if (rp[_i] == bcol) { \ 109 bap = ap + bs2*_i + bs*cidx + ridx; \ 110 if (addv == ADD_VALUES) *bap += value; \ 111 else *bap = value; \ 112 goto a_noinsert; \ 113 } \ 114 } \ 115 if (a->nonew == 1) goto a_noinsert; \ 116 else if (a->nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%d, %d) into matrix", row, col); \ 117 if (nrow >= rmax) { \ 118 /* there is no extra room in row, therefore enlarge */ \ 119 int new_nz = ai[a->mbs] + CHUNKSIZE,len,*new_i,*new_j; \ 120 MatScalar *new_a; \ 121 \ 122 if (a->nonew == -2) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%d, %d) in the matrix", row, col); \ 123 \ 124 /* malloc new storage space */ \ 125 len = new_nz*(sizeof(int)+bs2*sizeof(MatScalar))+(a->mbs+1)*sizeof(int); \ 126 ierr = PetscMalloc(len,&new_a);CHKERRQ(ierr); \ 127 new_j = (int*)(new_a + bs2*new_nz); \ 128 new_i = new_j + new_nz; \ 129 \ 130 /* copy over old data into new slots */ \ 131 for (ii=0; ii<brow+1; ii++) {new_i[ii] = ai[ii];} \ 132 for (ii=brow+1; ii<a->mbs+1; ii++) {new_i[ii] = ai[ii]+CHUNKSIZE;} \ 133 ierr = PetscMemcpy(new_j,aj,(ai[brow]+nrow)*sizeof(int));CHKERRQ(ierr); \ 134 len = (new_nz - CHUNKSIZE - ai[brow] - nrow); \ 135 ierr = PetscMemcpy(new_j+ai[brow]+nrow+CHUNKSIZE,aj+ai[brow]+nrow,len*sizeof(int));CHKERRQ(ierr); \ 136 ierr = PetscMemcpy(new_a,aa,(ai[brow]+nrow)*bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 137 ierr = PetscMemzero(new_a+bs2*(ai[brow]+nrow),bs2*CHUNKSIZE*sizeof(PetscScalar));CHKERRQ(ierr); \ 138 ierr = PetscMemcpy(new_a+bs2*(ai[brow]+nrow+CHUNKSIZE), \ 139 aa+bs2*(ai[brow]+nrow),bs2*len*sizeof(MatScalar));CHKERRQ(ierr); \ 140 /* free up old matrix storage */ \ 141 ierr = PetscFree(a->a);CHKERRQ(ierr); \ 142 if (!a->singlemalloc) { \ 143 ierr = PetscFree(a->i);CHKERRQ(ierr); \ 144 ierr = PetscFree(a->j);CHKERRQ(ierr);\ 145 } \ 146 aa = a->a = new_a; ai = a->i = new_i; aj = a->j = new_j; \ 147 a->singlemalloc = PETSC_TRUE; \ 148 \ 149 rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; \ 150 rmax = aimax[brow] = aimax[brow] + CHUNKSIZE; \ 151 PetscLogObjectMemory(A,CHUNKSIZE*(sizeof(int) + bs2*sizeof(MatScalar))); \ 152 a->s_maxnz += bs2*CHUNKSIZE; \ 153 a->reallocs++; \ 154 a->s_nz++; \ 155 } \ 156 N = nrow++ - 1; \ 157 /* shift up all the later entries in this row */ \ 158 for (ii=N; ii>=_i; ii--) { \ 159 rp[ii+1] = rp[ii]; \ 160 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 161 } \ 162 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr); } \ 163 rp[_i] = bcol; \ 164 ap[bs2*_i + bs*cidx + ridx] = value; \ 165 a_noinsert:; \ 166 ailen[brow] = nrow; \ 167 } 168 #ifndef MatSetValues_SeqBAIJ_B_Private 169 #define MatSetValues_SeqSBAIJ_B_Private(row,col,value,addv) \ 170 { \ 171 brow = row/bs; \ 172 rp = bj + bi[brow]; ap = ba + bs2*bi[brow]; \ 173 rmax = bimax[brow]; nrow = bilen[brow]; \ 174 bcol = col/bs; \ 175 ridx = row % bs; cidx = col % bs; \ 176 low = 0; high = nrow; \ 177 while (high-low > 3) { \ 178 t = (low+high)/2; \ 179 if (rp[t] > bcol) high = t; \ 180 else low = t; \ 181 } \ 182 for (_i=low; _i<high; _i++) { \ 183 if (rp[_i] > bcol) break; \ 184 if (rp[_i] == bcol) { \ 185 bap = ap + bs2*_i + bs*cidx + ridx; \ 186 if (addv == ADD_VALUES) *bap += value; \ 187 else *bap = value; \ 188 goto b_noinsert; \ 189 } \ 190 } \ 191 if (b->nonew == 1) goto b_noinsert; \ 192 else if (b->nonew == -1) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%d, %d) into matrix", row, col); \ 193 if (nrow >= rmax) { \ 194 /* there is no extra room in row, therefore enlarge */ \ 195 int new_nz = bi[b->mbs] + CHUNKSIZE,len,*new_i,*new_j; \ 196 MatScalar *new_a; \ 197 \ 198 if (b->nonew == -2) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%d, %d) in the matrix", row, col); \ 199 \ 200 /* malloc new storage space */ \ 201 len = new_nz*(sizeof(int)+bs2*sizeof(MatScalar))+(b->mbs+1)*sizeof(int); \ 202 ierr = PetscMalloc(len,&new_a);CHKERRQ(ierr); \ 203 new_j = (int*)(new_a + bs2*new_nz); \ 204 new_i = new_j + new_nz; \ 205 \ 206 /* copy over old data into new slots */ \ 207 for (ii=0; ii<brow+1; ii++) {new_i[ii] = bi[ii];} \ 208 for (ii=brow+1; ii<b->mbs+1; ii++) {new_i[ii] = bi[ii]+CHUNKSIZE;} \ 209 ierr = PetscMemcpy(new_j,bj,(bi[brow]+nrow)*sizeof(int));CHKERRQ(ierr); \ 210 len = (new_nz - CHUNKSIZE - bi[brow] - nrow); \ 211 ierr = PetscMemcpy(new_j+bi[brow]+nrow+CHUNKSIZE,bj+bi[brow]+nrow,len*sizeof(int));CHKERRQ(ierr); \ 212 ierr = PetscMemcpy(new_a,ba,(bi[brow]+nrow)*bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 213 ierr = PetscMemzero(new_a+bs2*(bi[brow]+nrow),bs2*CHUNKSIZE*sizeof(MatScalar));CHKERRQ(ierr); \ 214 ierr = PetscMemcpy(new_a+bs2*(bi[brow]+nrow+CHUNKSIZE), \ 215 ba+bs2*(bi[brow]+nrow),bs2*len*sizeof(MatScalar));CHKERRQ(ierr); \ 216 /* free up old matrix storage */ \ 217 ierr = PetscFree(b->a);CHKERRQ(ierr); \ 218 if (!b->singlemalloc) { \ 219 ierr = PetscFree(b->i);CHKERRQ(ierr); \ 220 ierr = PetscFree(b->j);CHKERRQ(ierr); \ 221 } \ 222 ba = b->a = new_a; bi = b->i = new_i; bj = b->j = new_j; \ 223 b->singlemalloc = PETSC_TRUE; \ 224 \ 225 rp = bj + bi[brow]; ap = ba + bs2*bi[brow]; \ 226 rmax = bimax[brow] = bimax[brow] + CHUNKSIZE; \ 227 PetscLogObjectMemory(B,CHUNKSIZE*(sizeof(int) + bs2*sizeof(MatScalar))); \ 228 b->maxnz += bs2*CHUNKSIZE; \ 229 b->reallocs++; \ 230 b->nz++; \ 231 } \ 232 N = nrow++ - 1; \ 233 /* shift up all the later entries in this row */ \ 234 for (ii=N; ii>=_i; ii--) { \ 235 rp[ii+1] = rp[ii]; \ 236 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 237 } \ 238 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr);} \ 239 rp[_i] = bcol; \ 240 ap[bs2*_i + bs*cidx + ridx] = value; \ 241 b_noinsert:; \ 242 bilen[brow] = nrow; \ 243 } 244 #endif 245 246 #if defined(PETSC_USE_MAT_SINGLE) 247 #undef __FUNCT__ 248 #define __FUNCT__ "MatSetValues_MPISBAIJ" 249 int MatSetValues_MPISBAIJ(Mat mat,int m,const int im[],int n,const int in[],const PetscScalar v[],InsertMode addv) 250 { 251 Mat_MPISBAIJ *b = (Mat_MPISBAIJ*)mat->data; 252 int ierr,i,N = m*n; 253 MatScalar *vsingle; 254 255 PetscFunctionBegin; 256 if (N > b->setvalueslen) { 257 if (b->setvaluescopy) {ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr);} 258 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 259 b->setvalueslen = N; 260 } 261 vsingle = b->setvaluescopy; 262 263 for (i=0; i<N; i++) { 264 vsingle[i] = v[i]; 265 } 266 ierr = MatSetValues_MPISBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 267 PetscFunctionReturn(0); 268 } 269 270 #undef __FUNCT__ 271 #define __FUNCT__ "MatSetValuesBlocked_MPISBAIJ" 272 int MatSetValuesBlocked_MPISBAIJ(Mat mat,int m,const int im[],int n,const int in[],const PetscScalar v[],InsertMode addv) 273 { 274 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data; 275 int ierr,i,N = m*n*b->bs2; 276 MatScalar *vsingle; 277 278 PetscFunctionBegin; 279 if (N > b->setvalueslen) { 280 if (b->setvaluescopy) {ierr = PetscFree(b->setvaluescopy);CHKERRQ(ierr);} 281 ierr = PetscMalloc(N*sizeof(MatScalar),&b->setvaluescopy);CHKERRQ(ierr); 282 b->setvalueslen = N; 283 } 284 vsingle = b->setvaluescopy; 285 for (i=0; i<N; i++) { 286 vsingle[i] = v[i]; 287 } 288 ierr = MatSetValuesBlocked_MPISBAIJ_MatScalar(mat,m,im,n,in,vsingle,addv);CHKERRQ(ierr); 289 PetscFunctionReturn(0); 290 } 291 292 #undef __FUNCT__ 293 #define __FUNCT__ "MatSetValues_MPISBAIJ_HT" 294 int MatSetValues_MPISBAIJ_HT(Mat mat,int m,const int im[],int n,const int in[],const PetscScalar v[],InsertMode addv) 295 { 296 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data; 297 int ierr,i,N = m*n; 298 MatScalar *vsingle; 299 300 PetscFunctionBegin; 301 SETERRQ(1,"Function not yet written for SBAIJ format"); 302 /* PetscFunctionReturn(0); */ 303 } 304 305 #undef __FUNCT__ 306 #define __FUNCT__ "MatSetValuesBlocked_MPISBAIJ_HT" 307 int MatSetValuesBlocked_MPISBAIJ_HT(Mat mat,int m,const int im[],int n,const int in[],const PetscScalar v[],InsertMode addv) 308 { 309 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)mat->data; 310 int ierr,i,N = m*n*b->bs2; 311 MatScalar *vsingle; 312 313 PetscFunctionBegin; 314 SETERRQ(1,"Function not yet written for SBAIJ format"); 315 /* PetscFunctionReturn(0); */ 316 } 317 #endif 318 319 /* Only add/insert a(i,j) with i<=j (blocks). 320 Any a(i,j) with i>j input by user is ingored. 321 */ 322 #undef __FUNCT__ 323 #define __FUNCT__ "MatSetValues_MPIBAIJ_MatScalar" 324 int MatSetValues_MPISBAIJ_MatScalar(Mat mat,int m,const int im[],int n,const int in[],const MatScalar v[],InsertMode addv) 325 { 326 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 327 MatScalar value; 328 PetscTruth roworiented = baij->roworiented; 329 int ierr,i,j,row,col; 330 int rstart_orig=baij->rstart_bs; 331 int rend_orig=baij->rend_bs,cstart_orig=baij->cstart_bs; 332 int cend_orig=baij->cend_bs,bs=baij->bs; 333 334 /* Some Variables required in the macro */ 335 Mat A = baij->A; 336 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)(A)->data; 337 int *aimax=a->imax,*ai=a->i,*ailen=a->ilen,*aj=a->j; 338 MatScalar *aa=a->a; 339 340 Mat B = baij->B; 341 Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(B)->data; 342 int *bimax=b->imax,*bi=b->i,*bilen=b->ilen,*bj=b->j; 343 MatScalar *ba=b->a; 344 345 int *rp,ii,nrow,_i,rmax,N,brow,bcol; 346 int low,high,t,ridx,cidx,bs2=a->bs2; 347 MatScalar *ap,*bap; 348 349 /* for stash */ 350 int n_loc, *in_loc=0; 351 MatScalar *v_loc=0; 352 353 PetscFunctionBegin; 354 355 if(!baij->donotstash){ 356 ierr = PetscMalloc(n*sizeof(int),&in_loc);CHKERRQ(ierr); 357 ierr = PetscMalloc(n*sizeof(MatScalar),&v_loc);CHKERRQ(ierr); 358 } 359 360 for (i=0; i<m; i++) { 361 if (im[i] < 0) continue; 362 #if defined(PETSC_USE_BOPT_g) 363 if (im[i] >= mat->M) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %d max %d",im[i],mat->M-1); 364 #endif 365 if (im[i] >= rstart_orig && im[i] < rend_orig) { /* this processor entry */ 366 row = im[i] - rstart_orig; /* local row index */ 367 for (j=0; j<n; j++) { 368 if (im[i]/bs > in[j]/bs) continue; /* ignore lower triangular blocks */ 369 if (in[j] >= cstart_orig && in[j] < cend_orig){ /* diag entry (A) */ 370 col = in[j] - cstart_orig; /* local col index */ 371 brow = row/bs; bcol = col/bs; 372 if (brow > bcol) continue; /* ignore lower triangular blocks of A */ 373 if (roworiented) value = v[i*n+j]; else value = v[i+j*m]; 374 MatSetValues_SeqSBAIJ_A_Private(row,col,value,addv); 375 /* ierr = MatSetValues_SeqBAIJ(baij->A,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 376 } else if (in[j] < 0) continue; 377 #if defined(PETSC_USE_BOPT_g) 378 else if (in[j] >= mat->N) {SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %d max %d",in[j],mat->N-1);} 379 #endif 380 else { /* off-diag entry (B) */ 381 if (mat->was_assembled) { 382 if (!baij->colmap) { 383 ierr = CreateColmap_MPISBAIJ_Private(mat);CHKERRQ(ierr); 384 } 385 #if defined (PETSC_USE_CTABLE) 386 ierr = PetscTableFind(baij->colmap,in[j]/bs + 1,&col);CHKERRQ(ierr); 387 col = col - 1; 388 #else 389 col = baij->colmap[in[j]/bs] - 1; 390 #endif 391 if (col < 0 && !((Mat_SeqSBAIJ*)(baij->A->data))->nonew) { 392 ierr = DisAssemble_MPISBAIJ(mat);CHKERRQ(ierr); 393 col = in[j]; 394 /* Reinitialize the variables required by MatSetValues_SeqBAIJ_B_Private() */ 395 B = baij->B; 396 b = (Mat_SeqBAIJ*)(B)->data; 397 bimax=b->imax;bi=b->i;bilen=b->ilen;bj=b->j; 398 ba=b->a; 399 } else col += in[j]%bs; 400 } else col = in[j]; 401 if (roworiented) value = v[i*n+j]; else value = v[i+j*m]; 402 MatSetValues_SeqSBAIJ_B_Private(row,col,value,addv); 403 /* ierr = MatSetValues_SeqBAIJ(baij->B,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 404 } 405 } 406 } else { /* off processor entry */ 407 if (!baij->donotstash) { 408 n_loc = 0; 409 for (j=0; j<n; j++){ 410 if (im[i]/bs > in[j]/bs) continue; /* ignore lower triangular blocks */ 411 in_loc[n_loc] = in[j]; 412 if (roworiented) { 413 v_loc[n_loc] = v[i*n+j]; 414 } else { 415 v_loc[n_loc] = v[j*m+i]; 416 } 417 n_loc++; 418 } 419 ierr = MatStashValuesRow_Private(&mat->stash,im[i],n_loc,in_loc,v_loc);CHKERRQ(ierr); 420 } 421 } 422 } 423 424 if(!baij->donotstash){ 425 ierr = PetscFree(in_loc);CHKERRQ(ierr); 426 ierr = PetscFree(v_loc);CHKERRQ(ierr); 427 } 428 PetscFunctionReturn(0); 429 } 430 431 #undef __FUNCT__ 432 #define __FUNCT__ "MatSetValuesBlocked_MPISBAIJ_MatScalar" 433 int MatSetValuesBlocked_MPISBAIJ_MatScalar(Mat mat,int m,const int im[],int n,const int in[],const MatScalar v[],InsertMode addv) 434 { 435 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 436 const MatScalar *value; 437 MatScalar *barray=baij->barray; 438 PetscTruth roworiented = baij->roworiented; 439 int ierr,i,j,ii,jj,row,col,rstart=baij->rstart; 440 int rend=baij->rend,cstart=baij->cstart,stepval; 441 int cend=baij->cend,bs=baij->bs,bs2=baij->bs2; 442 443 PetscFunctionBegin; 444 if(!barray) { 445 ierr = PetscMalloc(bs2*sizeof(MatScalar),&barray);CHKERRQ(ierr); 446 baij->barray = barray; 447 } 448 449 if (roworiented) { 450 stepval = (n-1)*bs; 451 } else { 452 stepval = (m-1)*bs; 453 } 454 for (i=0; i<m; i++) { 455 if (im[i] < 0) continue; 456 #if defined(PETSC_USE_BOPT_g) 457 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large, row %d max %d",im[i],baij->Mbs-1); 458 #endif 459 if (im[i] >= rstart && im[i] < rend) { 460 row = im[i] - rstart; 461 for (j=0; j<n; j++) { 462 /* If NumCol = 1 then a copy is not required */ 463 if ((roworiented) && (n == 1)) { 464 barray = (MatScalar*) v + i*bs2; 465 } else if((!roworiented) && (m == 1)) { 466 barray = (MatScalar*) v + j*bs2; 467 } else { /* Here a copy is required */ 468 if (roworiented) { 469 value = v + i*(stepval+bs)*bs + j*bs; 470 } else { 471 value = v + j*(stepval+bs)*bs + i*bs; 472 } 473 for (ii=0; ii<bs; ii++,value+=stepval) { 474 for (jj=0; jj<bs; jj++) { 475 *barray++ = *value++; 476 } 477 } 478 barray -=bs2; 479 } 480 481 if (in[j] >= cstart && in[j] < cend){ 482 col = in[j] - cstart; 483 ierr = MatSetValuesBlocked_SeqSBAIJ(baij->A,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 484 } 485 else if (in[j] < 0) continue; 486 #if defined(PETSC_USE_BOPT_g) 487 else if (in[j] >= baij->Nbs) {SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large, col %d max %d",in[j],baij->Nbs-1);} 488 #endif 489 else { 490 if (mat->was_assembled) { 491 if (!baij->colmap) { 492 ierr = CreateColmap_MPISBAIJ_Private(mat);CHKERRQ(ierr); 493 } 494 495 #if defined(PETSC_USE_BOPT_g) 496 #if defined (PETSC_USE_CTABLE) 497 { int data; 498 ierr = PetscTableFind(baij->colmap,in[j]+1,&data);CHKERRQ(ierr); 499 if ((data - 1) % bs) SETERRQ(PETSC_ERR_PLIB,"Incorrect colmap"); 500 } 501 #else 502 if ((baij->colmap[in[j]] - 1) % bs) SETERRQ(PETSC_ERR_PLIB,"Incorrect colmap"); 503 #endif 504 #endif 505 #if defined (PETSC_USE_CTABLE) 506 ierr = PetscTableFind(baij->colmap,in[j]+1,&col);CHKERRQ(ierr); 507 col = (col - 1)/bs; 508 #else 509 col = (baij->colmap[in[j]] - 1)/bs; 510 #endif 511 if (col < 0 && !((Mat_SeqBAIJ*)(baij->A->data))->nonew) { 512 ierr = DisAssemble_MPISBAIJ(mat);CHKERRQ(ierr); 513 col = in[j]; 514 } 515 } 516 else col = in[j]; 517 ierr = MatSetValuesBlocked_SeqBAIJ(baij->B,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 518 } 519 } 520 } else { 521 if (!baij->donotstash) { 522 if (roworiented) { 523 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 524 } else { 525 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 526 } 527 } 528 } 529 } 530 PetscFunctionReturn(0); 531 } 532 533 #define HASH_KEY 0.6180339887 534 #define HASH(size,key,tmp) (tmp = (key)*HASH_KEY,(int)((size)*(tmp-(int)tmp))) 535 /* #define HASH(size,key) ((int)((size)*fmod(((key)*HASH_KEY),1))) */ 536 /* #define HASH(size,key,tmp) ((int)((size)*fmod(((key)*HASH_KEY),1))) */ 537 #undef __FUNCT__ 538 #define __FUNCT__ "MatSetValues_MPISBAIJ_HT_MatScalar" 539 int MatSetValues_MPISBAIJ_HT_MatScalar(Mat mat,int m,const int im[],int n,const int in[],const MatScalar v[],InsertMode addv) 540 { 541 PetscFunctionBegin; 542 SETERRQ(1,"Function not yet written for SBAIJ format"); 543 /* PetscFunctionReturn(0); */ 544 } 545 546 #undef __FUNCT__ 547 #define __FUNCT__ "MatSetValuesBlocked_MPISBAIJ_HT_MatScalar" 548 int MatSetValuesBlocked_MPISBAIJ_HT_MatScalar(Mat mat,int m,const int im[],int n,const int in[],const MatScalar v[],InsertMode addv) 549 { 550 PetscFunctionBegin; 551 SETERRQ(1,"Function not yet written for SBAIJ format"); 552 /* PetscFunctionReturn(0); */ 553 } 554 555 #undef __FUNCT__ 556 #define __FUNCT__ "MatGetValues_MPISBAIJ" 557 int MatGetValues_MPISBAIJ(Mat mat,int m,const int idxm[],int n,const int idxn[],PetscScalar v[]) 558 { 559 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 560 int bs=baij->bs,ierr,i,j,bsrstart = baij->rstart*bs,bsrend = baij->rend*bs; 561 int bscstart = baij->cstart*bs,bscend = baij->cend*bs,row,col,data; 562 563 PetscFunctionBegin; 564 for (i=0; i<m; i++) { 565 if (idxm[i] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %d",idxm[i]); 566 if (idxm[i] >= mat->M) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %d max %d",idxm[i],mat->M-1); 567 if (idxm[i] >= bsrstart && idxm[i] < bsrend) { 568 row = idxm[i] - bsrstart; 569 for (j=0; j<n; j++) { 570 if (idxn[j] < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"Negative column %d",idxn[j]); 571 if (idxn[j] >= mat->N) SETERRQ2(PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %d max %d",idxn[j],mat->N-1); 572 if (idxn[j] >= bscstart && idxn[j] < bscend){ 573 col = idxn[j] - bscstart; 574 ierr = MatGetValues_SeqSBAIJ(baij->A,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 575 } else { 576 if (!baij->colmap) { 577 ierr = CreateColmap_MPISBAIJ_Private(mat);CHKERRQ(ierr); 578 } 579 #if defined (PETSC_USE_CTABLE) 580 ierr = PetscTableFind(baij->colmap,idxn[j]/bs+1,&data);CHKERRQ(ierr); 581 data --; 582 #else 583 data = baij->colmap[idxn[j]/bs]-1; 584 #endif 585 if((data < 0) || (baij->garray[data/bs] != idxn[j]/bs)) *(v+i*n+j) = 0.0; 586 else { 587 col = data + idxn[j]%bs; 588 ierr = MatGetValues_SeqBAIJ(baij->B,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 589 } 590 } 591 } 592 } else { 593 SETERRQ(PETSC_ERR_SUP,"Only local values currently supported"); 594 } 595 } 596 PetscFunctionReturn(0); 597 } 598 599 #undef __FUNCT__ 600 #define __FUNCT__ "MatNorm_MPISBAIJ" 601 int MatNorm_MPISBAIJ(Mat mat,NormType type,PetscReal *norm) 602 { 603 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 604 /* Mat_SeqSBAIJ *amat = (Mat_SeqSBAIJ*)baij->A->data; */ 605 /* Mat_SeqBAIJ *bmat = (Mat_SeqBAIJ*)baij->B->data; */ 606 int ierr; 607 PetscReal sum[2],*lnorm2; 608 609 PetscFunctionBegin; 610 if (baij->size == 1) { 611 ierr = MatNorm(baij->A,type,norm);CHKERRQ(ierr); 612 } else { 613 if (type == NORM_FROBENIUS) { 614 ierr = PetscMalloc(2*sizeof(PetscReal),&lnorm2);CHKERRQ(ierr); 615 ierr = MatNorm(baij->A,type,lnorm2);CHKERRQ(ierr); 616 *lnorm2 = (*lnorm2)*(*lnorm2); lnorm2++; /* squar power of norm(A) */ 617 ierr = MatNorm(baij->B,type,lnorm2);CHKERRQ(ierr); 618 *lnorm2 = (*lnorm2)*(*lnorm2); lnorm2--; /* squar power of norm(B) */ 619 /* 620 ierr = MPI_Comm_rank(mat->comm,&rank);CHKERRQ(ierr); 621 PetscSynchronizedPrintf(mat->comm,"[%d], lnorm2=%g, %g\n",rank,lnorm2[0],lnorm2[1]); 622 */ 623 ierr = MPI_Allreduce(lnorm2,&sum,2,MPIU_REAL,MPI_SUM,mat->comm);CHKERRQ(ierr); 624 /* 625 PetscSynchronizedPrintf(mat->comm,"[%d], sum=%g, %g\n",rank,sum[0],sum[1]); 626 PetscSynchronizedFlush(mat->comm); */ 627 628 *norm = sqrt(sum[0] + 2*sum[1]); 629 ierr = PetscFree(lnorm2);CHKERRQ(ierr); 630 } else { 631 SETERRQ(PETSC_ERR_SUP,"No support for this norm yet"); 632 } 633 } 634 PetscFunctionReturn(0); 635 } 636 637 /* 638 Creates the hash table, and sets the table 639 This table is created only once. 640 If new entried need to be added to the matrix 641 then the hash table has to be destroyed and 642 recreated. 643 */ 644 #undef __FUNCT__ 645 #define __FUNCT__ "MatCreateHashTable_MPISBAIJ_Private" 646 int MatCreateHashTable_MPISBAIJ_Private(Mat mat,PetscReal factor) 647 { 648 PetscFunctionBegin; 649 SETERRQ(1,"Function not yet written for SBAIJ format"); 650 /* PetscFunctionReturn(0); */ 651 } 652 653 #undef __FUNCT__ 654 #define __FUNCT__ "MatAssemblyBegin_MPISBAIJ" 655 int MatAssemblyBegin_MPISBAIJ(Mat mat,MatAssemblyType mode) 656 { 657 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 658 int ierr,nstash,reallocs; 659 InsertMode addv; 660 661 PetscFunctionBegin; 662 if (baij->donotstash) { 663 PetscFunctionReturn(0); 664 } 665 666 /* make sure all processors are either in INSERTMODE or ADDMODE */ 667 ierr = MPI_Allreduce(&mat->insertmode,&addv,1,MPI_INT,MPI_BOR,mat->comm);CHKERRQ(ierr); 668 if (addv == (ADD_VALUES|INSERT_VALUES)) { 669 SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Some processors inserted others added"); 670 } 671 mat->insertmode = addv; /* in case this processor had no cache */ 672 673 ierr = MatStashScatterBegin_Private(&mat->stash,baij->rowners_bs);CHKERRQ(ierr); 674 ierr = MatStashScatterBegin_Private(&mat->bstash,baij->rowners);CHKERRQ(ierr); 675 ierr = MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);CHKERRQ(ierr); 676 PetscLogInfo(0,"MatAssemblyBegin_MPISBAIJ:Stash has %d entries,uses %d mallocs.\n",nstash,reallocs); 677 ierr = MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);CHKERRQ(ierr); 678 PetscLogInfo(0,"MatAssemblyBegin_MPISBAIJ:Block-Stash has %d entries, uses %d mallocs.\n",nstash,reallocs); 679 PetscFunctionReturn(0); 680 } 681 682 #undef __FUNCT__ 683 #define __FUNCT__ "MatAssemblyEnd_MPISBAIJ" 684 int MatAssemblyEnd_MPISBAIJ(Mat mat,MatAssemblyType mode) 685 { 686 Mat_MPISBAIJ *baij=(Mat_MPISBAIJ*)mat->data; 687 Mat_SeqSBAIJ *a=(Mat_SeqSBAIJ*)baij->A->data; 688 Mat_SeqBAIJ *b=(Mat_SeqBAIJ*)baij->B->data; 689 int i,j,rstart,ncols,n,ierr,flg,bs2=baij->bs2; 690 int *row,*col,other_disassembled; 691 PetscTruth r1,r2,r3; 692 MatScalar *val; 693 InsertMode addv = mat->insertmode; 694 695 PetscFunctionBegin; 696 697 if (!baij->donotstash) { 698 while (1) { 699 ierr = MatStashScatterGetMesg_Private(&mat->stash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 700 /* 701 PetscSynchronizedPrintf(mat->comm,"[%d]: in AssemblyEnd, stash, flg=%d\n",rank,flg); 702 PetscSynchronizedFlush(mat->comm); 703 */ 704 if (!flg) break; 705 706 for (i=0; i<n;) { 707 /* Now identify the consecutive vals belonging to the same row */ 708 for (j=i,rstart=row[j]; j<n; j++) { if (row[j] != rstart) break; } 709 if (j < n) ncols = j-i; 710 else ncols = n-i; 711 /* Now assemble all these values with a single function call */ 712 ierr = MatSetValues_MPISBAIJ_MatScalar(mat,1,row+i,ncols,col+i,val+i,addv);CHKERRQ(ierr); 713 i = j; 714 } 715 } 716 ierr = MatStashScatterEnd_Private(&mat->stash);CHKERRQ(ierr); 717 /* Now process the block-stash. Since the values are stashed column-oriented, 718 set the roworiented flag to column oriented, and after MatSetValues() 719 restore the original flags */ 720 r1 = baij->roworiented; 721 r2 = a->roworiented; 722 r3 = b->roworiented; 723 baij->roworiented = PETSC_FALSE; 724 a->roworiented = PETSC_FALSE; 725 b->roworiented = PETSC_FALSE; 726 while (1) { 727 ierr = MatStashScatterGetMesg_Private(&mat->bstash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 728 if (!flg) break; 729 730 for (i=0; i<n;) { 731 /* Now identify the consecutive vals belonging to the same row */ 732 for (j=i,rstart=row[j]; j<n; j++) { if (row[j] != rstart) break; } 733 if (j < n) ncols = j-i; 734 else ncols = n-i; 735 ierr = MatSetValuesBlocked_MPISBAIJ_MatScalar(mat,1,row+i,ncols,col+i,val+i*bs2,addv);CHKERRQ(ierr); 736 i = j; 737 } 738 } 739 ierr = MatStashScatterEnd_Private(&mat->bstash);CHKERRQ(ierr); 740 baij->roworiented = r1; 741 a->roworiented = r2; 742 b->roworiented = r3; 743 } 744 745 ierr = MatAssemblyBegin(baij->A,mode);CHKERRQ(ierr); 746 ierr = MatAssemblyEnd(baij->A,mode);CHKERRQ(ierr); 747 748 /* determine if any processor has disassembled, if so we must 749 also disassemble ourselfs, in order that we may reassemble. */ 750 /* 751 if nonzero structure of submatrix B cannot change then we know that 752 no processor disassembled thus we can skip this stuff 753 */ 754 if (!((Mat_SeqBAIJ*)baij->B->data)->nonew) { 755 ierr = MPI_Allreduce(&mat->was_assembled,&other_disassembled,1,MPI_INT,MPI_PROD,mat->comm);CHKERRQ(ierr); 756 if (mat->was_assembled && !other_disassembled) { 757 ierr = DisAssemble_MPISBAIJ(mat);CHKERRQ(ierr); 758 } 759 } 760 761 if (!mat->was_assembled && mode == MAT_FINAL_ASSEMBLY) { 762 ierr = MatSetUpMultiply_MPISBAIJ(mat);CHKERRQ(ierr); /* setup Mvctx and sMvctx */ 763 } 764 ierr = MatAssemblyBegin(baij->B,mode);CHKERRQ(ierr); 765 ierr = MatAssemblyEnd(baij->B,mode);CHKERRQ(ierr); 766 767 #if defined(PETSC_USE_BOPT_g) 768 if (baij->ht && mode== MAT_FINAL_ASSEMBLY) { 769 PetscLogInfo(0,"MatAssemblyEnd_MPISBAIJ:Average Hash Table Search in MatSetValues = %5.2f\n",((PetscReal)baij->ht_total_ct)/baij->ht_insert_ct); 770 baij->ht_total_ct = 0; 771 baij->ht_insert_ct = 0; 772 } 773 #endif 774 if (baij->ht_flag && !baij->ht && mode == MAT_FINAL_ASSEMBLY) { 775 ierr = MatCreateHashTable_MPISBAIJ_Private(mat,baij->ht_fact);CHKERRQ(ierr); 776 mat->ops->setvalues = MatSetValues_MPISBAIJ_HT; 777 mat->ops->setvaluesblocked = MatSetValuesBlocked_MPISBAIJ_HT; 778 } 779 780 if (baij->rowvalues) { 781 ierr = PetscFree(baij->rowvalues);CHKERRQ(ierr); 782 baij->rowvalues = 0; 783 } 784 785 PetscFunctionReturn(0); 786 } 787 788 #undef __FUNCT__ 789 #define __FUNCT__ "MatView_MPISBAIJ_ASCIIorDraworSocket" 790 static int MatView_MPISBAIJ_ASCIIorDraworSocket(Mat mat,PetscViewer viewer) 791 { 792 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 793 int ierr,bs = baij->bs,size = baij->size,rank = baij->rank; 794 PetscTruth isascii,isdraw; 795 PetscViewer sviewer; 796 PetscViewerFormat format; 797 798 PetscFunctionBegin; 799 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&isascii);CHKERRQ(ierr); 800 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);CHKERRQ(ierr); 801 if (isascii) { 802 ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr); 803 if (format == PETSC_VIEWER_ASCII_INFO_DETAIL) { 804 MatInfo info; 805 ierr = MPI_Comm_rank(mat->comm,&rank);CHKERRQ(ierr); 806 ierr = MatGetInfo(mat,MAT_LOCAL,&info);CHKERRQ(ierr); 807 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] Local rows %d nz %d nz alloced %d bs %d mem %d\n", 808 rank,mat->m,(int)info.nz_used*bs,(int)info.nz_allocated*bs, 809 baij->bs,(int)info.memory);CHKERRQ(ierr); 810 ierr = MatGetInfo(baij->A,MAT_LOCAL,&info);CHKERRQ(ierr); 811 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] on-diagonal part: nz %d \n",rank,(int)info.nz_used*bs);CHKERRQ(ierr); 812 ierr = MatGetInfo(baij->B,MAT_LOCAL,&info);CHKERRQ(ierr); 813 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] off-diagonal part: nz %d \n",rank,(int)info.nz_used*bs);CHKERRQ(ierr); 814 ierr = PetscViewerFlush(viewer);CHKERRQ(ierr); 815 ierr = VecScatterView(baij->Mvctx,viewer);CHKERRQ(ierr); 816 PetscFunctionReturn(0); 817 } else if (format == PETSC_VIEWER_ASCII_INFO) { 818 ierr = PetscViewerASCIIPrintf(viewer," block size is %d\n",bs);CHKERRQ(ierr); 819 PetscFunctionReturn(0); 820 } 821 } 822 823 if (isdraw) { 824 PetscDraw draw; 825 PetscTruth isnull; 826 ierr = PetscViewerDrawGetDraw(viewer,0,&draw);CHKERRQ(ierr); 827 ierr = PetscDrawIsNull(draw,&isnull);CHKERRQ(ierr); if (isnull) PetscFunctionReturn(0); 828 } 829 830 if (size == 1) { 831 ierr = PetscObjectSetName((PetscObject)baij->A,mat->name);CHKERRQ(ierr); 832 ierr = MatView(baij->A,viewer);CHKERRQ(ierr); 833 } else { 834 /* assemble the entire matrix onto first processor. */ 835 Mat A; 836 Mat_SeqSBAIJ *Aloc; 837 Mat_SeqBAIJ *Bloc; 838 int M = mat->M,N = mat->N,*ai,*aj,col,i,j,k,*rvals,mbs = baij->mbs; 839 MatScalar *a; 840 841 if (!rank) { 842 ierr = MatCreateMPISBAIJ(mat->comm,baij->bs,M,N,M,N,0,PETSC_NULL,0,PETSC_NULL,&A);CHKERRQ(ierr); 843 } else { 844 ierr = MatCreateMPISBAIJ(mat->comm,baij->bs,0,0,M,N,0,PETSC_NULL,0,PETSC_NULL,&A);CHKERRQ(ierr); 845 } 846 PetscLogObjectParent(mat,A); 847 848 /* copy over the A part */ 849 Aloc = (Mat_SeqSBAIJ*)baij->A->data; 850 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 851 ierr = PetscMalloc(bs*sizeof(int),&rvals);CHKERRQ(ierr); 852 853 for (i=0; i<mbs; i++) { 854 rvals[0] = bs*(baij->rstart + i); 855 for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; } 856 for (j=ai[i]; j<ai[i+1]; j++) { 857 col = (baij->cstart+aj[j])*bs; 858 for (k=0; k<bs; k++) { 859 ierr = MatSetValues_MPISBAIJ_MatScalar(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 860 col++; a += bs; 861 } 862 } 863 } 864 /* copy over the B part */ 865 Bloc = (Mat_SeqBAIJ*)baij->B->data; 866 ai = Bloc->i; aj = Bloc->j; a = Bloc->a; 867 for (i=0; i<mbs; i++) { 868 rvals[0] = bs*(baij->rstart + i); 869 for (j=1; j<bs; j++) { rvals[j] = rvals[j-1] + 1; } 870 for (j=ai[i]; j<ai[i+1]; j++) { 871 col = baij->garray[aj[j]]*bs; 872 for (k=0; k<bs; k++) { 873 ierr = MatSetValues_MPISBAIJ_MatScalar(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 874 col++; a += bs; 875 } 876 } 877 } 878 ierr = PetscFree(rvals);CHKERRQ(ierr); 879 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 880 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 881 /* 882 Everyone has to call to draw the matrix since the graphics waits are 883 synchronized across all processors that share the PetscDraw object 884 */ 885 ierr = PetscViewerGetSingleton(viewer,&sviewer);CHKERRQ(ierr); 886 if (!rank) { 887 ierr = PetscObjectSetName((PetscObject)((Mat_MPISBAIJ*)(A->data))->A,mat->name);CHKERRQ(ierr); 888 ierr = MatView(((Mat_MPISBAIJ*)(A->data))->A,sviewer);CHKERRQ(ierr); 889 } 890 ierr = PetscViewerRestoreSingleton(viewer,&sviewer);CHKERRQ(ierr); 891 ierr = MatDestroy(A);CHKERRQ(ierr); 892 } 893 PetscFunctionReturn(0); 894 } 895 896 #undef __FUNCT__ 897 #define __FUNCT__ "MatView_MPISBAIJ" 898 int MatView_MPISBAIJ(Mat mat,PetscViewer viewer) 899 { 900 int ierr; 901 PetscTruth isascii,isdraw,issocket,isbinary; 902 903 PetscFunctionBegin; 904 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_ASCII,&isascii);CHKERRQ(ierr); 905 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_DRAW,&isdraw);CHKERRQ(ierr); 906 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_SOCKET,&issocket);CHKERRQ(ierr); 907 ierr = PetscTypeCompare((PetscObject)viewer,PETSC_VIEWER_BINARY,&isbinary);CHKERRQ(ierr); 908 if (isascii || isdraw || issocket || isbinary) { 909 ierr = MatView_MPISBAIJ_ASCIIorDraworSocket(mat,viewer);CHKERRQ(ierr); 910 } else { 911 SETERRQ1(1,"Viewer type %s not supported by MPISBAIJ matrices",((PetscObject)viewer)->type_name); 912 } 913 PetscFunctionReturn(0); 914 } 915 916 #undef __FUNCT__ 917 #define __FUNCT__ "MatDestroy_MPISBAIJ" 918 int MatDestroy_MPISBAIJ(Mat mat) 919 { 920 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 921 int ierr; 922 923 PetscFunctionBegin; 924 #if defined(PETSC_USE_LOG) 925 PetscLogObjectState((PetscObject)mat,"Rows=%d,Cols=%d",mat->M,mat->N); 926 #endif 927 ierr = MatStashDestroy_Private(&mat->stash);CHKERRQ(ierr); 928 ierr = MatStashDestroy_Private(&mat->bstash);CHKERRQ(ierr); 929 ierr = PetscFree(baij->rowners);CHKERRQ(ierr); 930 ierr = MatDestroy(baij->A);CHKERRQ(ierr); 931 ierr = MatDestroy(baij->B);CHKERRQ(ierr); 932 #if defined (PETSC_USE_CTABLE) 933 if (baij->colmap) {ierr = PetscTableDelete(baij->colmap);CHKERRQ(ierr);} 934 #else 935 if (baij->colmap) {ierr = PetscFree(baij->colmap);CHKERRQ(ierr);} 936 #endif 937 if (baij->garray) {ierr = PetscFree(baij->garray);CHKERRQ(ierr);} 938 if (baij->lvec) {ierr = VecDestroy(baij->lvec);CHKERRQ(ierr);} 939 if (baij->Mvctx) {ierr = VecScatterDestroy(baij->Mvctx);CHKERRQ(ierr);} 940 if (baij->slvec0) { 941 ierr = VecDestroy(baij->slvec0);CHKERRQ(ierr); 942 ierr = VecDestroy(baij->slvec0b);CHKERRQ(ierr); 943 } 944 if (baij->slvec1) { 945 ierr = VecDestroy(baij->slvec1);CHKERRQ(ierr); 946 ierr = VecDestroy(baij->slvec1a);CHKERRQ(ierr); 947 ierr = VecDestroy(baij->slvec1b);CHKERRQ(ierr); 948 } 949 if (baij->sMvctx) {ierr = VecScatterDestroy(baij->sMvctx);CHKERRQ(ierr);} 950 if (baij->rowvalues) {ierr = PetscFree(baij->rowvalues);CHKERRQ(ierr);} 951 if (baij->barray) {ierr = PetscFree(baij->barray);CHKERRQ(ierr);} 952 if (baij->hd) {ierr = PetscFree(baij->hd);CHKERRQ(ierr);} 953 #if defined(PETSC_USE_MAT_SINGLE) 954 if (baij->setvaluescopy) {ierr = PetscFree(baij->setvaluescopy);CHKERRQ(ierr);} 955 #endif 956 ierr = PetscFree(baij);CHKERRQ(ierr); 957 PetscFunctionReturn(0); 958 } 959 960 #undef __FUNCT__ 961 #define __FUNCT__ "MatMult_MPISBAIJ" 962 int MatMult_MPISBAIJ(Mat A,Vec xx,Vec yy) 963 { 964 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 965 int ierr,nt,mbs=a->mbs,bs=a->bs; 966 PetscScalar *x,*from,zero=0.0; 967 968 PetscFunctionBegin; 969 /* 970 PetscSynchronizedPrintf(A->comm," _1comm is called ...\n"); 971 PetscSynchronizedFlush(A->comm); 972 */ 973 ierr = VecGetLocalSize(xx,&nt);CHKERRQ(ierr); 974 if (nt != A->n) { 975 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx"); 976 } 977 ierr = VecGetLocalSize(yy,&nt);CHKERRQ(ierr); 978 if (nt != A->m) { 979 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible parition of A and yy"); 980 } 981 982 /* diagonal part */ 983 ierr = (*a->A->ops->mult)(a->A,xx,a->slvec1a);CHKERRQ(ierr); 984 ierr = VecSet(&zero,a->slvec1b);CHKERRQ(ierr); 985 986 /* subdiagonal part */ 987 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->slvec0b);CHKERRQ(ierr); 988 989 /* copy x into the vec slvec0 */ 990 ierr = VecGetArray(a->slvec0,&from);CHKERRQ(ierr); 991 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 992 ierr = PetscMemcpy(from,x,bs*mbs*sizeof(MatScalar));CHKERRQ(ierr); 993 ierr = VecRestoreArray(a->slvec0,&from);CHKERRQ(ierr); 994 995 ierr = VecScatterBegin(a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD,a->sMvctx);CHKERRQ(ierr); 996 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 997 ierr = VecScatterEnd(a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD,a->sMvctx);CHKERRQ(ierr); 998 999 /* supperdiagonal part */ 1000 ierr = (*a->B->ops->multadd)(a->B,a->slvec1b,a->slvec1a,yy);CHKERRQ(ierr); 1001 1002 PetscFunctionReturn(0); 1003 } 1004 1005 #undef __FUNCT__ 1006 #define __FUNCT__ "MatMult_MPISBAIJ_2comm" 1007 int MatMult_MPISBAIJ_2comm(Mat A,Vec xx,Vec yy) 1008 { 1009 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1010 int ierr,nt; 1011 1012 PetscFunctionBegin; 1013 ierr = VecGetLocalSize(xx,&nt);CHKERRQ(ierr); 1014 if (nt != A->n) { 1015 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx"); 1016 } 1017 ierr = VecGetLocalSize(yy,&nt);CHKERRQ(ierr); 1018 if (nt != A->m) { 1019 SETERRQ(PETSC_ERR_ARG_SIZ,"Incompatible parition of A and yy"); 1020 } 1021 1022 ierr = VecScatterBegin(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1023 /* do diagonal part */ 1024 ierr = (*a->A->ops->mult)(a->A,xx,yy);CHKERRQ(ierr); 1025 /* do supperdiagonal part */ 1026 ierr = VecScatterEnd(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1027 ierr = (*a->B->ops->multadd)(a->B,a->lvec,yy,yy);CHKERRQ(ierr); 1028 /* do subdiagonal part */ 1029 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 1030 ierr = VecScatterBegin(a->lvec,yy,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 1031 ierr = VecScatterEnd(a->lvec,yy,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 1032 1033 PetscFunctionReturn(0); 1034 } 1035 1036 #undef __FUNCT__ 1037 #define __FUNCT__ "MatMultAdd_MPISBAIJ" 1038 int MatMultAdd_MPISBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 1039 { 1040 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1041 int ierr,mbs=a->mbs,bs=a->bs; 1042 PetscScalar *x,*from,zero=0.0; 1043 1044 PetscFunctionBegin; 1045 /* 1046 PetscSynchronizedPrintf(A->comm," MatMultAdd is called ...\n"); 1047 PetscSynchronizedFlush(A->comm); 1048 */ 1049 /* diagonal part */ 1050 ierr = (*a->A->ops->multadd)(a->A,xx,yy,a->slvec1a);CHKERRQ(ierr); 1051 ierr = VecSet(&zero,a->slvec1b);CHKERRQ(ierr); 1052 1053 /* subdiagonal part */ 1054 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->slvec0b);CHKERRQ(ierr); 1055 1056 /* copy x into the vec slvec0 */ 1057 ierr = VecGetArray(a->slvec0,&from);CHKERRQ(ierr); 1058 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 1059 ierr = PetscMemcpy(from,x,bs*mbs*sizeof(MatScalar));CHKERRQ(ierr); 1060 ierr = VecRestoreArray(a->slvec0,&from);CHKERRQ(ierr); 1061 1062 ierr = VecScatterBegin(a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD,a->sMvctx);CHKERRQ(ierr); 1063 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 1064 ierr = VecScatterEnd(a->slvec0,a->slvec1,ADD_VALUES,SCATTER_FORWARD,a->sMvctx);CHKERRQ(ierr); 1065 1066 /* supperdiagonal part */ 1067 ierr = (*a->B->ops->multadd)(a->B,a->slvec1b,a->slvec1a,zz);CHKERRQ(ierr); 1068 1069 PetscFunctionReturn(0); 1070 } 1071 1072 #undef __FUNCT__ 1073 #define __FUNCT__ "MatMultAdd_MPISBAIJ_2comm" 1074 int MatMultAdd_MPISBAIJ_2comm(Mat A,Vec xx,Vec yy,Vec zz) 1075 { 1076 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1077 int ierr; 1078 1079 PetscFunctionBegin; 1080 ierr = VecScatterBegin(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1081 /* do diagonal part */ 1082 ierr = (*a->A->ops->multadd)(a->A,xx,yy,zz);CHKERRQ(ierr); 1083 /* do supperdiagonal part */ 1084 ierr = VecScatterEnd(xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD,a->Mvctx);CHKERRQ(ierr); 1085 ierr = (*a->B->ops->multadd)(a->B,a->lvec,zz,zz);CHKERRQ(ierr); 1086 1087 /* do subdiagonal part */ 1088 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 1089 ierr = VecScatterBegin(a->lvec,zz,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 1090 ierr = VecScatterEnd(a->lvec,zz,ADD_VALUES,SCATTER_REVERSE,a->Mvctx);CHKERRQ(ierr); 1091 1092 PetscFunctionReturn(0); 1093 } 1094 1095 #undef __FUNCT__ 1096 #define __FUNCT__ "MatMultTranspose_MPISBAIJ" 1097 int MatMultTranspose_MPISBAIJ(Mat A,Vec xx,Vec yy) 1098 { 1099 PetscFunctionBegin; 1100 SETERRQ(1,"Matrix is symmetric. Call MatMult()."); 1101 /* PetscFunctionReturn(0); */ 1102 } 1103 1104 #undef __FUNCT__ 1105 #define __FUNCT__ "MatMultTransposeAdd_MPISBAIJ" 1106 int MatMultTransposeAdd_MPISBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 1107 { 1108 PetscFunctionBegin; 1109 SETERRQ(1,"Matrix is symmetric. Call MatMultAdd()."); 1110 /* PetscFunctionReturn(0); */ 1111 } 1112 1113 /* 1114 This only works correctly for square matrices where the subblock A->A is the 1115 diagonal block 1116 */ 1117 #undef __FUNCT__ 1118 #define __FUNCT__ "MatGetDiagonal_MPISBAIJ" 1119 int MatGetDiagonal_MPISBAIJ(Mat A,Vec v) 1120 { 1121 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1122 int ierr; 1123 1124 PetscFunctionBegin; 1125 /* if (a->M != a->N) SETERRQ(PETSC_ERR_SUP,"Supports only square matrix where A->A is diag block"); */ 1126 ierr = MatGetDiagonal(a->A,v);CHKERRQ(ierr); 1127 PetscFunctionReturn(0); 1128 } 1129 1130 #undef __FUNCT__ 1131 #define __FUNCT__ "MatScale_MPISBAIJ" 1132 int MatScale_MPISBAIJ(const PetscScalar *aa,Mat A) 1133 { 1134 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1135 int ierr; 1136 1137 PetscFunctionBegin; 1138 ierr = MatScale(aa,a->A);CHKERRQ(ierr); 1139 ierr = MatScale(aa,a->B);CHKERRQ(ierr); 1140 PetscFunctionReturn(0); 1141 } 1142 1143 #undef __FUNCT__ 1144 #define __FUNCT__ "MatGetRow_MPISBAIJ" 1145 int MatGetRow_MPISBAIJ(Mat matin,int row,int *nz,int **idx,PetscScalar **v) 1146 { 1147 Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data; 1148 PetscScalar *vworkA,*vworkB,**pvA,**pvB,*v_p; 1149 int bs = mat->bs,bs2 = mat->bs2,i,ierr,*cworkA,*cworkB,**pcA,**pcB; 1150 int nztot,nzA,nzB,lrow,brstart = mat->rstart*bs,brend = mat->rend*bs; 1151 int *cmap,*idx_p,cstart = mat->cstart; 1152 1153 PetscFunctionBegin; 1154 if (mat->getrowactive == PETSC_TRUE) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"Already active"); 1155 mat->getrowactive = PETSC_TRUE; 1156 1157 if (!mat->rowvalues && (idx || v)) { 1158 /* 1159 allocate enough space to hold information from the longest row. 1160 */ 1161 Mat_SeqSBAIJ *Aa = (Mat_SeqSBAIJ*)mat->A->data; 1162 Mat_SeqBAIJ *Ba = (Mat_SeqBAIJ*)mat->B->data; 1163 int max = 1,mbs = mat->mbs,tmp; 1164 for (i=0; i<mbs; i++) { 1165 tmp = Aa->i[i+1] - Aa->i[i] + Ba->i[i+1] - Ba->i[i]; /* row length */ 1166 if (max < tmp) { max = tmp; } 1167 } 1168 ierr = PetscMalloc(max*bs2*(sizeof(int)+sizeof(PetscScalar)),&mat->rowvalues);CHKERRQ(ierr); 1169 mat->rowindices = (int*)(mat->rowvalues + max*bs2); 1170 } 1171 1172 if (row < brstart || row >= brend) SETERRQ(PETSC_ERR_SUP,"Only local rows") 1173 lrow = row - brstart; /* local row index */ 1174 1175 pvA = &vworkA; pcA = &cworkA; pvB = &vworkB; pcB = &cworkB; 1176 if (!v) {pvA = 0; pvB = 0;} 1177 if (!idx) {pcA = 0; if (!v) pcB = 0;} 1178 ierr = (*mat->A->ops->getrow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1179 ierr = (*mat->B->ops->getrow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1180 nztot = nzA + nzB; 1181 1182 cmap = mat->garray; 1183 if (v || idx) { 1184 if (nztot) { 1185 /* Sort by increasing column numbers, assuming A and B already sorted */ 1186 int imark = -1; 1187 if (v) { 1188 *v = v_p = mat->rowvalues; 1189 for (i=0; i<nzB; i++) { 1190 if (cmap[cworkB[i]/bs] < cstart) v_p[i] = vworkB[i]; 1191 else break; 1192 } 1193 imark = i; 1194 for (i=0; i<nzA; i++) v_p[imark+i] = vworkA[i]; 1195 for (i=imark; i<nzB; i++) v_p[nzA+i] = vworkB[i]; 1196 } 1197 if (idx) { 1198 *idx = idx_p = mat->rowindices; 1199 if (imark > -1) { 1200 for (i=0; i<imark; i++) { 1201 idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs; 1202 } 1203 } else { 1204 for (i=0; i<nzB; i++) { 1205 if (cmap[cworkB[i]/bs] < cstart) 1206 idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ; 1207 else break; 1208 } 1209 imark = i; 1210 } 1211 for (i=0; i<nzA; i++) idx_p[imark+i] = cstart*bs + cworkA[i]; 1212 for (i=imark; i<nzB; i++) idx_p[nzA+i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ; 1213 } 1214 } else { 1215 if (idx) *idx = 0; 1216 if (v) *v = 0; 1217 } 1218 } 1219 *nz = nztot; 1220 ierr = (*mat->A->ops->restorerow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1221 ierr = (*mat->B->ops->restorerow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1222 PetscFunctionReturn(0); 1223 } 1224 1225 #undef __FUNCT__ 1226 #define __FUNCT__ "MatRestoreRow_MPISBAIJ" 1227 int MatRestoreRow_MPISBAIJ(Mat mat,int row,int *nz,int **idx,PetscScalar **v) 1228 { 1229 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 1230 1231 PetscFunctionBegin; 1232 if (baij->getrowactive == PETSC_FALSE) { 1233 SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"MatGetRow not called"); 1234 } 1235 baij->getrowactive = PETSC_FALSE; 1236 PetscFunctionReturn(0); 1237 } 1238 1239 #undef __FUNCT__ 1240 #define __FUNCT__ "MatGetBlockSize_MPISBAIJ" 1241 int MatGetBlockSize_MPISBAIJ(Mat mat,int *bs) 1242 { 1243 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 1244 1245 PetscFunctionBegin; 1246 *bs = baij->bs; 1247 PetscFunctionReturn(0); 1248 } 1249 1250 #undef __FUNCT__ 1251 #define __FUNCT__ "MatZeroEntries_MPISBAIJ" 1252 int MatZeroEntries_MPISBAIJ(Mat A) 1253 { 1254 Mat_MPISBAIJ *l = (Mat_MPISBAIJ*)A->data; 1255 int ierr; 1256 1257 PetscFunctionBegin; 1258 ierr = MatZeroEntries(l->A);CHKERRQ(ierr); 1259 ierr = MatZeroEntries(l->B);CHKERRQ(ierr); 1260 PetscFunctionReturn(0); 1261 } 1262 1263 #undef __FUNCT__ 1264 #define __FUNCT__ "MatGetInfo_MPISBAIJ" 1265 int MatGetInfo_MPISBAIJ(Mat matin,MatInfoType flag,MatInfo *info) 1266 { 1267 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)matin->data; 1268 Mat A = a->A,B = a->B; 1269 int ierr; 1270 PetscReal isend[5],irecv[5]; 1271 1272 PetscFunctionBegin; 1273 info->block_size = (PetscReal)a->bs; 1274 ierr = MatGetInfo(A,MAT_LOCAL,info);CHKERRQ(ierr); 1275 isend[0] = info->nz_used; isend[1] = info->nz_allocated; isend[2] = info->nz_unneeded; 1276 isend[3] = info->memory; isend[4] = info->mallocs; 1277 ierr = MatGetInfo(B,MAT_LOCAL,info);CHKERRQ(ierr); 1278 isend[0] += info->nz_used; isend[1] += info->nz_allocated; isend[2] += info->nz_unneeded; 1279 isend[3] += info->memory; isend[4] += info->mallocs; 1280 if (flag == MAT_LOCAL) { 1281 info->nz_used = isend[0]; 1282 info->nz_allocated = isend[1]; 1283 info->nz_unneeded = isend[2]; 1284 info->memory = isend[3]; 1285 info->mallocs = isend[4]; 1286 } else if (flag == MAT_GLOBAL_MAX) { 1287 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPI_MAX,matin->comm);CHKERRQ(ierr); 1288 info->nz_used = irecv[0]; 1289 info->nz_allocated = irecv[1]; 1290 info->nz_unneeded = irecv[2]; 1291 info->memory = irecv[3]; 1292 info->mallocs = irecv[4]; 1293 } else if (flag == MAT_GLOBAL_SUM) { 1294 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPI_SUM,matin->comm);CHKERRQ(ierr); 1295 info->nz_used = irecv[0]; 1296 info->nz_allocated = irecv[1]; 1297 info->nz_unneeded = irecv[2]; 1298 info->memory = irecv[3]; 1299 info->mallocs = irecv[4]; 1300 } else { 1301 SETERRQ1(1,"Unknown MatInfoType argument %d",flag); 1302 } 1303 info->rows_global = (PetscReal)A->M; 1304 info->columns_global = (PetscReal)A->N; 1305 info->rows_local = (PetscReal)A->m; 1306 info->columns_local = (PetscReal)A->N; 1307 info->fill_ratio_given = 0; /* no parallel LU/ILU/Cholesky */ 1308 info->fill_ratio_needed = 0; 1309 info->factor_mallocs = 0; 1310 PetscFunctionReturn(0); 1311 } 1312 1313 #undef __FUNCT__ 1314 #define __FUNCT__ "MatSetOption_MPISBAIJ" 1315 int MatSetOption_MPISBAIJ(Mat A,MatOption op) 1316 { 1317 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1318 int ierr; 1319 1320 PetscFunctionBegin; 1321 switch (op) { 1322 case MAT_NO_NEW_NONZERO_LOCATIONS: 1323 case MAT_YES_NEW_NONZERO_LOCATIONS: 1324 case MAT_COLUMNS_UNSORTED: 1325 case MAT_COLUMNS_SORTED: 1326 case MAT_NEW_NONZERO_ALLOCATION_ERR: 1327 case MAT_KEEP_ZEROED_ROWS: 1328 case MAT_NEW_NONZERO_LOCATION_ERR: 1329 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1330 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1331 break; 1332 case MAT_ROW_ORIENTED: 1333 a->roworiented = PETSC_TRUE; 1334 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1335 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1336 break; 1337 case MAT_ROWS_SORTED: 1338 case MAT_ROWS_UNSORTED: 1339 case MAT_YES_NEW_DIAGONALS: 1340 PetscLogInfo(A,"Info:MatSetOption_MPIBAIJ:Option ignored\n"); 1341 break; 1342 case MAT_COLUMN_ORIENTED: 1343 a->roworiented = PETSC_FALSE; 1344 ierr = MatSetOption(a->A,op);CHKERRQ(ierr); 1345 ierr = MatSetOption(a->B,op);CHKERRQ(ierr); 1346 break; 1347 case MAT_IGNORE_OFF_PROC_ENTRIES: 1348 a->donotstash = PETSC_TRUE; 1349 break; 1350 case MAT_NO_NEW_DIAGONALS: 1351 SETERRQ(PETSC_ERR_SUP,"MAT_NO_NEW_DIAGONALS"); 1352 case MAT_USE_HASH_TABLE: 1353 a->ht_flag = PETSC_TRUE; 1354 break; 1355 case MAT_NOT_SYMMETRIC: 1356 case MAT_NOT_STRUCTURALLY_SYMMETRIC: 1357 case MAT_HERMITIAN: 1358 SETERRQ(PETSC_ERR_SUP,"Matrix must be symmetric"); 1359 case MAT_SYMMETRIC: 1360 case MAT_STRUCTURALLY_SYMMETRIC: 1361 case MAT_NOT_HERMITIAN: 1362 case MAT_SYMMETRY_ETERNAL: 1363 case MAT_NOT_SYMMETRY_ETERNAL: 1364 break; 1365 default: 1366 SETERRQ(PETSC_ERR_SUP,"unknown option"); 1367 } 1368 PetscFunctionReturn(0); 1369 } 1370 1371 #undef __FUNCT__ 1372 #define __FUNCT__ "MatTranspose_MPISBAIJ" 1373 int MatTranspose_MPISBAIJ(Mat A,Mat *B) 1374 { 1375 int ierr; 1376 PetscFunctionBegin; 1377 ierr = MatDuplicate(A,MAT_COPY_VALUES,B);CHKERRQ(ierr); 1378 PetscFunctionReturn(0); 1379 } 1380 1381 #undef __FUNCT__ 1382 #define __FUNCT__ "MatDiagonalScale_MPISBAIJ" 1383 int MatDiagonalScale_MPISBAIJ(Mat mat,Vec ll,Vec rr) 1384 { 1385 Mat_MPISBAIJ *baij = (Mat_MPISBAIJ*)mat->data; 1386 Mat a = baij->A,b = baij->B; 1387 int ierr,s1,s2,s3; 1388 1389 PetscFunctionBegin; 1390 if (ll != rr) { 1391 SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"For symmetric format, left and right scaling vectors must be same\n"); 1392 } 1393 ierr = MatGetLocalSize(mat,&s2,&s3);CHKERRQ(ierr); 1394 if (rr) { 1395 ierr = VecGetLocalSize(rr,&s1);CHKERRQ(ierr); 1396 if (s1!=s3) SETERRQ(PETSC_ERR_ARG_SIZ,"right vector non-conforming local size"); 1397 /* Overlap communication with computation. */ 1398 ierr = VecScatterBegin(rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD,baij->Mvctx);CHKERRQ(ierr); 1399 /*} if (ll) { */ 1400 ierr = VecGetLocalSize(ll,&s1);CHKERRQ(ierr); 1401 if (s1!=s2) SETERRQ(PETSC_ERR_ARG_SIZ,"left vector non-conforming local size"); 1402 ierr = (*b->ops->diagonalscale)(b,ll,PETSC_NULL);CHKERRQ(ierr); 1403 /* } */ 1404 /* scale the diagonal block */ 1405 ierr = (*a->ops->diagonalscale)(a,ll,rr);CHKERRQ(ierr); 1406 1407 /* if (rr) { */ 1408 /* Do a scatter end and then right scale the off-diagonal block */ 1409 ierr = VecScatterEnd(rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD,baij->Mvctx);CHKERRQ(ierr); 1410 ierr = (*b->ops->diagonalscale)(b,PETSC_NULL,baij->lvec);CHKERRQ(ierr); 1411 } 1412 1413 PetscFunctionReturn(0); 1414 } 1415 1416 #undef __FUNCT__ 1417 #define __FUNCT__ "MatZeroRows_MPISBAIJ" 1418 int MatZeroRows_MPISBAIJ(Mat A,IS is,const PetscScalar *diag) 1419 { 1420 PetscFunctionBegin; 1421 SETERRQ(PETSC_ERR_SUP,"No support for this function yet"); 1422 } 1423 1424 #undef __FUNCT__ 1425 #define __FUNCT__ "MatPrintHelp_MPISBAIJ" 1426 int MatPrintHelp_MPISBAIJ(Mat A) 1427 { 1428 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1429 MPI_Comm comm = A->comm; 1430 static int called = 0; 1431 int ierr; 1432 1433 PetscFunctionBegin; 1434 if (!a->rank) { 1435 ierr = MatPrintHelp_SeqSBAIJ(a->A);CHKERRQ(ierr); 1436 } 1437 if (called) {PetscFunctionReturn(0);} else called = 1; 1438 ierr = (*PetscHelpPrintf)(comm," Options for MATMPISBAIJ matrix format (the defaults):\n");CHKERRQ(ierr); 1439 ierr = (*PetscHelpPrintf)(comm," -mat_use_hash_table <factor>: Use hashtable for efficient matrix assembly\n");CHKERRQ(ierr); 1440 PetscFunctionReturn(0); 1441 } 1442 1443 #undef __FUNCT__ 1444 #define __FUNCT__ "MatSetUnfactored_MPISBAIJ" 1445 int MatSetUnfactored_MPISBAIJ(Mat A) 1446 { 1447 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 1448 int ierr; 1449 1450 PetscFunctionBegin; 1451 ierr = MatSetUnfactored(a->A);CHKERRQ(ierr); 1452 PetscFunctionReturn(0); 1453 } 1454 1455 static int MatDuplicate_MPISBAIJ(Mat,MatDuplicateOption,Mat *); 1456 1457 #undef __FUNCT__ 1458 #define __FUNCT__ "MatEqual_MPISBAIJ" 1459 int MatEqual_MPISBAIJ(Mat A,Mat B,PetscTruth *flag) 1460 { 1461 Mat_MPISBAIJ *matB = (Mat_MPISBAIJ*)B->data,*matA = (Mat_MPISBAIJ*)A->data; 1462 Mat a,b,c,d; 1463 PetscTruth flg; 1464 int ierr; 1465 1466 PetscFunctionBegin; 1467 a = matA->A; b = matA->B; 1468 c = matB->A; d = matB->B; 1469 1470 ierr = MatEqual(a,c,&flg);CHKERRQ(ierr); 1471 if (flg == PETSC_TRUE) { 1472 ierr = MatEqual(b,d,&flg);CHKERRQ(ierr); 1473 } 1474 ierr = MPI_Allreduce(&flg,flag,1,MPI_INT,MPI_LAND,A->comm);CHKERRQ(ierr); 1475 PetscFunctionReturn(0); 1476 } 1477 1478 #undef __FUNCT__ 1479 #define __FUNCT__ "MatSetUpPreallocation_MPISBAIJ" 1480 int MatSetUpPreallocation_MPISBAIJ(Mat A) 1481 { 1482 int ierr; 1483 1484 PetscFunctionBegin; 1485 ierr = MatMPISBAIJSetPreallocation(A,1,PETSC_DEFAULT,0,PETSC_DEFAULT,0);CHKERRQ(ierr); 1486 PetscFunctionReturn(0); 1487 } 1488 /* -------------------------------------------------------------------*/ 1489 static struct _MatOps MatOps_Values = { 1490 MatSetValues_MPISBAIJ, 1491 MatGetRow_MPISBAIJ, 1492 MatRestoreRow_MPISBAIJ, 1493 MatMult_MPISBAIJ, 1494 /* 4*/ MatMultAdd_MPISBAIJ, 1495 MatMultTranspose_MPISBAIJ, 1496 MatMultTransposeAdd_MPISBAIJ, 1497 0, 1498 0, 1499 0, 1500 /*10*/ 0, 1501 0, 1502 0, 1503 MatRelax_MPISBAIJ, 1504 MatTranspose_MPISBAIJ, 1505 /*15*/ MatGetInfo_MPISBAIJ, 1506 MatEqual_MPISBAIJ, 1507 MatGetDiagonal_MPISBAIJ, 1508 MatDiagonalScale_MPISBAIJ, 1509 MatNorm_MPISBAIJ, 1510 /*20*/ MatAssemblyBegin_MPISBAIJ, 1511 MatAssemblyEnd_MPISBAIJ, 1512 0, 1513 MatSetOption_MPISBAIJ, 1514 MatZeroEntries_MPISBAIJ, 1515 /*25*/ MatZeroRows_MPISBAIJ, 1516 0, 1517 0, 1518 0, 1519 0, 1520 /*30*/ MatSetUpPreallocation_MPISBAIJ, 1521 0, 1522 0, 1523 0, 1524 0, 1525 /*35*/ MatDuplicate_MPISBAIJ, 1526 0, 1527 0, 1528 0, 1529 0, 1530 /*40*/ 0, 1531 0, 1532 0, 1533 MatGetValues_MPISBAIJ, 1534 0, 1535 /*45*/ MatPrintHelp_MPISBAIJ, 1536 MatScale_MPISBAIJ, 1537 0, 1538 0, 1539 0, 1540 /*50*/ MatGetBlockSize_MPISBAIJ, 1541 0, 1542 0, 1543 0, 1544 0, 1545 /*55*/ 0, 1546 0, 1547 MatSetUnfactored_MPISBAIJ, 1548 0, 1549 MatSetValuesBlocked_MPISBAIJ, 1550 /*60*/ 0, 1551 0, 1552 0, 1553 MatGetPetscMaps_Petsc, 1554 0, 1555 /*65*/ 0, 1556 0, 1557 0, 1558 0, 1559 0, 1560 /*70*/ MatGetRowMax_MPISBAIJ, 1561 0, 1562 0, 1563 0, 1564 0, 1565 /*75*/ 0, 1566 0, 1567 0, 1568 0, 1569 0, 1570 /*80*/ 0, 1571 0, 1572 0, 1573 0, 1574 /*85*/ MatLoad_MPISBAIJ 1575 }; 1576 1577 1578 EXTERN_C_BEGIN 1579 #undef __FUNCT__ 1580 #define __FUNCT__ "MatGetDiagonalBlock_MPISBAIJ" 1581 int MatGetDiagonalBlock_MPISBAIJ(Mat A,PetscTruth *iscopy,MatReuse reuse,Mat *a) 1582 { 1583 PetscFunctionBegin; 1584 *a = ((Mat_MPISBAIJ *)A->data)->A; 1585 *iscopy = PETSC_FALSE; 1586 PetscFunctionReturn(0); 1587 } 1588 EXTERN_C_END 1589 1590 EXTERN_C_BEGIN 1591 #undef __FUNCT__ 1592 #define __FUNCT__ "MatMPISBAIJSetPreallocation_MPISBAIJ" 1593 int MatMPISBAIJSetPreallocation_MPISBAIJ(Mat B,int bs,int d_nz,int *d_nnz,int o_nz,int *o_nnz) 1594 { 1595 Mat_MPISBAIJ *b; 1596 int ierr,i,mbs,Mbs; 1597 1598 PetscFunctionBegin; 1599 ierr = PetscOptionsGetInt(B->prefix,"-mat_block_size",&bs,PETSC_NULL);CHKERRQ(ierr); 1600 1601 if (bs < 1) SETERRQ(PETSC_ERR_ARG_OUTOFRANGE,"Invalid block size specified, must be positive"); 1602 if (d_nz == PETSC_DECIDE || d_nz == PETSC_DEFAULT) d_nz = 3; 1603 if (o_nz == PETSC_DECIDE || o_nz == PETSC_DEFAULT) o_nz = 1; 1604 if (d_nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"d_nz cannot be less than 0: value %d",d_nz); 1605 if (o_nz < 0) SETERRQ1(PETSC_ERR_ARG_OUTOFRANGE,"o_nz cannot be less than 0: value %d",o_nz); 1606 if (d_nnz) { 1607 for (i=0; i<B->m/bs; i++) { 1608 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]); 1609 } 1610 } 1611 if (o_nnz) { 1612 for (i=0; i<B->m/bs; i++) { 1613 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]); 1614 } 1615 } 1616 B->preallocated = PETSC_TRUE; 1617 ierr = PetscSplitOwnershipBlock(B->comm,bs,&B->m,&B->M);CHKERRQ(ierr); 1618 ierr = PetscSplitOwnershipBlock(B->comm,bs,&B->n,&B->N);CHKERRQ(ierr); 1619 ierr = PetscMapCreateMPI(B->comm,B->m,B->M,&B->rmap);CHKERRQ(ierr); 1620 ierr = PetscMapCreateMPI(B->comm,B->m,B->M,&B->cmap);CHKERRQ(ierr); 1621 1622 b = (Mat_MPISBAIJ*)B->data; 1623 mbs = B->m/bs; 1624 Mbs = B->M/bs; 1625 if (mbs*bs != B->m) { 1626 SETERRQ2(PETSC_ERR_ARG_SIZ,"No of local rows %d must be divisible by blocksize %d",B->m,bs); 1627 } 1628 1629 b->bs = bs; 1630 b->bs2 = bs*bs; 1631 b->mbs = mbs; 1632 b->nbs = mbs; 1633 b->Mbs = Mbs; 1634 b->Nbs = Mbs; 1635 1636 ierr = MPI_Allgather(&b->mbs,1,MPI_INT,b->rowners+1,1,MPI_INT,B->comm);CHKERRQ(ierr); 1637 b->rowners[0] = 0; 1638 for (i=2; i<=b->size; i++) { 1639 b->rowners[i] += b->rowners[i-1]; 1640 } 1641 b->rstart = b->rowners[b->rank]; 1642 b->rend = b->rowners[b->rank+1]; 1643 b->cstart = b->rstart; 1644 b->cend = b->rend; 1645 for (i=0; i<=b->size; i++) { 1646 b->rowners_bs[i] = b->rowners[i]*bs; 1647 } 1648 b->rstart_bs = b-> rstart*bs; 1649 b->rend_bs = b->rend*bs; 1650 1651 b->cstart_bs = b->cstart*bs; 1652 b->cend_bs = b->cend*bs; 1653 1654 1655 ierr = MatCreateSeqSBAIJ(PETSC_COMM_SELF,bs,B->m,B->m,d_nz,d_nnz,&b->A);CHKERRQ(ierr); 1656 PetscLogObjectParent(B,b->A); 1657 ierr = MatCreateSeqBAIJ(PETSC_COMM_SELF,bs,B->m,B->M,o_nz,o_nnz,&b->B);CHKERRQ(ierr); 1658 PetscLogObjectParent(B,b->B); 1659 1660 /* build cache for off array entries formed */ 1661 ierr = MatStashCreate_Private(B->comm,bs,&B->bstash);CHKERRQ(ierr); 1662 1663 PetscFunctionReturn(0); 1664 } 1665 EXTERN_C_END 1666 1667 /*MC 1668 MATMPISBAIJ - MATMPISBAIJ = "mpisbaij" - A matrix type to be used for distributed symmetric sparse block matrices, 1669 based on block compressed sparse row format. Only the upper triangular portion of the matrix is stored. 1670 1671 Options Database Keys: 1672 . -mat_type mpisbaij - sets the matrix type to "mpisbaij" during a call to MatSetFromOptions() 1673 1674 Level: beginner 1675 1676 .seealso: MatCreateMPISBAIJ 1677 M*/ 1678 1679 EXTERN_C_BEGIN 1680 #undef __FUNCT__ 1681 #define __FUNCT__ "MatCreate_MPISBAIJ" 1682 int MatCreate_MPISBAIJ(Mat B) 1683 { 1684 Mat_MPISBAIJ *b; 1685 int ierr; 1686 PetscTruth flg; 1687 1688 PetscFunctionBegin; 1689 1690 ierr = PetscNew(Mat_MPISBAIJ,&b);CHKERRQ(ierr); 1691 B->data = (void*)b; 1692 ierr = PetscMemzero(b,sizeof(Mat_MPISBAIJ));CHKERRQ(ierr); 1693 ierr = PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));CHKERRQ(ierr); 1694 1695 B->ops->destroy = MatDestroy_MPISBAIJ; 1696 B->ops->view = MatView_MPISBAIJ; 1697 B->mapping = 0; 1698 B->factor = 0; 1699 B->assembled = PETSC_FALSE; 1700 1701 B->insertmode = NOT_SET_VALUES; 1702 ierr = MPI_Comm_rank(B->comm,&b->rank);CHKERRQ(ierr); 1703 ierr = MPI_Comm_size(B->comm,&b->size);CHKERRQ(ierr); 1704 1705 /* build local table of row and column ownerships */ 1706 ierr = PetscMalloc(3*(b->size+2)*sizeof(int),&b->rowners);CHKERRQ(ierr); 1707 b->cowners = b->rowners + b->size + 2; 1708 b->rowners_bs = b->cowners + b->size + 2; 1709 PetscLogObjectMemory(B,3*(b->size+2)*sizeof(int)+sizeof(struct _p_Mat)+sizeof(Mat_MPISBAIJ)); 1710 1711 /* build cache for off array entries formed */ 1712 ierr = MatStashCreate_Private(B->comm,1,&B->stash);CHKERRQ(ierr); 1713 b->donotstash = PETSC_FALSE; 1714 b->colmap = PETSC_NULL; 1715 b->garray = PETSC_NULL; 1716 b->roworiented = PETSC_TRUE; 1717 1718 #if defined(PETSC_USE_MAT_SINGLE) 1719 /* stuff for MatSetValues_XXX in single precision */ 1720 b->setvalueslen = 0; 1721 b->setvaluescopy = PETSC_NULL; 1722 #endif 1723 1724 /* stuff used in block assembly */ 1725 b->barray = 0; 1726 1727 /* stuff used for matrix vector multiply */ 1728 b->lvec = 0; 1729 b->Mvctx = 0; 1730 b->slvec0 = 0; 1731 b->slvec0b = 0; 1732 b->slvec1 = 0; 1733 b->slvec1a = 0; 1734 b->slvec1b = 0; 1735 b->sMvctx = 0; 1736 1737 /* stuff for MatGetRow() */ 1738 b->rowindices = 0; 1739 b->rowvalues = 0; 1740 b->getrowactive = PETSC_FALSE; 1741 1742 /* hash table stuff */ 1743 b->ht = 0; 1744 b->hd = 0; 1745 b->ht_size = 0; 1746 b->ht_flag = PETSC_FALSE; 1747 b->ht_fact = 0; 1748 b->ht_total_ct = 0; 1749 b->ht_insert_ct = 0; 1750 1751 ierr = PetscOptionsHasName(B->prefix,"-mat_use_hash_table",&flg);CHKERRQ(ierr); 1752 if (flg) { 1753 PetscReal fact = 1.39; 1754 ierr = MatSetOption(B,MAT_USE_HASH_TABLE);CHKERRQ(ierr); 1755 ierr = PetscOptionsGetReal(B->prefix,"-mat_use_hash_table",&fact,PETSC_NULL);CHKERRQ(ierr); 1756 if (fact <= 1.0) fact = 1.39; 1757 ierr = MatMPIBAIJSetHashTableFactor(B,fact);CHKERRQ(ierr); 1758 PetscLogInfo(0,"MatCreateMPISBAIJ:Hash table Factor used %5.2f\n",fact); 1759 } 1760 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatStoreValues_C", 1761 "MatStoreValues_MPISBAIJ", 1762 MatStoreValues_MPISBAIJ);CHKERRQ(ierr); 1763 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatRetrieveValues_C", 1764 "MatRetrieveValues_MPISBAIJ", 1765 MatRetrieveValues_MPISBAIJ);CHKERRQ(ierr); 1766 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatGetDiagonalBlock_C", 1767 "MatGetDiagonalBlock_MPISBAIJ", 1768 MatGetDiagonalBlock_MPISBAIJ);CHKERRQ(ierr); 1769 ierr = PetscObjectComposeFunctionDynamic((PetscObject)B,"MatMPISBAIJSetPreallocation_C", 1770 "MatMPISBAIJSetPreallocation_MPISBAIJ", 1771 MatMPISBAIJSetPreallocation_MPISBAIJ);CHKERRQ(ierr); 1772 PetscFunctionReturn(0); 1773 } 1774 EXTERN_C_END 1775 1776 /*MC 1777 MATSBAIJ - MATSBAIJ = "sbaij" - A matrix type to be used for symmetric block sparse matrices. 1778 1779 This matrix type is identical to MATSEQSBAIJ when constructed with a single process communicator, 1780 and MATMPISBAIJ otherwise. 1781 1782 Options Database Keys: 1783 . -mat_type sbaij - sets the matrix type to "sbaij" during a call to MatSetFromOptions() 1784 1785 Level: beginner 1786 1787 .seealso: MatCreateMPISBAIJ,MATSEQSBAIJ,MATMPISBAIJ 1788 M*/ 1789 1790 EXTERN_C_BEGIN 1791 #undef __FUNCT__ 1792 #define __FUNCT__ "MatCreate_SBAIJ" 1793 int MatCreate_SBAIJ(Mat A) { 1794 int ierr,size; 1795 1796 PetscFunctionBegin; 1797 ierr = PetscObjectChangeTypeName((PetscObject)A,MATSBAIJ);CHKERRQ(ierr); 1798 ierr = MPI_Comm_size(A->comm,&size);CHKERRQ(ierr); 1799 if (size == 1) { 1800 ierr = MatSetType(A,MATSEQSBAIJ);CHKERRQ(ierr); 1801 } else { 1802 ierr = MatSetType(A,MATMPISBAIJ);CHKERRQ(ierr); 1803 } 1804 PetscFunctionReturn(0); 1805 } 1806 EXTERN_C_END 1807 1808 #undef __FUNCT__ 1809 #define __FUNCT__ "MatMPISBAIJSetPreallocation" 1810 /*@C 1811 MatMPISBAIJSetPreallocation - For good matrix assembly performance 1812 the user should preallocate the matrix storage by setting the parameters 1813 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 1814 performance can be increased by more than a factor of 50. 1815 1816 Collective on Mat 1817 1818 Input Parameters: 1819 + A - the matrix 1820 . bs - size of blockk 1821 . d_nz - number of block nonzeros per block row in diagonal portion of local 1822 submatrix (same for all local rows) 1823 . d_nnz - array containing the number of block nonzeros in the various block rows 1824 in the upper triangular and diagonal part of the in diagonal portion of the local 1825 (possibly different for each block row) or PETSC_NULL. You must leave room 1826 for the diagonal entry even if it is zero. 1827 . o_nz - number of block nonzeros per block row in the off-diagonal portion of local 1828 submatrix (same for all local rows). 1829 - o_nnz - array containing the number of nonzeros in the various block rows of the 1830 off-diagonal portion of the local submatrix (possibly different for 1831 each block row) or PETSC_NULL. 1832 1833 1834 Options Database Keys: 1835 . -mat_no_unroll - uses code that does not unroll the loops in the 1836 block calculations (much slower) 1837 . -mat_block_size - size of the blocks to use 1838 1839 Notes: 1840 1841 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 1842 than it must be used on all processors that share the object for that argument. 1843 1844 Storage Information: 1845 For a square global matrix we define each processor's diagonal portion 1846 to be its local rows and the corresponding columns (a square submatrix); 1847 each processor's off-diagonal portion encompasses the remainder of the 1848 local matrix (a rectangular submatrix). 1849 1850 The user can specify preallocated storage for the diagonal part of 1851 the local submatrix with either d_nz or d_nnz (not both). Set 1852 d_nz=PETSC_DEFAULT and d_nnz=PETSC_NULL for PETSc to control dynamic 1853 memory allocation. Likewise, specify preallocated storage for the 1854 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 1855 1856 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 1857 the figure below we depict these three local rows and all columns (0-11). 1858 1859 .vb 1860 0 1 2 3 4 5 6 7 8 9 10 11 1861 ------------------- 1862 row 3 | o o o d d d o o o o o o 1863 row 4 | o o o d d d o o o o o o 1864 row 5 | o o o d d d o o o o o o 1865 ------------------- 1866 .ve 1867 1868 Thus, any entries in the d locations are stored in the d (diagonal) 1869 submatrix, and any entries in the o locations are stored in the 1870 o (off-diagonal) submatrix. Note that the d matrix is stored in 1871 MatSeqSBAIJ format and the o submatrix in MATSEQBAIJ format. 1872 1873 Now d_nz should indicate the number of block nonzeros per row in the upper triangular 1874 plus the diagonal part of the d matrix, 1875 and o_nz should indicate the number of block nonzeros per row in the o matrix. 1876 In general, for PDE problems in which most nonzeros are near the diagonal, 1877 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 1878 or you will get TERRIBLE performance; see the users' manual chapter on 1879 matrices. 1880 1881 Level: intermediate 1882 1883 .keywords: matrix, block, aij, compressed row, sparse, parallel 1884 1885 .seealso: MatCreate(), MatCreateSeqSBAIJ(), MatSetValues(), MatCreateMPIBAIJ() 1886 @*/ 1887 int MatMPISBAIJSetPreallocation(Mat B,int bs,int d_nz,const int d_nnz[],int o_nz,const int o_nnz[]) 1888 { 1889 int ierr,(*f)(Mat,int,int,const int[],int,const int[]); 1890 1891 PetscFunctionBegin; 1892 ierr = PetscObjectQueryFunction((PetscObject)B,"MatMPISBAIJSetPreallocation_C",(void (**)(void))&f);CHKERRQ(ierr); 1893 if (f) { 1894 ierr = (*f)(B,bs,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 1895 } 1896 PetscFunctionReturn(0); 1897 } 1898 1899 #undef __FUNCT__ 1900 #define __FUNCT__ "MatCreateMPISBAIJ" 1901 /*@C 1902 MatCreateMPISBAIJ - Creates a sparse parallel matrix in symmetric block AIJ format 1903 (block compressed row). For good matrix assembly performance 1904 the user should preallocate the matrix storage by setting the parameters 1905 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 1906 performance can be increased by more than a factor of 50. 1907 1908 Collective on MPI_Comm 1909 1910 Input Parameters: 1911 + comm - MPI communicator 1912 . bs - size of blockk 1913 . m - number of local rows (or PETSC_DECIDE to have calculated if M is given) 1914 This value should be the same as the local size used in creating the 1915 y vector for the matrix-vector product y = Ax. 1916 . n - number of local columns (or PETSC_DECIDE to have calculated if N is given) 1917 This value should be the same as the local size used in creating the 1918 x vector for the matrix-vector product y = Ax. 1919 . M - number of global rows (or PETSC_DETERMINE to have calculated if m is given) 1920 . N - number of global columns (or PETSC_DETERMINE to have calculated if n is given) 1921 . d_nz - number of block nonzeros per block row in diagonal portion of local 1922 submatrix (same for all local rows) 1923 . d_nnz - array containing the number of block nonzeros in the various block rows 1924 in the upper triangular portion of the in diagonal portion of the local 1925 (possibly different for each block block row) or PETSC_NULL. 1926 You must leave room for the diagonal entry even if it is zero. 1927 . o_nz - number of block nonzeros per block row in the off-diagonal portion of local 1928 submatrix (same for all local rows). 1929 - o_nnz - array containing the number of nonzeros in the various block rows of the 1930 off-diagonal portion of the local submatrix (possibly different for 1931 each block row) or PETSC_NULL. 1932 1933 Output Parameter: 1934 . A - the matrix 1935 1936 Options Database Keys: 1937 . -mat_no_unroll - uses code that does not unroll the loops in the 1938 block calculations (much slower) 1939 . -mat_block_size - size of the blocks to use 1940 . -mat_mpi - use the parallel matrix data structures even on one processor 1941 (defaults to using SeqBAIJ format on one processor) 1942 1943 Notes: 1944 The user MUST specify either the local or global matrix dimensions 1945 (possibly both). 1946 1947 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 1948 than it must be used on all processors that share the object for that argument. 1949 1950 Storage Information: 1951 For a square global matrix we define each processor's diagonal portion 1952 to be its local rows and the corresponding columns (a square submatrix); 1953 each processor's off-diagonal portion encompasses the remainder of the 1954 local matrix (a rectangular submatrix). 1955 1956 The user can specify preallocated storage for the diagonal part of 1957 the local submatrix with either d_nz or d_nnz (not both). Set 1958 d_nz=PETSC_DEFAULT and d_nnz=PETSC_NULL for PETSc to control dynamic 1959 memory allocation. Likewise, specify preallocated storage for the 1960 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 1961 1962 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 1963 the figure below we depict these three local rows and all columns (0-11). 1964 1965 .vb 1966 0 1 2 3 4 5 6 7 8 9 10 11 1967 ------------------- 1968 row 3 | o o o d d d o o o o o o 1969 row 4 | o o o d d d o o o o o o 1970 row 5 | o o o d d d o o o o o o 1971 ------------------- 1972 .ve 1973 1974 Thus, any entries in the d locations are stored in the d (diagonal) 1975 submatrix, and any entries in the o locations are stored in the 1976 o (off-diagonal) submatrix. Note that the d matrix is stored in 1977 MatSeqSBAIJ format and the o submatrix in MATSEQBAIJ format. 1978 1979 Now d_nz should indicate the number of block nonzeros per row in the upper triangular 1980 plus the diagonal part of the d matrix, 1981 and o_nz should indicate the number of block nonzeros per row in the o matrix. 1982 In general, for PDE problems in which most nonzeros are near the diagonal, 1983 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 1984 or you will get TERRIBLE performance; see the users' manual chapter on 1985 matrices. 1986 1987 Level: intermediate 1988 1989 .keywords: matrix, block, aij, compressed row, sparse, parallel 1990 1991 .seealso: MatCreate(), MatCreateSeqSBAIJ(), MatSetValues(), MatCreateMPIBAIJ() 1992 @*/ 1993 1994 int MatCreateMPISBAIJ(MPI_Comm comm,int bs,int m,int n,int M,int N,int d_nz,const int d_nnz[],int o_nz,const int o_nnz[],Mat *A) 1995 { 1996 int ierr,size; 1997 1998 PetscFunctionBegin; 1999 ierr = MatCreate(comm,m,n,M,N,A);CHKERRQ(ierr); 2000 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2001 if (size > 1) { 2002 ierr = MatSetType(*A,MATMPISBAIJ);CHKERRQ(ierr); 2003 ierr = MatMPISBAIJSetPreallocation(*A,bs,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 2004 } else { 2005 ierr = MatSetType(*A,MATSEQSBAIJ);CHKERRQ(ierr); 2006 ierr = MatSeqSBAIJSetPreallocation(*A,bs,d_nz,d_nnz);CHKERRQ(ierr); 2007 } 2008 PetscFunctionReturn(0); 2009 } 2010 2011 2012 #undef __FUNCT__ 2013 #define __FUNCT__ "MatDuplicate_MPISBAIJ" 2014 static int MatDuplicate_MPISBAIJ(Mat matin,MatDuplicateOption cpvalues,Mat *newmat) 2015 { 2016 Mat mat; 2017 Mat_MPISBAIJ *a,*oldmat = (Mat_MPISBAIJ*)matin->data; 2018 int ierr,len=0; 2019 2020 PetscFunctionBegin; 2021 *newmat = 0; 2022 ierr = MatCreate(matin->comm,matin->m,matin->n,matin->M,matin->N,&mat);CHKERRQ(ierr); 2023 ierr = MatSetType(mat,MATMPISBAIJ);CHKERRQ(ierr); 2024 mat->preallocated = PETSC_TRUE; 2025 a = (Mat_MPISBAIJ*)mat->data; 2026 a->bs = oldmat->bs; 2027 a->bs2 = oldmat->bs2; 2028 a->mbs = oldmat->mbs; 2029 a->nbs = oldmat->nbs; 2030 a->Mbs = oldmat->Mbs; 2031 a->Nbs = oldmat->Nbs; 2032 2033 a->rstart = oldmat->rstart; 2034 a->rend = oldmat->rend; 2035 a->cstart = oldmat->cstart; 2036 a->cend = oldmat->cend; 2037 a->size = oldmat->size; 2038 a->rank = oldmat->rank; 2039 a->donotstash = oldmat->donotstash; 2040 a->roworiented = oldmat->roworiented; 2041 a->rowindices = 0; 2042 a->rowvalues = 0; 2043 a->getrowactive = PETSC_FALSE; 2044 a->barray = 0; 2045 a->rstart_bs = oldmat->rstart_bs; 2046 a->rend_bs = oldmat->rend_bs; 2047 a->cstart_bs = oldmat->cstart_bs; 2048 a->cend_bs = oldmat->cend_bs; 2049 2050 /* hash table stuff */ 2051 a->ht = 0; 2052 a->hd = 0; 2053 a->ht_size = 0; 2054 a->ht_flag = oldmat->ht_flag; 2055 a->ht_fact = oldmat->ht_fact; 2056 a->ht_total_ct = 0; 2057 a->ht_insert_ct = 0; 2058 2059 ierr = PetscMalloc(3*(a->size+2)*sizeof(int),&a->rowners);CHKERRQ(ierr); 2060 PetscLogObjectMemory(mat,3*(a->size+2)*sizeof(int)+sizeof(struct _p_Mat)+sizeof(Mat_MPISBAIJ)); 2061 a->cowners = a->rowners + a->size + 2; 2062 a->rowners_bs = a->cowners + a->size + 2; 2063 ierr = PetscMemcpy(a->rowners,oldmat->rowners,3*(a->size+2)*sizeof(int));CHKERRQ(ierr); 2064 ierr = MatStashCreate_Private(matin->comm,1,&mat->stash);CHKERRQ(ierr); 2065 ierr = MatStashCreate_Private(matin->comm,oldmat->bs,&mat->bstash);CHKERRQ(ierr); 2066 if (oldmat->colmap) { 2067 #if defined (PETSC_USE_CTABLE) 2068 ierr = PetscTableCreateCopy(oldmat->colmap,&a->colmap);CHKERRQ(ierr); 2069 #else 2070 ierr = PetscMalloc((a->Nbs)*sizeof(int),&a->colmap);CHKERRQ(ierr); 2071 PetscLogObjectMemory(mat,(a->Nbs)*sizeof(int)); 2072 ierr = PetscMemcpy(a->colmap,oldmat->colmap,(a->Nbs)*sizeof(int));CHKERRQ(ierr); 2073 #endif 2074 } else a->colmap = 0; 2075 if (oldmat->garray && (len = ((Mat_SeqBAIJ*)(oldmat->B->data))->nbs)) { 2076 ierr = PetscMalloc(len*sizeof(int),&a->garray);CHKERRQ(ierr); 2077 PetscLogObjectMemory(mat,len*sizeof(int)); 2078 ierr = PetscMemcpy(a->garray,oldmat->garray,len*sizeof(int));CHKERRQ(ierr); 2079 } else a->garray = 0; 2080 2081 ierr = VecDuplicate(oldmat->lvec,&a->lvec);CHKERRQ(ierr); 2082 PetscLogObjectParent(mat,a->lvec); 2083 ierr = VecScatterCopy(oldmat->Mvctx,&a->Mvctx);CHKERRQ(ierr); 2084 2085 PetscLogObjectParent(mat,a->Mvctx); 2086 ierr = MatDuplicate(oldmat->A,cpvalues,&a->A);CHKERRQ(ierr); 2087 PetscLogObjectParent(mat,a->A); 2088 ierr = MatDuplicate(oldmat->B,cpvalues,&a->B);CHKERRQ(ierr); 2089 PetscLogObjectParent(mat,a->B); 2090 ierr = PetscFListDuplicate(mat->qlist,&matin->qlist);CHKERRQ(ierr); 2091 *newmat = mat; 2092 PetscFunctionReturn(0); 2093 } 2094 2095 #include "petscsys.h" 2096 2097 #undef __FUNCT__ 2098 #define __FUNCT__ "MatLoad_MPISBAIJ" 2099 int MatLoad_MPISBAIJ(PetscViewer viewer,const MatType type,Mat *newmat) 2100 { 2101 Mat A; 2102 int i,nz,ierr,j,rstart,rend,fd; 2103 PetscScalar *vals,*buf; 2104 MPI_Comm comm = ((PetscObject)viewer)->comm; 2105 MPI_Status status; 2106 int header[4],rank,size,*rowlengths = 0,M,N,m,*rowners,*browners,maxnz,*cols; 2107 int *locrowlens,*sndcounts = 0,*procsnz = 0,jj,*mycols,*ibuf; 2108 int tag = ((PetscObject)viewer)->tag,bs=1,Mbs,mbs,extra_rows; 2109 int *dlens,*odlens,*mask,*masked1,*masked2,rowcount,odcount; 2110 int dcount,kmax,k,nzcount,tmp; 2111 2112 PetscFunctionBegin; 2113 ierr = PetscOptionsGetInt(PETSC_NULL,"-matload_block_size",&bs,PETSC_NULL);CHKERRQ(ierr); 2114 2115 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2116 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 2117 if (!rank) { 2118 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 2119 ierr = PetscBinaryRead(fd,(char *)header,4,PETSC_INT);CHKERRQ(ierr); 2120 if (header[0] != MAT_FILE_COOKIE) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"not matrix object"); 2121 if (header[3] < 0) { 2122 SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"Matrix stored in special format, cannot load as MPISBAIJ"); 2123 } 2124 } 2125 2126 ierr = MPI_Bcast(header+1,3,MPI_INT,0,comm);CHKERRQ(ierr); 2127 M = header[1]; N = header[2]; 2128 2129 if (M != N) SETERRQ(PETSC_ERR_SUP,"Can only do square matrices"); 2130 2131 /* 2132 This code adds extra rows to make sure the number of rows is 2133 divisible by the blocksize 2134 */ 2135 Mbs = M/bs; 2136 extra_rows = bs - M + bs*(Mbs); 2137 if (extra_rows == bs) extra_rows = 0; 2138 else Mbs++; 2139 if (extra_rows &&!rank) { 2140 PetscLogInfo(0,"MatLoad_MPISBAIJ:Padding loaded matrix to match blocksize\n"); 2141 } 2142 2143 /* determine ownership of all rows */ 2144 mbs = Mbs/size + ((Mbs % size) > rank); 2145 m = mbs*bs; 2146 ierr = PetscMalloc(2*(size+2)*sizeof(int),&rowners);CHKERRQ(ierr); 2147 browners = rowners + size + 1; 2148 ierr = MPI_Allgather(&mbs,1,MPI_INT,rowners+1,1,MPI_INT,comm);CHKERRQ(ierr); 2149 rowners[0] = 0; 2150 for (i=2; i<=size; i++) rowners[i] += rowners[i-1]; 2151 for (i=0; i<=size; i++) browners[i] = rowners[i]*bs; 2152 rstart = rowners[rank]; 2153 rend = rowners[rank+1]; 2154 2155 /* distribute row lengths to all processors */ 2156 ierr = PetscMalloc((rend-rstart)*bs*sizeof(int),&locrowlens);CHKERRQ(ierr); 2157 if (!rank) { 2158 ierr = PetscMalloc((M+extra_rows)*sizeof(int),&rowlengths);CHKERRQ(ierr); 2159 ierr = PetscBinaryRead(fd,rowlengths,M,PETSC_INT);CHKERRQ(ierr); 2160 for (i=0; i<extra_rows; i++) rowlengths[M+i] = 1; 2161 ierr = PetscMalloc(size*sizeof(int),&sndcounts);CHKERRQ(ierr); 2162 for (i=0; i<size; i++) sndcounts[i] = browners[i+1] - browners[i]; 2163 ierr = MPI_Scatterv(rowlengths,sndcounts,browners,MPI_INT,locrowlens,(rend-rstart)*bs,MPI_INT,0,comm);CHKERRQ(ierr); 2164 ierr = PetscFree(sndcounts);CHKERRQ(ierr); 2165 } else { 2166 ierr = MPI_Scatterv(0,0,0,MPI_INT,locrowlens,(rend-rstart)*bs,MPI_INT,0,comm);CHKERRQ(ierr); 2167 } 2168 2169 if (!rank) { /* procs[0] */ 2170 /* calculate the number of nonzeros on each processor */ 2171 ierr = PetscMalloc(size*sizeof(int),&procsnz);CHKERRQ(ierr); 2172 ierr = PetscMemzero(procsnz,size*sizeof(int));CHKERRQ(ierr); 2173 for (i=0; i<size; i++) { 2174 for (j=rowners[i]*bs; j< rowners[i+1]*bs; j++) { 2175 procsnz[i] += rowlengths[j]; 2176 } 2177 } 2178 ierr = PetscFree(rowlengths);CHKERRQ(ierr); 2179 2180 /* determine max buffer needed and allocate it */ 2181 maxnz = 0; 2182 for (i=0; i<size; i++) { 2183 maxnz = PetscMax(maxnz,procsnz[i]); 2184 } 2185 ierr = PetscMalloc(maxnz*sizeof(int),&cols);CHKERRQ(ierr); 2186 2187 /* read in my part of the matrix column indices */ 2188 nz = procsnz[0]; 2189 ierr = PetscMalloc(nz*sizeof(int),&ibuf);CHKERRQ(ierr); 2190 mycols = ibuf; 2191 if (size == 1) nz -= extra_rows; 2192 ierr = PetscBinaryRead(fd,mycols,nz,PETSC_INT);CHKERRQ(ierr); 2193 if (size == 1) for (i=0; i< extra_rows; i++) { mycols[nz+i] = M+i; } 2194 2195 /* read in every ones (except the last) and ship off */ 2196 for (i=1; i<size-1; i++) { 2197 nz = procsnz[i]; 2198 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 2199 ierr = MPI_Send(cols,nz,MPI_INT,i,tag,comm);CHKERRQ(ierr); 2200 } 2201 /* read in the stuff for the last proc */ 2202 if (size != 1) { 2203 nz = procsnz[size-1] - extra_rows; /* the extra rows are not on the disk */ 2204 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 2205 for (i=0; i<extra_rows; i++) cols[nz+i] = M+i; 2206 ierr = MPI_Send(cols,nz+extra_rows,MPI_INT,size-1,tag,comm);CHKERRQ(ierr); 2207 } 2208 ierr = PetscFree(cols);CHKERRQ(ierr); 2209 } else { /* procs[i], i>0 */ 2210 /* determine buffer space needed for message */ 2211 nz = 0; 2212 for (i=0; i<m; i++) { 2213 nz += locrowlens[i]; 2214 } 2215 ierr = PetscMalloc(nz*sizeof(int),&ibuf);CHKERRQ(ierr); 2216 mycols = ibuf; 2217 /* receive message of column indices*/ 2218 ierr = MPI_Recv(mycols,nz,MPI_INT,0,tag,comm,&status);CHKERRQ(ierr); 2219 ierr = MPI_Get_count(&status,MPI_INT,&maxnz);CHKERRQ(ierr); 2220 if (maxnz != nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file"); 2221 } 2222 2223 /* loop over local rows, determining number of off diagonal entries */ 2224 ierr = PetscMalloc(2*(rend-rstart+1)*sizeof(int),&dlens);CHKERRQ(ierr); 2225 odlens = dlens + (rend-rstart); 2226 ierr = PetscMalloc(3*Mbs*sizeof(int),&mask);CHKERRQ(ierr); 2227 ierr = PetscMemzero(mask,3*Mbs*sizeof(int));CHKERRQ(ierr); 2228 masked1 = mask + Mbs; 2229 masked2 = masked1 + Mbs; 2230 rowcount = 0; nzcount = 0; 2231 for (i=0; i<mbs; i++) { 2232 dcount = 0; 2233 odcount = 0; 2234 for (j=0; j<bs; j++) { 2235 kmax = locrowlens[rowcount]; 2236 for (k=0; k<kmax; k++) { 2237 tmp = mycols[nzcount++]/bs; /* block col. index */ 2238 if (!mask[tmp]) { 2239 mask[tmp] = 1; 2240 if (tmp < rstart || tmp >= rend) masked2[odcount++] = tmp; /* entry in off-diag portion */ 2241 else masked1[dcount++] = tmp; /* entry in diag portion */ 2242 } 2243 } 2244 rowcount++; 2245 } 2246 2247 dlens[i] = dcount; /* d_nzz[i] */ 2248 odlens[i] = odcount; /* o_nzz[i] */ 2249 2250 /* zero out the mask elements we set */ 2251 for (j=0; j<dcount; j++) mask[masked1[j]] = 0; 2252 for (j=0; j<odcount; j++) mask[masked2[j]] = 0; 2253 } 2254 2255 /* create our matrix */ 2256 ierr = MatCreate(comm,m,m,PETSC_DETERMINE,PETSC_DETERMINE,&A);CHKERRQ(ierr); 2257 ierr = MatSetType(A,type);CHKERRQ(ierr); 2258 ierr = MatMPISBAIJSetPreallocation(A,bs,0,dlens,0,odlens);CHKERRQ(ierr); 2259 ierr = MatSetOption(A,MAT_COLUMNS_SORTED);CHKERRQ(ierr); 2260 2261 if (!rank) { 2262 ierr = PetscMalloc(maxnz*sizeof(PetscScalar),&buf);CHKERRQ(ierr); 2263 /* read in my part of the matrix numerical values */ 2264 nz = procsnz[0]; 2265 vals = buf; 2266 mycols = ibuf; 2267 if (size == 1) nz -= extra_rows; 2268 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2269 if (size == 1) for (i=0; i< extra_rows; i++) { vals[nz+i] = 1.0; } 2270 2271 /* insert into matrix */ 2272 jj = rstart*bs; 2273 for (i=0; i<m; i++) { 2274 ierr = MatSetValues(A,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 2275 mycols += locrowlens[i]; 2276 vals += locrowlens[i]; 2277 jj++; 2278 } 2279 2280 /* read in other processors (except the last one) and ship out */ 2281 for (i=1; i<size-1; i++) { 2282 nz = procsnz[i]; 2283 vals = buf; 2284 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2285 ierr = MPI_Send(vals,nz,MPIU_SCALAR,i,A->tag,comm);CHKERRQ(ierr); 2286 } 2287 /* the last proc */ 2288 if (size != 1){ 2289 nz = procsnz[i] - extra_rows; 2290 vals = buf; 2291 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 2292 for (i=0; i<extra_rows; i++) vals[nz+i] = 1.0; 2293 ierr = MPI_Send(vals,nz+extra_rows,MPIU_SCALAR,size-1,A->tag,comm);CHKERRQ(ierr); 2294 } 2295 ierr = PetscFree(procsnz);CHKERRQ(ierr); 2296 2297 } else { 2298 /* receive numeric values */ 2299 ierr = PetscMalloc(nz*sizeof(PetscScalar),&buf);CHKERRQ(ierr); 2300 2301 /* receive message of values*/ 2302 vals = buf; 2303 mycols = ibuf; 2304 ierr = MPI_Recv(vals,nz,MPIU_SCALAR,0,A->tag,comm,&status);CHKERRQ(ierr); 2305 ierr = MPI_Get_count(&status,MPIU_SCALAR,&maxnz);CHKERRQ(ierr); 2306 if (maxnz != nz) SETERRQ(PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file"); 2307 2308 /* insert into matrix */ 2309 jj = rstart*bs; 2310 for (i=0; i<m; i++) { 2311 ierr = MatSetValues_MPISBAIJ(A,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 2312 mycols += locrowlens[i]; 2313 vals += locrowlens[i]; 2314 jj++; 2315 } 2316 } 2317 2318 ierr = PetscFree(locrowlens);CHKERRQ(ierr); 2319 ierr = PetscFree(buf);CHKERRQ(ierr); 2320 ierr = PetscFree(ibuf);CHKERRQ(ierr); 2321 ierr = PetscFree(rowners);CHKERRQ(ierr); 2322 ierr = PetscFree(dlens);CHKERRQ(ierr); 2323 ierr = PetscFree(mask);CHKERRQ(ierr); 2324 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2325 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2326 *newmat = A; 2327 PetscFunctionReturn(0); 2328 } 2329 2330 #undef __FUNCT__ 2331 #define __FUNCT__ "MatMPISBAIJSetHashTableFactor" 2332 /*@ 2333 MatMPISBAIJSetHashTableFactor - Sets the factor required to compute the size of the HashTable. 2334 2335 Input Parameters: 2336 . mat - the matrix 2337 . fact - factor 2338 2339 Collective on Mat 2340 2341 Level: advanced 2342 2343 Notes: 2344 This can also be set by the command line option: -mat_use_hash_table fact 2345 2346 .keywords: matrix, hashtable, factor, HT 2347 2348 .seealso: MatSetOption() 2349 @*/ 2350 int MatMPISBAIJSetHashTableFactor(Mat mat,PetscReal fact) 2351 { 2352 PetscFunctionBegin; 2353 SETERRQ(1,"Function not yet written for SBAIJ format"); 2354 /* PetscFunctionReturn(0); */ 2355 } 2356 2357 #undef __FUNCT__ 2358 #define __FUNCT__ "MatGetRowMax_MPISBAIJ" 2359 int MatGetRowMax_MPISBAIJ(Mat A,Vec v) 2360 { 2361 Mat_MPISBAIJ *a = (Mat_MPISBAIJ*)A->data; 2362 Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(a->B)->data; 2363 PetscReal atmp; 2364 PetscReal *work,*svalues,*rvalues; 2365 int ierr,i,bs,mbs,*bi,*bj,brow,j,ncols,krow,kcol,col,row,Mbs,bcol; 2366 int rank,size,*rowners_bs,dest,count,source; 2367 PetscScalar *va; 2368 MatScalar *ba; 2369 MPI_Status stat; 2370 2371 PetscFunctionBegin; 2372 ierr = MatGetRowMax(a->A,v);CHKERRQ(ierr); 2373 ierr = VecGetArray(v,&va);CHKERRQ(ierr); 2374 2375 ierr = MPI_Comm_size(A->comm,&size);CHKERRQ(ierr); 2376 ierr = MPI_Comm_rank(A->comm,&rank);CHKERRQ(ierr); 2377 2378 bs = a->bs; 2379 mbs = a->mbs; 2380 Mbs = a->Mbs; 2381 ba = b->a; 2382 bi = b->i; 2383 bj = b->j; 2384 /* 2385 PetscSynchronizedPrintf(A->comm,"[%d] M: %d, bs: %d, mbs: %d \n",rank,bs*Mbs,bs,mbs); 2386 PetscSynchronizedFlush(A->comm); 2387 */ 2388 2389 /* find ownerships */ 2390 rowners_bs = a->rowners_bs; 2391 /* 2392 if (!rank){ 2393 for (i=0; i<size+1; i++) PetscPrintf(PETSC_COMM_SELF," rowners_bs[%d]: %d\n",i,rowners_bs[i]); 2394 } 2395 */ 2396 2397 /* each proc creates an array to be distributed */ 2398 ierr = PetscMalloc(bs*Mbs*sizeof(PetscReal),&work);CHKERRQ(ierr); 2399 ierr = PetscMemzero(work,bs*Mbs*sizeof(PetscReal));CHKERRQ(ierr); 2400 2401 /* row_max for B */ 2402 if (rank != size-1){ 2403 for (i=0; i<mbs; i++) { 2404 ncols = bi[1] - bi[0]; bi++; 2405 brow = bs*i; 2406 for (j=0; j<ncols; j++){ 2407 bcol = bs*(*bj); 2408 for (kcol=0; kcol<bs; kcol++){ 2409 col = bcol + kcol; /* local col index */ 2410 col += rowners_bs[rank+1]; /* global col index */ 2411 /* PetscPrintf(PETSC_COMM_SELF,"[%d], col: %d\n",rank,col); */ 2412 for (krow=0; krow<bs; krow++){ 2413 atmp = PetscAbsScalar(*ba); ba++; 2414 row = brow + krow; /* local row index */ 2415 /* printf("val[%d,%d]: %g\n",row,col,atmp); */ 2416 if (PetscRealPart(va[row]) < atmp) va[row] = atmp; 2417 if (work[col] < atmp) work[col] = atmp; 2418 } 2419 } 2420 bj++; 2421 } 2422 } 2423 /* 2424 PetscPrintf(PETSC_COMM_SELF,"[%d], work: ",rank); 2425 for (i=0; i<bs*Mbs; i++) PetscPrintf(PETSC_COMM_SELF,"%g ",work[i]); 2426 PetscPrintf(PETSC_COMM_SELF,"[%d]: \n"); 2427 */ 2428 2429 /* send values to its owners */ 2430 for (dest=rank+1; dest<size; dest++){ 2431 svalues = work + rowners_bs[dest]; 2432 count = rowners_bs[dest+1]-rowners_bs[dest]; 2433 ierr = MPI_Send(svalues,count,MPIU_REAL,dest,rank,A->comm);CHKERRQ(ierr); 2434 /* 2435 PetscSynchronizedPrintf(A->comm,"[%d] sends %d values to [%d]: %g, %g, %g, %g\n",rank,count,dest,svalues[0],svalues[1],svalues[2],svalues[3]); 2436 PetscSynchronizedFlush(A->comm); 2437 */ 2438 } 2439 } 2440 2441 /* receive values */ 2442 if (rank){ 2443 rvalues = work; 2444 count = rowners_bs[rank+1]-rowners_bs[rank]; 2445 for (source=0; source<rank; source++){ 2446 ierr = MPI_Recv(rvalues,count,MPIU_REAL,MPI_ANY_SOURCE,MPI_ANY_TAG,A->comm,&stat);CHKERRQ(ierr); 2447 /* process values */ 2448 for (i=0; i<count; i++){ 2449 if (PetscRealPart(va[i]) < rvalues[i]) va[i] = rvalues[i]; 2450 } 2451 /* 2452 PetscSynchronizedPrintf(A->comm,"[%d] received %d values from [%d]: %g, %g, %g, %g \n",rank,count,stat.MPI_SOURCE,rvalues[0],rvalues[1],rvalues[2],rvalues[3]); 2453 PetscSynchronizedFlush(A->comm); 2454 */ 2455 } 2456 } 2457 2458 ierr = VecRestoreArray(v,&va);CHKERRQ(ierr); 2459 ierr = PetscFree(work);CHKERRQ(ierr); 2460 PetscFunctionReturn(0); 2461 } 2462 2463 #undef __FUNCT__ 2464 #define __FUNCT__ "MatRelax_MPISBAIJ" 2465 int MatRelax_MPISBAIJ(Mat matin,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,int its,int lits,Vec xx) 2466 { 2467 Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data; 2468 int ierr,mbs=mat->mbs,bs=mat->bs; 2469 PetscScalar mone=-1.0,*x,*b,*ptr,zero=0.0; 2470 Vec bb1; 2471 2472 PetscFunctionBegin; 2473 if (its <= 0 || lits <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %d and local its %d both positive",its,lits); 2474 if (bs > 1) 2475 SETERRQ(PETSC_ERR_SUP,"SSOR for block size > 1 is not yet implemented"); 2476 2477 if ((flag & SOR_LOCAL_SYMMETRIC_SWEEP) == SOR_LOCAL_SYMMETRIC_SWEEP){ 2478 if ( flag & SOR_ZERO_INITIAL_GUESS ) { 2479 ierr = (*mat->A->ops->relax)(mat->A,bb,omega,flag,fshift,lits,lits,xx);CHKERRQ(ierr); 2480 its--; 2481 } 2482 2483 ierr = VecDuplicate(bb,&bb1);CHKERRQ(ierr); 2484 while (its--){ 2485 2486 /* lower triangular part: slvec0b = - B^T*xx */ 2487 ierr = (*mat->B->ops->multtranspose)(mat->B,xx,mat->slvec0b);CHKERRQ(ierr); 2488 2489 /* copy xx into slvec0a */ 2490 ierr = VecGetArray(mat->slvec0,&ptr);CHKERRQ(ierr); 2491 ierr = VecGetArray(xx,&x);CHKERRQ(ierr); 2492 ierr = PetscMemcpy(ptr,x,bs*mbs*sizeof(MatScalar));CHKERRQ(ierr); 2493 ierr = VecRestoreArray(mat->slvec0,&ptr);CHKERRQ(ierr); 2494 2495 ierr = VecScale(&mone,mat->slvec0);CHKERRQ(ierr); 2496 2497 /* copy bb into slvec1a */ 2498 ierr = VecGetArray(mat->slvec1,&ptr);CHKERRQ(ierr); 2499 ierr = VecGetArray(bb,&b);CHKERRQ(ierr); 2500 ierr = PetscMemcpy(ptr,b,bs*mbs*sizeof(MatScalar));CHKERRQ(ierr); 2501 ierr = VecRestoreArray(mat->slvec1,&ptr);CHKERRQ(ierr); 2502 2503 /* set slvec1b = 0 */ 2504 ierr = VecSet(&zero,mat->slvec1b);CHKERRQ(ierr); 2505 2506 ierr = VecScatterBegin(mat->slvec0,mat->slvec1,ADD_VALUES,SCATTER_FORWARD,mat->sMvctx);CHKERRQ(ierr); 2507 ierr = VecRestoreArray(xx,&x);CHKERRQ(ierr); 2508 ierr = VecRestoreArray(bb,&b);CHKERRQ(ierr); 2509 ierr = VecScatterEnd(mat->slvec0,mat->slvec1,ADD_VALUES,SCATTER_FORWARD,mat->sMvctx);CHKERRQ(ierr); 2510 2511 /* upper triangular part: bb1 = bb1 - B*x */ 2512 ierr = (*mat->B->ops->multadd)(mat->B,mat->slvec1b,mat->slvec1a,bb1);CHKERRQ(ierr); 2513 2514 /* local diagonal sweep */ 2515 ierr = (*mat->A->ops->relax)(mat->A,bb1,omega,SOR_SYMMETRIC_SWEEP,fshift,lits,lits,xx);CHKERRQ(ierr); 2516 } 2517 ierr = VecDestroy(bb1);CHKERRQ(ierr); 2518 } else { 2519 SETERRQ(PETSC_ERR_SUP,"MatSORType is not supported for SBAIJ matrix format"); 2520 } 2521 PetscFunctionReturn(0); 2522 } 2523 2524 #undef __FUNCT__ 2525 #define __FUNCT__ "MatRelax_MPISBAIJ_2comm" 2526 int MatRelax_MPISBAIJ_2comm(Mat matin,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,int its,int lits,Vec xx) 2527 { 2528 Mat_MPISBAIJ *mat = (Mat_MPISBAIJ*)matin->data; 2529 int ierr; 2530 PetscScalar mone=-1.0; 2531 Vec lvec1,bb1; 2532 2533 PetscFunctionBegin; 2534 if (its <= 0 || lits <= 0) SETERRQ2(PETSC_ERR_ARG_WRONG,"Relaxation requires global its %d and local its %d both positive",its,lits); 2535 if (mat->bs > 1) 2536 SETERRQ(PETSC_ERR_SUP,"SSOR for block size > 1 is not yet implemented"); 2537 2538 if ((flag & SOR_LOCAL_SYMMETRIC_SWEEP) == SOR_LOCAL_SYMMETRIC_SWEEP){ 2539 if ( flag & SOR_ZERO_INITIAL_GUESS ) { 2540 ierr = (*mat->A->ops->relax)(mat->A,bb,omega,flag,fshift,lits,lits,xx);CHKERRQ(ierr); 2541 its--; 2542 } 2543 2544 ierr = VecDuplicate(mat->lvec,&lvec1);CHKERRQ(ierr); 2545 ierr = VecDuplicate(bb,&bb1);CHKERRQ(ierr); 2546 while (its--){ 2547 ierr = VecScatterBegin(xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD,mat->Mvctx);CHKERRQ(ierr); 2548 2549 /* lower diagonal part: bb1 = bb - B^T*xx */ 2550 ierr = (*mat->B->ops->multtranspose)(mat->B,xx,lvec1);CHKERRQ(ierr); 2551 ierr = VecScale(&mone,lvec1);CHKERRQ(ierr); 2552 2553 ierr = VecScatterEnd(xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD,mat->Mvctx);CHKERRQ(ierr); 2554 ierr = VecCopy(bb,bb1);CHKERRQ(ierr); 2555 ierr = VecScatterBegin(lvec1,bb1,ADD_VALUES,SCATTER_REVERSE,mat->Mvctx);CHKERRQ(ierr); 2556 2557 /* upper diagonal part: bb1 = bb1 - B*x */ 2558 ierr = VecScale(&mone,mat->lvec);CHKERRQ(ierr); 2559 ierr = (*mat->B->ops->multadd)(mat->B,mat->lvec,bb1,bb1);CHKERRQ(ierr); 2560 2561 ierr = VecScatterEnd(lvec1,bb1,ADD_VALUES,SCATTER_REVERSE,mat->Mvctx);CHKERRQ(ierr); 2562 2563 /* diagonal sweep */ 2564 ierr = (*mat->A->ops->relax)(mat->A,bb1,omega,SOR_SYMMETRIC_SWEEP,fshift,lits,lits,xx);CHKERRQ(ierr); 2565 } 2566 ierr = VecDestroy(lvec1);CHKERRQ(ierr); 2567 ierr = VecDestroy(bb1);CHKERRQ(ierr); 2568 } else { 2569 SETERRQ(PETSC_ERR_SUP,"MatSORType is not supported for SBAIJ matrix format"); 2570 } 2571 PetscFunctionReturn(0); 2572 } 2573 2574