1 2 #include <../src/mat/impls/baij/mpi/mpibaij.h> /*I "petscmat.h" I*/ 3 #include <petscblaslapack.h> 4 5 extern PetscErrorCode MatSetUpMultiply_MPIBAIJ(Mat); 6 extern PetscErrorCode MatDisAssemble_MPIBAIJ(Mat); 7 extern PetscErrorCode MatGetValues_SeqBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt [],PetscScalar []); 8 extern PetscErrorCode MatSetValues_SeqBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt [],const PetscScalar [],InsertMode); 9 extern PetscErrorCode MatSetValuesBlocked_SeqBAIJ(Mat,PetscInt,const PetscInt[],PetscInt,const PetscInt[],const PetscScalar[],InsertMode); 10 extern PetscErrorCode MatGetRow_SeqBAIJ(Mat,PetscInt,PetscInt*,PetscInt*[],PetscScalar*[]); 11 extern PetscErrorCode MatRestoreRow_SeqBAIJ(Mat,PetscInt,PetscInt*,PetscInt*[],PetscScalar*[]); 12 extern PetscErrorCode MatZeroRows_SeqBAIJ(Mat,PetscInt,const PetscInt[],PetscScalar,Vec,Vec); 13 14 #undef __FUNCT__ 15 #define __FUNCT__ "MatGetRowMaxAbs_MPIBAIJ" 16 PetscErrorCode MatGetRowMaxAbs_MPIBAIJ(Mat A,Vec v,PetscInt idx[]) 17 { 18 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 19 PetscErrorCode ierr; 20 PetscInt i,*idxb = 0; 21 PetscScalar *va,*vb; 22 Vec vtmp; 23 24 PetscFunctionBegin; 25 ierr = MatGetRowMaxAbs(a->A,v,idx);CHKERRQ(ierr); 26 ierr = VecGetArray(v,&va);CHKERRQ(ierr); 27 if (idx) { 28 for (i=0; i<A->rmap->n; i++) { 29 if (PetscAbsScalar(va[i])) idx[i] += A->cmap->rstart; 30 } 31 } 32 33 ierr = VecCreateSeq(PETSC_COMM_SELF,A->rmap->n,&vtmp);CHKERRQ(ierr); 34 if (idx) {ierr = PetscMalloc(A->rmap->n*sizeof(PetscInt),&idxb);CHKERRQ(ierr);} 35 ierr = MatGetRowMaxAbs(a->B,vtmp,idxb);CHKERRQ(ierr); 36 ierr = VecGetArray(vtmp,&vb);CHKERRQ(ierr); 37 38 for (i=0; i<A->rmap->n; i++) { 39 if (PetscAbsScalar(va[i]) < PetscAbsScalar(vb[i])) { 40 va[i] = vb[i]; 41 if (idx) idx[i] = A->cmap->bs*a->garray[idxb[i]/A->cmap->bs] + (idxb[i] % A->cmap->bs); 42 } 43 } 44 45 ierr = VecRestoreArray(v,&va);CHKERRQ(ierr); 46 ierr = VecRestoreArray(vtmp,&vb);CHKERRQ(ierr); 47 ierr = PetscFree(idxb);CHKERRQ(ierr); 48 ierr = VecDestroy(&vtmp);CHKERRQ(ierr); 49 PetscFunctionReturn(0); 50 } 51 52 #undef __FUNCT__ 53 #define __FUNCT__ "MatStoreValues_MPIBAIJ" 54 PetscErrorCode MatStoreValues_MPIBAIJ(Mat mat) 55 { 56 Mat_MPIBAIJ *aij = (Mat_MPIBAIJ*)mat->data; 57 PetscErrorCode ierr; 58 59 PetscFunctionBegin; 60 ierr = MatStoreValues(aij->A);CHKERRQ(ierr); 61 ierr = MatStoreValues(aij->B);CHKERRQ(ierr); 62 PetscFunctionReturn(0); 63 } 64 65 #undef __FUNCT__ 66 #define __FUNCT__ "MatRetrieveValues_MPIBAIJ" 67 PetscErrorCode MatRetrieveValues_MPIBAIJ(Mat mat) 68 { 69 Mat_MPIBAIJ *aij = (Mat_MPIBAIJ*)mat->data; 70 PetscErrorCode ierr; 71 72 PetscFunctionBegin; 73 ierr = MatRetrieveValues(aij->A);CHKERRQ(ierr); 74 ierr = MatRetrieveValues(aij->B);CHKERRQ(ierr); 75 PetscFunctionReturn(0); 76 } 77 78 /* 79 Local utility routine that creates a mapping from the global column 80 number to the local number in the off-diagonal part of the local 81 storage of the matrix. This is done in a non scalable way since the 82 length of colmap equals the global matrix length. 83 */ 84 #undef __FUNCT__ 85 #define __FUNCT__ "MatCreateColmap_MPIBAIJ_Private" 86 PetscErrorCode MatCreateColmap_MPIBAIJ_Private(Mat mat) 87 { 88 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 89 Mat_SeqBAIJ *B = (Mat_SeqBAIJ*)baij->B->data; 90 PetscErrorCode ierr; 91 PetscInt nbs = B->nbs,i,bs=mat->rmap->bs; 92 93 PetscFunctionBegin; 94 #if defined(PETSC_USE_CTABLE) 95 ierr = PetscTableCreate(baij->nbs,baij->Nbs+1,&baij->colmap);CHKERRQ(ierr); 96 for (i=0; i<nbs; i++) { 97 ierr = PetscTableAdd(baij->colmap,baij->garray[i]+1,i*bs+1,INSERT_VALUES);CHKERRQ(ierr); 98 } 99 #else 100 ierr = PetscMalloc((baij->Nbs+1)*sizeof(PetscInt),&baij->colmap);CHKERRQ(ierr); 101 ierr = PetscLogObjectMemory((PetscObject)mat,baij->Nbs*sizeof(PetscInt));CHKERRQ(ierr); 102 ierr = PetscMemzero(baij->colmap,baij->Nbs*sizeof(PetscInt));CHKERRQ(ierr); 103 for (i=0; i<nbs; i++) baij->colmap[baij->garray[i]] = i*bs+1; 104 #endif 105 PetscFunctionReturn(0); 106 } 107 108 #define MatSetValues_SeqBAIJ_A_Private(row,col,value,addv) \ 109 { \ 110 \ 111 brow = row/bs; \ 112 rp = aj + ai[brow]; ap = aa + bs2*ai[brow]; \ 113 rmax = aimax[brow]; nrow = ailen[brow]; \ 114 bcol = col/bs; \ 115 ridx = row % bs; cidx = col % bs; \ 116 low = 0; high = nrow; \ 117 while (high-low > 3) { \ 118 t = (low+high)/2; \ 119 if (rp[t] > bcol) high = t; \ 120 else low = t; \ 121 } \ 122 for (_i=low; _i<high; _i++) { \ 123 if (rp[_i] > bcol) break; \ 124 if (rp[_i] == bcol) { \ 125 bap = ap + bs2*_i + bs*cidx + ridx; \ 126 if (addv == ADD_VALUES) *bap += value; \ 127 else *bap = value; \ 128 goto a_noinsert; \ 129 } \ 130 } \ 131 if (a->nonew == 1) goto a_noinsert; \ 132 if (a->nonew == -1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", row, col); \ 133 MatSeqXAIJReallocateAIJ(A,a->mbs,bs2,nrow,brow,bcol,rmax,aa,ai,aj,rp,ap,aimax,a->nonew,MatScalar); \ 134 N = nrow++ - 1; \ 135 /* shift up all the later entries in this row */ \ 136 for (ii=N; ii>=_i; ii--) { \ 137 rp[ii+1] = rp[ii]; \ 138 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 139 } \ 140 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr); } \ 141 rp[_i] = bcol; \ 142 ap[bs2*_i + bs*cidx + ridx] = value; \ 143 a_noinsert:; \ 144 ailen[brow] = nrow; \ 145 } 146 147 #define MatSetValues_SeqBAIJ_B_Private(row,col,value,addv) \ 148 { \ 149 brow = row/bs; \ 150 rp = bj + bi[brow]; ap = ba + bs2*bi[brow]; \ 151 rmax = bimax[brow]; nrow = bilen[brow]; \ 152 bcol = col/bs; \ 153 ridx = row % bs; cidx = col % bs; \ 154 low = 0; high = nrow; \ 155 while (high-low > 3) { \ 156 t = (low+high)/2; \ 157 if (rp[t] > bcol) high = t; \ 158 else low = t; \ 159 } \ 160 for (_i=low; _i<high; _i++) { \ 161 if (rp[_i] > bcol) break; \ 162 if (rp[_i] == bcol) { \ 163 bap = ap + bs2*_i + bs*cidx + ridx; \ 164 if (addv == ADD_VALUES) *bap += value; \ 165 else *bap = value; \ 166 goto b_noinsert; \ 167 } \ 168 } \ 169 if (b->nonew == 1) goto b_noinsert; \ 170 if (b->nonew == -1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", row, col); \ 171 MatSeqXAIJReallocateAIJ(B,b->mbs,bs2,nrow,brow,bcol,rmax,ba,bi,bj,rp,ap,bimax,b->nonew,MatScalar); \ 172 N = nrow++ - 1; \ 173 /* shift up all the later entries in this row */ \ 174 for (ii=N; ii>=_i; ii--) { \ 175 rp[ii+1] = rp[ii]; \ 176 ierr = PetscMemcpy(ap+bs2*(ii+1),ap+bs2*(ii),bs2*sizeof(MatScalar));CHKERRQ(ierr); \ 177 } \ 178 if (N>=_i) { ierr = PetscMemzero(ap+bs2*_i,bs2*sizeof(MatScalar));CHKERRQ(ierr);} \ 179 rp[_i] = bcol; \ 180 ap[bs2*_i + bs*cidx + ridx] = value; \ 181 b_noinsert:; \ 182 bilen[brow] = nrow; \ 183 } 184 185 #undef __FUNCT__ 186 #define __FUNCT__ "MatSetValues_MPIBAIJ" 187 PetscErrorCode MatSetValues_MPIBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 188 { 189 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 190 MatScalar value; 191 PetscBool roworiented = baij->roworiented; 192 PetscErrorCode ierr; 193 PetscInt i,j,row,col; 194 PetscInt rstart_orig=mat->rmap->rstart; 195 PetscInt rend_orig =mat->rmap->rend,cstart_orig=mat->cmap->rstart; 196 PetscInt cend_orig =mat->cmap->rend,bs=mat->rmap->bs; 197 198 /* Some Variables required in the macro */ 199 Mat A = baij->A; 200 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)(A)->data; 201 PetscInt *aimax=a->imax,*ai=a->i,*ailen=a->ilen,*aj=a->j; 202 MatScalar *aa =a->a; 203 204 Mat B = baij->B; 205 Mat_SeqBAIJ *b = (Mat_SeqBAIJ*)(B)->data; 206 PetscInt *bimax=b->imax,*bi=b->i,*bilen=b->ilen,*bj=b->j; 207 MatScalar *ba =b->a; 208 209 PetscInt *rp,ii,nrow,_i,rmax,N,brow,bcol; 210 PetscInt low,high,t,ridx,cidx,bs2=a->bs2; 211 MatScalar *ap,*bap; 212 213 PetscFunctionBegin; 214 if (v) PetscValidScalarPointer(v,6); 215 for (i=0; i<m; i++) { 216 if (im[i] < 0) continue; 217 #if defined(PETSC_USE_DEBUG) 218 if (im[i] >= mat->rmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",im[i],mat->rmap->N-1); 219 #endif 220 if (im[i] >= rstart_orig && im[i] < rend_orig) { 221 row = im[i] - rstart_orig; 222 for (j=0; j<n; j++) { 223 if (in[j] >= cstart_orig && in[j] < cend_orig) { 224 col = in[j] - cstart_orig; 225 if (roworiented) value = v[i*n+j]; 226 else value = v[i+j*m]; 227 MatSetValues_SeqBAIJ_A_Private(row,col,value,addv); 228 /* ierr = MatSetValues_SeqBAIJ(baij->A,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 229 } else if (in[j] < 0) continue; 230 #if defined(PETSC_USE_DEBUG) 231 else if (in[j] >= mat->cmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",in[j],mat->cmap->N-1); 232 #endif 233 else { 234 if (mat->was_assembled) { 235 if (!baij->colmap) { 236 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 237 } 238 #if defined(PETSC_USE_CTABLE) 239 ierr = PetscTableFind(baij->colmap,in[j]/bs + 1,&col);CHKERRQ(ierr); 240 col = col - 1; 241 #else 242 col = baij->colmap[in[j]/bs] - 1; 243 #endif 244 if (col < 0 && !((Mat_SeqBAIJ*)(baij->B->data))->nonew) { 245 ierr = MatDisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 246 col = in[j]; 247 /* Reinitialize the variables required by MatSetValues_SeqBAIJ_B_Private() */ 248 B = baij->B; 249 b = (Mat_SeqBAIJ*)(B)->data; 250 bimax=b->imax;bi=b->i;bilen=b->ilen;bj=b->j; 251 ba =b->a; 252 } else if (col < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", im[i], in[j]); 253 else col += in[j]%bs; 254 } else col = in[j]; 255 if (roworiented) value = v[i*n+j]; 256 else value = v[i+j*m]; 257 MatSetValues_SeqBAIJ_B_Private(row,col,value,addv); 258 /* ierr = MatSetValues_SeqBAIJ(baij->B,1,&row,1,&col,&value,addv);CHKERRQ(ierr); */ 259 } 260 } 261 } else { 262 if (mat->nooffprocentries) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Setting off process row %D even though MatSetOption(,MAT_NO_OFF_PROC_ENTRIES,PETSC_TRUE) was set",im[i]); 263 if (!baij->donotstash) { 264 mat->assembled = PETSC_FALSE; 265 if (roworiented) { 266 ierr = MatStashValuesRow_Private(&mat->stash,im[i],n,in,v+i*n,PETSC_FALSE);CHKERRQ(ierr); 267 } else { 268 ierr = MatStashValuesCol_Private(&mat->stash,im[i],n,in,v+i,m,PETSC_FALSE);CHKERRQ(ierr); 269 } 270 } 271 } 272 } 273 PetscFunctionReturn(0); 274 } 275 276 #undef __FUNCT__ 277 #define __FUNCT__ "MatSetValuesBlocked_MPIBAIJ" 278 PetscErrorCode MatSetValuesBlocked_MPIBAIJ(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 279 { 280 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 281 const PetscScalar *value; 282 MatScalar *barray = baij->barray; 283 PetscBool roworiented = baij->roworiented; 284 PetscErrorCode ierr; 285 PetscInt i,j,ii,jj,row,col,rstart=baij->rstartbs; 286 PetscInt rend=baij->rendbs,cstart=baij->cstartbs,stepval; 287 PetscInt cend=baij->cendbs,bs=mat->rmap->bs,bs2=baij->bs2; 288 289 PetscFunctionBegin; 290 if (!barray) { 291 ierr = PetscMalloc(bs2*sizeof(MatScalar),&barray);CHKERRQ(ierr); 292 baij->barray = barray; 293 } 294 295 if (roworiented) stepval = (n-1)*bs; 296 else stepval = (m-1)*bs; 297 298 for (i=0; i<m; i++) { 299 if (im[i] < 0) continue; 300 #if defined(PETSC_USE_DEBUG) 301 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large, row %D max %D",im[i],baij->Mbs-1); 302 #endif 303 if (im[i] >= rstart && im[i] < rend) { 304 row = im[i] - rstart; 305 for (j=0; j<n; j++) { 306 /* If NumCol = 1 then a copy is not required */ 307 if ((roworiented) && (n == 1)) { 308 barray = (MatScalar*)v + i*bs2; 309 } else if ((!roworiented) && (m == 1)) { 310 barray = (MatScalar*)v + j*bs2; 311 } else { /* Here a copy is required */ 312 if (roworiented) { 313 value = v + (i*(stepval+bs) + j)*bs; 314 } else { 315 value = v + (j*(stepval+bs) + i)*bs; 316 } 317 for (ii=0; ii<bs; ii++,value+=bs+stepval) { 318 for (jj=0; jj<bs; jj++) barray[jj] = value[jj]; 319 barray += bs; 320 } 321 barray -= bs2; 322 } 323 324 if (in[j] >= cstart && in[j] < cend) { 325 col = in[j] - cstart; 326 ierr = MatSetValuesBlocked_SeqBAIJ(baij->A,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 327 } else if (in[j] < 0) continue; 328 #if defined(PETSC_USE_DEBUG) 329 else if (in[j] >= baij->Nbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column too large, col %D max %D",in[j],baij->Nbs-1); 330 #endif 331 else { 332 if (mat->was_assembled) { 333 if (!baij->colmap) { 334 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 335 } 336 337 #if defined(PETSC_USE_DEBUG) 338 #if defined(PETSC_USE_CTABLE) 339 { PetscInt data; 340 ierr = PetscTableFind(baij->colmap,in[j]+1,&data);CHKERRQ(ierr); 341 if ((data - 1) % bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Incorrect colmap"); 342 } 343 #else 344 if ((baij->colmap[in[j]] - 1) % bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Incorrect colmap"); 345 #endif 346 #endif 347 #if defined(PETSC_USE_CTABLE) 348 ierr = PetscTableFind(baij->colmap,in[j]+1,&col);CHKERRQ(ierr); 349 col = (col - 1)/bs; 350 #else 351 col = (baij->colmap[in[j]] - 1)/bs; 352 #endif 353 if (col < 0 && !((Mat_SeqBAIJ*)(baij->B->data))->nonew) { 354 ierr = MatDisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 355 col = in[j]; 356 } else if (col < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Inserting a new nonzero (%D, %D) into matrix", bs*im[i], bs*in[j]); 357 } else col = in[j]; 358 ierr = MatSetValuesBlocked_SeqBAIJ(baij->B,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 359 } 360 } 361 } else { 362 if (mat->nooffprocentries) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Setting off process row %D even though MatSetOption(,MAT_NO_OFF_PROC_ENTRIES,PETSC_TRUE) was set",im[i]); 363 if (!baij->donotstash) { 364 if (roworiented) { 365 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 366 } else { 367 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 368 } 369 } 370 } 371 } 372 PetscFunctionReturn(0); 373 } 374 375 #define HASH_KEY 0.6180339887 376 #define HASH(size,key,tmp) (tmp = (key)*HASH_KEY,(PetscInt)((size)*(tmp-(PetscInt)tmp))) 377 /* #define HASH(size,key) ((PetscInt)((size)*fmod(((key)*HASH_KEY),1))) */ 378 /* #define HASH(size,key,tmp) ((PetscInt)((size)*fmod(((key)*HASH_KEY),1))) */ 379 #undef __FUNCT__ 380 #define __FUNCT__ "MatSetValues_MPIBAIJ_HT" 381 PetscErrorCode MatSetValues_MPIBAIJ_HT(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 382 { 383 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 384 PetscBool roworiented = baij->roworiented; 385 PetscErrorCode ierr; 386 PetscInt i,j,row,col; 387 PetscInt rstart_orig=mat->rmap->rstart; 388 PetscInt rend_orig =mat->rmap->rend,Nbs=baij->Nbs; 389 PetscInt h1,key,size=baij->ht_size,bs=mat->rmap->bs,*HT=baij->ht,idx; 390 PetscReal tmp; 391 MatScalar **HD = baij->hd,value; 392 #if defined(PETSC_USE_DEBUG) 393 PetscInt total_ct=baij->ht_total_ct,insert_ct=baij->ht_insert_ct; 394 #endif 395 396 PetscFunctionBegin; 397 if (v) PetscValidScalarPointer(v,6); 398 for (i=0; i<m; i++) { 399 #if defined(PETSC_USE_DEBUG) 400 if (im[i] < 0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative row"); 401 if (im[i] >= mat->rmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",im[i],mat->rmap->N-1); 402 #endif 403 row = im[i]; 404 if (row >= rstart_orig && row < rend_orig) { 405 for (j=0; j<n; j++) { 406 col = in[j]; 407 if (roworiented) value = v[i*n+j]; 408 else value = v[i+j*m]; 409 /* Look up PetscInto the Hash Table */ 410 key = (row/bs)*Nbs+(col/bs)+1; 411 h1 = HASH(size,key,tmp); 412 413 414 idx = h1; 415 #if defined(PETSC_USE_DEBUG) 416 insert_ct++; 417 total_ct++; 418 if (HT[idx] != key) { 419 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++,total_ct++) ; 420 if (idx == size) { 421 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++,total_ct++) ; 422 if (idx == h1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"(%D,%D) has no entry in the hash table", row, col); 423 } 424 } 425 #else 426 if (HT[idx] != key) { 427 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++) ; 428 if (idx == size) { 429 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++) ; 430 if (idx == h1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"(%D,%D) has no entry in the hash table", row, col); 431 } 432 } 433 #endif 434 /* A HASH table entry is found, so insert the values at the correct address */ 435 if (addv == ADD_VALUES) *(HD[idx]+ (col % bs)*bs + (row % bs)) += value; 436 else *(HD[idx]+ (col % bs)*bs + (row % bs)) = value; 437 } 438 } else if (!baij->donotstash) { 439 if (roworiented) { 440 ierr = MatStashValuesRow_Private(&mat->stash,im[i],n,in,v+i*n,PETSC_FALSE);CHKERRQ(ierr); 441 } else { 442 ierr = MatStashValuesCol_Private(&mat->stash,im[i],n,in,v+i,m,PETSC_FALSE);CHKERRQ(ierr); 443 } 444 } 445 } 446 #if defined(PETSC_USE_DEBUG) 447 baij->ht_total_ct = total_ct; 448 baij->ht_insert_ct = insert_ct; 449 #endif 450 PetscFunctionReturn(0); 451 } 452 453 #undef __FUNCT__ 454 #define __FUNCT__ "MatSetValuesBlocked_MPIBAIJ_HT" 455 PetscErrorCode MatSetValuesBlocked_MPIBAIJ_HT(Mat mat,PetscInt m,const PetscInt im[],PetscInt n,const PetscInt in[],const PetscScalar v[],InsertMode addv) 456 { 457 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 458 PetscBool roworiented = baij->roworiented; 459 PetscErrorCode ierr; 460 PetscInt i,j,ii,jj,row,col; 461 PetscInt rstart=baij->rstartbs; 462 PetscInt rend =mat->rmap->rend,stepval,bs=mat->rmap->bs,bs2=baij->bs2,nbs2=n*bs2; 463 PetscInt h1,key,size=baij->ht_size,idx,*HT=baij->ht,Nbs=baij->Nbs; 464 PetscReal tmp; 465 MatScalar **HD = baij->hd,*baij_a; 466 const PetscScalar *v_t,*value; 467 #if defined(PETSC_USE_DEBUG) 468 PetscInt total_ct=baij->ht_total_ct,insert_ct=baij->ht_insert_ct; 469 #endif 470 471 PetscFunctionBegin; 472 if (roworiented) stepval = (n-1)*bs; 473 else stepval = (m-1)*bs; 474 475 for (i=0; i<m; i++) { 476 #if defined(PETSC_USE_DEBUG) 477 if (im[i] < 0) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %D",im[i]); 478 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",im[i],baij->Mbs-1); 479 #endif 480 row = im[i]; 481 v_t = v + i*nbs2; 482 if (row >= rstart && row < rend) { 483 for (j=0; j<n; j++) { 484 col = in[j]; 485 486 /* Look up into the Hash Table */ 487 key = row*Nbs+col+1; 488 h1 = HASH(size,key,tmp); 489 490 idx = h1; 491 #if defined(PETSC_USE_DEBUG) 492 total_ct++; 493 insert_ct++; 494 if (HT[idx] != key) { 495 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++,total_ct++) ; 496 if (idx == size) { 497 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++,total_ct++) ; 498 if (idx == h1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"(%D,%D) has no entry in the hash table", row, col); 499 } 500 } 501 #else 502 if (HT[idx] != key) { 503 for (idx=h1; (idx<size) && (HT[idx]!=key); idx++) ; 504 if (idx == size) { 505 for (idx=0; (idx<h1) && (HT[idx]!=key); idx++) ; 506 if (idx == h1) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"(%D,%D) has no entry in the hash table", row, col); 507 } 508 } 509 #endif 510 baij_a = HD[idx]; 511 if (roworiented) { 512 /*value = v + i*(stepval+bs)*bs + j*bs;*/ 513 /* value = v + (i*(stepval+bs)+j)*bs; */ 514 value = v_t; 515 v_t += bs; 516 if (addv == ADD_VALUES) { 517 for (ii=0; ii<bs; ii++,value+=stepval) { 518 for (jj=ii; jj<bs2; jj+=bs) { 519 baij_a[jj] += *value++; 520 } 521 } 522 } else { 523 for (ii=0; ii<bs; ii++,value+=stepval) { 524 for (jj=ii; jj<bs2; jj+=bs) { 525 baij_a[jj] = *value++; 526 } 527 } 528 } 529 } else { 530 value = v + j*(stepval+bs)*bs + i*bs; 531 if (addv == ADD_VALUES) { 532 for (ii=0; ii<bs; ii++,value+=stepval,baij_a+=bs) { 533 for (jj=0; jj<bs; jj++) { 534 baij_a[jj] += *value++; 535 } 536 } 537 } else { 538 for (ii=0; ii<bs; ii++,value+=stepval,baij_a+=bs) { 539 for (jj=0; jj<bs; jj++) { 540 baij_a[jj] = *value++; 541 } 542 } 543 } 544 } 545 } 546 } else { 547 if (!baij->donotstash) { 548 if (roworiented) { 549 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 550 } else { 551 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 552 } 553 } 554 } 555 } 556 #if defined(PETSC_USE_DEBUG) 557 baij->ht_total_ct = total_ct; 558 baij->ht_insert_ct = insert_ct; 559 #endif 560 PetscFunctionReturn(0); 561 } 562 563 #undef __FUNCT__ 564 #define __FUNCT__ "MatGetValues_MPIBAIJ" 565 PetscErrorCode MatGetValues_MPIBAIJ(Mat mat,PetscInt m,const PetscInt idxm[],PetscInt n,const PetscInt idxn[],PetscScalar v[]) 566 { 567 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 568 PetscErrorCode ierr; 569 PetscInt bs = mat->rmap->bs,i,j,bsrstart = mat->rmap->rstart,bsrend = mat->rmap->rend; 570 PetscInt bscstart = mat->cmap->rstart,bscend = mat->cmap->rend,row,col,data; 571 572 PetscFunctionBegin; 573 for (i=0; i<m; i++) { 574 if (idxm[i] < 0) continue; /* SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative row: %D",idxm[i]);*/ 575 if (idxm[i] >= mat->rmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large: row %D max %D",idxm[i],mat->rmap->N-1); 576 if (idxm[i] >= bsrstart && idxm[i] < bsrend) { 577 row = idxm[i] - bsrstart; 578 for (j=0; j<n; j++) { 579 if (idxn[j] < 0) continue; /* SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Negative column: %D",idxn[j]); */ 580 if (idxn[j] >= mat->cmap->N) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column too large: col %D max %D",idxn[j],mat->cmap->N-1); 581 if (idxn[j] >= bscstart && idxn[j] < bscend) { 582 col = idxn[j] - bscstart; 583 ierr = MatGetValues_SeqBAIJ(baij->A,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 584 } else { 585 if (!baij->colmap) { 586 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 587 } 588 #if defined(PETSC_USE_CTABLE) 589 ierr = PetscTableFind(baij->colmap,idxn[j]/bs+1,&data);CHKERRQ(ierr); 590 data--; 591 #else 592 data = baij->colmap[idxn[j]/bs]-1; 593 #endif 594 if ((data < 0) || (baij->garray[data/bs] != idxn[j]/bs)) *(v+i*n+j) = 0.0; 595 else { 596 col = data + idxn[j]%bs; 597 ierr = MatGetValues_SeqBAIJ(baij->B,1,&row,1,&col,v+i*n+j);CHKERRQ(ierr); 598 } 599 } 600 } 601 } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only local values currently supported"); 602 } 603 PetscFunctionReturn(0); 604 } 605 606 #undef __FUNCT__ 607 #define __FUNCT__ "MatNorm_MPIBAIJ" 608 PetscErrorCode MatNorm_MPIBAIJ(Mat mat,NormType type,PetscReal *nrm) 609 { 610 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 611 Mat_SeqBAIJ *amat = (Mat_SeqBAIJ*)baij->A->data,*bmat = (Mat_SeqBAIJ*)baij->B->data; 612 PetscErrorCode ierr; 613 PetscInt i,j,bs2=baij->bs2,bs=baij->A->rmap->bs,nz,row,col; 614 PetscReal sum = 0.0; 615 MatScalar *v; 616 617 PetscFunctionBegin; 618 if (baij->size == 1) { 619 ierr = MatNorm(baij->A,type,nrm);CHKERRQ(ierr); 620 } else { 621 if (type == NORM_FROBENIUS) { 622 v = amat->a; 623 nz = amat->nz*bs2; 624 for (i=0; i<nz; i++) { 625 sum += PetscRealPart(PetscConj(*v)*(*v)); v++; 626 } 627 v = bmat->a; 628 nz = bmat->nz*bs2; 629 for (i=0; i<nz; i++) { 630 sum += PetscRealPart(PetscConj(*v)*(*v)); v++; 631 } 632 ierr = MPI_Allreduce(&sum,nrm,1,MPIU_REAL,MPIU_SUM,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 633 *nrm = PetscSqrtReal(*nrm); 634 } else if (type == NORM_1) { /* max column sum */ 635 PetscReal *tmp,*tmp2; 636 PetscInt *jj,*garray=baij->garray,cstart=baij->rstartbs; 637 ierr = PetscMalloc2(mat->cmap->N,PetscReal,&tmp,mat->cmap->N,PetscReal,&tmp2);CHKERRQ(ierr); 638 ierr = PetscMemzero(tmp,mat->cmap->N*sizeof(PetscReal));CHKERRQ(ierr); 639 v = amat->a; jj = amat->j; 640 for (i=0; i<amat->nz; i++) { 641 for (j=0; j<bs; j++) { 642 col = bs*(cstart + *jj) + j; /* column index */ 643 for (row=0; row<bs; row++) { 644 tmp[col] += PetscAbsScalar(*v); v++; 645 } 646 } 647 jj++; 648 } 649 v = bmat->a; jj = bmat->j; 650 for (i=0; i<bmat->nz; i++) { 651 for (j=0; j<bs; j++) { 652 col = bs*garray[*jj] + j; 653 for (row=0; row<bs; row++) { 654 tmp[col] += PetscAbsScalar(*v); v++; 655 } 656 } 657 jj++; 658 } 659 ierr = MPI_Allreduce(tmp,tmp2,mat->cmap->N,MPIU_REAL,MPIU_SUM,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 660 *nrm = 0.0; 661 for (j=0; j<mat->cmap->N; j++) { 662 if (tmp2[j] > *nrm) *nrm = tmp2[j]; 663 } 664 ierr = PetscFree2(tmp,tmp2);CHKERRQ(ierr); 665 } else if (type == NORM_INFINITY) { /* max row sum */ 666 PetscReal *sums; 667 ierr = PetscMalloc(bs*sizeof(PetscReal),&sums);CHKERRQ(ierr); 668 sum = 0.0; 669 for (j=0; j<amat->mbs; j++) { 670 for (row=0; row<bs; row++) sums[row] = 0.0; 671 v = amat->a + bs2*amat->i[j]; 672 nz = amat->i[j+1]-amat->i[j]; 673 for (i=0; i<nz; i++) { 674 for (col=0; col<bs; col++) { 675 for (row=0; row<bs; row++) { 676 sums[row] += PetscAbsScalar(*v); v++; 677 } 678 } 679 } 680 v = bmat->a + bs2*bmat->i[j]; 681 nz = bmat->i[j+1]-bmat->i[j]; 682 for (i=0; i<nz; i++) { 683 for (col=0; col<bs; col++) { 684 for (row=0; row<bs; row++) { 685 sums[row] += PetscAbsScalar(*v); v++; 686 } 687 } 688 } 689 for (row=0; row<bs; row++) { 690 if (sums[row] > sum) sum = sums[row]; 691 } 692 } 693 ierr = MPI_Allreduce(&sum,nrm,1,MPIU_REAL,MPIU_MAX,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 694 ierr = PetscFree(sums);CHKERRQ(ierr); 695 } else SETERRQ(PetscObjectComm((PetscObject)mat),PETSC_ERR_SUP,"No support for this norm yet"); 696 } 697 PetscFunctionReturn(0); 698 } 699 700 /* 701 Creates the hash table, and sets the table 702 This table is created only once. 703 If new entried need to be added to the matrix 704 then the hash table has to be destroyed and 705 recreated. 706 */ 707 #undef __FUNCT__ 708 #define __FUNCT__ "MatCreateHashTable_MPIBAIJ_Private" 709 PetscErrorCode MatCreateHashTable_MPIBAIJ_Private(Mat mat,PetscReal factor) 710 { 711 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 712 Mat A = baij->A,B=baij->B; 713 Mat_SeqBAIJ *a = (Mat_SeqBAIJ*)A->data,*b=(Mat_SeqBAIJ*)B->data; 714 PetscInt i,j,k,nz=a->nz+b->nz,h1,*ai=a->i,*aj=a->j,*bi=b->i,*bj=b->j; 715 PetscErrorCode ierr; 716 PetscInt ht_size,bs2=baij->bs2,rstart=baij->rstartbs; 717 PetscInt cstart=baij->cstartbs,*garray=baij->garray,row,col,Nbs=baij->Nbs; 718 PetscInt *HT,key; 719 MatScalar **HD; 720 PetscReal tmp; 721 #if defined(PETSC_USE_INFO) 722 PetscInt ct=0,max=0; 723 #endif 724 725 PetscFunctionBegin; 726 if (baij->ht) PetscFunctionReturn(0); 727 728 baij->ht_size = (PetscInt)(factor*nz); 729 ht_size = baij->ht_size; 730 731 /* Allocate Memory for Hash Table */ 732 ierr = PetscMalloc2(ht_size,MatScalar*,&baij->hd,ht_size,PetscInt,&baij->ht);CHKERRQ(ierr); 733 ierr = PetscMemzero(baij->hd,ht_size*sizeof(MatScalar*));CHKERRQ(ierr); 734 ierr = PetscMemzero(baij->ht,ht_size*sizeof(PetscInt));CHKERRQ(ierr); 735 HD = baij->hd; 736 HT = baij->ht; 737 738 /* Loop Over A */ 739 for (i=0; i<a->mbs; i++) { 740 for (j=ai[i]; j<ai[i+1]; j++) { 741 row = i+rstart; 742 col = aj[j]+cstart; 743 744 key = row*Nbs + col + 1; 745 h1 = HASH(ht_size,key,tmp); 746 for (k=0; k<ht_size; k++) { 747 if (!HT[(h1+k)%ht_size]) { 748 HT[(h1+k)%ht_size] = key; 749 HD[(h1+k)%ht_size] = a->a + j*bs2; 750 break; 751 #if defined(PETSC_USE_INFO) 752 } else { 753 ct++; 754 #endif 755 } 756 } 757 #if defined(PETSC_USE_INFO) 758 if (k> max) max = k; 759 #endif 760 } 761 } 762 /* Loop Over B */ 763 for (i=0; i<b->mbs; i++) { 764 for (j=bi[i]; j<bi[i+1]; j++) { 765 row = i+rstart; 766 col = garray[bj[j]]; 767 key = row*Nbs + col + 1; 768 h1 = HASH(ht_size,key,tmp); 769 for (k=0; k<ht_size; k++) { 770 if (!HT[(h1+k)%ht_size]) { 771 HT[(h1+k)%ht_size] = key; 772 HD[(h1+k)%ht_size] = b->a + j*bs2; 773 break; 774 #if defined(PETSC_USE_INFO) 775 } else { 776 ct++; 777 #endif 778 } 779 } 780 #if defined(PETSC_USE_INFO) 781 if (k> max) max = k; 782 #endif 783 } 784 } 785 786 /* Print Summary */ 787 #if defined(PETSC_USE_INFO) 788 for (i=0,j=0; i<ht_size; i++) { 789 if (HT[i]) j++; 790 } 791 ierr = PetscInfo2(mat,"Average Search = %5.2f,max search = %D\n",(!j)? 0.0:((PetscReal)(ct+j))/j,max);CHKERRQ(ierr); 792 #endif 793 PetscFunctionReturn(0); 794 } 795 796 #undef __FUNCT__ 797 #define __FUNCT__ "MatAssemblyBegin_MPIBAIJ" 798 PetscErrorCode MatAssemblyBegin_MPIBAIJ(Mat mat,MatAssemblyType mode) 799 { 800 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 801 PetscErrorCode ierr; 802 PetscInt nstash,reallocs; 803 InsertMode addv; 804 805 PetscFunctionBegin; 806 if (baij->donotstash || mat->nooffprocentries) PetscFunctionReturn(0); 807 808 /* make sure all processors are either in INSERTMODE or ADDMODE */ 809 ierr = MPI_Allreduce((PetscEnum*)&mat->insertmode,(PetscEnum*)&addv,1,MPIU_ENUM,MPI_BOR,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 810 if (addv == (ADD_VALUES|INSERT_VALUES)) SETERRQ(PetscObjectComm((PetscObject)mat),PETSC_ERR_ARG_WRONGSTATE,"Some processors inserted others added"); 811 mat->insertmode = addv; /* in case this processor had no cache */ 812 813 ierr = MatStashScatterBegin_Private(mat,&mat->stash,mat->rmap->range);CHKERRQ(ierr); 814 ierr = MatStashScatterBegin_Private(mat,&mat->bstash,baij->rangebs);CHKERRQ(ierr); 815 ierr = MatStashGetInfo_Private(&mat->stash,&nstash,&reallocs);CHKERRQ(ierr); 816 ierr = PetscInfo2(mat,"Stash has %D entries,uses %D mallocs.\n",nstash,reallocs);CHKERRQ(ierr); 817 ierr = MatStashGetInfo_Private(&mat->bstash,&nstash,&reallocs);CHKERRQ(ierr); 818 ierr = PetscInfo2(mat,"Block-Stash has %D entries, uses %D mallocs.\n",nstash,reallocs);CHKERRQ(ierr); 819 PetscFunctionReturn(0); 820 } 821 822 #undef __FUNCT__ 823 #define __FUNCT__ "MatAssemblyEnd_MPIBAIJ" 824 PetscErrorCode MatAssemblyEnd_MPIBAIJ(Mat mat,MatAssemblyType mode) 825 { 826 Mat_MPIBAIJ *baij=(Mat_MPIBAIJ*)mat->data; 827 Mat_SeqBAIJ *a =(Mat_SeqBAIJ*)baij->A->data; 828 PetscErrorCode ierr; 829 PetscInt i,j,rstart,ncols,flg,bs2=baij->bs2; 830 PetscInt *row,*col; 831 PetscBool r1,r2,r3,other_disassembled; 832 MatScalar *val; 833 InsertMode addv = mat->insertmode; 834 PetscMPIInt n; 835 836 PetscFunctionBegin; 837 /* do not use 'b=(Mat_SeqBAIJ*)baij->B->data' as B can be reset in disassembly */ 838 if (!baij->donotstash && !mat->nooffprocentries) { 839 while (1) { 840 ierr = MatStashScatterGetMesg_Private(&mat->stash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 841 if (!flg) break; 842 843 for (i=0; i<n;) { 844 /* Now identify the consecutive vals belonging to the same row */ 845 for (j=i,rstart=row[j]; j<n; j++) { 846 if (row[j] != rstart) break; 847 } 848 if (j < n) ncols = j-i; 849 else ncols = n-i; 850 /* Now assemble all these values with a single function call */ 851 ierr = MatSetValues_MPIBAIJ(mat,1,row+i,ncols,col+i,val+i,addv);CHKERRQ(ierr); 852 i = j; 853 } 854 } 855 ierr = MatStashScatterEnd_Private(&mat->stash);CHKERRQ(ierr); 856 /* Now process the block-stash. Since the values are stashed column-oriented, 857 set the roworiented flag to column oriented, and after MatSetValues() 858 restore the original flags */ 859 r1 = baij->roworiented; 860 r2 = a->roworiented; 861 r3 = ((Mat_SeqBAIJ*)baij->B->data)->roworiented; 862 863 baij->roworiented = PETSC_FALSE; 864 a->roworiented = PETSC_FALSE; 865 866 (((Mat_SeqBAIJ*)baij->B->data))->roworiented = PETSC_FALSE; /* b->roworiented */ 867 while (1) { 868 ierr = MatStashScatterGetMesg_Private(&mat->bstash,&n,&row,&col,&val,&flg);CHKERRQ(ierr); 869 if (!flg) break; 870 871 for (i=0; i<n;) { 872 /* Now identify the consecutive vals belonging to the same row */ 873 for (j=i,rstart=row[j]; j<n; j++) { 874 if (row[j] != rstart) break; 875 } 876 if (j < n) ncols = j-i; 877 else ncols = n-i; 878 ierr = MatSetValuesBlocked_MPIBAIJ(mat,1,row+i,ncols,col+i,val+i*bs2,addv);CHKERRQ(ierr); 879 i = j; 880 } 881 } 882 ierr = MatStashScatterEnd_Private(&mat->bstash);CHKERRQ(ierr); 883 884 baij->roworiented = r1; 885 a->roworiented = r2; 886 887 ((Mat_SeqBAIJ*)baij->B->data)->roworiented = r3; /* b->roworiented */ 888 } 889 890 ierr = MatAssemblyBegin(baij->A,mode);CHKERRQ(ierr); 891 ierr = MatAssemblyEnd(baij->A,mode);CHKERRQ(ierr); 892 893 /* determine if any processor has disassembled, if so we must 894 also disassemble ourselfs, in order that we may reassemble. */ 895 /* 896 if nonzero structure of submatrix B cannot change then we know that 897 no processor disassembled thus we can skip this stuff 898 */ 899 if (!((Mat_SeqBAIJ*)baij->B->data)->nonew) { 900 ierr = MPI_Allreduce(&mat->was_assembled,&other_disassembled,1,MPIU_BOOL,MPI_PROD,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 901 if (mat->was_assembled && !other_disassembled) { 902 ierr = MatDisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 903 } 904 } 905 906 if (!mat->was_assembled && mode == MAT_FINAL_ASSEMBLY) { 907 ierr = MatSetUpMultiply_MPIBAIJ(mat);CHKERRQ(ierr); 908 } 909 ierr = MatSetOption(baij->B,MAT_CHECK_COMPRESSED_ROW,PETSC_FALSE);CHKERRQ(ierr); 910 ierr = MatAssemblyBegin(baij->B,mode);CHKERRQ(ierr); 911 ierr = MatAssemblyEnd(baij->B,mode);CHKERRQ(ierr); 912 913 #if defined(PETSC_USE_INFO) 914 if (baij->ht && mode== MAT_FINAL_ASSEMBLY) { 915 ierr = PetscInfo1(mat,"Average Hash Table Search in MatSetValues = %5.2f\n",((PetscReal)baij->ht_total_ct)/baij->ht_insert_ct);CHKERRQ(ierr); 916 917 baij->ht_total_ct = 0; 918 baij->ht_insert_ct = 0; 919 } 920 #endif 921 if (baij->ht_flag && !baij->ht && mode == MAT_FINAL_ASSEMBLY) { 922 ierr = MatCreateHashTable_MPIBAIJ_Private(mat,baij->ht_fact);CHKERRQ(ierr); 923 924 mat->ops->setvalues = MatSetValues_MPIBAIJ_HT; 925 mat->ops->setvaluesblocked = MatSetValuesBlocked_MPIBAIJ_HT; 926 } 927 928 ierr = PetscFree2(baij->rowvalues,baij->rowindices);CHKERRQ(ierr); 929 930 baij->rowvalues = 0; 931 PetscFunctionReturn(0); 932 } 933 934 #include <petscdraw.h> 935 #undef __FUNCT__ 936 #define __FUNCT__ "MatView_MPIBAIJ_ASCIIorDraworSocket" 937 static PetscErrorCode MatView_MPIBAIJ_ASCIIorDraworSocket(Mat mat,PetscViewer viewer) 938 { 939 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 940 PetscErrorCode ierr; 941 PetscMPIInt size = baij->size,rank = baij->rank; 942 PetscInt bs = mat->rmap->bs; 943 PetscBool iascii,isdraw; 944 PetscViewer sviewer; 945 PetscViewerFormat format; 946 947 PetscFunctionBegin; 948 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&iascii);CHKERRQ(ierr); 949 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERDRAW,&isdraw);CHKERRQ(ierr); 950 if (iascii) { 951 ierr = PetscViewerGetFormat(viewer,&format);CHKERRQ(ierr); 952 if (format == PETSC_VIEWER_ASCII_INFO_DETAIL) { 953 MatInfo info; 954 ierr = MPI_Comm_rank(PetscObjectComm((PetscObject)mat),&rank);CHKERRQ(ierr); 955 ierr = MatGetInfo(mat,MAT_LOCAL,&info);CHKERRQ(ierr); 956 ierr = PetscViewerASCIISynchronizedAllow(viewer,PETSC_TRUE);CHKERRQ(ierr); 957 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] Local rows %D nz %D nz alloced %D bs %D mem %D\n", 958 rank,mat->rmap->n,(PetscInt)info.nz_used,(PetscInt)info.nz_allocated,mat->rmap->bs,(PetscInt)info.memory);CHKERRQ(ierr); 959 ierr = MatGetInfo(baij->A,MAT_LOCAL,&info);CHKERRQ(ierr); 960 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] on-diagonal part: nz %D \n",rank,(PetscInt)info.nz_used);CHKERRQ(ierr); 961 ierr = MatGetInfo(baij->B,MAT_LOCAL,&info);CHKERRQ(ierr); 962 ierr = PetscViewerASCIISynchronizedPrintf(viewer,"[%d] off-diagonal part: nz %D \n",rank,(PetscInt)info.nz_used);CHKERRQ(ierr); 963 ierr = PetscViewerFlush(viewer);CHKERRQ(ierr); 964 ierr = PetscViewerASCIISynchronizedAllow(viewer,PETSC_FALSE);CHKERRQ(ierr); 965 ierr = PetscViewerASCIIPrintf(viewer,"Information on VecScatter used in matrix-vector product: \n");CHKERRQ(ierr); 966 ierr = VecScatterView(baij->Mvctx,viewer);CHKERRQ(ierr); 967 PetscFunctionReturn(0); 968 } else if (format == PETSC_VIEWER_ASCII_INFO) { 969 ierr = PetscViewerASCIIPrintf(viewer," block size is %D\n",bs);CHKERRQ(ierr); 970 PetscFunctionReturn(0); 971 } else if (format == PETSC_VIEWER_ASCII_FACTOR_INFO) { 972 PetscFunctionReturn(0); 973 } 974 } 975 976 if (isdraw) { 977 PetscDraw draw; 978 PetscBool isnull; 979 ierr = PetscViewerDrawGetDraw(viewer,0,&draw);CHKERRQ(ierr); 980 ierr = PetscDrawIsNull(draw,&isnull);CHKERRQ(ierr); if (isnull) PetscFunctionReturn(0); 981 } 982 983 if (size == 1) { 984 ierr = PetscObjectSetName((PetscObject)baij->A,((PetscObject)mat)->name);CHKERRQ(ierr); 985 ierr = MatView(baij->A,viewer);CHKERRQ(ierr); 986 } else { 987 /* assemble the entire matrix onto first processor. */ 988 Mat A; 989 Mat_SeqBAIJ *Aloc; 990 PetscInt M = mat->rmap->N,N = mat->cmap->N,*ai,*aj,col,i,j,k,*rvals,mbs = baij->mbs; 991 MatScalar *a; 992 993 /* Here we are creating a temporary matrix, so will assume MPIBAIJ is acceptable */ 994 /* Perhaps this should be the type of mat? */ 995 ierr = MatCreate(PetscObjectComm((PetscObject)mat),&A);CHKERRQ(ierr); 996 if (!rank) { 997 ierr = MatSetSizes(A,M,N,M,N);CHKERRQ(ierr); 998 } else { 999 ierr = MatSetSizes(A,0,0,M,N);CHKERRQ(ierr); 1000 } 1001 ierr = MatSetType(A,MATMPIBAIJ);CHKERRQ(ierr); 1002 ierr = MatMPIBAIJSetPreallocation(A,mat->rmap->bs,0,NULL,0,NULL);CHKERRQ(ierr); 1003 ierr = MatSetOption(A,MAT_NEW_NONZERO_LOCATION_ERR,PETSC_FALSE);CHKERRQ(ierr); 1004 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)A);CHKERRQ(ierr); 1005 1006 /* copy over the A part */ 1007 Aloc = (Mat_SeqBAIJ*)baij->A->data; 1008 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1009 ierr = PetscMalloc(bs*sizeof(PetscInt),&rvals);CHKERRQ(ierr); 1010 1011 for (i=0; i<mbs; i++) { 1012 rvals[0] = bs*(baij->rstartbs + i); 1013 for (j=1; j<bs; j++) rvals[j] = rvals[j-1] + 1; 1014 for (j=ai[i]; j<ai[i+1]; j++) { 1015 col = (baij->cstartbs+aj[j])*bs; 1016 for (k=0; k<bs; k++) { 1017 ierr = MatSetValues_MPIBAIJ(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 1018 col++; a += bs; 1019 } 1020 } 1021 } 1022 /* copy over the B part */ 1023 Aloc = (Mat_SeqBAIJ*)baij->B->data; 1024 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1025 for (i=0; i<mbs; i++) { 1026 rvals[0] = bs*(baij->rstartbs + i); 1027 for (j=1; j<bs; j++) rvals[j] = rvals[j-1] + 1; 1028 for (j=ai[i]; j<ai[i+1]; j++) { 1029 col = baij->garray[aj[j]]*bs; 1030 for (k=0; k<bs; k++) { 1031 ierr = MatSetValues_MPIBAIJ(A,bs,rvals,1,&col,a,INSERT_VALUES);CHKERRQ(ierr); 1032 col++; a += bs; 1033 } 1034 } 1035 } 1036 ierr = PetscFree(rvals);CHKERRQ(ierr); 1037 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1038 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1039 /* 1040 Everyone has to call to draw the matrix since the graphics waits are 1041 synchronized across all processors that share the PetscDraw object 1042 */ 1043 ierr = PetscViewerGetSingleton(viewer,&sviewer);CHKERRQ(ierr); 1044 if (!rank) { 1045 ierr = PetscObjectSetName((PetscObject)((Mat_MPIBAIJ*)(A->data))->A,((PetscObject)mat)->name);CHKERRQ(ierr); 1046 /* Set the type name to MATMPIBAIJ so that the correct type can be printed out by PetscObjectPrintClassNamePrefixType() in MatView_SeqBAIJ_ASCII()*/ 1047 PetscStrcpy(((PetscObject)((Mat_MPIBAIJ*)(A->data))->A)->type_name,MATMPIBAIJ); 1048 ierr = MatView(((Mat_MPIBAIJ*)(A->data))->A,sviewer);CHKERRQ(ierr); 1049 } 1050 ierr = PetscViewerRestoreSingleton(viewer,&sviewer);CHKERRQ(ierr); 1051 ierr = MatDestroy(&A);CHKERRQ(ierr); 1052 } 1053 PetscFunctionReturn(0); 1054 } 1055 1056 #undef __FUNCT__ 1057 #define __FUNCT__ "MatView_MPIBAIJ_Binary" 1058 static PetscErrorCode MatView_MPIBAIJ_Binary(Mat mat,PetscViewer viewer) 1059 { 1060 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)mat->data; 1061 Mat_SeqBAIJ *A = (Mat_SeqBAIJ*)a->A->data; 1062 Mat_SeqBAIJ *B = (Mat_SeqBAIJ*)a->B->data; 1063 PetscErrorCode ierr; 1064 PetscInt i,*row_lens,*crow_lens,bs = mat->rmap->bs,j,k,bs2=a->bs2,header[4],nz,rlen; 1065 PetscInt *range=0,nzmax,*column_indices,cnt,col,*garray = a->garray,cstart = mat->cmap->rstart/bs,len,pcnt,l,ll; 1066 int fd; 1067 PetscScalar *column_values; 1068 FILE *file; 1069 PetscMPIInt rank,size,tag = ((PetscObject)viewer)->tag; 1070 PetscInt message_count,flowcontrolcount; 1071 1072 PetscFunctionBegin; 1073 ierr = MPI_Comm_rank(PetscObjectComm((PetscObject)mat),&rank);CHKERRQ(ierr); 1074 ierr = MPI_Comm_size(PetscObjectComm((PetscObject)mat),&size);CHKERRQ(ierr); 1075 nz = bs2*(A->nz + B->nz); 1076 rlen = mat->rmap->n; 1077 if (!rank) { 1078 header[0] = MAT_FILE_CLASSID; 1079 header[1] = mat->rmap->N; 1080 header[2] = mat->cmap->N; 1081 1082 ierr = MPI_Reduce(&nz,&header[3],1,MPIU_INT,MPI_SUM,0,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1083 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 1084 ierr = PetscBinaryWrite(fd,header,4,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1085 /* get largest number of rows any processor has */ 1086 range = mat->rmap->range; 1087 for (i=1; i<size; i++) { 1088 rlen = PetscMax(rlen,range[i+1] - range[i]); 1089 } 1090 } else { 1091 ierr = MPI_Reduce(&nz,0,1,MPIU_INT,MPI_SUM,0,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1092 } 1093 1094 ierr = PetscMalloc((rlen/bs)*sizeof(PetscInt),&crow_lens);CHKERRQ(ierr); 1095 /* compute lengths of each row */ 1096 for (i=0; i<a->mbs; i++) { 1097 crow_lens[i] = A->i[i+1] - A->i[i] + B->i[i+1] - B->i[i]; 1098 } 1099 /* store the row lengths to the file */ 1100 ierr = PetscViewerFlowControlStart(viewer,&message_count,&flowcontrolcount);CHKERRQ(ierr); 1101 if (!rank) { 1102 MPI_Status status; 1103 ierr = PetscMalloc(rlen*sizeof(PetscInt),&row_lens);CHKERRQ(ierr); 1104 rlen = (range[1] - range[0])/bs; 1105 for (i=0; i<rlen; i++) { 1106 for (j=0; j<bs; j++) { 1107 row_lens[i*bs+j] = bs*crow_lens[i]; 1108 } 1109 } 1110 ierr = PetscBinaryWrite(fd,row_lens,bs*rlen,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1111 for (i=1; i<size; i++) { 1112 rlen = (range[i+1] - range[i])/bs; 1113 ierr = PetscViewerFlowControlStepMaster(viewer,i,&message_count,flowcontrolcount);CHKERRQ(ierr); 1114 ierr = MPI_Recv(crow_lens,rlen,MPIU_INT,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1115 for (k=0; k<rlen; k++) { 1116 for (j=0; j<bs; j++) { 1117 row_lens[k*bs+j] = bs*crow_lens[k]; 1118 } 1119 } 1120 ierr = PetscBinaryWrite(fd,row_lens,bs*rlen,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1121 } 1122 ierr = PetscViewerFlowControlEndMaster(viewer,&message_count);CHKERRQ(ierr); 1123 ierr = PetscFree(row_lens);CHKERRQ(ierr); 1124 } else { 1125 ierr = PetscViewerFlowControlStepWorker(viewer,rank,&message_count);CHKERRQ(ierr); 1126 ierr = MPI_Send(crow_lens,mat->rmap->n/bs,MPIU_INT,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1127 ierr = PetscViewerFlowControlEndWorker(viewer,&message_count);CHKERRQ(ierr); 1128 } 1129 ierr = PetscFree(crow_lens);CHKERRQ(ierr); 1130 1131 /* load up the local column indices. Include for all rows not just one for each block row since process 0 does not have the 1132 information needed to make it for each row from a block row. This does require more communication but still not more than 1133 the communication needed for the nonzero values */ 1134 nzmax = nz; /* space a largest processor needs */ 1135 ierr = MPI_Reduce(&nz,&nzmax,1,MPIU_INT,MPI_MAX,0,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1136 ierr = PetscMalloc(nzmax*sizeof(PetscInt),&column_indices);CHKERRQ(ierr); 1137 cnt = 0; 1138 for (i=0; i<a->mbs; i++) { 1139 pcnt = cnt; 1140 for (j=B->i[i]; j<B->i[i+1]; j++) { 1141 if ((col = garray[B->j[j]]) > cstart) break; 1142 for (l=0; l<bs; l++) { 1143 column_indices[cnt++] = bs*col+l; 1144 } 1145 } 1146 for (k=A->i[i]; k<A->i[i+1]; k++) { 1147 for (l=0; l<bs; l++) { 1148 column_indices[cnt++] = bs*(A->j[k] + cstart)+l; 1149 } 1150 } 1151 for (; j<B->i[i+1]; j++) { 1152 for (l=0; l<bs; l++) { 1153 column_indices[cnt++] = bs*garray[B->j[j]]+l; 1154 } 1155 } 1156 len = cnt - pcnt; 1157 for (k=1; k<bs; k++) { 1158 ierr = PetscMemcpy(&column_indices[cnt],&column_indices[pcnt],len*sizeof(PetscInt));CHKERRQ(ierr); 1159 cnt += len; 1160 } 1161 } 1162 if (cnt != nz) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_LIB,"Internal PETSc error: cnt = %D nz = %D",cnt,nz); 1163 1164 /* store the columns to the file */ 1165 ierr = PetscViewerFlowControlStart(viewer,&message_count,&flowcontrolcount);CHKERRQ(ierr); 1166 if (!rank) { 1167 MPI_Status status; 1168 ierr = PetscBinaryWrite(fd,column_indices,nz,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1169 for (i=1; i<size; i++) { 1170 ierr = PetscViewerFlowControlStepMaster(viewer,i,&message_count,flowcontrolcount);CHKERRQ(ierr); 1171 ierr = MPI_Recv(&cnt,1,MPIU_INT,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1172 ierr = MPI_Recv(column_indices,cnt,MPIU_INT,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1173 ierr = PetscBinaryWrite(fd,column_indices,cnt,PETSC_INT,PETSC_TRUE);CHKERRQ(ierr); 1174 } 1175 ierr = PetscViewerFlowControlEndMaster(viewer,&message_count);CHKERRQ(ierr); 1176 } else { 1177 ierr = PetscViewerFlowControlStepWorker(viewer,rank,&message_count);CHKERRQ(ierr); 1178 ierr = MPI_Send(&cnt,1,MPIU_INT,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1179 ierr = MPI_Send(column_indices,cnt,MPIU_INT,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1180 ierr = PetscViewerFlowControlEndWorker(viewer,&message_count);CHKERRQ(ierr); 1181 } 1182 ierr = PetscFree(column_indices);CHKERRQ(ierr); 1183 1184 /* load up the numerical values */ 1185 ierr = PetscMalloc(nzmax*sizeof(PetscScalar),&column_values);CHKERRQ(ierr); 1186 cnt = 0; 1187 for (i=0; i<a->mbs; i++) { 1188 rlen = bs*(B->i[i+1] - B->i[i] + A->i[i+1] - A->i[i]); 1189 for (j=B->i[i]; j<B->i[i+1]; j++) { 1190 if (garray[B->j[j]] > cstart) break; 1191 for (l=0; l<bs; l++) { 1192 for (ll=0; ll<bs; ll++) { 1193 column_values[cnt + l*rlen + ll] = B->a[bs2*j+l+bs*ll]; 1194 } 1195 } 1196 cnt += bs; 1197 } 1198 for (k=A->i[i]; k<A->i[i+1]; k++) { 1199 for (l=0; l<bs; l++) { 1200 for (ll=0; ll<bs; ll++) { 1201 column_values[cnt + l*rlen + ll] = A->a[bs2*k+l+bs*ll]; 1202 } 1203 } 1204 cnt += bs; 1205 } 1206 for (; j<B->i[i+1]; j++) { 1207 for (l=0; l<bs; l++) { 1208 for (ll=0; ll<bs; ll++) { 1209 column_values[cnt + l*rlen + ll] = B->a[bs2*j+l+bs*ll]; 1210 } 1211 } 1212 cnt += bs; 1213 } 1214 cnt += (bs-1)*rlen; 1215 } 1216 if (cnt != nz) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Internal PETSc error: cnt = %D nz = %D",cnt,nz); 1217 1218 /* store the column values to the file */ 1219 ierr = PetscViewerFlowControlStart(viewer,&message_count,&flowcontrolcount);CHKERRQ(ierr); 1220 if (!rank) { 1221 MPI_Status status; 1222 ierr = PetscBinaryWrite(fd,column_values,nz,PETSC_SCALAR,PETSC_TRUE);CHKERRQ(ierr); 1223 for (i=1; i<size; i++) { 1224 ierr = PetscViewerFlowControlStepMaster(viewer,i,&message_count,flowcontrolcount);CHKERRQ(ierr); 1225 ierr = MPI_Recv(&cnt,1,MPIU_INT,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1226 ierr = MPI_Recv(column_values,cnt,MPIU_SCALAR,i,tag,PetscObjectComm((PetscObject)mat),&status);CHKERRQ(ierr); 1227 ierr = PetscBinaryWrite(fd,column_values,cnt,PETSC_SCALAR,PETSC_TRUE);CHKERRQ(ierr); 1228 } 1229 ierr = PetscViewerFlowControlEndMaster(viewer,&message_count);CHKERRQ(ierr); 1230 } else { 1231 ierr = PetscViewerFlowControlStepWorker(viewer,rank,&message_count);CHKERRQ(ierr); 1232 ierr = MPI_Send(&nz,1,MPIU_INT,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1233 ierr = MPI_Send(column_values,nz,MPIU_SCALAR,0,tag,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 1234 ierr = PetscViewerFlowControlEndWorker(viewer,&message_count);CHKERRQ(ierr); 1235 } 1236 ierr = PetscFree(column_values);CHKERRQ(ierr); 1237 1238 ierr = PetscViewerBinaryGetInfoPointer(viewer,&file);CHKERRQ(ierr); 1239 if (file) { 1240 fprintf(file,"-matload_block_size %d\n",(int)mat->rmap->bs); 1241 } 1242 PetscFunctionReturn(0); 1243 } 1244 1245 #undef __FUNCT__ 1246 #define __FUNCT__ "MatView_MPIBAIJ" 1247 PetscErrorCode MatView_MPIBAIJ(Mat mat,PetscViewer viewer) 1248 { 1249 PetscErrorCode ierr; 1250 PetscBool iascii,isdraw,issocket,isbinary; 1251 1252 PetscFunctionBegin; 1253 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERASCII,&iascii);CHKERRQ(ierr); 1254 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERDRAW,&isdraw);CHKERRQ(ierr); 1255 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERSOCKET,&issocket);CHKERRQ(ierr); 1256 ierr = PetscObjectTypeCompare((PetscObject)viewer,PETSCVIEWERBINARY,&isbinary);CHKERRQ(ierr); 1257 if (iascii || isdraw || issocket) { 1258 ierr = MatView_MPIBAIJ_ASCIIorDraworSocket(mat,viewer);CHKERRQ(ierr); 1259 } else if (isbinary) { 1260 ierr = MatView_MPIBAIJ_Binary(mat,viewer);CHKERRQ(ierr); 1261 } 1262 PetscFunctionReturn(0); 1263 } 1264 1265 #undef __FUNCT__ 1266 #define __FUNCT__ "MatDestroy_MPIBAIJ" 1267 PetscErrorCode MatDestroy_MPIBAIJ(Mat mat) 1268 { 1269 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1270 PetscErrorCode ierr; 1271 1272 PetscFunctionBegin; 1273 #if defined(PETSC_USE_LOG) 1274 PetscLogObjectState((PetscObject)mat,"Rows=%D,Cols=%D",mat->rmap->N,mat->cmap->N); 1275 #endif 1276 ierr = MatStashDestroy_Private(&mat->stash);CHKERRQ(ierr); 1277 ierr = MatStashDestroy_Private(&mat->bstash);CHKERRQ(ierr); 1278 ierr = MatDestroy(&baij->A);CHKERRQ(ierr); 1279 ierr = MatDestroy(&baij->B);CHKERRQ(ierr); 1280 #if defined(PETSC_USE_CTABLE) 1281 ierr = PetscTableDestroy(&baij->colmap);CHKERRQ(ierr); 1282 #else 1283 ierr = PetscFree(baij->colmap);CHKERRQ(ierr); 1284 #endif 1285 ierr = PetscFree(baij->garray);CHKERRQ(ierr); 1286 ierr = VecDestroy(&baij->lvec);CHKERRQ(ierr); 1287 ierr = VecScatterDestroy(&baij->Mvctx);CHKERRQ(ierr); 1288 ierr = PetscFree2(baij->rowvalues,baij->rowindices);CHKERRQ(ierr); 1289 ierr = PetscFree(baij->barray);CHKERRQ(ierr); 1290 ierr = PetscFree2(baij->hd,baij->ht);CHKERRQ(ierr); 1291 ierr = PetscFree(baij->rangebs);CHKERRQ(ierr); 1292 ierr = PetscFree(mat->data);CHKERRQ(ierr); 1293 1294 ierr = PetscObjectChangeTypeName((PetscObject)mat,0);CHKERRQ(ierr); 1295 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatStoreValues_C",NULL);CHKERRQ(ierr); 1296 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatRetrieveValues_C",NULL);CHKERRQ(ierr); 1297 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatGetDiagonalBlock_C",NULL);CHKERRQ(ierr); 1298 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatMPIBAIJSetPreallocation_C",NULL);CHKERRQ(ierr); 1299 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatMPIBAIJSetPreallocationCSR_C",NULL);CHKERRQ(ierr); 1300 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatDiagonalScaleLocal_C",NULL);CHKERRQ(ierr); 1301 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatSetHashTableFactor_C",NULL);CHKERRQ(ierr); 1302 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatConvert_mpibaij_mpisbaij_C",NULL);CHKERRQ(ierr); 1303 ierr = PetscObjectComposeFunction((PetscObject)mat,"MatConvert_mpibaij_mpibstrm_C",NULL);CHKERRQ(ierr); 1304 PetscFunctionReturn(0); 1305 } 1306 1307 #undef __FUNCT__ 1308 #define __FUNCT__ "MatMult_MPIBAIJ" 1309 PetscErrorCode MatMult_MPIBAIJ(Mat A,Vec xx,Vec yy) 1310 { 1311 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1312 PetscErrorCode ierr; 1313 PetscInt nt; 1314 1315 PetscFunctionBegin; 1316 ierr = VecGetLocalSize(xx,&nt);CHKERRQ(ierr); 1317 if (nt != A->cmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Incompatible partition of A and xx"); 1318 ierr = VecGetLocalSize(yy,&nt);CHKERRQ(ierr); 1319 if (nt != A->rmap->n) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Incompatible parition of A and yy"); 1320 ierr = VecScatterBegin(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1321 ierr = (*a->A->ops->mult)(a->A,xx,yy);CHKERRQ(ierr); 1322 ierr = VecScatterEnd(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1323 ierr = (*a->B->ops->multadd)(a->B,a->lvec,yy,yy);CHKERRQ(ierr); 1324 PetscFunctionReturn(0); 1325 } 1326 1327 #undef __FUNCT__ 1328 #define __FUNCT__ "MatMultAdd_MPIBAIJ" 1329 PetscErrorCode MatMultAdd_MPIBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 1330 { 1331 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1332 PetscErrorCode ierr; 1333 1334 PetscFunctionBegin; 1335 ierr = VecScatterBegin(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1336 ierr = (*a->A->ops->multadd)(a->A,xx,yy,zz);CHKERRQ(ierr); 1337 ierr = VecScatterEnd(a->Mvctx,xx,a->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1338 ierr = (*a->B->ops->multadd)(a->B,a->lvec,zz,zz);CHKERRQ(ierr); 1339 PetscFunctionReturn(0); 1340 } 1341 1342 #undef __FUNCT__ 1343 #define __FUNCT__ "MatMultTranspose_MPIBAIJ" 1344 PetscErrorCode MatMultTranspose_MPIBAIJ(Mat A,Vec xx,Vec yy) 1345 { 1346 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1347 PetscErrorCode ierr; 1348 PetscBool merged; 1349 1350 PetscFunctionBegin; 1351 ierr = VecScatterGetMerged(a->Mvctx,&merged);CHKERRQ(ierr); 1352 /* do nondiagonal part */ 1353 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 1354 if (!merged) { 1355 /* send it on its way */ 1356 ierr = VecScatterBegin(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1357 /* do local part */ 1358 ierr = (*a->A->ops->multtranspose)(a->A,xx,yy);CHKERRQ(ierr); 1359 /* receive remote parts: note this assumes the values are not actually */ 1360 /* inserted in yy until the next line */ 1361 ierr = VecScatterEnd(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1362 } else { 1363 /* do local part */ 1364 ierr = (*a->A->ops->multtranspose)(a->A,xx,yy);CHKERRQ(ierr); 1365 /* send it on its way */ 1366 ierr = VecScatterBegin(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1367 /* values actually were received in the Begin() but we need to call this nop */ 1368 ierr = VecScatterEnd(a->Mvctx,a->lvec,yy,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1369 } 1370 PetscFunctionReturn(0); 1371 } 1372 1373 #undef __FUNCT__ 1374 #define __FUNCT__ "MatMultTransposeAdd_MPIBAIJ" 1375 PetscErrorCode MatMultTransposeAdd_MPIBAIJ(Mat A,Vec xx,Vec yy,Vec zz) 1376 { 1377 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1378 PetscErrorCode ierr; 1379 1380 PetscFunctionBegin; 1381 /* do nondiagonal part */ 1382 ierr = (*a->B->ops->multtranspose)(a->B,xx,a->lvec);CHKERRQ(ierr); 1383 /* send it on its way */ 1384 ierr = VecScatterBegin(a->Mvctx,a->lvec,zz,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1385 /* do local part */ 1386 ierr = (*a->A->ops->multtransposeadd)(a->A,xx,yy,zz);CHKERRQ(ierr); 1387 /* receive remote parts: note this assumes the values are not actually */ 1388 /* inserted in yy until the next line, which is true for my implementation*/ 1389 /* but is not perhaps always true. */ 1390 ierr = VecScatterEnd(a->Mvctx,a->lvec,zz,ADD_VALUES,SCATTER_REVERSE);CHKERRQ(ierr); 1391 PetscFunctionReturn(0); 1392 } 1393 1394 /* 1395 This only works correctly for square matrices where the subblock A->A is the 1396 diagonal block 1397 */ 1398 #undef __FUNCT__ 1399 #define __FUNCT__ "MatGetDiagonal_MPIBAIJ" 1400 PetscErrorCode MatGetDiagonal_MPIBAIJ(Mat A,Vec v) 1401 { 1402 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1403 PetscErrorCode ierr; 1404 1405 PetscFunctionBegin; 1406 if (A->rmap->N != A->cmap->N) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Supports only square matrix where A->A is diag block"); 1407 ierr = MatGetDiagonal(a->A,v);CHKERRQ(ierr); 1408 PetscFunctionReturn(0); 1409 } 1410 1411 #undef __FUNCT__ 1412 #define __FUNCT__ "MatScale_MPIBAIJ" 1413 PetscErrorCode MatScale_MPIBAIJ(Mat A,PetscScalar aa) 1414 { 1415 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1416 PetscErrorCode ierr; 1417 1418 PetscFunctionBegin; 1419 ierr = MatScale(a->A,aa);CHKERRQ(ierr); 1420 ierr = MatScale(a->B,aa);CHKERRQ(ierr); 1421 PetscFunctionReturn(0); 1422 } 1423 1424 #undef __FUNCT__ 1425 #define __FUNCT__ "MatGetRow_MPIBAIJ" 1426 PetscErrorCode MatGetRow_MPIBAIJ(Mat matin,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v) 1427 { 1428 Mat_MPIBAIJ *mat = (Mat_MPIBAIJ*)matin->data; 1429 PetscScalar *vworkA,*vworkB,**pvA,**pvB,*v_p; 1430 PetscErrorCode ierr; 1431 PetscInt bs = matin->rmap->bs,bs2 = mat->bs2,i,*cworkA,*cworkB,**pcA,**pcB; 1432 PetscInt nztot,nzA,nzB,lrow,brstart = matin->rmap->rstart,brend = matin->rmap->rend; 1433 PetscInt *cmap,*idx_p,cstart = mat->cstartbs; 1434 1435 PetscFunctionBegin; 1436 if (row < brstart || row >= brend) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Only local rows"); 1437 if (mat->getrowactive) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Already active"); 1438 mat->getrowactive = PETSC_TRUE; 1439 1440 if (!mat->rowvalues && (idx || v)) { 1441 /* 1442 allocate enough space to hold information from the longest row. 1443 */ 1444 Mat_SeqBAIJ *Aa = (Mat_SeqBAIJ*)mat->A->data,*Ba = (Mat_SeqBAIJ*)mat->B->data; 1445 PetscInt max = 1,mbs = mat->mbs,tmp; 1446 for (i=0; i<mbs; i++) { 1447 tmp = Aa->i[i+1] - Aa->i[i] + Ba->i[i+1] - Ba->i[i]; 1448 if (max < tmp) max = tmp; 1449 } 1450 ierr = PetscMalloc2(max*bs2,PetscScalar,&mat->rowvalues,max*bs2,PetscInt,&mat->rowindices);CHKERRQ(ierr); 1451 } 1452 lrow = row - brstart; 1453 1454 pvA = &vworkA; pcA = &cworkA; pvB = &vworkB; pcB = &cworkB; 1455 if (!v) {pvA = 0; pvB = 0;} 1456 if (!idx) {pcA = 0; if (!v) pcB = 0;} 1457 ierr = (*mat->A->ops->getrow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1458 ierr = (*mat->B->ops->getrow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1459 nztot = nzA + nzB; 1460 1461 cmap = mat->garray; 1462 if (v || idx) { 1463 if (nztot) { 1464 /* Sort by increasing column numbers, assuming A and B already sorted */ 1465 PetscInt imark = -1; 1466 if (v) { 1467 *v = v_p = mat->rowvalues; 1468 for (i=0; i<nzB; i++) { 1469 if (cmap[cworkB[i]/bs] < cstart) v_p[i] = vworkB[i]; 1470 else break; 1471 } 1472 imark = i; 1473 for (i=0; i<nzA; i++) v_p[imark+i] = vworkA[i]; 1474 for (i=imark; i<nzB; i++) v_p[nzA+i] = vworkB[i]; 1475 } 1476 if (idx) { 1477 *idx = idx_p = mat->rowindices; 1478 if (imark > -1) { 1479 for (i=0; i<imark; i++) { 1480 idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs; 1481 } 1482 } else { 1483 for (i=0; i<nzB; i++) { 1484 if (cmap[cworkB[i]/bs] < cstart) idx_p[i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs; 1485 else break; 1486 } 1487 imark = i; 1488 } 1489 for (i=0; i<nzA; i++) idx_p[imark+i] = cstart*bs + cworkA[i]; 1490 for (i=imark; i<nzB; i++) idx_p[nzA+i] = cmap[cworkB[i]/bs]*bs + cworkB[i]%bs ; 1491 } 1492 } else { 1493 if (idx) *idx = 0; 1494 if (v) *v = 0; 1495 } 1496 } 1497 *nz = nztot; 1498 ierr = (*mat->A->ops->restorerow)(mat->A,lrow,&nzA,pcA,pvA);CHKERRQ(ierr); 1499 ierr = (*mat->B->ops->restorerow)(mat->B,lrow,&nzB,pcB,pvB);CHKERRQ(ierr); 1500 PetscFunctionReturn(0); 1501 } 1502 1503 #undef __FUNCT__ 1504 #define __FUNCT__ "MatRestoreRow_MPIBAIJ" 1505 PetscErrorCode MatRestoreRow_MPIBAIJ(Mat mat,PetscInt row,PetscInt *nz,PetscInt **idx,PetscScalar **v) 1506 { 1507 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1508 1509 PetscFunctionBegin; 1510 if (!baij->getrowactive) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"MatGetRow not called"); 1511 baij->getrowactive = PETSC_FALSE; 1512 PetscFunctionReturn(0); 1513 } 1514 1515 #undef __FUNCT__ 1516 #define __FUNCT__ "MatZeroEntries_MPIBAIJ" 1517 PetscErrorCode MatZeroEntries_MPIBAIJ(Mat A) 1518 { 1519 Mat_MPIBAIJ *l = (Mat_MPIBAIJ*)A->data; 1520 PetscErrorCode ierr; 1521 1522 PetscFunctionBegin; 1523 ierr = MatZeroEntries(l->A);CHKERRQ(ierr); 1524 ierr = MatZeroEntries(l->B);CHKERRQ(ierr); 1525 PetscFunctionReturn(0); 1526 } 1527 1528 #undef __FUNCT__ 1529 #define __FUNCT__ "MatGetInfo_MPIBAIJ" 1530 PetscErrorCode MatGetInfo_MPIBAIJ(Mat matin,MatInfoType flag,MatInfo *info) 1531 { 1532 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)matin->data; 1533 Mat A = a->A,B = a->B; 1534 PetscErrorCode ierr; 1535 PetscReal isend[5],irecv[5]; 1536 1537 PetscFunctionBegin; 1538 info->block_size = (PetscReal)matin->rmap->bs; 1539 1540 ierr = MatGetInfo(A,MAT_LOCAL,info);CHKERRQ(ierr); 1541 1542 isend[0] = info->nz_used; isend[1] = info->nz_allocated; isend[2] = info->nz_unneeded; 1543 isend[3] = info->memory; isend[4] = info->mallocs; 1544 1545 ierr = MatGetInfo(B,MAT_LOCAL,info);CHKERRQ(ierr); 1546 1547 isend[0] += info->nz_used; isend[1] += info->nz_allocated; isend[2] += info->nz_unneeded; 1548 isend[3] += info->memory; isend[4] += info->mallocs; 1549 1550 if (flag == MAT_LOCAL) { 1551 info->nz_used = isend[0]; 1552 info->nz_allocated = isend[1]; 1553 info->nz_unneeded = isend[2]; 1554 info->memory = isend[3]; 1555 info->mallocs = isend[4]; 1556 } else if (flag == MAT_GLOBAL_MAX) { 1557 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPIU_MAX,PetscObjectComm((PetscObject)matin));CHKERRQ(ierr); 1558 1559 info->nz_used = irecv[0]; 1560 info->nz_allocated = irecv[1]; 1561 info->nz_unneeded = irecv[2]; 1562 info->memory = irecv[3]; 1563 info->mallocs = irecv[4]; 1564 } else if (flag == MAT_GLOBAL_SUM) { 1565 ierr = MPI_Allreduce(isend,irecv,5,MPIU_REAL,MPIU_SUM,PetscObjectComm((PetscObject)matin));CHKERRQ(ierr); 1566 1567 info->nz_used = irecv[0]; 1568 info->nz_allocated = irecv[1]; 1569 info->nz_unneeded = irecv[2]; 1570 info->memory = irecv[3]; 1571 info->mallocs = irecv[4]; 1572 } else SETERRQ1(PetscObjectComm((PetscObject)matin),PETSC_ERR_ARG_WRONG,"Unknown MatInfoType argument %d",(int)flag); 1573 info->fill_ratio_given = 0; /* no parallel LU/ILU/Cholesky */ 1574 info->fill_ratio_needed = 0; 1575 info->factor_mallocs = 0; 1576 PetscFunctionReturn(0); 1577 } 1578 1579 #undef __FUNCT__ 1580 #define __FUNCT__ "MatSetOption_MPIBAIJ" 1581 PetscErrorCode MatSetOption_MPIBAIJ(Mat A,MatOption op,PetscBool flg) 1582 { 1583 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1584 PetscErrorCode ierr; 1585 1586 PetscFunctionBegin; 1587 switch (op) { 1588 case MAT_NEW_NONZERO_LOCATIONS: 1589 case MAT_NEW_NONZERO_ALLOCATION_ERR: 1590 case MAT_UNUSED_NONZERO_LOCATION_ERR: 1591 case MAT_KEEP_NONZERO_PATTERN: 1592 case MAT_NEW_NONZERO_LOCATION_ERR: 1593 ierr = MatSetOption(a->A,op,flg);CHKERRQ(ierr); 1594 ierr = MatSetOption(a->B,op,flg);CHKERRQ(ierr); 1595 break; 1596 case MAT_ROW_ORIENTED: 1597 a->roworiented = flg; 1598 1599 ierr = MatSetOption(a->A,op,flg);CHKERRQ(ierr); 1600 ierr = MatSetOption(a->B,op,flg);CHKERRQ(ierr); 1601 break; 1602 case MAT_NEW_DIAGONALS: 1603 ierr = PetscInfo1(A,"Option %s ignored\n",MatOptions[op]);CHKERRQ(ierr); 1604 break; 1605 case MAT_IGNORE_OFF_PROC_ENTRIES: 1606 a->donotstash = flg; 1607 break; 1608 case MAT_USE_HASH_TABLE: 1609 a->ht_flag = flg; 1610 break; 1611 case MAT_SYMMETRIC: 1612 case MAT_STRUCTURALLY_SYMMETRIC: 1613 case MAT_HERMITIAN: 1614 case MAT_SYMMETRY_ETERNAL: 1615 ierr = MatSetOption(a->A,op,flg);CHKERRQ(ierr); 1616 break; 1617 default: 1618 SETERRQ1(PetscObjectComm((PetscObject)A),PETSC_ERR_SUP,"unknown option %d",op); 1619 } 1620 PetscFunctionReturn(0); 1621 } 1622 1623 #undef __FUNCT__ 1624 #define __FUNCT__ "MatTranspose_MPIBAIJ" 1625 PetscErrorCode MatTranspose_MPIBAIJ(Mat A,MatReuse reuse,Mat *matout) 1626 { 1627 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)A->data; 1628 Mat_SeqBAIJ *Aloc; 1629 Mat B; 1630 PetscErrorCode ierr; 1631 PetscInt M =A->rmap->N,N=A->cmap->N,*ai,*aj,i,*rvals,j,k,col; 1632 PetscInt bs=A->rmap->bs,mbs=baij->mbs; 1633 MatScalar *a; 1634 1635 PetscFunctionBegin; 1636 if (reuse == MAT_REUSE_MATRIX && A == *matout && M != N) SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_SIZ,"Square matrix only for in-place"); 1637 if (reuse == MAT_INITIAL_MATRIX || *matout == A) { 1638 ierr = MatCreate(PetscObjectComm((PetscObject)A),&B);CHKERRQ(ierr); 1639 ierr = MatSetSizes(B,A->cmap->n,A->rmap->n,N,M);CHKERRQ(ierr); 1640 ierr = MatSetType(B,((PetscObject)A)->type_name);CHKERRQ(ierr); 1641 /* Do not know preallocation information, but must set block size */ 1642 ierr = MatMPIBAIJSetPreallocation(B,A->rmap->bs,PETSC_DECIDE,NULL,PETSC_DECIDE,NULL);CHKERRQ(ierr); 1643 } else { 1644 B = *matout; 1645 } 1646 1647 /* copy over the A part */ 1648 Aloc = (Mat_SeqBAIJ*)baij->A->data; 1649 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1650 ierr = PetscMalloc(bs*sizeof(PetscInt),&rvals);CHKERRQ(ierr); 1651 1652 for (i=0; i<mbs; i++) { 1653 rvals[0] = bs*(baij->rstartbs + i); 1654 for (j=1; j<bs; j++) rvals[j] = rvals[j-1] + 1; 1655 for (j=ai[i]; j<ai[i+1]; j++) { 1656 col = (baij->cstartbs+aj[j])*bs; 1657 for (k=0; k<bs; k++) { 1658 ierr = MatSetValues_MPIBAIJ(B,1,&col,bs,rvals,a,INSERT_VALUES);CHKERRQ(ierr); 1659 1660 col++; a += bs; 1661 } 1662 } 1663 } 1664 /* copy over the B part */ 1665 Aloc = (Mat_SeqBAIJ*)baij->B->data; 1666 ai = Aloc->i; aj = Aloc->j; a = Aloc->a; 1667 for (i=0; i<mbs; i++) { 1668 rvals[0] = bs*(baij->rstartbs + i); 1669 for (j=1; j<bs; j++) rvals[j] = rvals[j-1] + 1; 1670 for (j=ai[i]; j<ai[i+1]; j++) { 1671 col = baij->garray[aj[j]]*bs; 1672 for (k=0; k<bs; k++) { 1673 ierr = MatSetValues_MPIBAIJ(B,1,&col,bs,rvals,a,INSERT_VALUES);CHKERRQ(ierr); 1674 col++; 1675 a += bs; 1676 } 1677 } 1678 } 1679 ierr = PetscFree(rvals);CHKERRQ(ierr); 1680 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1681 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1682 1683 if (reuse == MAT_INITIAL_MATRIX || *matout != A) *matout = B; 1684 else { 1685 ierr = MatHeaderMerge(A,B);CHKERRQ(ierr); 1686 } 1687 PetscFunctionReturn(0); 1688 } 1689 1690 #undef __FUNCT__ 1691 #define __FUNCT__ "MatDiagonalScale_MPIBAIJ" 1692 PetscErrorCode MatDiagonalScale_MPIBAIJ(Mat mat,Vec ll,Vec rr) 1693 { 1694 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 1695 Mat a = baij->A,b = baij->B; 1696 PetscErrorCode ierr; 1697 PetscInt s1,s2,s3; 1698 1699 PetscFunctionBegin; 1700 ierr = MatGetLocalSize(mat,&s2,&s3);CHKERRQ(ierr); 1701 if (rr) { 1702 ierr = VecGetLocalSize(rr,&s1);CHKERRQ(ierr); 1703 if (s1!=s3) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"right vector non-conforming local size"); 1704 /* Overlap communication with computation. */ 1705 ierr = VecScatterBegin(baij->Mvctx,rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1706 } 1707 if (ll) { 1708 ierr = VecGetLocalSize(ll,&s1);CHKERRQ(ierr); 1709 if (s1!=s2) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"left vector non-conforming local size"); 1710 ierr = (*b->ops->diagonalscale)(b,ll,NULL);CHKERRQ(ierr); 1711 } 1712 /* scale the diagonal block */ 1713 ierr = (*a->ops->diagonalscale)(a,ll,rr);CHKERRQ(ierr); 1714 1715 if (rr) { 1716 /* Do a scatter end and then right scale the off-diagonal block */ 1717 ierr = VecScatterEnd(baij->Mvctx,rr,baij->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 1718 ierr = (*b->ops->diagonalscale)(b,NULL,baij->lvec);CHKERRQ(ierr); 1719 } 1720 PetscFunctionReturn(0); 1721 } 1722 1723 #undef __FUNCT__ 1724 #define __FUNCT__ "MatZeroRows_MPIBAIJ" 1725 PetscErrorCode MatZeroRows_MPIBAIJ(Mat A,PetscInt N,const PetscInt rows[],PetscScalar diag,Vec x,Vec b) 1726 { 1727 Mat_MPIBAIJ *l = (Mat_MPIBAIJ*)A->data; 1728 PetscErrorCode ierr; 1729 PetscMPIInt imdex,size = l->size,n,rank = l->rank; 1730 PetscInt i,*owners = A->rmap->range; 1731 PetscInt *nprocs,j,idx,nsends,row; 1732 PetscInt nmax,*svalues,*starts,*owner,nrecvs; 1733 PetscInt *rvalues,tag = ((PetscObject)A)->tag,count,base,slen,*source,lastidx = -1; 1734 PetscInt *lens,*lrows,*values,rstart_bs=A->rmap->rstart; 1735 MPI_Comm comm; 1736 MPI_Request *send_waits,*recv_waits; 1737 MPI_Status recv_status,*send_status; 1738 const PetscScalar *xx; 1739 PetscScalar *bb; 1740 #if defined(PETSC_DEBUG) 1741 PetscBool found = PETSC_FALSE; 1742 #endif 1743 1744 PetscFunctionBegin; 1745 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 1746 /* first count number of contributors to each processor */ 1747 ierr = PetscMalloc(2*size*sizeof(PetscInt),&nprocs);CHKERRQ(ierr); 1748 ierr = PetscMemzero(nprocs,2*size*sizeof(PetscInt));CHKERRQ(ierr); 1749 ierr = PetscMalloc((N+1)*sizeof(PetscInt),&owner);CHKERRQ(ierr); /* see note*/ 1750 j = 0; 1751 for (i=0; i<N; i++) { 1752 if (lastidx > (idx = rows[i])) j = 0; 1753 lastidx = idx; 1754 for (; j<size; j++) { 1755 if (idx >= owners[j] && idx < owners[j+1]) { 1756 nprocs[2*j]++; 1757 nprocs[2*j+1] = 1; 1758 owner[i] = j; 1759 #if defined(PETSC_DEBUG) 1760 found = PETSC_TRUE; 1761 #endif 1762 break; 1763 } 1764 } 1765 #if defined(PETSC_DEBUG) 1766 if (!found) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Index out of range"); 1767 found = PETSC_FALSE; 1768 #endif 1769 } 1770 nsends = 0; for (i=0; i<size; i++) nsends += nprocs[2*i+1]; 1771 1772 if (A->nooffproczerorows) { 1773 if (nsends > 1) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"You called MatSetOption(,MAT_NO_OFF_PROC_ZERO_ROWS,PETSC_TRUE) but set an off process zero row"); 1774 nrecvs = nsends; 1775 nmax = N; 1776 } else { 1777 /* inform other processors of number of messages and max length*/ 1778 ierr = PetscMaxSum(comm,nprocs,&nmax,&nrecvs);CHKERRQ(ierr); 1779 } 1780 1781 /* post receives: */ 1782 ierr = PetscMalloc((nrecvs+1)*(nmax+1)*sizeof(PetscInt),&rvalues);CHKERRQ(ierr); 1783 ierr = PetscMalloc((nrecvs+1)*sizeof(MPI_Request),&recv_waits);CHKERRQ(ierr); 1784 for (i=0; i<nrecvs; i++) { 1785 ierr = MPI_Irecv(rvalues+nmax*i,nmax,MPIU_INT,MPI_ANY_SOURCE,tag,comm,recv_waits+i);CHKERRQ(ierr); 1786 } 1787 1788 /* do sends: 1789 1) starts[i] gives the starting index in svalues for stuff going to 1790 the ith processor 1791 */ 1792 ierr = PetscMalloc((N+1)*sizeof(PetscInt),&svalues);CHKERRQ(ierr); 1793 ierr = PetscMalloc((nsends+1)*sizeof(MPI_Request),&send_waits);CHKERRQ(ierr); 1794 ierr = PetscMalloc((size+1)*sizeof(PetscInt),&starts);CHKERRQ(ierr); 1795 starts[0] = 0; 1796 for (i=1; i<size; i++) starts[i] = starts[i-1] + nprocs[2*i-2]; 1797 for (i=0; i<N; i++) { 1798 svalues[starts[owner[i]]++] = rows[i]; 1799 } 1800 1801 starts[0] = 0; 1802 for (i=1; i<size+1; i++) starts[i] = starts[i-1] + nprocs[2*i-2]; 1803 count = 0; 1804 for (i=0; i<size; i++) { 1805 if (nprocs[2*i+1]) { 1806 ierr = MPI_Isend(svalues+starts[i],nprocs[2*i],MPIU_INT,i,tag,comm,send_waits+count++);CHKERRQ(ierr); 1807 } 1808 } 1809 ierr = PetscFree(starts);CHKERRQ(ierr); 1810 1811 base = owners[rank]; 1812 1813 /* wait on receives */ 1814 ierr = PetscMalloc2(nrecvs+1,PetscInt,&lens,nrecvs+1,PetscInt,&source);CHKERRQ(ierr); 1815 count = nrecvs; 1816 slen = 0; 1817 while (count) { 1818 ierr = MPI_Waitany(nrecvs,recv_waits,&imdex,&recv_status);CHKERRQ(ierr); 1819 /* unpack receives into our local space */ 1820 ierr = MPI_Get_count(&recv_status,MPIU_INT,&n);CHKERRQ(ierr); 1821 1822 source[imdex] = recv_status.MPI_SOURCE; 1823 lens[imdex] = n; 1824 slen += n; 1825 count--; 1826 } 1827 ierr = PetscFree(recv_waits);CHKERRQ(ierr); 1828 1829 /* move the data into the send scatter */ 1830 ierr = PetscMalloc((slen+1)*sizeof(PetscInt),&lrows);CHKERRQ(ierr); 1831 count = 0; 1832 for (i=0; i<nrecvs; i++) { 1833 values = rvalues + i*nmax; 1834 for (j=0; j<lens[i]; j++) { 1835 lrows[count++] = values[j] - base; 1836 } 1837 } 1838 ierr = PetscFree(rvalues);CHKERRQ(ierr); 1839 ierr = PetscFree2(lens,source);CHKERRQ(ierr); 1840 ierr = PetscFree(owner);CHKERRQ(ierr); 1841 ierr = PetscFree(nprocs);CHKERRQ(ierr); 1842 1843 /* fix right hand side if needed */ 1844 if (x && b) { 1845 ierr = VecGetArrayRead(x,&xx);CHKERRQ(ierr); 1846 ierr = VecGetArray(b,&bb);CHKERRQ(ierr); 1847 for (i=0; i<slen; i++) { 1848 bb[lrows[i]] = diag*xx[lrows[i]]; 1849 } 1850 ierr = VecRestoreArrayRead(x,&xx);CHKERRQ(ierr); 1851 ierr = VecRestoreArray(b,&bb);CHKERRQ(ierr); 1852 } 1853 1854 /* actually zap the local rows */ 1855 /* 1856 Zero the required rows. If the "diagonal block" of the matrix 1857 is square and the user wishes to set the diagonal we use separate 1858 code so that MatSetValues() is not called for each diagonal allocating 1859 new memory, thus calling lots of mallocs and slowing things down. 1860 1861 */ 1862 /* must zero l->B before l->A because the (diag) case below may put values into l->B*/ 1863 ierr = MatZeroRows_SeqBAIJ(l->B,slen,lrows,0.0,0,0);CHKERRQ(ierr); 1864 if ((diag != 0.0) && (l->A->rmap->N == l->A->cmap->N)) { 1865 ierr = MatZeroRows_SeqBAIJ(l->A,slen,lrows,diag,0,0);CHKERRQ(ierr); 1866 } else if (diag != 0.0) { 1867 ierr = MatZeroRows_SeqBAIJ(l->A,slen,lrows,0.0,0,0);CHKERRQ(ierr); 1868 if (((Mat_SeqBAIJ*)l->A->data)->nonew) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"MatZeroRows() on rectangular matrices cannot be used with the Mat options \n\ 1869 MAT_NEW_NONZERO_LOCATIONS,MAT_NEW_NONZERO_LOCATION_ERR,MAT_NEW_NONZERO_ALLOCATION_ERR"); 1870 for (i=0; i<slen; i++) { 1871 row = lrows[i] + rstart_bs; 1872 ierr = MatSetValues(A,1,&row,1,&row,&diag,INSERT_VALUES);CHKERRQ(ierr); 1873 } 1874 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1875 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1876 } else { 1877 ierr = MatZeroRows_SeqBAIJ(l->A,slen,lrows,0.0,0,0);CHKERRQ(ierr); 1878 } 1879 1880 ierr = PetscFree(lrows);CHKERRQ(ierr); 1881 1882 /* wait on sends */ 1883 if (nsends) { 1884 ierr = PetscMalloc(nsends*sizeof(MPI_Status),&send_status);CHKERRQ(ierr); 1885 ierr = MPI_Waitall(nsends,send_waits,send_status);CHKERRQ(ierr); 1886 ierr = PetscFree(send_status);CHKERRQ(ierr); 1887 } 1888 ierr = PetscFree(send_waits);CHKERRQ(ierr); 1889 ierr = PetscFree(svalues);CHKERRQ(ierr); 1890 PetscFunctionReturn(0); 1891 } 1892 1893 #undef __FUNCT__ 1894 #define __FUNCT__ "MatSetUnfactored_MPIBAIJ" 1895 PetscErrorCode MatSetUnfactored_MPIBAIJ(Mat A) 1896 { 1897 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1898 PetscErrorCode ierr; 1899 1900 PetscFunctionBegin; 1901 ierr = MatSetUnfactored(a->A);CHKERRQ(ierr); 1902 PetscFunctionReturn(0); 1903 } 1904 1905 static PetscErrorCode MatDuplicate_MPIBAIJ(Mat,MatDuplicateOption,Mat*); 1906 1907 #undef __FUNCT__ 1908 #define __FUNCT__ "MatEqual_MPIBAIJ" 1909 PetscErrorCode MatEqual_MPIBAIJ(Mat A,Mat B,PetscBool *flag) 1910 { 1911 Mat_MPIBAIJ *matB = (Mat_MPIBAIJ*)B->data,*matA = (Mat_MPIBAIJ*)A->data; 1912 Mat a,b,c,d; 1913 PetscBool flg; 1914 PetscErrorCode ierr; 1915 1916 PetscFunctionBegin; 1917 a = matA->A; b = matA->B; 1918 c = matB->A; d = matB->B; 1919 1920 ierr = MatEqual(a,c,&flg);CHKERRQ(ierr); 1921 if (flg) { 1922 ierr = MatEqual(b,d,&flg);CHKERRQ(ierr); 1923 } 1924 ierr = MPI_Allreduce(&flg,flag,1,MPIU_BOOL,MPI_LAND,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 1925 PetscFunctionReturn(0); 1926 } 1927 1928 #undef __FUNCT__ 1929 #define __FUNCT__ "MatCopy_MPIBAIJ" 1930 PetscErrorCode MatCopy_MPIBAIJ(Mat A,Mat B,MatStructure str) 1931 { 1932 PetscErrorCode ierr; 1933 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1934 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)B->data; 1935 1936 PetscFunctionBegin; 1937 /* If the two matrices don't have the same copy implementation, they aren't compatible for fast copy. */ 1938 if ((str != SAME_NONZERO_PATTERN) || (A->ops->copy != B->ops->copy)) { 1939 ierr = MatCopy_Basic(A,B,str);CHKERRQ(ierr); 1940 } else { 1941 ierr = MatCopy(a->A,b->A,str);CHKERRQ(ierr); 1942 ierr = MatCopy(a->B,b->B,str);CHKERRQ(ierr); 1943 } 1944 PetscFunctionReturn(0); 1945 } 1946 1947 #undef __FUNCT__ 1948 #define __FUNCT__ "MatSetUp_MPIBAIJ" 1949 PetscErrorCode MatSetUp_MPIBAIJ(Mat A) 1950 { 1951 PetscErrorCode ierr; 1952 1953 PetscFunctionBegin; 1954 ierr = MatMPIBAIJSetPreallocation(A,A->rmap->bs,PETSC_DEFAULT,0,PETSC_DEFAULT,0);CHKERRQ(ierr); 1955 PetscFunctionReturn(0); 1956 } 1957 1958 #undef __FUNCT__ 1959 #define __FUNCT__ "MatAXPY_MPIBAIJ" 1960 PetscErrorCode MatAXPY_MPIBAIJ(Mat Y,PetscScalar a,Mat X,MatStructure str) 1961 { 1962 PetscErrorCode ierr; 1963 Mat_MPIBAIJ *xx=(Mat_MPIBAIJ*)X->data,*yy=(Mat_MPIBAIJ*)Y->data; 1964 PetscBLASInt bnz,one=1; 1965 Mat_SeqBAIJ *x,*y; 1966 1967 PetscFunctionBegin; 1968 if (str == SAME_NONZERO_PATTERN) { 1969 PetscScalar alpha = a; 1970 x = (Mat_SeqBAIJ*)xx->A->data; 1971 y = (Mat_SeqBAIJ*)yy->A->data; 1972 ierr = PetscBLASIntCast(x->nz,&bnz);CHKERRQ(ierr); 1973 PetscStackCallBLAS("BLASaxpy",BLASaxpy_(&bnz,&alpha,x->a,&one,y->a,&one)); 1974 x = (Mat_SeqBAIJ*)xx->B->data; 1975 y = (Mat_SeqBAIJ*)yy->B->data; 1976 ierr = PetscBLASIntCast(x->nz,&bnz);CHKERRQ(ierr); 1977 PetscStackCallBLAS("BLASaxpy",BLASaxpy_(&bnz,&alpha,x->a,&one,y->a,&one)); 1978 } else { 1979 ierr = MatAXPY_Basic(Y,a,X,str);CHKERRQ(ierr); 1980 } 1981 PetscFunctionReturn(0); 1982 } 1983 1984 #undef __FUNCT__ 1985 #define __FUNCT__ "MatRealPart_MPIBAIJ" 1986 PetscErrorCode MatRealPart_MPIBAIJ(Mat A) 1987 { 1988 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 1989 PetscErrorCode ierr; 1990 1991 PetscFunctionBegin; 1992 ierr = MatRealPart(a->A);CHKERRQ(ierr); 1993 ierr = MatRealPart(a->B);CHKERRQ(ierr); 1994 PetscFunctionReturn(0); 1995 } 1996 1997 #undef __FUNCT__ 1998 #define __FUNCT__ "MatImaginaryPart_MPIBAIJ" 1999 PetscErrorCode MatImaginaryPart_MPIBAIJ(Mat A) 2000 { 2001 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 2002 PetscErrorCode ierr; 2003 2004 PetscFunctionBegin; 2005 ierr = MatImaginaryPart(a->A);CHKERRQ(ierr); 2006 ierr = MatImaginaryPart(a->B);CHKERRQ(ierr); 2007 PetscFunctionReturn(0); 2008 } 2009 2010 #undef __FUNCT__ 2011 #define __FUNCT__ "MatGetSubMatrix_MPIBAIJ" 2012 PetscErrorCode MatGetSubMatrix_MPIBAIJ(Mat mat,IS isrow,IS iscol,MatReuse call,Mat *newmat) 2013 { 2014 PetscErrorCode ierr; 2015 IS iscol_local; 2016 PetscInt csize; 2017 2018 PetscFunctionBegin; 2019 ierr = ISGetLocalSize(iscol,&csize);CHKERRQ(ierr); 2020 if (call == MAT_REUSE_MATRIX) { 2021 ierr = PetscObjectQuery((PetscObject)*newmat,"ISAllGather",(PetscObject*)&iscol_local);CHKERRQ(ierr); 2022 if (!iscol_local) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Submatrix passed in was not used before, cannot reuse"); 2023 } else { 2024 ierr = ISAllGather(iscol,&iscol_local);CHKERRQ(ierr); 2025 } 2026 ierr = MatGetSubMatrix_MPIBAIJ_Private(mat,isrow,iscol_local,csize,call,newmat);CHKERRQ(ierr); 2027 if (call == MAT_INITIAL_MATRIX) { 2028 ierr = PetscObjectCompose((PetscObject)*newmat,"ISAllGather",(PetscObject)iscol_local);CHKERRQ(ierr); 2029 ierr = ISDestroy(&iscol_local);CHKERRQ(ierr); 2030 } 2031 PetscFunctionReturn(0); 2032 } 2033 extern PetscErrorCode MatGetSubMatrices_MPIBAIJ_local(Mat,PetscInt,const IS[],const IS[],MatReuse,PetscBool*,PetscBool*,Mat*); 2034 #undef __FUNCT__ 2035 #define __FUNCT__ "MatGetSubMatrix_MPIBAIJ_Private" 2036 /* 2037 Not great since it makes two copies of the submatrix, first an SeqBAIJ 2038 in local and then by concatenating the local matrices the end result. 2039 Writing it directly would be much like MatGetSubMatrices_MPIBAIJ() 2040 */ 2041 PetscErrorCode MatGetSubMatrix_MPIBAIJ_Private(Mat mat,IS isrow,IS iscol,PetscInt csize,MatReuse call,Mat *newmat) 2042 { 2043 PetscErrorCode ierr; 2044 PetscMPIInt rank,size; 2045 PetscInt i,m,n,rstart,row,rend,nz,*cwork,j,bs; 2046 PetscInt *ii,*jj,nlocal,*dlens,*olens,dlen,olen,jend,mglobal,ncol,nrow; 2047 Mat M,Mreuse; 2048 MatScalar *vwork,*aa; 2049 MPI_Comm comm; 2050 IS isrow_new, iscol_new; 2051 PetscBool idflag,allrows, allcols; 2052 Mat_SeqBAIJ *aij; 2053 2054 PetscFunctionBegin; 2055 ierr = PetscObjectGetComm((PetscObject)mat,&comm);CHKERRQ(ierr); 2056 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 2057 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 2058 /* The compression and expansion should be avoided. Doesn't point 2059 out errors, might change the indices, hence buggey */ 2060 ierr = ISCompressIndicesGeneral(mat->rmap->N,mat->rmap->n,mat->rmap->bs,1,&isrow,&isrow_new);CHKERRQ(ierr); 2061 ierr = ISCompressIndicesGeneral(mat->cmap->N,mat->cmap->n,mat->cmap->bs,1,&iscol,&iscol_new);CHKERRQ(ierr); 2062 2063 /* Check for special case: each processor gets entire matrix columns */ 2064 ierr = ISIdentity(iscol,&idflag);CHKERRQ(ierr); 2065 ierr = ISGetLocalSize(iscol,&ncol);CHKERRQ(ierr); 2066 if (idflag && ncol == mat->cmap->N) allcols = PETSC_TRUE; 2067 else allcols = PETSC_FALSE; 2068 2069 ierr = ISIdentity(isrow,&idflag);CHKERRQ(ierr); 2070 ierr = ISGetLocalSize(isrow,&nrow);CHKERRQ(ierr); 2071 if (idflag && nrow == mat->rmap->N) allrows = PETSC_TRUE; 2072 else allrows = PETSC_FALSE; 2073 2074 if (call == MAT_REUSE_MATRIX) { 2075 ierr = PetscObjectQuery((PetscObject)*newmat,"SubMatrix",(PetscObject*)&Mreuse);CHKERRQ(ierr); 2076 if (!Mreuse) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Submatrix passed in was not used before, cannot reuse"); 2077 ierr = MatGetSubMatrices_MPIBAIJ_local(mat,1,&isrow_new,&iscol_new,MAT_REUSE_MATRIX,&allrows,&allcols,&Mreuse);CHKERRQ(ierr); 2078 } else { 2079 ierr = MatGetSubMatrices_MPIBAIJ_local(mat,1,&isrow_new,&iscol_new,MAT_INITIAL_MATRIX,&allrows,&allcols,&Mreuse);CHKERRQ(ierr); 2080 } 2081 ierr = ISDestroy(&isrow_new);CHKERRQ(ierr); 2082 ierr = ISDestroy(&iscol_new);CHKERRQ(ierr); 2083 /* 2084 m - number of local rows 2085 n - number of columns (same on all processors) 2086 rstart - first row in new global matrix generated 2087 */ 2088 ierr = MatGetBlockSize(mat,&bs);CHKERRQ(ierr); 2089 ierr = MatGetSize(Mreuse,&m,&n);CHKERRQ(ierr); 2090 m = m/bs; 2091 n = n/bs; 2092 2093 if (call == MAT_INITIAL_MATRIX) { 2094 aij = (Mat_SeqBAIJ*)(Mreuse)->data; 2095 ii = aij->i; 2096 jj = aij->j; 2097 2098 /* 2099 Determine the number of non-zeros in the diagonal and off-diagonal 2100 portions of the matrix in order to do correct preallocation 2101 */ 2102 2103 /* first get start and end of "diagonal" columns */ 2104 if (csize == PETSC_DECIDE) { 2105 ierr = ISGetSize(isrow,&mglobal);CHKERRQ(ierr); 2106 if (mglobal == n*bs) { /* square matrix */ 2107 nlocal = m; 2108 } else { 2109 nlocal = n/size + ((n % size) > rank); 2110 } 2111 } else { 2112 nlocal = csize/bs; 2113 } 2114 ierr = MPI_Scan(&nlocal,&rend,1,MPIU_INT,MPI_SUM,comm);CHKERRQ(ierr); 2115 rstart = rend - nlocal; 2116 if (rank == size - 1 && rend != n) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Local column sizes %D do not add up to total number of columns %D",rend,n); 2117 2118 /* next, compute all the lengths */ 2119 ierr = PetscMalloc2(m+1,PetscInt,&dlens,m+1,PetscInt,&olens);CHKERRQ(ierr); 2120 for (i=0; i<m; i++) { 2121 jend = ii[i+1] - ii[i]; 2122 olen = 0; 2123 dlen = 0; 2124 for (j=0; j<jend; j++) { 2125 if (*jj < rstart || *jj >= rend) olen++; 2126 else dlen++; 2127 jj++; 2128 } 2129 olens[i] = olen; 2130 dlens[i] = dlen; 2131 } 2132 ierr = MatCreate(comm,&M);CHKERRQ(ierr); 2133 ierr = MatSetSizes(M,bs*m,bs*nlocal,PETSC_DECIDE,bs*n);CHKERRQ(ierr); 2134 ierr = MatSetType(M,((PetscObject)mat)->type_name);CHKERRQ(ierr); 2135 ierr = MatMPIBAIJSetPreallocation(M,bs,0,dlens,0,olens);CHKERRQ(ierr); 2136 ierr = PetscFree2(dlens,olens);CHKERRQ(ierr); 2137 } else { 2138 PetscInt ml,nl; 2139 2140 M = *newmat; 2141 ierr = MatGetLocalSize(M,&ml,&nl);CHKERRQ(ierr); 2142 if (ml != m) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Previous matrix must be same size/layout as request"); 2143 ierr = MatZeroEntries(M);CHKERRQ(ierr); 2144 /* 2145 The next two lines are needed so we may call MatSetValues_MPIAIJ() below directly, 2146 rather than the slower MatSetValues(). 2147 */ 2148 M->was_assembled = PETSC_TRUE; 2149 M->assembled = PETSC_FALSE; 2150 } 2151 ierr = MatSetOption(M,MAT_ROW_ORIENTED,PETSC_FALSE);CHKERRQ(ierr); 2152 ierr = MatGetOwnershipRange(M,&rstart,&rend);CHKERRQ(ierr); 2153 aij = (Mat_SeqBAIJ*)(Mreuse)->data; 2154 ii = aij->i; 2155 jj = aij->j; 2156 aa = aij->a; 2157 for (i=0; i<m; i++) { 2158 row = rstart/bs + i; 2159 nz = ii[i+1] - ii[i]; 2160 cwork = jj; jj += nz; 2161 vwork = aa; aa += nz*bs*bs; 2162 ierr = MatSetValuesBlocked_MPIBAIJ(M,1,&row,nz,cwork,vwork,INSERT_VALUES);CHKERRQ(ierr); 2163 } 2164 2165 ierr = MatAssemblyBegin(M,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2166 ierr = MatAssemblyEnd(M,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2167 *newmat = M; 2168 2169 /* save submatrix used in processor for next request */ 2170 if (call == MAT_INITIAL_MATRIX) { 2171 ierr = PetscObjectCompose((PetscObject)M,"SubMatrix",(PetscObject)Mreuse);CHKERRQ(ierr); 2172 ierr = PetscObjectDereference((PetscObject)Mreuse);CHKERRQ(ierr); 2173 } 2174 PetscFunctionReturn(0); 2175 } 2176 2177 #undef __FUNCT__ 2178 #define __FUNCT__ "MatPermute_MPIBAIJ" 2179 PetscErrorCode MatPermute_MPIBAIJ(Mat A,IS rowp,IS colp,Mat *B) 2180 { 2181 MPI_Comm comm,pcomm; 2182 PetscInt first,rlocal_size,clocal_size,nrows; 2183 const PetscInt *rows; 2184 PetscMPIInt size; 2185 IS crowp,growp,irowp,lrowp,lcolp; 2186 PetscErrorCode ierr; 2187 2188 PetscFunctionBegin; 2189 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 2190 /* make a collective version of 'rowp' */ 2191 ierr = PetscObjectGetComm((PetscObject)rowp,&pcomm);CHKERRQ(ierr); 2192 if (pcomm==comm) { 2193 crowp = rowp; 2194 } else { 2195 ierr = ISGetSize(rowp,&nrows);CHKERRQ(ierr); 2196 ierr = ISGetIndices(rowp,&rows);CHKERRQ(ierr); 2197 ierr = ISCreateGeneral(comm,nrows,rows,PETSC_COPY_VALUES,&crowp);CHKERRQ(ierr); 2198 ierr = ISRestoreIndices(rowp,&rows);CHKERRQ(ierr); 2199 } 2200 /* collect the global row permutation and invert it */ 2201 ierr = ISAllGather(crowp,&growp);CHKERRQ(ierr); 2202 ierr = ISSetPermutation(growp);CHKERRQ(ierr); 2203 if (pcomm!=comm) { 2204 ierr = ISDestroy(&crowp);CHKERRQ(ierr); 2205 } 2206 ierr = ISInvertPermutation(growp,PETSC_DECIDE,&irowp);CHKERRQ(ierr); 2207 ierr = ISDestroy(&growp);CHKERRQ(ierr); 2208 /* get the local target indices */ 2209 ierr = MatGetOwnershipRange(A,&first,NULL);CHKERRQ(ierr); 2210 ierr = MatGetLocalSize(A,&rlocal_size,&clocal_size);CHKERRQ(ierr); 2211 ierr = ISGetIndices(irowp,&rows);CHKERRQ(ierr); 2212 ierr = ISCreateGeneral(MPI_COMM_SELF,rlocal_size,rows+first,PETSC_COPY_VALUES,&lrowp);CHKERRQ(ierr); 2213 ierr = ISRestoreIndices(irowp,&rows);CHKERRQ(ierr); 2214 ierr = ISDestroy(&irowp);CHKERRQ(ierr); 2215 /* the column permutation is so much easier; 2216 make a local version of 'colp' and invert it */ 2217 ierr = PetscObjectGetComm((PetscObject)colp,&pcomm);CHKERRQ(ierr); 2218 ierr = MPI_Comm_size(pcomm,&size);CHKERRQ(ierr); 2219 if (size==1) { 2220 lcolp = colp; 2221 } else { 2222 ierr = ISAllGather(colp,&lcolp);CHKERRQ(ierr); 2223 } 2224 ierr = ISSetPermutation(lcolp);CHKERRQ(ierr); 2225 /* now we just get the submatrix */ 2226 ierr = MatGetSubMatrix_MPIBAIJ_Private(A,lrowp,lcolp,clocal_size,MAT_INITIAL_MATRIX,B);CHKERRQ(ierr); 2227 if (size>1) { 2228 ierr = ISDestroy(&lcolp);CHKERRQ(ierr); 2229 } 2230 /* clean up */ 2231 ierr = ISDestroy(&lrowp);CHKERRQ(ierr); 2232 PetscFunctionReturn(0); 2233 } 2234 2235 #undef __FUNCT__ 2236 #define __FUNCT__ "MatGetGhosts_MPIBAIJ" 2237 PetscErrorCode MatGetGhosts_MPIBAIJ(Mat mat,PetscInt *nghosts,const PetscInt *ghosts[]) 2238 { 2239 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*) mat->data; 2240 Mat_SeqBAIJ *B = (Mat_SeqBAIJ*)baij->B->data; 2241 2242 PetscFunctionBegin; 2243 if (nghosts) *nghosts = B->nbs; 2244 if (ghosts) *ghosts = baij->garray; 2245 PetscFunctionReturn(0); 2246 } 2247 2248 extern PetscErrorCode MatCreateColmap_MPIBAIJ_Private(Mat); 2249 2250 #undef __FUNCT__ 2251 #define __FUNCT__ "MatFDColoringCreate_MPIBAIJ" 2252 /* 2253 This routine is almost identical to MatFDColoringCreate_MPIBAIJ()! 2254 */ 2255 PetscErrorCode MatFDColoringCreate_MPIBAIJ(Mat mat,ISColoring iscoloring,MatFDColoring c) 2256 { 2257 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 2258 PetscErrorCode ierr; 2259 PetscMPIInt size,*ncolsonproc,*disp,nn; 2260 PetscInt bs,i,n,nrows,j,k,m,ncols,col; 2261 const PetscInt *is,*rows = 0,*A_ci,*A_cj,*B_ci,*B_cj,*ltog; 2262 PetscInt nis = iscoloring->n,nctot,*cols; 2263 PetscInt *rowhit,M,cstart,cend,colb; 2264 PetscInt *columnsforrow,l; 2265 IS *isa; 2266 PetscBool done,flg; 2267 ISLocalToGlobalMapping map = mat->cmap->bmapping; 2268 PetscInt ctype=c->ctype; 2269 2270 PetscFunctionBegin; 2271 if (!mat->assembled) SETERRQ(PetscObjectComm((PetscObject)mat),PETSC_ERR_ARG_WRONGSTATE,"Matrix must be assembled first; MatAssemblyBegin/End();"); 2272 if (ctype == IS_COLORING_GHOSTED && !map) SETERRQ(PetscObjectComm((PetscObject)mat),PETSC_ERR_ARG_INCOMP,"When using ghosted differencing matrix must have local to global mapping provided with MatSetLocalToGlobalMappingBlock"); 2273 2274 if (map) {ierr = ISLocalToGlobalMappingGetIndices(map,<og);CHKERRQ(ierr);} 2275 else ltog = NULL; 2276 ierr = ISColoringGetIS(iscoloring,PETSC_IGNORE,&isa);CHKERRQ(ierr); 2277 ierr = MatGetBlockSize(mat,&bs);CHKERRQ(ierr); 2278 2279 M = mat->rmap->n/bs; 2280 cstart = mat->cmap->rstart/bs; 2281 cend = mat->cmap->rend/bs; 2282 c->M = mat->rmap->N/bs; /* set the global rows and columns and local rows */ 2283 c->N = mat->cmap->N/bs; 2284 c->m = mat->rmap->n/bs; 2285 c->rstart = mat->rmap->rstart/bs; 2286 2287 c->ncolors = nis; 2288 ierr = PetscMalloc(nis*sizeof(PetscInt),&c->ncolumns);CHKERRQ(ierr); 2289 ierr = PetscMalloc(nis*sizeof(PetscInt*),&c->columns);CHKERRQ(ierr); 2290 ierr = PetscMalloc(nis*sizeof(PetscInt),&c->nrows);CHKERRQ(ierr); 2291 ierr = PetscMalloc(nis*sizeof(PetscInt*),&c->rows);CHKERRQ(ierr); 2292 ierr = PetscMalloc(nis*sizeof(PetscInt*),&c->columnsforrow);CHKERRQ(ierr); 2293 ierr = PetscLogObjectMemory((PetscObject)c,5*nis*sizeof(PetscInt));CHKERRQ(ierr); 2294 2295 /* Allow access to data structures of local part of matrix */ 2296 if (!baij->colmap) { 2297 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 2298 } 2299 ierr = MatGetColumnIJ(baij->A,0,PETSC_FALSE,PETSC_FALSE,&ncols,&A_ci,&A_cj,&done);CHKERRQ(ierr); 2300 ierr = MatGetColumnIJ(baij->B,0,PETSC_FALSE,PETSC_FALSE,&ncols,&B_ci,&B_cj,&done);CHKERRQ(ierr); 2301 2302 ierr = PetscMalloc((M+1)*sizeof(PetscInt),&rowhit);CHKERRQ(ierr); 2303 ierr = PetscMalloc((M+1)*sizeof(PetscInt),&columnsforrow);CHKERRQ(ierr); 2304 2305 for (i=0; i<nis; i++) { 2306 ierr = ISGetLocalSize(isa[i],&n);CHKERRQ(ierr); 2307 ierr = ISGetIndices(isa[i],&is);CHKERRQ(ierr); 2308 2309 c->ncolumns[i] = n; 2310 if (n) { 2311 ierr = PetscMalloc(n*sizeof(PetscInt),&c->columns[i]);CHKERRQ(ierr); 2312 ierr = PetscLogObjectMemory((PetscObject)c,n*sizeof(PetscInt));CHKERRQ(ierr); 2313 ierr = PetscMemcpy(c->columns[i],is,n*sizeof(PetscInt));CHKERRQ(ierr); 2314 } else { 2315 c->columns[i] = 0; 2316 } 2317 2318 if (ctype == IS_COLORING_GLOBAL) { 2319 /* Determine the total (parallel) number of columns of this color */ 2320 ierr = MPI_Comm_size(PetscObjectComm((PetscObject)mat),&size);CHKERRQ(ierr); 2321 ierr = PetscMalloc2(size,PetscMPIInt,&ncolsonproc,size,PetscMPIInt,&disp);CHKERRQ(ierr); 2322 2323 ierr = PetscMPIIntCast(n,&nn);CHKERRQ(ierr); 2324 ierr = MPI_Allgather(&nn,1,MPI_INT,ncolsonproc,1,MPI_INT,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 2325 nctot = 0; for (j=0; j<size; j++) nctot += ncolsonproc[j]; 2326 if (!nctot) { 2327 ierr = PetscInfo(mat,"Coloring of matrix has some unneeded colors with no corresponding rows\n");CHKERRQ(ierr); 2328 } 2329 2330 disp[0] = 0; 2331 for (j=1; j<size; j++) { 2332 disp[j] = disp[j-1] + ncolsonproc[j-1]; 2333 } 2334 2335 /* Get complete list of columns for color on each processor */ 2336 ierr = PetscMalloc((nctot+1)*sizeof(PetscInt),&cols);CHKERRQ(ierr); 2337 ierr = MPI_Allgatherv((void*)is,n,MPIU_INT,cols,ncolsonproc,disp,MPIU_INT,PetscObjectComm((PetscObject)mat));CHKERRQ(ierr); 2338 ierr = PetscFree2(ncolsonproc,disp);CHKERRQ(ierr); 2339 } else if (ctype == IS_COLORING_GHOSTED) { 2340 /* Determine local number of columns of this color on this process, including ghost points */ 2341 nctot = n; 2342 ierr = PetscMalloc((nctot+1)*sizeof(PetscInt),&cols);CHKERRQ(ierr); 2343 ierr = PetscMemcpy(cols,is,n*sizeof(PetscInt));CHKERRQ(ierr); 2344 } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Not provided for this MatFDColoring type"); 2345 2346 /* 2347 Mark all rows affect by these columns 2348 */ 2349 /* Temporary option to allow for debugging/testing */ 2350 flg = PETSC_FALSE; 2351 ierr = PetscOptionsGetBool(NULL,"-matfdcoloring_slow",&flg,NULL);CHKERRQ(ierr); 2352 if (!flg) { /*-----------------------------------------------------------------------------*/ 2353 /* crude, fast version */ 2354 ierr = PetscMemzero(rowhit,M*sizeof(PetscInt));CHKERRQ(ierr); 2355 /* loop over columns*/ 2356 for (j=0; j<nctot; j++) { 2357 if (ctype == IS_COLORING_GHOSTED) { 2358 col = ltog[cols[j]]; 2359 } else { 2360 col = cols[j]; 2361 } 2362 if (col >= cstart && col < cend) { 2363 /* column is in diagonal block of matrix */ 2364 rows = A_cj + A_ci[col-cstart]; 2365 m = A_ci[col-cstart+1] - A_ci[col-cstart]; 2366 } else { 2367 #if defined(PETSC_USE_CTABLE) 2368 ierr = PetscTableFind(baij->colmap,col+1,&colb);CHKERRQ(ierr); 2369 colb--; 2370 #else 2371 colb = baij->colmap[col] - 1; 2372 #endif 2373 if (colb == -1) { 2374 m = 0; 2375 } else { 2376 colb = colb/bs; 2377 rows = B_cj + B_ci[colb]; 2378 m = B_ci[colb+1] - B_ci[colb]; 2379 } 2380 } 2381 /* loop over columns marking them in rowhit */ 2382 for (k=0; k<m; k++) { 2383 rowhit[*rows++] = col + 1; 2384 } 2385 } 2386 2387 /* count the number of hits */ 2388 nrows = 0; 2389 for (j=0; j<M; j++) { 2390 if (rowhit[j]) nrows++; 2391 } 2392 c->nrows[i] = nrows; 2393 ierr = PetscMalloc((nrows+1)*sizeof(PetscInt),&c->rows[i]);CHKERRQ(ierr); 2394 ierr = PetscMalloc((nrows+1)*sizeof(PetscInt),&c->columnsforrow[i]);CHKERRQ(ierr); 2395 ierr = PetscLogObjectMemory((PetscObject)c,2*(nrows+1)*sizeof(PetscInt));CHKERRQ(ierr); 2396 nrows = 0; 2397 for (j=0; j<M; j++) { 2398 if (rowhit[j]) { 2399 c->rows[i][nrows] = j; 2400 c->columnsforrow[i][nrows] = rowhit[j] - 1; 2401 nrows++; 2402 } 2403 } 2404 } else { /*-------------------------------------------------------------------------------*/ 2405 /* slow version, using rowhit as a linked list */ 2406 PetscInt currentcol,fm,mfm; 2407 rowhit[M] = M; 2408 nrows = 0; 2409 /* loop over columns*/ 2410 for (j=0; j<nctot; j++) { 2411 if (ctype == IS_COLORING_GHOSTED) { 2412 col = ltog[cols[j]]; 2413 } else { 2414 col = cols[j]; 2415 } 2416 if (col >= cstart && col < cend) { 2417 /* column is in diagonal block of matrix */ 2418 rows = A_cj + A_ci[col-cstart]; 2419 m = A_ci[col-cstart+1] - A_ci[col-cstart]; 2420 } else { 2421 #if defined(PETSC_USE_CTABLE) 2422 ierr = PetscTableFind(baij->colmap,col+1,&colb);CHKERRQ(ierr); 2423 colb--; 2424 #else 2425 colb = baij->colmap[col] - 1; 2426 #endif 2427 if (colb == -1) { 2428 m = 0; 2429 } else { 2430 colb = colb/bs; 2431 rows = B_cj + B_ci[colb]; 2432 m = B_ci[colb+1] - B_ci[colb]; 2433 } 2434 } 2435 2436 /* loop over columns marking them in rowhit */ 2437 fm = M; /* fm points to first entry in linked list */ 2438 for (k=0; k<m; k++) { 2439 currentcol = *rows++; 2440 /* is it already in the list? */ 2441 do { 2442 mfm = fm; 2443 fm = rowhit[fm]; 2444 } while (fm < currentcol); 2445 /* not in list so add it */ 2446 if (fm != currentcol) { 2447 nrows++; 2448 columnsforrow[currentcol] = col; 2449 /* next three lines insert new entry into linked list */ 2450 rowhit[mfm] = currentcol; 2451 rowhit[currentcol] = fm; 2452 fm = currentcol; 2453 /* fm points to present position in list since we know the columns are sorted */ 2454 } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Invalid coloring of matrix detected"); 2455 } 2456 } 2457 c->nrows[i] = nrows; 2458 ierr = PetscMalloc((nrows+1)*sizeof(PetscInt),&c->rows[i]);CHKERRQ(ierr); 2459 ierr = PetscMalloc((nrows+1)*sizeof(PetscInt),&c->columnsforrow[i]);CHKERRQ(ierr); 2460 ierr = PetscLogObjectMemory((PetscObject)c,(nrows+1)*sizeof(PetscInt));CHKERRQ(ierr); 2461 /* now store the linked list of rows into c->rows[i] */ 2462 nrows = 0; 2463 fm = rowhit[M]; 2464 do { 2465 c->rows[i][nrows] = fm; 2466 c->columnsforrow[i][nrows++] = columnsforrow[fm]; 2467 fm = rowhit[fm]; 2468 } while (fm < M); 2469 } /* ---------------------------------------------------------------------------------------*/ 2470 ierr = PetscFree(cols);CHKERRQ(ierr); 2471 } 2472 2473 /* Optimize by adding the vscale, and scaleforrow[][] fields */ 2474 /* 2475 vscale will contain the "diagonal" on processor scalings followed by the off processor 2476 */ 2477 if (ctype == IS_COLORING_GLOBAL) { 2478 PetscInt *garray; 2479 ierr = PetscMalloc(baij->B->cmap->n*sizeof(PetscInt),&garray);CHKERRQ(ierr); 2480 for (i=0; i<baij->B->cmap->n/bs; i++) { 2481 for (j=0; j<bs; j++) { 2482 garray[i*bs+j] = bs*baij->garray[i]+j; 2483 } 2484 } 2485 ierr = VecCreateGhost(PetscObjectComm((PetscObject)mat),baij->A->rmap->n,PETSC_DETERMINE,baij->B->cmap->n,garray,&c->vscale);CHKERRQ(ierr); 2486 ierr = PetscFree(garray);CHKERRQ(ierr); 2487 ierr = PetscMalloc(c->ncolors*sizeof(PetscInt*),&c->vscaleforrow);CHKERRQ(ierr); 2488 for (k=0; k<c->ncolors; k++) { 2489 ierr = PetscMalloc((c->nrows[k]+1)*sizeof(PetscInt),&c->vscaleforrow[k]);CHKERRQ(ierr); 2490 for (l=0; l<c->nrows[k]; l++) { 2491 col = c->columnsforrow[k][l]; 2492 if (col >= cstart && col < cend) { 2493 /* column is in diagonal block of matrix */ 2494 colb = col - cstart; 2495 } else { 2496 /* column is in "off-processor" part */ 2497 #if defined(PETSC_USE_CTABLE) 2498 ierr = PetscTableFind(baij->colmap,col+1,&colb);CHKERRQ(ierr); 2499 colb--; 2500 #else 2501 colb = baij->colmap[col] - 1; 2502 #endif 2503 colb = colb/bs; 2504 colb += cend - cstart; 2505 } 2506 c->vscaleforrow[k][l] = colb; 2507 } 2508 } 2509 } else if (ctype == IS_COLORING_GHOSTED) { 2510 /* Get gtol mapping */ 2511 PetscInt N = mat->cmap->N,nlocal,*gtol; 2512 ierr = PetscMalloc((N+1)*sizeof(PetscInt),>ol);CHKERRQ(ierr); 2513 for (i=0; i<N; i++) gtol[i] = -1; 2514 ierr = ISLocalToGlobalMappingGetSize(map,&nlocal);CHKERRQ(ierr); 2515 for (i=0; i<nlocal; i++) gtol[ltog[i]] = i; 2516 2517 c->vscale = 0; /* will be created in MatFDColoringApply() */ 2518 ierr = PetscMalloc(c->ncolors*sizeof(PetscInt*),&c->vscaleforrow);CHKERRQ(ierr); 2519 for (k=0; k<c->ncolors; k++) { 2520 ierr = PetscMalloc((c->nrows[k]+1)*sizeof(PetscInt),&c->vscaleforrow[k]);CHKERRQ(ierr); 2521 for (l=0; l<c->nrows[k]; l++) { 2522 col = c->columnsforrow[k][l]; /* global column index */ 2523 2524 c->vscaleforrow[k][l] = gtol[col]; /* local column index */ 2525 } 2526 } 2527 ierr = PetscFree(gtol);CHKERRQ(ierr); 2528 } 2529 ierr = ISColoringRestoreIS(iscoloring,&isa);CHKERRQ(ierr); 2530 2531 ierr = PetscFree(rowhit);CHKERRQ(ierr); 2532 ierr = PetscFree(columnsforrow);CHKERRQ(ierr); 2533 ierr = MatRestoreColumnIJ(baij->A,0,PETSC_FALSE,PETSC_FALSE,&ncols,&A_ci,&A_cj,&done);CHKERRQ(ierr); 2534 ierr = MatRestoreColumnIJ(baij->B,0,PETSC_FALSE,PETSC_FALSE,&ncols,&B_ci,&B_cj,&done);CHKERRQ(ierr); 2535 if (map) {ierr = ISLocalToGlobalMappingRestoreIndices(map,<og);CHKERRQ(ierr);} 2536 PetscFunctionReturn(0); 2537 } 2538 2539 #undef __FUNCT__ 2540 #define __FUNCT__ "MatGetSeqNonzeroStructure_MPIBAIJ" 2541 PetscErrorCode MatGetSeqNonzeroStructure_MPIBAIJ(Mat A,Mat *newmat) 2542 { 2543 Mat B; 2544 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 2545 Mat_SeqBAIJ *ad = (Mat_SeqBAIJ*)a->A->data,*bd = (Mat_SeqBAIJ*)a->B->data; 2546 Mat_SeqAIJ *b; 2547 PetscErrorCode ierr; 2548 PetscMPIInt size,rank,*recvcounts = 0,*displs = 0; 2549 PetscInt sendcount,i,*rstarts = A->rmap->range,n,cnt,j,bs = A->rmap->bs; 2550 PetscInt m,*garray = a->garray,*lens,*jsendbuf,*a_jsendbuf,*b_jsendbuf; 2551 2552 PetscFunctionBegin; 2553 ierr = MPI_Comm_size(PetscObjectComm((PetscObject)A),&size);CHKERRQ(ierr); 2554 ierr = MPI_Comm_rank(PetscObjectComm((PetscObject)A),&rank);CHKERRQ(ierr); 2555 2556 /* ---------------------------------------------------------------- 2557 Tell every processor the number of nonzeros per row 2558 */ 2559 ierr = PetscMalloc((A->rmap->N/bs)*sizeof(PetscInt),&lens);CHKERRQ(ierr); 2560 for (i=A->rmap->rstart/bs; i<A->rmap->rend/bs; i++) { 2561 lens[i] = ad->i[i-A->rmap->rstart/bs+1] - ad->i[i-A->rmap->rstart/bs] + bd->i[i-A->rmap->rstart/bs+1] - bd->i[i-A->rmap->rstart/bs]; 2562 } 2563 sendcount = A->rmap->rend/bs - A->rmap->rstart/bs; 2564 ierr = PetscMalloc(2*size*sizeof(PetscMPIInt),&recvcounts);CHKERRQ(ierr); 2565 displs = recvcounts + size; 2566 for (i=0; i<size; i++) { 2567 recvcounts[i] = A->rmap->range[i+1]/bs - A->rmap->range[i]/bs; 2568 displs[i] = A->rmap->range[i]/bs; 2569 } 2570 #if defined(PETSC_HAVE_MPI_IN_PLACE) 2571 ierr = MPI_Allgatherv(MPI_IN_PLACE,0,MPI_DATATYPE_NULL,lens,recvcounts,displs,MPIU_INT,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2572 #else 2573 ierr = MPI_Allgatherv(lens+A->rmap->rstart/bs,sendcount,MPIU_INT,lens,recvcounts,displs,MPIU_INT,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2574 #endif 2575 /* --------------------------------------------------------------- 2576 Create the sequential matrix of the same type as the local block diagonal 2577 */ 2578 ierr = MatCreate(PETSC_COMM_SELF,&B);CHKERRQ(ierr); 2579 ierr = MatSetSizes(B,A->rmap->N/bs,A->cmap->N/bs,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 2580 ierr = MatSetType(B,MATSEQAIJ);CHKERRQ(ierr); 2581 ierr = MatSeqAIJSetPreallocation(B,0,lens);CHKERRQ(ierr); 2582 b = (Mat_SeqAIJ*)B->data; 2583 2584 /*-------------------------------------------------------------------- 2585 Copy my part of matrix column indices over 2586 */ 2587 sendcount = ad->nz + bd->nz; 2588 jsendbuf = b->j + b->i[rstarts[rank]/bs]; 2589 a_jsendbuf = ad->j; 2590 b_jsendbuf = bd->j; 2591 n = A->rmap->rend/bs - A->rmap->rstart/bs; 2592 cnt = 0; 2593 for (i=0; i<n; i++) { 2594 2595 /* put in lower diagonal portion */ 2596 m = bd->i[i+1] - bd->i[i]; 2597 while (m > 0) { 2598 /* is it above diagonal (in bd (compressed) numbering) */ 2599 if (garray[*b_jsendbuf] > A->rmap->rstart/bs + i) break; 2600 jsendbuf[cnt++] = garray[*b_jsendbuf++]; 2601 m--; 2602 } 2603 2604 /* put in diagonal portion */ 2605 for (j=ad->i[i]; j<ad->i[i+1]; j++) { 2606 jsendbuf[cnt++] = A->rmap->rstart/bs + *a_jsendbuf++; 2607 } 2608 2609 /* put in upper diagonal portion */ 2610 while (m-- > 0) { 2611 jsendbuf[cnt++] = garray[*b_jsendbuf++]; 2612 } 2613 } 2614 if (cnt != sendcount) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Corrupted PETSc matrix: nz given %D actual nz %D",sendcount,cnt); 2615 2616 /*-------------------------------------------------------------------- 2617 Gather all column indices to all processors 2618 */ 2619 for (i=0; i<size; i++) { 2620 recvcounts[i] = 0; 2621 for (j=A->rmap->range[i]/bs; j<A->rmap->range[i+1]/bs; j++) { 2622 recvcounts[i] += lens[j]; 2623 } 2624 } 2625 displs[0] = 0; 2626 for (i=1; i<size; i++) { 2627 displs[i] = displs[i-1] + recvcounts[i-1]; 2628 } 2629 #if defined(PETSC_HAVE_MPI_IN_PLACE) 2630 ierr = MPI_Allgatherv(MPI_IN_PLACE,0,MPI_DATATYPE_NULL,b->j,recvcounts,displs,MPIU_INT,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2631 #else 2632 ierr = MPI_Allgatherv(jsendbuf,sendcount,MPIU_INT,b->j,recvcounts,displs,MPIU_INT,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2633 #endif 2634 /*-------------------------------------------------------------------- 2635 Assemble the matrix into useable form (note numerical values not yet set) 2636 */ 2637 /* set the b->ilen (length of each row) values */ 2638 ierr = PetscMemcpy(b->ilen,lens,(A->rmap->N/bs)*sizeof(PetscInt));CHKERRQ(ierr); 2639 /* set the b->i indices */ 2640 b->i[0] = 0; 2641 for (i=1; i<=A->rmap->N/bs; i++) { 2642 b->i[i] = b->i[i-1] + lens[i-1]; 2643 } 2644 ierr = PetscFree(lens);CHKERRQ(ierr); 2645 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2646 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2647 ierr = PetscFree(recvcounts);CHKERRQ(ierr); 2648 2649 if (A->symmetric) { 2650 ierr = MatSetOption(B,MAT_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 2651 } else if (A->hermitian) { 2652 ierr = MatSetOption(B,MAT_HERMITIAN,PETSC_TRUE);CHKERRQ(ierr); 2653 } else if (A->structurally_symmetric) { 2654 ierr = MatSetOption(B,MAT_STRUCTURALLY_SYMMETRIC,PETSC_TRUE);CHKERRQ(ierr); 2655 } 2656 *newmat = B; 2657 PetscFunctionReturn(0); 2658 } 2659 2660 #undef __FUNCT__ 2661 #define __FUNCT__ "MatSOR_MPIBAIJ" 2662 PetscErrorCode MatSOR_MPIBAIJ(Mat matin,Vec bb,PetscReal omega,MatSORType flag,PetscReal fshift,PetscInt its,PetscInt lits,Vec xx) 2663 { 2664 Mat_MPIBAIJ *mat = (Mat_MPIBAIJ*)matin->data; 2665 PetscErrorCode ierr; 2666 Vec bb1 = 0; 2667 2668 PetscFunctionBegin; 2669 if (flag == SOR_APPLY_UPPER) { 2670 ierr = (*mat->A->ops->sor)(mat->A,bb,omega,flag,fshift,lits,1,xx);CHKERRQ(ierr); 2671 PetscFunctionReturn(0); 2672 } 2673 2674 if (its > 1 || ~flag & SOR_ZERO_INITIAL_GUESS) { 2675 ierr = VecDuplicate(bb,&bb1);CHKERRQ(ierr); 2676 } 2677 2678 if ((flag & SOR_LOCAL_SYMMETRIC_SWEEP) == SOR_LOCAL_SYMMETRIC_SWEEP) { 2679 if (flag & SOR_ZERO_INITIAL_GUESS) { 2680 ierr = (*mat->A->ops->sor)(mat->A,bb,omega,flag,fshift,lits,1,xx);CHKERRQ(ierr); 2681 its--; 2682 } 2683 2684 while (its--) { 2685 ierr = VecScatterBegin(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2686 ierr = VecScatterEnd(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2687 2688 /* update rhs: bb1 = bb - B*x */ 2689 ierr = VecScale(mat->lvec,-1.0);CHKERRQ(ierr); 2690 ierr = (*mat->B->ops->multadd)(mat->B,mat->lvec,bb,bb1);CHKERRQ(ierr); 2691 2692 /* local sweep */ 2693 ierr = (*mat->A->ops->sor)(mat->A,bb1,omega,SOR_SYMMETRIC_SWEEP,fshift,lits,1,xx);CHKERRQ(ierr); 2694 } 2695 } else if (flag & SOR_LOCAL_FORWARD_SWEEP) { 2696 if (flag & SOR_ZERO_INITIAL_GUESS) { 2697 ierr = (*mat->A->ops->sor)(mat->A,bb,omega,flag,fshift,lits,1,xx);CHKERRQ(ierr); 2698 its--; 2699 } 2700 while (its--) { 2701 ierr = VecScatterBegin(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2702 ierr = VecScatterEnd(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2703 2704 /* update rhs: bb1 = bb - B*x */ 2705 ierr = VecScale(mat->lvec,-1.0);CHKERRQ(ierr); 2706 ierr = (*mat->B->ops->multadd)(mat->B,mat->lvec,bb,bb1);CHKERRQ(ierr); 2707 2708 /* local sweep */ 2709 ierr = (*mat->A->ops->sor)(mat->A,bb1,omega,SOR_FORWARD_SWEEP,fshift,lits,1,xx);CHKERRQ(ierr); 2710 } 2711 } else if (flag & SOR_LOCAL_BACKWARD_SWEEP) { 2712 if (flag & SOR_ZERO_INITIAL_GUESS) { 2713 ierr = (*mat->A->ops->sor)(mat->A,bb,omega,flag,fshift,lits,1,xx);CHKERRQ(ierr); 2714 its--; 2715 } 2716 while (its--) { 2717 ierr = VecScatterBegin(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2718 ierr = VecScatterEnd(mat->Mvctx,xx,mat->lvec,INSERT_VALUES,SCATTER_FORWARD);CHKERRQ(ierr); 2719 2720 /* update rhs: bb1 = bb - B*x */ 2721 ierr = VecScale(mat->lvec,-1.0);CHKERRQ(ierr); 2722 ierr = (*mat->B->ops->multadd)(mat->B,mat->lvec,bb,bb1);CHKERRQ(ierr); 2723 2724 /* local sweep */ 2725 ierr = (*mat->A->ops->sor)(mat->A,bb1,omega,SOR_BACKWARD_SWEEP,fshift,lits,1,xx);CHKERRQ(ierr); 2726 } 2727 } else SETERRQ(PetscObjectComm((PetscObject)matin),PETSC_ERR_SUP,"Parallel version of SOR requested not supported"); 2728 2729 ierr = VecDestroy(&bb1);CHKERRQ(ierr); 2730 PetscFunctionReturn(0); 2731 } 2732 2733 extern PetscErrorCode MatFDColoringApply_BAIJ(Mat,MatFDColoring,Vec,MatStructure*,void*); 2734 2735 #undef __FUNCT__ 2736 #define __FUNCT__ "MatGetColumnNorms_MPIBAIJ" 2737 PetscErrorCode MatGetColumnNorms_MPIBAIJ(Mat A,NormType type,PetscReal *norms) 2738 { 2739 PetscErrorCode ierr; 2740 Mat_MPIBAIJ *aij = (Mat_MPIBAIJ*)A->data; 2741 PetscInt N,i,*garray = aij->garray; 2742 PetscInt ib,jb,bs = A->rmap->bs; 2743 Mat_SeqBAIJ *a_aij = (Mat_SeqBAIJ*) aij->A->data; 2744 MatScalar *a_val = a_aij->a; 2745 Mat_SeqBAIJ *b_aij = (Mat_SeqBAIJ*) aij->B->data; 2746 MatScalar *b_val = b_aij->a; 2747 PetscReal *work; 2748 2749 PetscFunctionBegin; 2750 ierr = MatGetSize(A,NULL,&N);CHKERRQ(ierr); 2751 ierr = PetscMalloc(N*sizeof(PetscReal),&work);CHKERRQ(ierr); 2752 ierr = PetscMemzero(work,N*sizeof(PetscReal));CHKERRQ(ierr); 2753 if (type == NORM_2) { 2754 for (i=a_aij->i[0]; i<a_aij->i[aij->A->rmap->n/bs]; i++) { 2755 for (jb=0; jb<bs; jb++) { 2756 for (ib=0; ib<bs; ib++) { 2757 work[A->cmap->rstart + a_aij->j[i] * bs + jb] += PetscAbsScalar(*a_val * *a_val); 2758 a_val++; 2759 } 2760 } 2761 } 2762 for (i=b_aij->i[0]; i<b_aij->i[aij->B->rmap->n/bs]; i++) { 2763 for (jb=0; jb<bs; jb++) { 2764 for (ib=0; ib<bs; ib++) { 2765 work[garray[b_aij->j[i]] * bs + jb] += PetscAbsScalar(*b_val * *b_val); 2766 b_val++; 2767 } 2768 } 2769 } 2770 } else if (type == NORM_1) { 2771 for (i=a_aij->i[0]; i<a_aij->i[aij->A->rmap->n/bs]; i++) { 2772 for (jb=0; jb<bs; jb++) { 2773 for (ib=0; ib<bs; ib++) { 2774 work[A->cmap->rstart + a_aij->j[i] * bs + jb] += PetscAbsScalar(*a_val); 2775 a_val++; 2776 } 2777 } 2778 } 2779 for (i=b_aij->i[0]; i<b_aij->i[aij->B->rmap->n/bs]; i++) { 2780 for (jb=0; jb<bs; jb++) { 2781 for (ib=0; ib<bs; ib++) { 2782 work[garray[b_aij->j[i]] * bs + jb] += PetscAbsScalar(*b_val); 2783 b_val++; 2784 } 2785 } 2786 } 2787 } else if (type == NORM_INFINITY) { 2788 for (i=a_aij->i[0]; i<a_aij->i[aij->A->rmap->n/bs]; i++) { 2789 for (jb=0; jb<bs; jb++) { 2790 for (ib=0; ib<bs; ib++) { 2791 int col = A->cmap->rstart + a_aij->j[i] * bs + jb; 2792 work[col] = PetscMax(PetscAbsScalar(*a_val), work[col]); 2793 a_val++; 2794 } 2795 } 2796 } 2797 for (i=b_aij->i[0]; i<b_aij->i[aij->B->rmap->n/bs]; i++) { 2798 for (jb=0; jb<bs; jb++) { 2799 for (ib=0; ib<bs; ib++) { 2800 int col = garray[b_aij->j[i]] * bs + jb; 2801 work[col] = PetscMax(PetscAbsScalar(*b_val), work[col]); 2802 b_val++; 2803 } 2804 } 2805 } 2806 } else SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_ARG_WRONG,"Unknown NormType"); 2807 if (type == NORM_INFINITY) { 2808 ierr = MPI_Allreduce(work,norms,N,MPIU_REAL,MPIU_MAX,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2809 } else { 2810 ierr = MPI_Allreduce(work,norms,N,MPIU_REAL,MPIU_SUM,PetscObjectComm((PetscObject)A));CHKERRQ(ierr); 2811 } 2812 ierr = PetscFree(work);CHKERRQ(ierr); 2813 if (type == NORM_2) { 2814 for (i=0; i<N; i++) norms[i] = PetscSqrtReal(norms[i]); 2815 } 2816 PetscFunctionReturn(0); 2817 } 2818 2819 #undef __FUNCT__ 2820 #define __FUNCT__ "MatInvertBlockDiagonal_MPIBAIJ" 2821 PetscErrorCode MatInvertBlockDiagonal_MPIBAIJ(Mat A,const PetscScalar **values) 2822 { 2823 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*) A->data; 2824 PetscErrorCode ierr; 2825 2826 PetscFunctionBegin; 2827 ierr = MatInvertBlockDiagonal(a->A,values);CHKERRQ(ierr); 2828 PetscFunctionReturn(0); 2829 } 2830 2831 2832 /* -------------------------------------------------------------------*/ 2833 static struct _MatOps MatOps_Values = {MatSetValues_MPIBAIJ, 2834 MatGetRow_MPIBAIJ, 2835 MatRestoreRow_MPIBAIJ, 2836 MatMult_MPIBAIJ, 2837 /* 4*/ MatMultAdd_MPIBAIJ, 2838 MatMultTranspose_MPIBAIJ, 2839 MatMultTransposeAdd_MPIBAIJ, 2840 0, 2841 0, 2842 0, 2843 /*10*/ 0, 2844 0, 2845 0, 2846 MatSOR_MPIBAIJ, 2847 MatTranspose_MPIBAIJ, 2848 /*15*/ MatGetInfo_MPIBAIJ, 2849 MatEqual_MPIBAIJ, 2850 MatGetDiagonal_MPIBAIJ, 2851 MatDiagonalScale_MPIBAIJ, 2852 MatNorm_MPIBAIJ, 2853 /*20*/ MatAssemblyBegin_MPIBAIJ, 2854 MatAssemblyEnd_MPIBAIJ, 2855 MatSetOption_MPIBAIJ, 2856 MatZeroEntries_MPIBAIJ, 2857 /*24*/ MatZeroRows_MPIBAIJ, 2858 0, 2859 0, 2860 0, 2861 0, 2862 /*29*/ MatSetUp_MPIBAIJ, 2863 0, 2864 0, 2865 0, 2866 0, 2867 /*34*/ MatDuplicate_MPIBAIJ, 2868 0, 2869 0, 2870 0, 2871 0, 2872 /*39*/ MatAXPY_MPIBAIJ, 2873 MatGetSubMatrices_MPIBAIJ, 2874 MatIncreaseOverlap_MPIBAIJ, 2875 MatGetValues_MPIBAIJ, 2876 MatCopy_MPIBAIJ, 2877 /*44*/ 0, 2878 MatScale_MPIBAIJ, 2879 0, 2880 0, 2881 0, 2882 /*49*/ 0, 2883 0, 2884 0, 2885 0, 2886 0, 2887 /*54*/ MatFDColoringCreate_MPIBAIJ, 2888 0, 2889 MatSetUnfactored_MPIBAIJ, 2890 MatPermute_MPIBAIJ, 2891 MatSetValuesBlocked_MPIBAIJ, 2892 /*59*/ MatGetSubMatrix_MPIBAIJ, 2893 MatDestroy_MPIBAIJ, 2894 MatView_MPIBAIJ, 2895 0, 2896 0, 2897 /*64*/ 0, 2898 0, 2899 0, 2900 0, 2901 0, 2902 /*69*/ MatGetRowMaxAbs_MPIBAIJ, 2903 0, 2904 0, 2905 0, 2906 0, 2907 /*74*/ 0, 2908 MatFDColoringApply_BAIJ, 2909 0, 2910 0, 2911 0, 2912 /*79*/ 0, 2913 0, 2914 0, 2915 0, 2916 MatLoad_MPIBAIJ, 2917 /*84*/ 0, 2918 0, 2919 0, 2920 0, 2921 0, 2922 /*89*/ 0, 2923 0, 2924 0, 2925 0, 2926 0, 2927 /*94*/ 0, 2928 0, 2929 0, 2930 0, 2931 0, 2932 /*99*/ 0, 2933 0, 2934 0, 2935 0, 2936 0, 2937 /*104*/0, 2938 MatRealPart_MPIBAIJ, 2939 MatImaginaryPart_MPIBAIJ, 2940 0, 2941 0, 2942 /*109*/0, 2943 0, 2944 0, 2945 0, 2946 0, 2947 /*114*/MatGetSeqNonzeroStructure_MPIBAIJ, 2948 0, 2949 MatGetGhosts_MPIBAIJ, 2950 0, 2951 0, 2952 /*119*/0, 2953 0, 2954 0, 2955 0, 2956 MatGetMultiProcBlock_MPIBAIJ, 2957 /*124*/0, 2958 MatGetColumnNorms_MPIBAIJ, 2959 MatInvertBlockDiagonal_MPIBAIJ, 2960 0, 2961 0, 2962 /*129*/ 0, 2963 0, 2964 0, 2965 0, 2966 0, 2967 /*134*/ 0, 2968 0, 2969 0, 2970 0, 2971 0, 2972 /*139*/ 0, 2973 0 2974 }; 2975 2976 #undef __FUNCT__ 2977 #define __FUNCT__ "MatGetDiagonalBlock_MPIBAIJ" 2978 PetscErrorCode MatGetDiagonalBlock_MPIBAIJ(Mat A,Mat *a) 2979 { 2980 PetscFunctionBegin; 2981 *a = ((Mat_MPIBAIJ*)A->data)->A; 2982 PetscFunctionReturn(0); 2983 } 2984 2985 PETSC_EXTERN PetscErrorCode MatConvert_MPIBAIJ_MPISBAIJ(Mat, MatType,MatReuse,Mat*); 2986 2987 #undef __FUNCT__ 2988 #define __FUNCT__ "MatMPIBAIJSetPreallocationCSR_MPIBAIJ" 2989 PetscErrorCode MatMPIBAIJSetPreallocationCSR_MPIBAIJ(Mat B,PetscInt bs,const PetscInt ii[],const PetscInt jj[],const PetscScalar V[]) 2990 { 2991 PetscInt m,rstart,cstart,cend; 2992 PetscInt i,j,d,nz,nz_max=0,*d_nnz=0,*o_nnz=0; 2993 const PetscInt *JJ =0; 2994 PetscScalar *values=0; 2995 PetscErrorCode ierr; 2996 2997 PetscFunctionBegin; 2998 ierr = PetscLayoutSetBlockSize(B->rmap,bs);CHKERRQ(ierr); 2999 ierr = PetscLayoutSetBlockSize(B->cmap,bs);CHKERRQ(ierr); 3000 ierr = PetscLayoutSetUp(B->rmap);CHKERRQ(ierr); 3001 ierr = PetscLayoutSetUp(B->cmap);CHKERRQ(ierr); 3002 ierr = PetscLayoutGetBlockSize(B->rmap,&bs);CHKERRQ(ierr); 3003 m = B->rmap->n/bs; 3004 rstart = B->rmap->rstart/bs; 3005 cstart = B->cmap->rstart/bs; 3006 cend = B->cmap->rend/bs; 3007 3008 if (ii[0]) SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"ii[0] must be 0 but it is %D",ii[0]); 3009 ierr = PetscMalloc2(m,PetscInt,&d_nnz,m,PetscInt,&o_nnz);CHKERRQ(ierr); 3010 for (i=0; i<m; i++) { 3011 nz = ii[i+1] - ii[i]; 3012 if (nz < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Local row %D has a negative number of columns %D",i,nz); 3013 nz_max = PetscMax(nz_max,nz); 3014 JJ = jj + ii[i]; 3015 for (j=0; j<nz; j++) { 3016 if (*JJ >= cstart) break; 3017 JJ++; 3018 } 3019 d = 0; 3020 for (; j<nz; j++) { 3021 if (*JJ++ >= cend) break; 3022 d++; 3023 } 3024 d_nnz[i] = d; 3025 o_nnz[i] = nz - d; 3026 } 3027 ierr = MatMPIBAIJSetPreallocation(B,bs,0,d_nnz,0,o_nnz);CHKERRQ(ierr); 3028 ierr = PetscFree2(d_nnz,o_nnz);CHKERRQ(ierr); 3029 3030 values = (PetscScalar*)V; 3031 if (!values) { 3032 ierr = PetscMalloc(bs*bs*nz_max*sizeof(PetscScalar),&values);CHKERRQ(ierr); 3033 ierr = PetscMemzero(values,bs*bs*nz_max*sizeof(PetscScalar));CHKERRQ(ierr); 3034 } 3035 for (i=0; i<m; i++) { 3036 PetscInt row = i + rstart; 3037 PetscInt ncols = ii[i+1] - ii[i]; 3038 const PetscInt *icols = jj + ii[i]; 3039 const PetscScalar *svals = values + (V ? (bs*bs*ii[i]) : 0); 3040 ierr = MatSetValuesBlocked_MPIBAIJ(B,1,&row,ncols,icols,svals,INSERT_VALUES);CHKERRQ(ierr); 3041 } 3042 3043 if (!V) { ierr = PetscFree(values);CHKERRQ(ierr); } 3044 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3045 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3046 ierr = MatSetOption(B,MAT_NEW_NONZERO_LOCATION_ERR,PETSC_TRUE);CHKERRQ(ierr); 3047 PetscFunctionReturn(0); 3048 } 3049 3050 #undef __FUNCT__ 3051 #define __FUNCT__ "MatMPIBAIJSetPreallocationCSR" 3052 /*@C 3053 MatMPIBAIJSetPreallocationCSR - Allocates memory for a sparse parallel matrix in BAIJ format 3054 (the default parallel PETSc format). 3055 3056 Collective on MPI_Comm 3057 3058 Input Parameters: 3059 + A - the matrix 3060 . bs - the block size 3061 . i - the indices into j for the start of each local row (starts with zero) 3062 . j - the column indices for each local row (starts with zero) these must be sorted for each row 3063 - v - optional values in the matrix 3064 3065 Level: developer 3066 3067 .keywords: matrix, aij, compressed row, sparse, parallel 3068 3069 .seealso: MatCreate(), MatCreateSeqAIJ(), MatSetValues(), MatMPIBAIJSetPreallocation(), MatCreateAIJ(), MPIAIJ 3070 @*/ 3071 PetscErrorCode MatMPIBAIJSetPreallocationCSR(Mat B,PetscInt bs,const PetscInt i[],const PetscInt j[], const PetscScalar v[]) 3072 { 3073 PetscErrorCode ierr; 3074 3075 PetscFunctionBegin; 3076 PetscValidHeaderSpecific(B,MAT_CLASSID,1); 3077 PetscValidType(B,1); 3078 PetscValidLogicalCollectiveInt(B,bs,2); 3079 ierr = PetscTryMethod(B,"MatMPIBAIJSetPreallocationCSR_C",(Mat,PetscInt,const PetscInt[],const PetscInt[],const PetscScalar[]),(B,bs,i,j,v));CHKERRQ(ierr); 3080 PetscFunctionReturn(0); 3081 } 3082 3083 #undef __FUNCT__ 3084 #define __FUNCT__ "MatMPIBAIJSetPreallocation_MPIBAIJ" 3085 PetscErrorCode MatMPIBAIJSetPreallocation_MPIBAIJ(Mat B,PetscInt bs,PetscInt d_nz,const PetscInt *d_nnz,PetscInt o_nz,const PetscInt *o_nnz) 3086 { 3087 Mat_MPIBAIJ *b; 3088 PetscErrorCode ierr; 3089 PetscInt i; 3090 3091 PetscFunctionBegin; 3092 ierr = PetscLayoutSetBlockSize(B->rmap,bs);CHKERRQ(ierr); 3093 ierr = PetscLayoutSetBlockSize(B->cmap,bs);CHKERRQ(ierr); 3094 ierr = PetscLayoutSetUp(B->rmap);CHKERRQ(ierr); 3095 ierr = PetscLayoutSetUp(B->cmap);CHKERRQ(ierr); 3096 ierr = PetscLayoutGetBlockSize(B->rmap,&bs);CHKERRQ(ierr); 3097 3098 if (d_nnz) { 3099 for (i=0; i<B->rmap->n/bs; i++) { 3100 if (d_nnz[i] < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"d_nnz cannot be less than -1: local row %D value %D",i,d_nnz[i]); 3101 } 3102 } 3103 if (o_nnz) { 3104 for (i=0; i<B->rmap->n/bs; i++) { 3105 if (o_nnz[i] < 0) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"o_nnz cannot be less than -1: local row %D value %D",i,o_nnz[i]); 3106 } 3107 } 3108 3109 b = (Mat_MPIBAIJ*)B->data; 3110 b->bs2 = bs*bs; 3111 b->mbs = B->rmap->n/bs; 3112 b->nbs = B->cmap->n/bs; 3113 b->Mbs = B->rmap->N/bs; 3114 b->Nbs = B->cmap->N/bs; 3115 3116 for (i=0; i<=b->size; i++) { 3117 b->rangebs[i] = B->rmap->range[i]/bs; 3118 } 3119 b->rstartbs = B->rmap->rstart/bs; 3120 b->rendbs = B->rmap->rend/bs; 3121 b->cstartbs = B->cmap->rstart/bs; 3122 b->cendbs = B->cmap->rend/bs; 3123 3124 if (!B->preallocated) { 3125 ierr = MatCreate(PETSC_COMM_SELF,&b->A);CHKERRQ(ierr); 3126 ierr = MatSetSizes(b->A,B->rmap->n,B->cmap->n,B->rmap->n,B->cmap->n);CHKERRQ(ierr); 3127 ierr = MatSetType(b->A,MATSEQBAIJ);CHKERRQ(ierr); 3128 ierr = PetscLogObjectParent((PetscObject)B,(PetscObject)b->A);CHKERRQ(ierr); 3129 ierr = MatCreate(PETSC_COMM_SELF,&b->B);CHKERRQ(ierr); 3130 ierr = MatSetSizes(b->B,B->rmap->n,B->cmap->N,B->rmap->n,B->cmap->N);CHKERRQ(ierr); 3131 ierr = MatSetType(b->B,MATSEQBAIJ);CHKERRQ(ierr); 3132 ierr = PetscLogObjectParent((PetscObject)B,(PetscObject)b->B);CHKERRQ(ierr); 3133 ierr = MatStashCreate_Private(PetscObjectComm((PetscObject)B),bs,&B->bstash);CHKERRQ(ierr); 3134 } 3135 3136 ierr = MatSeqBAIJSetPreallocation(b->A,bs,d_nz,d_nnz);CHKERRQ(ierr); 3137 ierr = MatSeqBAIJSetPreallocation(b->B,bs,o_nz,o_nnz);CHKERRQ(ierr); 3138 B->preallocated = PETSC_TRUE; 3139 PetscFunctionReturn(0); 3140 } 3141 3142 extern PetscErrorCode MatDiagonalScaleLocal_MPIBAIJ(Mat,Vec); 3143 extern PetscErrorCode MatSetHashTableFactor_MPIBAIJ(Mat,PetscReal); 3144 3145 #undef __FUNCT__ 3146 #define __FUNCT__ "MatConvert_MPIBAIJ_MPIAdj" 3147 PETSC_EXTERN PetscErrorCode MatConvert_MPIBAIJ_MPIAdj(Mat B, MatType newtype,MatReuse reuse,Mat *adj) 3148 { 3149 Mat_MPIBAIJ *b = (Mat_MPIBAIJ*)B->data; 3150 PetscErrorCode ierr; 3151 Mat_SeqBAIJ *d = (Mat_SeqBAIJ*) b->A->data,*o = (Mat_SeqBAIJ*) b->B->data; 3152 PetscInt M = B->rmap->n/B->rmap->bs,i,*ii,*jj,cnt,j,k,rstart = B->rmap->rstart/B->rmap->bs; 3153 const PetscInt *id = d->i, *jd = d->j, *io = o->i, *jo = o->j, *garray = b->garray; 3154 3155 PetscFunctionBegin; 3156 ierr = PetscMalloc((M+1)*sizeof(PetscInt),&ii);CHKERRQ(ierr); 3157 ii[0] = 0; 3158 for (i=0; i<M; i++) { 3159 if ((id[i+1] - id[i]) < 0) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Indices wrong %D %D %D",i,id[i],id[i+1]); 3160 if ((io[i+1] - io[i]) < 0) SETERRQ3(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Indices wrong %D %D %D",i,io[i],io[i+1]); 3161 ii[i+1] = ii[i] + id[i+1] - id[i] + io[i+1] - io[i]; 3162 /* remove one from count of matrix has diagonal */ 3163 for (j=id[i]; j<id[i+1]; j++) { 3164 if (jd[j] == i) {ii[i+1]--;break;} 3165 } 3166 } 3167 ierr = PetscMalloc(ii[M]*sizeof(PetscInt),&jj);CHKERRQ(ierr); 3168 cnt = 0; 3169 for (i=0; i<M; i++) { 3170 for (j=io[i]; j<io[i+1]; j++) { 3171 if (garray[jo[j]] > rstart) break; 3172 jj[cnt++] = garray[jo[j]]; 3173 } 3174 for (k=id[i]; k<id[i+1]; k++) { 3175 if (jd[k] != i) { 3176 jj[cnt++] = rstart + jd[k]; 3177 } 3178 } 3179 for (; j<io[i+1]; j++) { 3180 jj[cnt++] = garray[jo[j]]; 3181 } 3182 } 3183 ierr = MatCreateMPIAdj(PetscObjectComm((PetscObject)B),M,B->cmap->N/B->rmap->bs,ii,jj,NULL,adj);CHKERRQ(ierr); 3184 PetscFunctionReturn(0); 3185 } 3186 3187 #include <../src/mat/impls/aij/mpi/mpiaij.h> 3188 3189 PETSC_EXTERN PetscErrorCode MatConvert_SeqBAIJ_SeqAIJ(Mat,MatType,MatReuse,Mat*); 3190 3191 #undef __FUNCT__ 3192 #define __FUNCT__ "MatConvert_MPIBAIJ_MPIAIJ" 3193 PETSC_EXTERN PetscErrorCode MatConvert_MPIBAIJ_MPIAIJ(Mat A,MatType newtype,MatReuse reuse,Mat *newmat) 3194 { 3195 PetscErrorCode ierr; 3196 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 3197 Mat B; 3198 Mat_MPIAIJ *b; 3199 3200 PetscFunctionBegin; 3201 if (!A->assembled) SETERRQ(PetscObjectComm((PetscObject)A),PETSC_ERR_SUP,"Matrix must be assembled"); 3202 3203 ierr = MatCreate(PetscObjectComm((PetscObject)A),&B);CHKERRQ(ierr); 3204 ierr = MatSetSizes(B,A->rmap->n,A->cmap->n,A->rmap->N,A->cmap->N);CHKERRQ(ierr); 3205 ierr = MatSetType(B,MATMPIAIJ);CHKERRQ(ierr); 3206 ierr = MatSeqAIJSetPreallocation(B,0,NULL);CHKERRQ(ierr); 3207 ierr = MatMPIAIJSetPreallocation(B,0,NULL,0,NULL);CHKERRQ(ierr); 3208 b = (Mat_MPIAIJ*) B->data; 3209 3210 ierr = MatDestroy(&b->A);CHKERRQ(ierr); 3211 ierr = MatDestroy(&b->B);CHKERRQ(ierr); 3212 ierr = MatDisAssemble_MPIBAIJ(A);CHKERRQ(ierr); 3213 ierr = MatConvert_SeqBAIJ_SeqAIJ(a->A, MATSEQAIJ, MAT_INITIAL_MATRIX, &b->A);CHKERRQ(ierr); 3214 ierr = MatConvert_SeqBAIJ_SeqAIJ(a->B, MATSEQAIJ, MAT_INITIAL_MATRIX, &b->B);CHKERRQ(ierr); 3215 ierr = MatAssemblyBegin(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3216 ierr = MatAssemblyEnd(B,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3217 ierr = MatAssemblyBegin(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3218 ierr = MatAssemblyEnd(A,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3219 if (reuse == MAT_REUSE_MATRIX) { 3220 ierr = MatHeaderReplace(A,B);CHKERRQ(ierr); 3221 } else { 3222 *newmat = B; 3223 } 3224 PetscFunctionReturn(0); 3225 } 3226 3227 #if defined(PETSC_HAVE_MUMPS) 3228 PETSC_EXTERN PetscErrorCode MatGetFactor_baij_mumps(Mat,MatFactorType,Mat*); 3229 #endif 3230 3231 /*MC 3232 MATMPIBAIJ - MATMPIBAIJ = "mpibaij" - A matrix type to be used for distributed block sparse matrices. 3233 3234 Options Database Keys: 3235 + -mat_type mpibaij - sets the matrix type to "mpibaij" during a call to MatSetFromOptions() 3236 . -mat_block_size <bs> - set the blocksize used to store the matrix 3237 - -mat_use_hash_table <fact> 3238 3239 Level: beginner 3240 3241 .seealso: MatCreateMPIBAIJ 3242 M*/ 3243 3244 PETSC_EXTERN PetscErrorCode MatConvert_MPIBAIJ_MPIBSTRM(Mat,MatType,MatReuse,Mat*); 3245 3246 #undef __FUNCT__ 3247 #define __FUNCT__ "MatCreate_MPIBAIJ" 3248 PETSC_EXTERN PetscErrorCode MatCreate_MPIBAIJ(Mat B) 3249 { 3250 Mat_MPIBAIJ *b; 3251 PetscErrorCode ierr; 3252 PetscBool flg; 3253 3254 PetscFunctionBegin; 3255 ierr = PetscNewLog(B,Mat_MPIBAIJ,&b);CHKERRQ(ierr); 3256 B->data = (void*)b; 3257 3258 ierr = PetscMemcpy(B->ops,&MatOps_Values,sizeof(struct _MatOps));CHKERRQ(ierr); 3259 B->assembled = PETSC_FALSE; 3260 3261 B->insertmode = NOT_SET_VALUES; 3262 ierr = MPI_Comm_rank(PetscObjectComm((PetscObject)B),&b->rank);CHKERRQ(ierr); 3263 ierr = MPI_Comm_size(PetscObjectComm((PetscObject)B),&b->size);CHKERRQ(ierr); 3264 3265 /* build local table of row and column ownerships */ 3266 ierr = PetscMalloc((b->size+1)*sizeof(PetscInt),&b->rangebs);CHKERRQ(ierr); 3267 3268 /* build cache for off array entries formed */ 3269 ierr = MatStashCreate_Private(PetscObjectComm((PetscObject)B),1,&B->stash);CHKERRQ(ierr); 3270 3271 b->donotstash = PETSC_FALSE; 3272 b->colmap = NULL; 3273 b->garray = NULL; 3274 b->roworiented = PETSC_TRUE; 3275 3276 /* stuff used in block assembly */ 3277 b->barray = 0; 3278 3279 /* stuff used for matrix vector multiply */ 3280 b->lvec = 0; 3281 b->Mvctx = 0; 3282 3283 /* stuff for MatGetRow() */ 3284 b->rowindices = 0; 3285 b->rowvalues = 0; 3286 b->getrowactive = PETSC_FALSE; 3287 3288 /* hash table stuff */ 3289 b->ht = 0; 3290 b->hd = 0; 3291 b->ht_size = 0; 3292 b->ht_flag = PETSC_FALSE; 3293 b->ht_fact = 0; 3294 b->ht_total_ct = 0; 3295 b->ht_insert_ct = 0; 3296 3297 /* stuff for MatGetSubMatrices_MPIBAIJ_local() */ 3298 b->ijonly = PETSC_FALSE; 3299 3300 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)B),NULL,"Options for loading MPIBAIJ matrix 1","Mat");CHKERRQ(ierr); 3301 ierr = PetscOptionsBool("-mat_use_hash_table","Use hash table to save memory in constructing matrix","MatSetOption",PETSC_FALSE,&flg,NULL);CHKERRQ(ierr); 3302 if (flg) { 3303 PetscReal fact = 1.39; 3304 ierr = MatSetOption(B,MAT_USE_HASH_TABLE,PETSC_TRUE);CHKERRQ(ierr); 3305 ierr = PetscOptionsReal("-mat_use_hash_table","Use hash table factor","MatMPIBAIJSetHashTableFactor",fact,&fact,NULL);CHKERRQ(ierr); 3306 if (fact <= 1.0) fact = 1.39; 3307 ierr = MatMPIBAIJSetHashTableFactor(B,fact);CHKERRQ(ierr); 3308 ierr = PetscInfo1(B,"Hash table Factor used %5.2f\n",fact);CHKERRQ(ierr); 3309 } 3310 ierr = PetscOptionsEnd();CHKERRQ(ierr); 3311 3312 #if defined(PETSC_HAVE_MUMPS) 3313 ierr = PetscObjectComposeFunction((PetscObject)B,"MatGetFactor_mumps_C",MatGetFactor_baij_mumps);CHKERRQ(ierr); 3314 #endif 3315 ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_mpibaij_mpiadj_C",MatConvert_MPIBAIJ_MPIAdj);CHKERRQ(ierr); 3316 ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_mpibaij_mpiaij_C",MatConvert_MPIBAIJ_MPIAIJ);CHKERRQ(ierr); 3317 ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_mpibaij_mpisbaij_C",MatConvert_MPIBAIJ_MPISBAIJ);CHKERRQ(ierr); 3318 ierr = PetscObjectComposeFunction((PetscObject)B,"MatStoreValues_C",MatStoreValues_MPIBAIJ);CHKERRQ(ierr); 3319 ierr = PetscObjectComposeFunction((PetscObject)B,"MatRetrieveValues_C",MatRetrieveValues_MPIBAIJ);CHKERRQ(ierr); 3320 ierr = PetscObjectComposeFunction((PetscObject)B,"MatGetDiagonalBlock_C",MatGetDiagonalBlock_MPIBAIJ);CHKERRQ(ierr); 3321 ierr = PetscObjectComposeFunction((PetscObject)B,"MatMPIBAIJSetPreallocation_C",MatMPIBAIJSetPreallocation_MPIBAIJ);CHKERRQ(ierr); 3322 ierr = PetscObjectComposeFunction((PetscObject)B,"MatMPIBAIJSetPreallocationCSR_C",MatMPIBAIJSetPreallocationCSR_MPIBAIJ);CHKERRQ(ierr); 3323 ierr = PetscObjectComposeFunction((PetscObject)B,"MatDiagonalScaleLocal_C",MatDiagonalScaleLocal_MPIBAIJ);CHKERRQ(ierr); 3324 ierr = PetscObjectComposeFunction((PetscObject)B,"MatSetHashTableFactor_C",MatSetHashTableFactor_MPIBAIJ);CHKERRQ(ierr); 3325 ierr = PetscObjectComposeFunction((PetscObject)B,"MatConvert_mpibaij_mpibstrm_C",MatConvert_MPIBAIJ_MPIBSTRM);CHKERRQ(ierr); 3326 ierr = PetscObjectChangeTypeName((PetscObject)B,MATMPIBAIJ);CHKERRQ(ierr); 3327 PetscFunctionReturn(0); 3328 } 3329 3330 /*MC 3331 MATBAIJ - MATBAIJ = "baij" - A matrix type to be used for block sparse matrices. 3332 3333 This matrix type is identical to MATSEQBAIJ when constructed with a single process communicator, 3334 and MATMPIBAIJ otherwise. 3335 3336 Options Database Keys: 3337 . -mat_type baij - sets the matrix type to "baij" during a call to MatSetFromOptions() 3338 3339 Level: beginner 3340 3341 .seealso: MatCreateBAIJ(),MATSEQBAIJ,MATMPIBAIJ, MatMPIBAIJSetPreallocation(), MatMPIBAIJSetPreallocationCSR() 3342 M*/ 3343 3344 #undef __FUNCT__ 3345 #define __FUNCT__ "MatMPIBAIJSetPreallocation" 3346 /*@C 3347 MatMPIBAIJSetPreallocation - Allocates memory for a sparse parallel matrix in block AIJ format 3348 (block compressed row). For good matrix assembly performance 3349 the user should preallocate the matrix storage by setting the parameters 3350 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 3351 performance can be increased by more than a factor of 50. 3352 3353 Collective on Mat 3354 3355 Input Parameters: 3356 + A - the matrix 3357 . bs - size of block 3358 . d_nz - number of block nonzeros per block row in diagonal portion of local 3359 submatrix (same for all local rows) 3360 . d_nnz - array containing the number of block nonzeros in the various block rows 3361 of the in diagonal portion of the local (possibly different for each block 3362 row) or NULL. If you plan to factor the matrix you must leave room for the diagonal entry and 3363 set it even if it is zero. 3364 . o_nz - number of block nonzeros per block row in the off-diagonal portion of local 3365 submatrix (same for all local rows). 3366 - o_nnz - array containing the number of nonzeros in the various block rows of the 3367 off-diagonal portion of the local submatrix (possibly different for 3368 each block row) or NULL. 3369 3370 If the *_nnz parameter is given then the *_nz parameter is ignored 3371 3372 Options Database Keys: 3373 + -mat_block_size - size of the blocks to use 3374 - -mat_use_hash_table <fact> 3375 3376 Notes: 3377 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 3378 than it must be used on all processors that share the object for that argument. 3379 3380 Storage Information: 3381 For a square global matrix we define each processor's diagonal portion 3382 to be its local rows and the corresponding columns (a square submatrix); 3383 each processor's off-diagonal portion encompasses the remainder of the 3384 local matrix (a rectangular submatrix). 3385 3386 The user can specify preallocated storage for the diagonal part of 3387 the local submatrix with either d_nz or d_nnz (not both). Set 3388 d_nz=PETSC_DEFAULT and d_nnz=NULL for PETSc to control dynamic 3389 memory allocation. Likewise, specify preallocated storage for the 3390 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 3391 3392 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 3393 the figure below we depict these three local rows and all columns (0-11). 3394 3395 .vb 3396 0 1 2 3 4 5 6 7 8 9 10 11 3397 -------------------------- 3398 row 3 |o o o d d d o o o o o o 3399 row 4 |o o o d d d o o o o o o 3400 row 5 |o o o d d d o o o o o o 3401 -------------------------- 3402 .ve 3403 3404 Thus, any entries in the d locations are stored in the d (diagonal) 3405 submatrix, and any entries in the o locations are stored in the 3406 o (off-diagonal) submatrix. Note that the d and the o submatrices are 3407 stored simply in the MATSEQBAIJ format for compressed row storage. 3408 3409 Now d_nz should indicate the number of block nonzeros per row in the d matrix, 3410 and o_nz should indicate the number of block nonzeros per row in the o matrix. 3411 In general, for PDE problems in which most nonzeros are near the diagonal, 3412 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 3413 or you will get TERRIBLE performance; see the users' manual chapter on 3414 matrices. 3415 3416 You can call MatGetInfo() to get information on how effective the preallocation was; 3417 for example the fields mallocs,nz_allocated,nz_used,nz_unneeded; 3418 You can also run with the option -info and look for messages with the string 3419 malloc in them to see if additional memory allocation was needed. 3420 3421 Level: intermediate 3422 3423 .keywords: matrix, block, aij, compressed row, sparse, parallel 3424 3425 .seealso: MatCreate(), MatCreateSeqBAIJ(), MatSetValues(), MatCreateBAIJ(), MatMPIBAIJSetPreallocationCSR(), PetscSplitOwnership() 3426 @*/ 3427 PetscErrorCode MatMPIBAIJSetPreallocation(Mat B,PetscInt bs,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[]) 3428 { 3429 PetscErrorCode ierr; 3430 3431 PetscFunctionBegin; 3432 PetscValidHeaderSpecific(B,MAT_CLASSID,1); 3433 PetscValidType(B,1); 3434 PetscValidLogicalCollectiveInt(B,bs,2); 3435 ierr = PetscTryMethod(B,"MatMPIBAIJSetPreallocation_C",(Mat,PetscInt,PetscInt,const PetscInt[],PetscInt,const PetscInt[]),(B,bs,d_nz,d_nnz,o_nz,o_nnz));CHKERRQ(ierr); 3436 PetscFunctionReturn(0); 3437 } 3438 3439 #undef __FUNCT__ 3440 #define __FUNCT__ "MatCreateBAIJ" 3441 /*@C 3442 MatCreateBAIJ - Creates a sparse parallel matrix in block AIJ format 3443 (block compressed row). For good matrix assembly performance 3444 the user should preallocate the matrix storage by setting the parameters 3445 d_nz (or d_nnz) and o_nz (or o_nnz). By setting these parameters accurately, 3446 performance can be increased by more than a factor of 50. 3447 3448 Collective on MPI_Comm 3449 3450 Input Parameters: 3451 + comm - MPI communicator 3452 . bs - size of blockk 3453 . m - number of local rows (or PETSC_DECIDE to have calculated if M is given) 3454 This value should be the same as the local size used in creating the 3455 y vector for the matrix-vector product y = Ax. 3456 . n - number of local columns (or PETSC_DECIDE to have calculated if N is given) 3457 This value should be the same as the local size used in creating the 3458 x vector for the matrix-vector product y = Ax. 3459 . M - number of global rows (or PETSC_DETERMINE to have calculated if m is given) 3460 . N - number of global columns (or PETSC_DETERMINE to have calculated if n is given) 3461 . d_nz - number of nonzero blocks per block row in diagonal portion of local 3462 submatrix (same for all local rows) 3463 . d_nnz - array containing the number of nonzero blocks in the various block rows 3464 of the in diagonal portion of the local (possibly different for each block 3465 row) or NULL. If you plan to factor the matrix you must leave room for the diagonal entry 3466 and set it even if it is zero. 3467 . o_nz - number of nonzero blocks per block row in the off-diagonal portion of local 3468 submatrix (same for all local rows). 3469 - o_nnz - array containing the number of nonzero blocks in the various block rows of the 3470 off-diagonal portion of the local submatrix (possibly different for 3471 each block row) or NULL. 3472 3473 Output Parameter: 3474 . A - the matrix 3475 3476 Options Database Keys: 3477 + -mat_block_size - size of the blocks to use 3478 - -mat_use_hash_table <fact> 3479 3480 It is recommended that one use the MatCreate(), MatSetType() and/or MatSetFromOptions(), 3481 MatXXXXSetPreallocation() paradgm instead of this routine directly. 3482 [MatXXXXSetPreallocation() is, for example, MatSeqAIJSetPreallocation] 3483 3484 Notes: 3485 If the *_nnz parameter is given then the *_nz parameter is ignored 3486 3487 A nonzero block is any block that as 1 or more nonzeros in it 3488 3489 The user MUST specify either the local or global matrix dimensions 3490 (possibly both). 3491 3492 If PETSC_DECIDE or PETSC_DETERMINE is used for a particular argument on one processor 3493 than it must be used on all processors that share the object for that argument. 3494 3495 Storage Information: 3496 For a square global matrix we define each processor's diagonal portion 3497 to be its local rows and the corresponding columns (a square submatrix); 3498 each processor's off-diagonal portion encompasses the remainder of the 3499 local matrix (a rectangular submatrix). 3500 3501 The user can specify preallocated storage for the diagonal part of 3502 the local submatrix with either d_nz or d_nnz (not both). Set 3503 d_nz=PETSC_DEFAULT and d_nnz=NULL for PETSc to control dynamic 3504 memory allocation. Likewise, specify preallocated storage for the 3505 off-diagonal part of the local submatrix with o_nz or o_nnz (not both). 3506 3507 Consider a processor that owns rows 3, 4 and 5 of a parallel matrix. In 3508 the figure below we depict these three local rows and all columns (0-11). 3509 3510 .vb 3511 0 1 2 3 4 5 6 7 8 9 10 11 3512 -------------------------- 3513 row 3 |o o o d d d o o o o o o 3514 row 4 |o o o d d d o o o o o o 3515 row 5 |o o o d d d o o o o o o 3516 -------------------------- 3517 .ve 3518 3519 Thus, any entries in the d locations are stored in the d (diagonal) 3520 submatrix, and any entries in the o locations are stored in the 3521 o (off-diagonal) submatrix. Note that the d and the o submatrices are 3522 stored simply in the MATSEQBAIJ format for compressed row storage. 3523 3524 Now d_nz should indicate the number of block nonzeros per row in the d matrix, 3525 and o_nz should indicate the number of block nonzeros per row in the o matrix. 3526 In general, for PDE problems in which most nonzeros are near the diagonal, 3527 one expects d_nz >> o_nz. For large problems you MUST preallocate memory 3528 or you will get TERRIBLE performance; see the users' manual chapter on 3529 matrices. 3530 3531 Level: intermediate 3532 3533 .keywords: matrix, block, aij, compressed row, sparse, parallel 3534 3535 .seealso: MatCreate(), MatCreateSeqBAIJ(), MatSetValues(), MatCreateBAIJ(), MatMPIBAIJSetPreallocation(), MatMPIBAIJSetPreallocationCSR() 3536 @*/ 3537 PetscErrorCode MatCreateBAIJ(MPI_Comm comm,PetscInt bs,PetscInt m,PetscInt n,PetscInt M,PetscInt N,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[],Mat *A) 3538 { 3539 PetscErrorCode ierr; 3540 PetscMPIInt size; 3541 3542 PetscFunctionBegin; 3543 ierr = MatCreate(comm,A);CHKERRQ(ierr); 3544 ierr = MatSetSizes(*A,m,n,M,N);CHKERRQ(ierr); 3545 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 3546 if (size > 1) { 3547 ierr = MatSetType(*A,MATMPIBAIJ);CHKERRQ(ierr); 3548 ierr = MatMPIBAIJSetPreallocation(*A,bs,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 3549 } else { 3550 ierr = MatSetType(*A,MATSEQBAIJ);CHKERRQ(ierr); 3551 ierr = MatSeqBAIJSetPreallocation(*A,bs,d_nz,d_nnz);CHKERRQ(ierr); 3552 } 3553 PetscFunctionReturn(0); 3554 } 3555 3556 #undef __FUNCT__ 3557 #define __FUNCT__ "MatDuplicate_MPIBAIJ" 3558 static PetscErrorCode MatDuplicate_MPIBAIJ(Mat matin,MatDuplicateOption cpvalues,Mat *newmat) 3559 { 3560 Mat mat; 3561 Mat_MPIBAIJ *a,*oldmat = (Mat_MPIBAIJ*)matin->data; 3562 PetscErrorCode ierr; 3563 PetscInt len=0; 3564 3565 PetscFunctionBegin; 3566 *newmat = 0; 3567 ierr = MatCreate(PetscObjectComm((PetscObject)matin),&mat);CHKERRQ(ierr); 3568 ierr = MatSetSizes(mat,matin->rmap->n,matin->cmap->n,matin->rmap->N,matin->cmap->N);CHKERRQ(ierr); 3569 ierr = MatSetType(mat,((PetscObject)matin)->type_name);CHKERRQ(ierr); 3570 ierr = PetscMemcpy(mat->ops,matin->ops,sizeof(struct _MatOps));CHKERRQ(ierr); 3571 3572 mat->factortype = matin->factortype; 3573 mat->preallocated = PETSC_TRUE; 3574 mat->assembled = PETSC_TRUE; 3575 mat->insertmode = NOT_SET_VALUES; 3576 3577 a = (Mat_MPIBAIJ*)mat->data; 3578 mat->rmap->bs = matin->rmap->bs; 3579 a->bs2 = oldmat->bs2; 3580 a->mbs = oldmat->mbs; 3581 a->nbs = oldmat->nbs; 3582 a->Mbs = oldmat->Mbs; 3583 a->Nbs = oldmat->Nbs; 3584 3585 ierr = PetscLayoutReference(matin->rmap,&mat->rmap);CHKERRQ(ierr); 3586 ierr = PetscLayoutReference(matin->cmap,&mat->cmap);CHKERRQ(ierr); 3587 3588 a->size = oldmat->size; 3589 a->rank = oldmat->rank; 3590 a->donotstash = oldmat->donotstash; 3591 a->roworiented = oldmat->roworiented; 3592 a->rowindices = 0; 3593 a->rowvalues = 0; 3594 a->getrowactive = PETSC_FALSE; 3595 a->barray = 0; 3596 a->rstartbs = oldmat->rstartbs; 3597 a->rendbs = oldmat->rendbs; 3598 a->cstartbs = oldmat->cstartbs; 3599 a->cendbs = oldmat->cendbs; 3600 3601 /* hash table stuff */ 3602 a->ht = 0; 3603 a->hd = 0; 3604 a->ht_size = 0; 3605 a->ht_flag = oldmat->ht_flag; 3606 a->ht_fact = oldmat->ht_fact; 3607 a->ht_total_ct = 0; 3608 a->ht_insert_ct = 0; 3609 3610 ierr = PetscMemcpy(a->rangebs,oldmat->rangebs,(a->size+1)*sizeof(PetscInt));CHKERRQ(ierr); 3611 if (oldmat->colmap) { 3612 #if defined(PETSC_USE_CTABLE) 3613 ierr = PetscTableCreateCopy(oldmat->colmap,&a->colmap);CHKERRQ(ierr); 3614 #else 3615 ierr = PetscMalloc((a->Nbs)*sizeof(PetscInt),&a->colmap);CHKERRQ(ierr); 3616 ierr = PetscLogObjectMemory((PetscObject)mat,(a->Nbs)*sizeof(PetscInt));CHKERRQ(ierr); 3617 ierr = PetscMemcpy(a->colmap,oldmat->colmap,(a->Nbs)*sizeof(PetscInt));CHKERRQ(ierr); 3618 #endif 3619 } else a->colmap = 0; 3620 3621 if (oldmat->garray && (len = ((Mat_SeqBAIJ*)(oldmat->B->data))->nbs)) { 3622 ierr = PetscMalloc(len*sizeof(PetscInt),&a->garray);CHKERRQ(ierr); 3623 ierr = PetscLogObjectMemory((PetscObject)mat,len*sizeof(PetscInt));CHKERRQ(ierr); 3624 ierr = PetscMemcpy(a->garray,oldmat->garray,len*sizeof(PetscInt));CHKERRQ(ierr); 3625 } else a->garray = 0; 3626 3627 ierr = MatStashCreate_Private(PetscObjectComm((PetscObject)matin),matin->rmap->bs,&mat->bstash);CHKERRQ(ierr); 3628 ierr = VecDuplicate(oldmat->lvec,&a->lvec);CHKERRQ(ierr); 3629 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)a->lvec);CHKERRQ(ierr); 3630 ierr = VecScatterCopy(oldmat->Mvctx,&a->Mvctx);CHKERRQ(ierr); 3631 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)a->Mvctx);CHKERRQ(ierr); 3632 3633 ierr = MatDuplicate(oldmat->A,cpvalues,&a->A);CHKERRQ(ierr); 3634 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)a->A);CHKERRQ(ierr); 3635 ierr = MatDuplicate(oldmat->B,cpvalues,&a->B);CHKERRQ(ierr); 3636 ierr = PetscLogObjectParent((PetscObject)mat,(PetscObject)a->B);CHKERRQ(ierr); 3637 ierr = PetscFunctionListDuplicate(((PetscObject)matin)->qlist,&((PetscObject)mat)->qlist);CHKERRQ(ierr); 3638 *newmat = mat; 3639 PetscFunctionReturn(0); 3640 } 3641 3642 #undef __FUNCT__ 3643 #define __FUNCT__ "MatLoad_MPIBAIJ" 3644 PetscErrorCode MatLoad_MPIBAIJ(Mat newmat,PetscViewer viewer) 3645 { 3646 PetscErrorCode ierr; 3647 int fd; 3648 PetscInt i,nz,j,rstart,rend; 3649 PetscScalar *vals,*buf; 3650 MPI_Comm comm; 3651 MPI_Status status; 3652 PetscMPIInt rank,size,maxnz; 3653 PetscInt header[4],*rowlengths = 0,M,N,m,*rowners,*cols; 3654 PetscInt *locrowlens = NULL,*procsnz = NULL,*browners = NULL; 3655 PetscInt jj,*mycols,*ibuf,bs=1,Mbs,mbs,extra_rows,mmax; 3656 PetscMPIInt tag = ((PetscObject)viewer)->tag; 3657 PetscInt *dlens = NULL,*odlens = NULL,*mask = NULL,*masked1 = NULL,*masked2 = NULL,rowcount,odcount; 3658 PetscInt dcount,kmax,k,nzcount,tmp,mend,sizesset=1,grows,gcols; 3659 3660 PetscFunctionBegin; 3661 ierr = PetscObjectGetComm((PetscObject)viewer,&comm);CHKERRQ(ierr); 3662 ierr = PetscOptionsBegin(comm,NULL,"Options for loading MPIBAIJ matrix 2","Mat");CHKERRQ(ierr); 3663 ierr = PetscOptionsInt("-matload_block_size","Set the blocksize used to store the matrix","MatLoad",bs,&bs,NULL);CHKERRQ(ierr); 3664 ierr = PetscOptionsEnd();CHKERRQ(ierr); 3665 3666 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 3667 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 3668 if (!rank) { 3669 ierr = PetscViewerBinaryGetDescriptor(viewer,&fd);CHKERRQ(ierr); 3670 ierr = PetscBinaryRead(fd,(char*)header,4,PETSC_INT);CHKERRQ(ierr); 3671 if (header[0] != MAT_FILE_CLASSID) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED,"not matrix object"); 3672 } 3673 3674 if (newmat->rmap->n < 0 && newmat->rmap->N < 0 && newmat->cmap->n < 0 && newmat->cmap->N < 0) sizesset = 0; 3675 3676 ierr = MPI_Bcast(header+1,3,MPIU_INT,0,comm);CHKERRQ(ierr); 3677 M = header[1]; N = header[2]; 3678 3679 /* If global rows/cols are set to PETSC_DECIDE, set it to the sizes given in the file */ 3680 if (sizesset && newmat->rmap->N < 0) newmat->rmap->N = M; 3681 if (sizesset && newmat->cmap->N < 0) newmat->cmap->N = N; 3682 3683 /* If global sizes are set, check if they are consistent with that given in the file */ 3684 if (sizesset) { 3685 ierr = MatGetSize(newmat,&grows,&gcols);CHKERRQ(ierr); 3686 } 3687 if (sizesset && newmat->rmap->N != grows) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED, "Inconsistent # of rows:Matrix in file has (%d) and input matrix has (%d)",M,grows); 3688 if (sizesset && newmat->cmap->N != gcols) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED, "Inconsistent # of cols:Matrix in file has (%d) and input matrix has (%d)",N,gcols); 3689 3690 if (M != N) SETERRQ(PetscObjectComm((PetscObject)viewer),PETSC_ERR_SUP,"Can only do square matrices"); 3691 3692 /* 3693 This code adds extra rows to make sure the number of rows is 3694 divisible by the blocksize 3695 */ 3696 Mbs = M/bs; 3697 extra_rows = bs - M + bs*Mbs; 3698 if (extra_rows == bs) extra_rows = 0; 3699 else Mbs++; 3700 if (extra_rows && !rank) { 3701 ierr = PetscInfo(viewer,"Padding loaded matrix to match blocksize\n");CHKERRQ(ierr); 3702 } 3703 3704 /* determine ownership of all rows */ 3705 if (newmat->rmap->n < 0) { /* PETSC_DECIDE */ 3706 mbs = Mbs/size + ((Mbs % size) > rank); 3707 m = mbs*bs; 3708 } else { /* User set */ 3709 m = newmat->rmap->n; 3710 mbs = m/bs; 3711 } 3712 ierr = PetscMalloc2(size+1,PetscInt,&rowners,size+1,PetscInt,&browners);CHKERRQ(ierr); 3713 ierr = MPI_Allgather(&mbs,1,MPIU_INT,rowners+1,1,MPIU_INT,comm);CHKERRQ(ierr); 3714 3715 /* process 0 needs enough room for process with most rows */ 3716 if (!rank) { 3717 mmax = rowners[1]; 3718 for (i=2; i<=size; i++) { 3719 mmax = PetscMax(mmax,rowners[i]); 3720 } 3721 mmax*=bs; 3722 } else mmax = -1; /* unused, but compiler warns anyway */ 3723 3724 rowners[0] = 0; 3725 for (i=2; i<=size; i++) rowners[i] += rowners[i-1]; 3726 for (i=0; i<=size; i++) browners[i] = rowners[i]*bs; 3727 rstart = rowners[rank]; 3728 rend = rowners[rank+1]; 3729 3730 /* distribute row lengths to all processors */ 3731 ierr = PetscMalloc(m*sizeof(PetscInt),&locrowlens);CHKERRQ(ierr); 3732 if (!rank) { 3733 mend = m; 3734 if (size == 1) mend = mend - extra_rows; 3735 ierr = PetscBinaryRead(fd,locrowlens,mend,PETSC_INT);CHKERRQ(ierr); 3736 for (j=mend; j<m; j++) locrowlens[j] = 1; 3737 ierr = PetscMalloc(mmax*sizeof(PetscInt),&rowlengths);CHKERRQ(ierr); 3738 ierr = PetscMalloc(size*sizeof(PetscInt),&procsnz);CHKERRQ(ierr); 3739 ierr = PetscMemzero(procsnz,size*sizeof(PetscInt));CHKERRQ(ierr); 3740 for (j=0; j<m; j++) { 3741 procsnz[0] += locrowlens[j]; 3742 } 3743 for (i=1; i<size; i++) { 3744 mend = browners[i+1] - browners[i]; 3745 if (i == size-1) mend = mend - extra_rows; 3746 ierr = PetscBinaryRead(fd,rowlengths,mend,PETSC_INT);CHKERRQ(ierr); 3747 for (j=mend; j<browners[i+1] - browners[i]; j++) rowlengths[j] = 1; 3748 /* calculate the number of nonzeros on each processor */ 3749 for (j=0; j<browners[i+1]-browners[i]; j++) { 3750 procsnz[i] += rowlengths[j]; 3751 } 3752 ierr = MPI_Send(rowlengths,browners[i+1]-browners[i],MPIU_INT,i,tag,comm);CHKERRQ(ierr); 3753 } 3754 ierr = PetscFree(rowlengths);CHKERRQ(ierr); 3755 } else { 3756 ierr = MPI_Recv(locrowlens,m,MPIU_INT,0,tag,comm,&status);CHKERRQ(ierr); 3757 } 3758 3759 if (!rank) { 3760 /* determine max buffer needed and allocate it */ 3761 maxnz = procsnz[0]; 3762 for (i=1; i<size; i++) { 3763 maxnz = PetscMax(maxnz,procsnz[i]); 3764 } 3765 ierr = PetscMalloc(maxnz*sizeof(PetscInt),&cols);CHKERRQ(ierr); 3766 3767 /* read in my part of the matrix column indices */ 3768 nz = procsnz[0]; 3769 ierr = PetscMalloc((nz+1)*sizeof(PetscInt),&ibuf);CHKERRQ(ierr); 3770 mycols = ibuf; 3771 if (size == 1) nz -= extra_rows; 3772 ierr = PetscBinaryRead(fd,mycols,nz,PETSC_INT);CHKERRQ(ierr); 3773 if (size == 1) { 3774 for (i=0; i< extra_rows; i++) mycols[nz+i] = M+i; 3775 } 3776 3777 /* read in every ones (except the last) and ship off */ 3778 for (i=1; i<size-1; i++) { 3779 nz = procsnz[i]; 3780 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 3781 ierr = MPI_Send(cols,nz,MPIU_INT,i,tag,comm);CHKERRQ(ierr); 3782 } 3783 /* read in the stuff for the last proc */ 3784 if (size != 1) { 3785 nz = procsnz[size-1] - extra_rows; /* the extra rows are not on the disk */ 3786 ierr = PetscBinaryRead(fd,cols,nz,PETSC_INT);CHKERRQ(ierr); 3787 for (i=0; i<extra_rows; i++) cols[nz+i] = M+i; 3788 ierr = MPI_Send(cols,nz+extra_rows,MPIU_INT,size-1,tag,comm);CHKERRQ(ierr); 3789 } 3790 ierr = PetscFree(cols);CHKERRQ(ierr); 3791 } else { 3792 /* determine buffer space needed for message */ 3793 nz = 0; 3794 for (i=0; i<m; i++) { 3795 nz += locrowlens[i]; 3796 } 3797 ierr = PetscMalloc((nz+1)*sizeof(PetscInt),&ibuf);CHKERRQ(ierr); 3798 mycols = ibuf; 3799 /* receive message of column indices*/ 3800 ierr = MPI_Recv(mycols,nz,MPIU_INT,0,tag,comm,&status);CHKERRQ(ierr); 3801 ierr = MPI_Get_count(&status,MPIU_INT,&maxnz);CHKERRQ(ierr); 3802 if (maxnz != nz) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_FILE_UNEXPECTED,"something is wrong with file"); 3803 } 3804 3805 /* loop over local rows, determining number of off diagonal entries */ 3806 ierr = PetscMalloc2(rend-rstart,PetscInt,&dlens,rend-rstart,PetscInt,&odlens);CHKERRQ(ierr); 3807 ierr = PetscMalloc3(Mbs,PetscInt,&mask,Mbs,PetscInt,&masked1,Mbs,PetscInt,&masked2);CHKERRQ(ierr); 3808 ierr = PetscMemzero(mask,Mbs*sizeof(PetscInt));CHKERRQ(ierr); 3809 ierr = PetscMemzero(masked1,Mbs*sizeof(PetscInt));CHKERRQ(ierr); 3810 ierr = PetscMemzero(masked2,Mbs*sizeof(PetscInt));CHKERRQ(ierr); 3811 rowcount = 0; nzcount = 0; 3812 for (i=0; i<mbs; i++) { 3813 dcount = 0; 3814 odcount = 0; 3815 for (j=0; j<bs; j++) { 3816 kmax = locrowlens[rowcount]; 3817 for (k=0; k<kmax; k++) { 3818 tmp = mycols[nzcount++]/bs; 3819 if (!mask[tmp]) { 3820 mask[tmp] = 1; 3821 if (tmp < rstart || tmp >= rend) masked2[odcount++] = tmp; 3822 else masked1[dcount++] = tmp; 3823 } 3824 } 3825 rowcount++; 3826 } 3827 3828 dlens[i] = dcount; 3829 odlens[i] = odcount; 3830 3831 /* zero out the mask elements we set */ 3832 for (j=0; j<dcount; j++) mask[masked1[j]] = 0; 3833 for (j=0; j<odcount; j++) mask[masked2[j]] = 0; 3834 } 3835 3836 3837 if (!sizesset) { 3838 ierr = MatSetSizes(newmat,m,m,M+extra_rows,N+extra_rows);CHKERRQ(ierr); 3839 } 3840 ierr = MatMPIBAIJSetPreallocation(newmat,bs,0,dlens,0,odlens);CHKERRQ(ierr); 3841 3842 if (!rank) { 3843 ierr = PetscMalloc((maxnz+1)*sizeof(PetscScalar),&buf);CHKERRQ(ierr); 3844 /* read in my part of the matrix numerical values */ 3845 nz = procsnz[0]; 3846 vals = buf; 3847 mycols = ibuf; 3848 if (size == 1) nz -= extra_rows; 3849 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 3850 if (size == 1) { 3851 for (i=0; i< extra_rows; i++) vals[nz+i] = 1.0; 3852 } 3853 3854 /* insert into matrix */ 3855 jj = rstart*bs; 3856 for (i=0; i<m; i++) { 3857 ierr = MatSetValues_MPIBAIJ(newmat,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 3858 mycols += locrowlens[i]; 3859 vals += locrowlens[i]; 3860 jj++; 3861 } 3862 /* read in other processors (except the last one) and ship out */ 3863 for (i=1; i<size-1; i++) { 3864 nz = procsnz[i]; 3865 vals = buf; 3866 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 3867 ierr = MPIULong_Send(vals,nz,MPIU_SCALAR,i,((PetscObject)newmat)->tag,comm);CHKERRQ(ierr); 3868 } 3869 /* the last proc */ 3870 if (size != 1) { 3871 nz = procsnz[i] - extra_rows; 3872 vals = buf; 3873 ierr = PetscBinaryRead(fd,vals,nz,PETSC_SCALAR);CHKERRQ(ierr); 3874 for (i=0; i<extra_rows; i++) vals[nz+i] = 1.0; 3875 ierr = MPIULong_Send(vals,nz+extra_rows,MPIU_SCALAR,size-1,((PetscObject)newmat)->tag,comm);CHKERRQ(ierr); 3876 } 3877 ierr = PetscFree(procsnz);CHKERRQ(ierr); 3878 } else { 3879 /* receive numeric values */ 3880 ierr = PetscMalloc((nz+1)*sizeof(PetscScalar),&buf);CHKERRQ(ierr); 3881 3882 /* receive message of values*/ 3883 vals = buf; 3884 mycols = ibuf; 3885 ierr = MPIULong_Recv(vals,nz,MPIU_SCALAR,0,((PetscObject)newmat)->tag,comm);CHKERRQ(ierr); 3886 3887 /* insert into matrix */ 3888 jj = rstart*bs; 3889 for (i=0; i<m; i++) { 3890 ierr = MatSetValues_MPIBAIJ(newmat,1,&jj,locrowlens[i],mycols,vals,INSERT_VALUES);CHKERRQ(ierr); 3891 mycols += locrowlens[i]; 3892 vals += locrowlens[i]; 3893 jj++; 3894 } 3895 } 3896 ierr = PetscFree(locrowlens);CHKERRQ(ierr); 3897 ierr = PetscFree(buf);CHKERRQ(ierr); 3898 ierr = PetscFree(ibuf);CHKERRQ(ierr); 3899 ierr = PetscFree2(rowners,browners);CHKERRQ(ierr); 3900 ierr = PetscFree2(dlens,odlens);CHKERRQ(ierr); 3901 ierr = PetscFree3(mask,masked1,masked2);CHKERRQ(ierr); 3902 ierr = MatAssemblyBegin(newmat,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3903 ierr = MatAssemblyEnd(newmat,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 3904 PetscFunctionReturn(0); 3905 } 3906 3907 #undef __FUNCT__ 3908 #define __FUNCT__ "MatMPIBAIJSetHashTableFactor" 3909 /*@ 3910 MatMPIBAIJSetHashTableFactor - Sets the factor required to compute the size of the HashTable. 3911 3912 Input Parameters: 3913 . mat - the matrix 3914 . fact - factor 3915 3916 Not Collective, each process can use a different factor 3917 3918 Level: advanced 3919 3920 Notes: 3921 This can also be set by the command line option: -mat_use_hash_table <fact> 3922 3923 .keywords: matrix, hashtable, factor, HT 3924 3925 .seealso: MatSetOption() 3926 @*/ 3927 PetscErrorCode MatMPIBAIJSetHashTableFactor(Mat mat,PetscReal fact) 3928 { 3929 PetscErrorCode ierr; 3930 3931 PetscFunctionBegin; 3932 ierr = PetscTryMethod(mat,"MatSetHashTableFactor_C",(Mat,PetscReal),(mat,fact));CHKERRQ(ierr); 3933 PetscFunctionReturn(0); 3934 } 3935 3936 #undef __FUNCT__ 3937 #define __FUNCT__ "MatSetHashTableFactor_MPIBAIJ" 3938 PetscErrorCode MatSetHashTableFactor_MPIBAIJ(Mat mat,PetscReal fact) 3939 { 3940 Mat_MPIBAIJ *baij; 3941 3942 PetscFunctionBegin; 3943 baij = (Mat_MPIBAIJ*)mat->data; 3944 baij->ht_fact = fact; 3945 PetscFunctionReturn(0); 3946 } 3947 3948 #undef __FUNCT__ 3949 #define __FUNCT__ "MatMPIBAIJGetSeqBAIJ" 3950 PetscErrorCode MatMPIBAIJGetSeqBAIJ(Mat A,Mat *Ad,Mat *Ao,const PetscInt *colmap[]) 3951 { 3952 Mat_MPIBAIJ *a = (Mat_MPIBAIJ*)A->data; 3953 3954 PetscFunctionBegin; 3955 *Ad = a->A; 3956 *Ao = a->B; 3957 *colmap = a->garray; 3958 PetscFunctionReturn(0); 3959 } 3960 3961 /* 3962 Special version for direct calls from Fortran (to eliminate two function call overheads 3963 */ 3964 #if defined(PETSC_HAVE_FORTRAN_CAPS) 3965 #define matmpibaijsetvaluesblocked_ MATMPIBAIJSETVALUESBLOCKED 3966 #elif !defined(PETSC_HAVE_FORTRAN_UNDERSCORE) 3967 #define matmpibaijsetvaluesblocked_ matmpibaijsetvaluesblocked 3968 #endif 3969 3970 #undef __FUNCT__ 3971 #define __FUNCT__ "matmpibiajsetvaluesblocked" 3972 /*@C 3973 MatMPIBAIJSetValuesBlocked - Direct Fortran call to replace call to MatSetValuesBlocked() 3974 3975 Collective on Mat 3976 3977 Input Parameters: 3978 + mat - the matrix 3979 . min - number of input rows 3980 . im - input rows 3981 . nin - number of input columns 3982 . in - input columns 3983 . v - numerical values input 3984 - addvin - INSERT_VALUES or ADD_VALUES 3985 3986 Notes: This has a complete copy of MatSetValuesBlocked_MPIBAIJ() which is terrible code un-reuse. 3987 3988 Level: advanced 3989 3990 .seealso: MatSetValuesBlocked() 3991 @*/ 3992 PetscErrorCode matmpibaijsetvaluesblocked_(Mat *matin,PetscInt *min,const PetscInt im[],PetscInt *nin,const PetscInt in[],const MatScalar v[],InsertMode *addvin) 3993 { 3994 /* convert input arguments to C version */ 3995 Mat mat = *matin; 3996 PetscInt m = *min, n = *nin; 3997 InsertMode addv = *addvin; 3998 3999 Mat_MPIBAIJ *baij = (Mat_MPIBAIJ*)mat->data; 4000 const MatScalar *value; 4001 MatScalar *barray = baij->barray; 4002 PetscBool roworiented = baij->roworiented; 4003 PetscErrorCode ierr; 4004 PetscInt i,j,ii,jj,row,col,rstart=baij->rstartbs; 4005 PetscInt rend=baij->rendbs,cstart=baij->cstartbs,stepval; 4006 PetscInt cend=baij->cendbs,bs=mat->rmap->bs,bs2=baij->bs2; 4007 4008 PetscFunctionBegin; 4009 /* tasks normally handled by MatSetValuesBlocked() */ 4010 if (mat->insertmode == NOT_SET_VALUES) mat->insertmode = addv; 4011 #if defined(PETSC_USE_DEBUG) 4012 else if (mat->insertmode != addv) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Cannot mix add values and insert values"); 4013 if (mat->factortype) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Not for factored matrix"); 4014 #endif 4015 if (mat->assembled) { 4016 mat->was_assembled = PETSC_TRUE; 4017 mat->assembled = PETSC_FALSE; 4018 } 4019 ierr = PetscLogEventBegin(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 4020 4021 4022 if (!barray) { 4023 ierr = PetscMalloc(bs2*sizeof(MatScalar),&barray);CHKERRQ(ierr); 4024 baij->barray = barray; 4025 } 4026 4027 if (roworiented) stepval = (n-1)*bs; 4028 else stepval = (m-1)*bs; 4029 4030 for (i=0; i<m; i++) { 4031 if (im[i] < 0) continue; 4032 #if defined(PETSC_USE_DEBUG) 4033 if (im[i] >= baij->Mbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Row too large, row %D max %D",im[i],baij->Mbs-1); 4034 #endif 4035 if (im[i] >= rstart && im[i] < rend) { 4036 row = im[i] - rstart; 4037 for (j=0; j<n; j++) { 4038 /* If NumCol = 1 then a copy is not required */ 4039 if ((roworiented) && (n == 1)) { 4040 barray = (MatScalar*)v + i*bs2; 4041 } else if ((!roworiented) && (m == 1)) { 4042 barray = (MatScalar*)v + j*bs2; 4043 } else { /* Here a copy is required */ 4044 if (roworiented) { 4045 value = v + i*(stepval+bs)*bs + j*bs; 4046 } else { 4047 value = v + j*(stepval+bs)*bs + i*bs; 4048 } 4049 for (ii=0; ii<bs; ii++,value+=stepval) { 4050 for (jj=0; jj<bs; jj++) { 4051 *barray++ = *value++; 4052 } 4053 } 4054 barray -=bs2; 4055 } 4056 4057 if (in[j] >= cstart && in[j] < cend) { 4058 col = in[j] - cstart; 4059 ierr = MatSetValuesBlocked_SeqBAIJ(baij->A,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 4060 } else if (in[j] < 0) continue; 4061 #if defined(PETSC_USE_DEBUG) 4062 else if (in[j] >= baij->Nbs) SETERRQ2(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"Column too large, col %D max %D",in[j],baij->Nbs-1); 4063 #endif 4064 else { 4065 if (mat->was_assembled) { 4066 if (!baij->colmap) { 4067 ierr = MatCreateColmap_MPIBAIJ_Private(mat);CHKERRQ(ierr); 4068 } 4069 4070 #if defined(PETSC_USE_DEBUG) 4071 #if defined(PETSC_USE_CTABLE) 4072 { PetscInt data; 4073 ierr = PetscTableFind(baij->colmap,in[j]+1,&data);CHKERRQ(ierr); 4074 if ((data - 1) % bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Incorrect colmap"); 4075 } 4076 #else 4077 if ((baij->colmap[in[j]] - 1) % bs) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_PLIB,"Incorrect colmap"); 4078 #endif 4079 #endif 4080 #if defined(PETSC_USE_CTABLE) 4081 ierr = PetscTableFind(baij->colmap,in[j]+1,&col);CHKERRQ(ierr); 4082 col = (col - 1)/bs; 4083 #else 4084 col = (baij->colmap[in[j]] - 1)/bs; 4085 #endif 4086 if (col < 0 && !((Mat_SeqBAIJ*)(baij->A->data))->nonew) { 4087 ierr = MatDisAssemble_MPIBAIJ(mat);CHKERRQ(ierr); 4088 col = in[j]; 4089 } 4090 } else col = in[j]; 4091 ierr = MatSetValuesBlocked_SeqBAIJ(baij->B,1,&row,1,&col,barray,addv);CHKERRQ(ierr); 4092 } 4093 } 4094 } else { 4095 if (!baij->donotstash) { 4096 if (roworiented) { 4097 ierr = MatStashValuesRowBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 4098 } else { 4099 ierr = MatStashValuesColBlocked_Private(&mat->bstash,im[i],n,in,v,m,n,i);CHKERRQ(ierr); 4100 } 4101 } 4102 } 4103 } 4104 4105 /* task normally handled by MatSetValuesBlocked() */ 4106 ierr = PetscLogEventEnd(MAT_SetValues,mat,0,0,0);CHKERRQ(ierr); 4107 PetscFunctionReturn(0); 4108 } 4109 4110 #undef __FUNCT__ 4111 #define __FUNCT__ "MatCreateMPIBAIJWithArrays" 4112 /*@ 4113 MatCreateMPIBAIJWithArrays - creates a MPI BAIJ matrix using arrays that contain in standard 4114 CSR format the local rows. 4115 4116 Collective on MPI_Comm 4117 4118 Input Parameters: 4119 + comm - MPI communicator 4120 . bs - the block size, only a block size of 1 is supported 4121 . m - number of local rows (Cannot be PETSC_DECIDE) 4122 . n - This value should be the same as the local size used in creating the 4123 x vector for the matrix-vector product y = Ax. (or PETSC_DECIDE to have 4124 calculated if N is given) For square matrices n is almost always m. 4125 . M - number of global rows (or PETSC_DETERMINE to have calculated if m is given) 4126 . N - number of global columns (or PETSC_DETERMINE to have calculated if n is given) 4127 . i - row indices 4128 . j - column indices 4129 - a - matrix values 4130 4131 Output Parameter: 4132 . mat - the matrix 4133 4134 Level: intermediate 4135 4136 Notes: 4137 The i, j, and a arrays ARE copied by this routine into the internal format used by PETSc; 4138 thus you CANNOT change the matrix entries by changing the values of a[] after you have 4139 called this routine. Use MatCreateMPIAIJWithSplitArrays() to avoid needing to copy the arrays. 4140 4141 The i and j indices are 0 based, and i indices are indices corresponding to the local j array. 4142 4143 .keywords: matrix, aij, compressed row, sparse, parallel 4144 4145 .seealso: MatCreate(), MatCreateSeqAIJ(), MatSetValues(), MatMPIAIJSetPreallocation(), MatMPIAIJSetPreallocationCSR(), 4146 MPIAIJ, MatCreateAIJ(), MatCreateMPIAIJWithSplitArrays() 4147 @*/ 4148 PetscErrorCode MatCreateMPIBAIJWithArrays(MPI_Comm comm,PetscInt bs,PetscInt m,PetscInt n,PetscInt M,PetscInt N,const PetscInt i[],const PetscInt j[],const PetscScalar a[],Mat *mat) 4149 { 4150 PetscErrorCode ierr; 4151 4152 PetscFunctionBegin; 4153 if (i[0]) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"i (row indices) must start with 0"); 4154 if (m < 0) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_ARG_OUTOFRANGE,"local number of rows (m) cannot be PETSC_DECIDE, or negative"); 4155 ierr = MatCreate(comm,mat);CHKERRQ(ierr); 4156 ierr = MatSetSizes(*mat,m,n,M,N);CHKERRQ(ierr); 4157 ierr = MatSetType(*mat,MATMPISBAIJ);CHKERRQ(ierr); 4158 ierr = MatMPIBAIJSetPreallocationCSR(*mat,bs,i,j,a);CHKERRQ(ierr); 4159 PetscFunctionReturn(0); 4160 } 4161