1 2 #include <../src/mat/impls/baij/seq/baij.h> 3 #include <../src/mat/impls/sbaij/seq/sbaij.h> 4 #include <petsc/private/kernels/blockinvert.h> 5 #include <petscis.h> 6 7 PetscErrorCode MatGetInertia_SeqSBAIJ(Mat F,PetscInt *nneg,PetscInt *nzero,PetscInt *npos) 8 { 9 Mat_SeqSBAIJ *fact=(Mat_SeqSBAIJ*)F->data; 10 MatScalar *dd=fact->a; 11 PetscInt mbs=fact->mbs,bs=F->rmap->bs,i,nneg_tmp,npos_tmp,*fi=fact->diag; 12 13 PetscFunctionBegin; 14 PetscCheck(bs == 1,PETSC_COMM_SELF,PETSC_ERR_SUP,"No support for bs: %" PetscInt_FMT " >1 yet",bs); 15 16 nneg_tmp = 0; npos_tmp = 0; 17 for (i=0; i<mbs; i++) { 18 if (PetscRealPart(dd[*fi]) > 0.0) npos_tmp++; 19 else if (PetscRealPart(dd[*fi]) < 0.0) nneg_tmp++; 20 fi++; 21 } 22 if (nneg) *nneg = nneg_tmp; 23 if (npos) *npos = npos_tmp; 24 if (nzero) *nzero = mbs - nneg_tmp - npos_tmp; 25 PetscFunctionReturn(0); 26 } 27 28 /* 29 Symbolic U^T*D*U factorization for SBAIJ format. Modified from SSF of YSMP. 30 Use Modified Sparse Row (MSR) storage for u and ju. See page 85, "Iterative Methods ..." by Saad. 31 */ 32 PetscErrorCode MatCholeskyFactorSymbolic_SeqSBAIJ_MSR(Mat F,Mat A,IS perm,const MatFactorInfo *info) 33 { 34 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)A->data,*b; 35 const PetscInt *rip,*ai,*aj; 36 PetscInt i,mbs = a->mbs,*jutmp,bs = A->rmap->bs,bs2=a->bs2; 37 PetscInt m,reallocs = 0,prow; 38 PetscInt *jl,*q,jmin,jmax,juidx,nzk,qm,*iu,*ju,k,j,vj,umax,maxadd; 39 PetscReal f = info->fill; 40 PetscBool perm_identity; 41 42 PetscFunctionBegin; 43 /* check whether perm is the identity mapping */ 44 PetscCall(ISIdentity(perm,&perm_identity)); 45 PetscCall(ISGetIndices(perm,&rip)); 46 47 if (perm_identity) { /* without permutation */ 48 a->permute = PETSC_FALSE; 49 50 ai = a->i; aj = a->j; 51 } else { /* non-trivial permutation */ 52 a->permute = PETSC_TRUE; 53 54 PetscCall(MatReorderingSeqSBAIJ(A,perm)); 55 56 ai = a->inew; aj = a->jnew; 57 } 58 59 /* initialization */ 60 PetscCall(PetscMalloc1(mbs+1,&iu)); 61 umax = (PetscInt)(f*ai[mbs] + 1); umax += mbs + 1; 62 PetscCall(PetscMalloc1(umax,&ju)); 63 iu[0] = mbs+1; 64 juidx = mbs + 1; /* index for ju */ 65 /* jl linked list for pivot row -- linked list for col index */ 66 PetscCall(PetscMalloc2(mbs,&jl,mbs,&q)); 67 for (i=0; i<mbs; i++) { 68 jl[i] = mbs; 69 q[i] = 0; 70 } 71 72 /* for each row k */ 73 for (k=0; k<mbs; k++) { 74 for (i=0; i<mbs; i++) q[i] = 0; /* to be removed! */ 75 nzk = 0; /* num. of nz blocks in k-th block row with diagonal block excluded */ 76 q[k] = mbs; 77 /* initialize nonzero structure of k-th row to row rip[k] of A */ 78 jmin = ai[rip[k]] +1; /* exclude diag[k] */ 79 jmax = ai[rip[k]+1]; 80 for (j=jmin; j<jmax; j++) { 81 vj = rip[aj[j]]; /* col. value */ 82 if (vj > k) { 83 qm = k; 84 do { 85 m = qm; qm = q[m]; 86 } while (qm < vj); 87 PetscCheck(qm != vj,PETSC_COMM_SELF,PETSC_ERR_PLIB,"Duplicate entry in A"); 88 nzk++; 89 q[m] = vj; 90 q[vj] = qm; 91 } /* if (vj > k) */ 92 } /* for (j=jmin; j<jmax; j++) */ 93 94 /* modify nonzero structure of k-th row by computing fill-in 95 for each row i to be merged in */ 96 prow = k; 97 prow = jl[prow]; /* next pivot row (== mbs for symbolic factorization) */ 98 99 while (prow < k) { 100 /* merge row prow into k-th row */ 101 jmin = iu[prow] + 1; jmax = iu[prow+1]; 102 qm = k; 103 for (j=jmin; j<jmax; j++) { 104 vj = ju[j]; 105 do { 106 m = qm; qm = q[m]; 107 } while (qm < vj); 108 if (qm != vj) { 109 nzk++; q[m] = vj; q[vj] = qm; qm = vj; 110 } 111 } 112 prow = jl[prow]; /* next pivot row */ 113 } 114 115 /* add k to row list for first nonzero element in k-th row */ 116 if (nzk > 0) { 117 i = q[k]; /* col value of first nonzero element in U(k, k+1:mbs-1) */ 118 jl[k] = jl[i]; jl[i] = k; 119 } 120 iu[k+1] = iu[k] + nzk; 121 122 /* allocate more space to ju if needed */ 123 if (iu[k+1] > umax) { 124 /* estimate how much additional space we will need */ 125 /* use the strategy suggested by David Hysom <hysom@perch-t.icase.edu> */ 126 /* just double the memory each time */ 127 maxadd = umax; 128 if (maxadd < nzk) maxadd = (mbs-k)*(nzk+1)/2; 129 umax += maxadd; 130 131 /* allocate a longer ju */ 132 PetscCall(PetscMalloc1(umax,&jutmp)); 133 PetscCall(PetscArraycpy(jutmp,ju,iu[k])); 134 PetscCall(PetscFree(ju)); 135 ju = jutmp; 136 reallocs++; /* count how many times we realloc */ 137 } 138 139 /* save nonzero structure of k-th row in ju */ 140 i=k; 141 while (nzk--) { 142 i = q[i]; 143 ju[juidx++] = i; 144 } 145 } 146 147 #if defined(PETSC_USE_INFO) 148 if (ai[mbs] != 0) { 149 PetscReal af = ((PetscReal)iu[mbs])/((PetscReal)ai[mbs]); 150 PetscCall(PetscInfo(A,"Reallocs %" PetscInt_FMT " Fill ratio:given %g needed %g\n",reallocs,(double)f,(double)af)); 151 PetscCall(PetscInfo(A,"Run with -pc_factor_fill %g or use \n",(double)af)); 152 PetscCall(PetscInfo(A,"PCFactorSetFill(pc,%g);\n",(double)af)); 153 PetscCall(PetscInfo(A,"for best performance.\n")); 154 } else { 155 PetscCall(PetscInfo(A,"Empty matrix\n")); 156 } 157 #endif 158 159 PetscCall(ISRestoreIndices(perm,&rip)); 160 PetscCall(PetscFree2(jl,q)); 161 162 /* put together the new matrix */ 163 PetscCall(MatSeqSBAIJSetPreallocation(F,bs,MAT_SKIP_ALLOCATION,NULL)); 164 165 /* PetscCall(PetscLogObjectParent((PetscObject)B,(PetscObject)iperm)); */ 166 b = (Mat_SeqSBAIJ*)(F)->data; 167 b->singlemalloc = PETSC_FALSE; 168 b->free_a = PETSC_TRUE; 169 b->free_ij = PETSC_TRUE; 170 171 PetscCall(PetscMalloc1((iu[mbs]+1)*bs2,&b->a)); 172 b->j = ju; 173 b->i = iu; 174 b->diag = NULL; 175 b->ilen = NULL; 176 b->imax = NULL; 177 b->row = perm; 178 179 b->pivotinblocks = PETSC_FALSE; /* need to get from MatFactorInfo */ 180 181 PetscCall(PetscObjectReference((PetscObject)perm)); 182 183 b->icol = perm; 184 PetscCall(PetscObjectReference((PetscObject)perm)); 185 PetscCall(PetscMalloc1(bs*mbs+bs,&b->solve_work)); 186 /* In b structure: Free imax, ilen, old a, old j. 187 Allocate idnew, solve_work, new a, new j */ 188 PetscCall(PetscLogObjectMemory((PetscObject)F,(iu[mbs]-mbs)*(sizeof(PetscInt)+sizeof(MatScalar)))); 189 b->maxnz = b->nz = iu[mbs]; 190 191 (F)->info.factor_mallocs = reallocs; 192 (F)->info.fill_ratio_given = f; 193 if (ai[mbs] != 0) { 194 (F)->info.fill_ratio_needed = ((PetscReal)iu[mbs])/((PetscReal)ai[mbs]); 195 } else { 196 (F)->info.fill_ratio_needed = 0.0; 197 } 198 PetscCall(MatSeqSBAIJSetNumericFactorization_inplace(F,perm_identity)); 199 PetscFunctionReturn(0); 200 } 201 /* 202 Symbolic U^T*D*U factorization for SBAIJ format. 203 See MatICCFactorSymbolic_SeqAIJ() for description of its data structure. 204 */ 205 #include <petscbt.h> 206 #include <../src/mat/utils/freespace.h> 207 PetscErrorCode MatCholeskyFactorSymbolic_SeqSBAIJ(Mat fact,Mat A,IS perm,const MatFactorInfo *info) 208 { 209 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)A->data; 210 Mat_SeqSBAIJ *b; 211 PetscBool perm_identity,missing; 212 PetscReal fill = info->fill; 213 const PetscInt *rip,*ai=a->i,*aj=a->j; 214 PetscInt i,mbs=a->mbs,bs=A->rmap->bs,reallocs=0,prow; 215 PetscInt *jl,jmin,jmax,nzk,*ui,k,j,*il,nextprow; 216 PetscInt nlnk,*lnk,ncols,*cols,*uj,**ui_ptr,*uj_ptr,*udiag; 217 PetscFreeSpaceList free_space=NULL,current_space=NULL; 218 PetscBT lnkbt; 219 220 PetscFunctionBegin; 221 PetscCheck(A->rmap->n == A->cmap->n,PETSC_COMM_SELF,PETSC_ERR_ARG_WRONG,"Must be square matrix, rows %" PetscInt_FMT " columns %" PetscInt_FMT,A->rmap->n,A->cmap->n); 222 PetscCall(MatMissingDiagonal(A,&missing,&i)); 223 PetscCheck(!missing,PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Matrix is missing diagonal entry %" PetscInt_FMT,i); 224 if (bs > 1) { 225 PetscCall(MatCholeskyFactorSymbolic_SeqSBAIJ_inplace(fact,A,perm,info)); 226 PetscFunctionReturn(0); 227 } 228 229 /* check whether perm is the identity mapping */ 230 PetscCall(ISIdentity(perm,&perm_identity)); 231 PetscCheck(perm_identity,PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix reordering is not supported for sbaij matrix. Use aij format"); 232 a->permute = PETSC_FALSE; 233 PetscCall(ISGetIndices(perm,&rip)); 234 235 /* initialization */ 236 PetscCall(PetscMalloc1(mbs+1,&ui)); 237 PetscCall(PetscMalloc1(mbs+1,&udiag)); 238 ui[0] = 0; 239 240 /* jl: linked list for storing indices of the pivot rows 241 il: il[i] points to the 1st nonzero entry of U(i,k:mbs-1) */ 242 PetscCall(PetscMalloc4(mbs,&ui_ptr,mbs,&il,mbs,&jl,mbs,&cols)); 243 for (i=0; i<mbs; i++) { 244 jl[i] = mbs; il[i] = 0; 245 } 246 247 /* create and initialize a linked list for storing column indices of the active row k */ 248 nlnk = mbs + 1; 249 PetscCall(PetscLLCreate(mbs,mbs,nlnk,lnk,lnkbt)); 250 251 /* initial FreeSpace size is fill*(ai[mbs]+1) */ 252 PetscCall(PetscFreeSpaceGet(PetscRealIntMultTruncate(fill,ai[mbs]+1),&free_space)); 253 current_space = free_space; 254 255 for (k=0; k<mbs; k++) { /* for each active row k */ 256 /* initialize lnk by the column indices of row rip[k] of A */ 257 nzk = 0; 258 ncols = ai[k+1] - ai[k]; 259 PetscCheck(ncols,PETSC_COMM_SELF,PETSC_ERR_MAT_CH_ZRPVT,"Empty row %" PetscInt_FMT " in matrix ",k); 260 for (j=0; j<ncols; j++) { 261 i = *(aj + ai[k] + j); 262 cols[j] = i; 263 } 264 PetscCall(PetscLLAdd(ncols,cols,mbs,&nlnk,lnk,lnkbt)); 265 nzk += nlnk; 266 267 /* update lnk by computing fill-in for each pivot row to be merged in */ 268 prow = jl[k]; /* 1st pivot row */ 269 270 while (prow < k) { 271 nextprow = jl[prow]; 272 /* merge prow into k-th row */ 273 jmin = il[prow] + 1; /* index of the 2nd nzero entry in U(prow,k:mbs-1) */ 274 jmax = ui[prow+1]; 275 ncols = jmax-jmin; 276 uj_ptr = ui_ptr[prow] + jmin - ui[prow]; /* points to the 2nd nzero entry in U(prow,k:mbs-1) */ 277 PetscCall(PetscLLAddSorted(ncols,uj_ptr,mbs,&nlnk,lnk,lnkbt)); 278 nzk += nlnk; 279 280 /* update il and jl for prow */ 281 if (jmin < jmax) { 282 il[prow] = jmin; 283 j = *uj_ptr; jl[prow] = jl[j]; jl[j] = prow; 284 } 285 prow = nextprow; 286 } 287 288 /* if free space is not available, make more free space */ 289 if (current_space->local_remaining<nzk) { 290 i = mbs - k + 1; /* num of unfactored rows */ 291 i = PetscIntMultTruncate(i,PetscMin(nzk, i-1)); /* i*nzk, i*(i-1): estimated and max additional space needed */ 292 PetscCall(PetscFreeSpaceGet(i,¤t_space)); 293 reallocs++; 294 } 295 296 /* copy data into free space, then initialize lnk */ 297 PetscCall(PetscLLClean(mbs,mbs,nzk,lnk,current_space->array,lnkbt)); 298 299 /* add the k-th row into il and jl */ 300 if (nzk > 1) { 301 i = current_space->array[1]; /* col value of the first nonzero element in U(k, k+1:mbs-1) */ 302 jl[k] = jl[i]; jl[i] = k; 303 il[k] = ui[k] + 1; 304 } 305 ui_ptr[k] = current_space->array; 306 307 current_space->array += nzk; 308 current_space->local_used += nzk; 309 current_space->local_remaining -= nzk; 310 311 ui[k+1] = ui[k] + nzk; 312 } 313 314 PetscCall(ISRestoreIndices(perm,&rip)); 315 PetscCall(PetscFree4(ui_ptr,il,jl,cols)); 316 317 /* destroy list of free space and other temporary array(s) */ 318 PetscCall(PetscMalloc1(ui[mbs]+1,&uj)); 319 PetscCall(PetscFreeSpaceContiguous_Cholesky(&free_space,uj,mbs,ui,udiag)); /* store matrix factor */ 320 PetscCall(PetscLLDestroy(lnk,lnkbt)); 321 322 /* put together the new matrix in MATSEQSBAIJ format */ 323 PetscCall(MatSeqSBAIJSetPreallocation(fact,bs,MAT_SKIP_ALLOCATION,NULL)); 324 325 b = (Mat_SeqSBAIJ*)fact->data; 326 b->singlemalloc = PETSC_FALSE; 327 b->free_a = PETSC_TRUE; 328 b->free_ij = PETSC_TRUE; 329 330 PetscCall(PetscMalloc1(ui[mbs]+1,&b->a)); 331 332 b->j = uj; 333 b->i = ui; 334 b->diag = udiag; 335 b->free_diag = PETSC_TRUE; 336 b->ilen = NULL; 337 b->imax = NULL; 338 b->row = perm; 339 b->icol = perm; 340 341 PetscCall(PetscObjectReference((PetscObject)perm)); 342 PetscCall(PetscObjectReference((PetscObject)perm)); 343 344 b->pivotinblocks = PETSC_FALSE; /* need to get from MatFactorInfo */ 345 346 PetscCall(PetscMalloc1(mbs+1,&b->solve_work)); 347 PetscCall(PetscLogObjectMemory((PetscObject)fact,ui[mbs]*(sizeof(PetscInt)+sizeof(MatScalar)))); 348 349 b->maxnz = b->nz = ui[mbs]; 350 351 fact->info.factor_mallocs = reallocs; 352 fact->info.fill_ratio_given = fill; 353 if (ai[mbs] != 0) { 354 fact->info.fill_ratio_needed = ((PetscReal)ui[mbs])/ai[mbs]; 355 } else { 356 fact->info.fill_ratio_needed = 0.0; 357 } 358 #if defined(PETSC_USE_INFO) 359 if (ai[mbs] != 0) { 360 PetscReal af = fact->info.fill_ratio_needed; 361 PetscCall(PetscInfo(A,"Reallocs %" PetscInt_FMT " Fill ratio:given %g needed %g\n",reallocs,(double)fill,(double)af)); 362 PetscCall(PetscInfo(A,"Run with -pc_factor_fill %g or use \n",(double)af)); 363 PetscCall(PetscInfo(A,"PCFactorSetFill(pc,%g) for best performance.\n",(double)af)); 364 } else { 365 PetscCall(PetscInfo(A,"Empty matrix\n")); 366 } 367 #endif 368 fact->ops->choleskyfactornumeric = MatCholeskyFactorNumeric_SeqSBAIJ_1_NaturalOrdering; 369 PetscFunctionReturn(0); 370 } 371 372 PetscErrorCode MatCholeskyFactorSymbolic_SeqSBAIJ_inplace(Mat fact,Mat A,IS perm,const MatFactorInfo *info) 373 { 374 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)A->data; 375 Mat_SeqSBAIJ *b; 376 PetscBool perm_identity,missing; 377 PetscReal fill = info->fill; 378 const PetscInt *rip,*ai,*aj; 379 PetscInt i,mbs=a->mbs,bs=A->rmap->bs,reallocs=0,prow,d; 380 PetscInt *jl,jmin,jmax,nzk,*ui,k,j,*il,nextprow; 381 PetscInt nlnk,*lnk,ncols,*cols,*uj,**ui_ptr,*uj_ptr; 382 PetscFreeSpaceList free_space=NULL,current_space=NULL; 383 PetscBT lnkbt; 384 385 PetscFunctionBegin; 386 PetscCall(MatMissingDiagonal(A,&missing,&d)); 387 PetscCheck(!missing,PETSC_COMM_SELF,PETSC_ERR_ARG_WRONGSTATE,"Matrix is missing diagonal entry %" PetscInt_FMT,d); 388 389 /* 390 This code originally uses Modified Sparse Row (MSR) storage 391 (see page 85, "Iterative Methods ..." by Saad) for the output matrix B - bad choise! 392 Then it is rewritten so the factor B takes seqsbaij format. However the associated 393 MatCholeskyFactorNumeric_() have not been modified for the cases of bs>1 or !perm_identity, 394 thus the original code in MSR format is still used for these cases. 395 The code below should replace MatCholeskyFactorSymbolic_SeqSBAIJ_MSR() whenever 396 MatCholeskyFactorNumeric_() is modified for using sbaij symbolic factor. 397 */ 398 if (bs > 1) { 399 PetscCall(MatCholeskyFactorSymbolic_SeqSBAIJ_MSR(fact,A,perm,info)); 400 PetscFunctionReturn(0); 401 } 402 403 /* check whether perm is the identity mapping */ 404 PetscCall(ISIdentity(perm,&perm_identity)); 405 PetscCheck(perm_identity,PETSC_COMM_SELF,PETSC_ERR_SUP,"Matrix reordering is not supported for sbaij matrix. Use aij format"); 406 a->permute = PETSC_FALSE; 407 ai = a->i; aj = a->j; 408 PetscCall(ISGetIndices(perm,&rip)); 409 410 /* initialization */ 411 PetscCall(PetscMalloc1(mbs+1,&ui)); 412 ui[0] = 0; 413 414 /* jl: linked list for storing indices of the pivot rows 415 il: il[i] points to the 1st nonzero entry of U(i,k:mbs-1) */ 416 PetscCall(PetscMalloc4(mbs,&ui_ptr,mbs,&il,mbs,&jl,mbs,&cols)); 417 for (i=0; i<mbs; i++) { 418 jl[i] = mbs; il[i] = 0; 419 } 420 421 /* create and initialize a linked list for storing column indices of the active row k */ 422 nlnk = mbs + 1; 423 PetscCall(PetscLLCreate(mbs,mbs,nlnk,lnk,lnkbt)); 424 425 /* initial FreeSpace size is fill*(ai[mbs]+1) */ 426 PetscCall(PetscFreeSpaceGet(PetscRealIntMultTruncate(fill,ai[mbs]+1),&free_space)); 427 current_space = free_space; 428 429 for (k=0; k<mbs; k++) { /* for each active row k */ 430 /* initialize lnk by the column indices of row rip[k] of A */ 431 nzk = 0; 432 ncols = ai[rip[k]+1] - ai[rip[k]]; 433 for (j=0; j<ncols; j++) { 434 i = *(aj + ai[rip[k]] + j); 435 cols[j] = rip[i]; 436 } 437 PetscCall(PetscLLAdd(ncols,cols,mbs,&nlnk,lnk,lnkbt)); 438 nzk += nlnk; 439 440 /* update lnk by computing fill-in for each pivot row to be merged in */ 441 prow = jl[k]; /* 1st pivot row */ 442 443 while (prow < k) { 444 nextprow = jl[prow]; 445 /* merge prow into k-th row */ 446 jmin = il[prow] + 1; /* index of the 2nd nzero entry in U(prow,k:mbs-1) */ 447 jmax = ui[prow+1]; 448 ncols = jmax-jmin; 449 uj_ptr = ui_ptr[prow] + jmin - ui[prow]; /* points to the 2nd nzero entry in U(prow,k:mbs-1) */ 450 PetscCall(PetscLLAddSorted(ncols,uj_ptr,mbs,&nlnk,lnk,lnkbt)); 451 nzk += nlnk; 452 453 /* update il and jl for prow */ 454 if (jmin < jmax) { 455 il[prow] = jmin; 456 457 j = *uj_ptr; jl[prow] = jl[j]; jl[j] = prow; 458 } 459 prow = nextprow; 460 } 461 462 /* if free space is not available, make more free space */ 463 if (current_space->local_remaining<nzk) { 464 i = mbs - k + 1; /* num of unfactored rows */ 465 i = PetscMin(PetscIntMultTruncate(i,nzk), PetscIntMultTruncate(i,i-1)); /* i*nzk, i*(i-1): estimated and max additional space needed */ 466 PetscCall(PetscFreeSpaceGet(i,¤t_space)); 467 reallocs++; 468 } 469 470 /* copy data into free space, then initialize lnk */ 471 PetscCall(PetscLLClean(mbs,mbs,nzk,lnk,current_space->array,lnkbt)); 472 473 /* add the k-th row into il and jl */ 474 if (nzk-1 > 0) { 475 i = current_space->array[1]; /* col value of the first nonzero element in U(k, k+1:mbs-1) */ 476 jl[k] = jl[i]; jl[i] = k; 477 il[k] = ui[k] + 1; 478 } 479 ui_ptr[k] = current_space->array; 480 481 current_space->array += nzk; 482 current_space->local_used += nzk; 483 current_space->local_remaining -= nzk; 484 485 ui[k+1] = ui[k] + nzk; 486 } 487 488 PetscCall(ISRestoreIndices(perm,&rip)); 489 PetscCall(PetscFree4(ui_ptr,il,jl,cols)); 490 491 /* destroy list of free space and other temporary array(s) */ 492 PetscCall(PetscMalloc1(ui[mbs]+1,&uj)); 493 PetscCall(PetscFreeSpaceContiguous(&free_space,uj)); 494 PetscCall(PetscLLDestroy(lnk,lnkbt)); 495 496 /* put together the new matrix in MATSEQSBAIJ format */ 497 PetscCall(MatSeqSBAIJSetPreallocation(fact,bs,MAT_SKIP_ALLOCATION,NULL)); 498 499 b = (Mat_SeqSBAIJ*)fact->data; 500 b->singlemalloc = PETSC_FALSE; 501 b->free_a = PETSC_TRUE; 502 b->free_ij = PETSC_TRUE; 503 504 PetscCall(PetscMalloc1(ui[mbs]+1,&b->a)); 505 506 b->j = uj; 507 b->i = ui; 508 b->diag = NULL; 509 b->ilen = NULL; 510 b->imax = NULL; 511 b->row = perm; 512 513 b->pivotinblocks = PETSC_FALSE; /* need to get from MatFactorInfo */ 514 515 PetscCall(PetscObjectReference((PetscObject)perm)); 516 b->icol = perm; 517 PetscCall(PetscObjectReference((PetscObject)perm)); 518 PetscCall(PetscMalloc1(mbs+1,&b->solve_work)); 519 PetscCall(PetscLogObjectMemory((PetscObject)fact,(ui[mbs]-mbs)*(sizeof(PetscInt)+sizeof(MatScalar)))); 520 b->maxnz = b->nz = ui[mbs]; 521 522 fact->info.factor_mallocs = reallocs; 523 fact->info.fill_ratio_given = fill; 524 if (ai[mbs] != 0) { 525 fact->info.fill_ratio_needed = ((PetscReal)ui[mbs])/ai[mbs]; 526 } else { 527 fact->info.fill_ratio_needed = 0.0; 528 } 529 #if defined(PETSC_USE_INFO) 530 if (ai[mbs] != 0) { 531 PetscReal af = fact->info.fill_ratio_needed; 532 PetscCall(PetscInfo(A,"Reallocs %" PetscInt_FMT " Fill ratio:given %g needed %g\n",reallocs,(double)fill,(double)af)); 533 PetscCall(PetscInfo(A,"Run with -pc_factor_fill %g or use \n",(double)af)); 534 PetscCall(PetscInfo(A,"PCFactorSetFill(pc,%g) for best performance.\n",(double)af)); 535 } else { 536 PetscCall(PetscInfo(A,"Empty matrix\n")); 537 } 538 #endif 539 PetscCall(MatSeqSBAIJSetNumericFactorization_inplace(fact,perm_identity)); 540 PetscFunctionReturn(0); 541 } 542 543 PetscErrorCode MatCholeskyFactorNumeric_SeqSBAIJ_N(Mat C,Mat A,const MatFactorInfo *info) 544 { 545 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)A->data,*b = (Mat_SeqSBAIJ*)C->data; 546 IS perm = b->row; 547 const PetscInt *ai,*aj,*perm_ptr,mbs=a->mbs,*bi=b->i,*bj=b->j; 548 PetscInt i,j; 549 PetscInt *a2anew,k,k1,jmin,jmax,*jl,*il,vj,nexti,ili; 550 PetscInt bs =A->rmap->bs,bs2 = a->bs2; 551 MatScalar *ba = b->a,*aa,*ap,*dk,*uik; 552 MatScalar *u,*diag,*rtmp,*rtmp_ptr; 553 MatScalar *work; 554 PetscInt *pivots; 555 PetscBool allowzeropivot,zeropivotdetected; 556 557 PetscFunctionBegin; 558 /* initialization */ 559 PetscCall(PetscCalloc1(bs2*mbs,&rtmp)); 560 PetscCall(PetscMalloc2(mbs,&il,mbs,&jl)); 561 allowzeropivot = PetscNot(A->erroriffailure); 562 563 il[0] = 0; 564 for (i=0; i<mbs; i++) jl[i] = mbs; 565 566 PetscCall(PetscMalloc3(bs2,&dk,bs2,&uik,bs,&work)); 567 PetscCall(PetscMalloc1(bs,&pivots)); 568 569 PetscCall(ISGetIndices(perm,&perm_ptr)); 570 571 /* check permutation */ 572 if (!a->permute) { 573 ai = a->i; aj = a->j; aa = a->a; 574 } else { 575 ai = a->inew; aj = a->jnew; 576 PetscCall(PetscMalloc1(bs2*ai[mbs],&aa)); 577 PetscCall(PetscArraycpy(aa,a->a,bs2*ai[mbs])); 578 PetscCall(PetscMalloc1(ai[mbs],&a2anew)); 579 PetscCall(PetscArraycpy(a2anew,a->a2anew,ai[mbs])); 580 581 for (i=0; i<mbs; i++) { 582 jmin = ai[i]; jmax = ai[i+1]; 583 for (j=jmin; j<jmax; j++) { 584 while (a2anew[j] != j) { 585 k = a2anew[j]; a2anew[j] = a2anew[k]; a2anew[k] = k; 586 for (k1=0; k1<bs2; k1++) { 587 dk[k1] = aa[k*bs2+k1]; 588 aa[k*bs2+k1] = aa[j*bs2+k1]; 589 aa[j*bs2+k1] = dk[k1]; 590 } 591 } 592 /* transform columnoriented blocks that lie in the lower triangle to roworiented blocks */ 593 if (i > aj[j]) { 594 ap = aa + j*bs2; /* ptr to the beginning of j-th block of aa */ 595 for (k=0; k<bs2; k++) dk[k] = ap[k]; /* dk <- j-th block of aa */ 596 for (k=0; k<bs; k++) { /* j-th block of aa <- dk^T */ 597 for (k1=0; k1<bs; k1++) *ap++ = dk[k + bs*k1]; 598 } 599 } 600 } 601 } 602 PetscCall(PetscFree(a2anew)); 603 } 604 605 /* for each row k */ 606 for (k = 0; k<mbs; k++) { 607 608 /*initialize k-th row with elements nonzero in row perm(k) of A */ 609 jmin = ai[perm_ptr[k]]; jmax = ai[perm_ptr[k]+1]; 610 611 ap = aa + jmin*bs2; 612 for (j = jmin; j < jmax; j++) { 613 vj = perm_ptr[aj[j]]; /* block col. index */ 614 rtmp_ptr = rtmp + vj*bs2; 615 for (i=0; i<bs2; i++) *rtmp_ptr++ = *ap++; 616 } 617 618 /* modify k-th row by adding in those rows i with U(i,k) != 0 */ 619 PetscCall(PetscArraycpy(dk,rtmp+k*bs2,bs2)); 620 i = jl[k]; /* first row to be added to k_th row */ 621 622 while (i < k) { 623 nexti = jl[i]; /* next row to be added to k_th row */ 624 625 /* compute multiplier */ 626 ili = il[i]; /* index of first nonzero element in U(i,k:bms-1) */ 627 628 /* uik = -inv(Di)*U_bar(i,k) */ 629 diag = ba + i*bs2; 630 u = ba + ili*bs2; 631 PetscCall(PetscArrayzero(uik,bs2)); 632 PetscKernel_A_gets_A_minus_B_times_C(bs,uik,diag,u); 633 634 /* update D(k) += -U(i,k)^T * U_bar(i,k) */ 635 PetscKernel_A_gets_A_plus_Btranspose_times_C(bs,dk,uik,u); 636 PetscCall(PetscLogFlops(4.0*bs*bs2)); 637 638 /* update -U(i,k) */ 639 PetscCall(PetscArraycpy(ba+ili*bs2,uik,bs2)); 640 641 /* add multiple of row i to k-th row ... */ 642 jmin = ili + 1; jmax = bi[i+1]; 643 if (jmin < jmax) { 644 for (j=jmin; j<jmax; j++) { 645 /* rtmp += -U(i,k)^T * U_bar(i,j) */ 646 rtmp_ptr = rtmp + bj[j]*bs2; 647 u = ba + j*bs2; 648 PetscKernel_A_gets_A_plus_Btranspose_times_C(bs,rtmp_ptr,uik,u); 649 } 650 PetscCall(PetscLogFlops(2.0*bs*bs2*(jmax-jmin))); 651 652 /* ... add i to row list for next nonzero entry */ 653 il[i] = jmin; /* update il(i) in column k+1, ... mbs-1 */ 654 j = bj[jmin]; 655 jl[i] = jl[j]; jl[j] = i; /* update jl */ 656 } 657 i = nexti; 658 } 659 660 /* save nonzero entries in k-th row of U ... */ 661 662 /* invert diagonal block */ 663 diag = ba+k*bs2; 664 PetscCall(PetscArraycpy(diag,dk,bs2)); 665 666 PetscCall(PetscKernel_A_gets_inverse_A(bs,diag,pivots,work,allowzeropivot,&zeropivotdetected)); 667 if (zeropivotdetected) C->factorerrortype = MAT_FACTOR_NUMERIC_ZEROPIVOT; 668 669 jmin = bi[k]; jmax = bi[k+1]; 670 if (jmin < jmax) { 671 for (j=jmin; j<jmax; j++) { 672 vj = bj[j]; /* block col. index of U */ 673 u = ba + j*bs2; 674 rtmp_ptr = rtmp + vj*bs2; 675 for (k1=0; k1<bs2; k1++) { 676 *u++ = *rtmp_ptr; 677 *rtmp_ptr++ = 0.0; 678 } 679 } 680 681 /* ... add k to row list for first nonzero entry in k-th row */ 682 il[k] = jmin; 683 i = bj[jmin]; 684 jl[k] = jl[i]; jl[i] = k; 685 } 686 } 687 688 PetscCall(PetscFree(rtmp)); 689 PetscCall(PetscFree2(il,jl)); 690 PetscCall(PetscFree3(dk,uik,work)); 691 PetscCall(PetscFree(pivots)); 692 if (a->permute) PetscCall(PetscFree(aa)); 693 694 PetscCall(ISRestoreIndices(perm,&perm_ptr)); 695 696 C->ops->solve = MatSolve_SeqSBAIJ_N_inplace; 697 C->ops->solvetranspose = MatSolve_SeqSBAIJ_N_inplace; 698 C->ops->forwardsolve = MatForwardSolve_SeqSBAIJ_N_inplace; 699 C->ops->backwardsolve = MatBackwardSolve_SeqSBAIJ_N_inplace; 700 701 C->assembled = PETSC_TRUE; 702 C->preallocated = PETSC_TRUE; 703 704 PetscCall(PetscLogFlops(1.3333*bs*bs2*b->mbs)); /* from inverting diagonal blocks */ 705 PetscFunctionReturn(0); 706 } 707 708 PetscErrorCode MatCholeskyFactorNumeric_SeqSBAIJ_N_NaturalOrdering(Mat C,Mat A,const MatFactorInfo *info) 709 { 710 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)A->data,*b = (Mat_SeqSBAIJ*)C->data; 711 PetscInt i,j,mbs=a->mbs,*bi=b->i,*bj=b->j; 712 PetscInt *ai,*aj,k,k1,jmin,jmax,*jl,*il,vj,nexti,ili; 713 PetscInt bs =A->rmap->bs,bs2 = a->bs2; 714 MatScalar *ba = b->a,*aa,*ap,*dk,*uik; 715 MatScalar *u,*diag,*rtmp,*rtmp_ptr; 716 MatScalar *work; 717 PetscInt *pivots; 718 PetscBool allowzeropivot,zeropivotdetected; 719 720 PetscFunctionBegin; 721 PetscCall(PetscCalloc1(bs2*mbs,&rtmp)); 722 PetscCall(PetscMalloc2(mbs,&il,mbs,&jl)); 723 il[0] = 0; 724 for (i=0; i<mbs; i++) jl[i] = mbs; 725 726 PetscCall(PetscMalloc3(bs2,&dk,bs2,&uik,bs,&work)); 727 PetscCall(PetscMalloc1(bs,&pivots)); 728 allowzeropivot = PetscNot(A->erroriffailure); 729 730 ai = a->i; aj = a->j; aa = a->a; 731 732 /* for each row k */ 733 for (k = 0; k<mbs; k++) { 734 735 /*initialize k-th row with elements nonzero in row k of A */ 736 jmin = ai[k]; jmax = ai[k+1]; 737 ap = aa + jmin*bs2; 738 for (j = jmin; j < jmax; j++) { 739 vj = aj[j]; /* block col. index */ 740 rtmp_ptr = rtmp + vj*bs2; 741 for (i=0; i<bs2; i++) *rtmp_ptr++ = *ap++; 742 } 743 744 /* modify k-th row by adding in those rows i with U(i,k) != 0 */ 745 PetscCall(PetscArraycpy(dk,rtmp+k*bs2,bs2)); 746 i = jl[k]; /* first row to be added to k_th row */ 747 748 while (i < k) { 749 nexti = jl[i]; /* next row to be added to k_th row */ 750 751 /* compute multiplier */ 752 ili = il[i]; /* index of first nonzero element in U(i,k:bms-1) */ 753 754 /* uik = -inv(Di)*U_bar(i,k) */ 755 diag = ba + i*bs2; 756 u = ba + ili*bs2; 757 PetscCall(PetscArrayzero(uik,bs2)); 758 PetscKernel_A_gets_A_minus_B_times_C(bs,uik,diag,u); 759 760 /* update D(k) += -U(i,k)^T * U_bar(i,k) */ 761 PetscKernel_A_gets_A_plus_Btranspose_times_C(bs,dk,uik,u); 762 PetscCall(PetscLogFlops(2.0*bs*bs2)); 763 764 /* update -U(i,k) */ 765 PetscCall(PetscArraycpy(ba+ili*bs2,uik,bs2)); 766 767 /* add multiple of row i to k-th row ... */ 768 jmin = ili + 1; jmax = bi[i+1]; 769 if (jmin < jmax) { 770 for (j=jmin; j<jmax; j++) { 771 /* rtmp += -U(i,k)^T * U_bar(i,j) */ 772 rtmp_ptr = rtmp + bj[j]*bs2; 773 u = ba + j*bs2; 774 PetscKernel_A_gets_A_plus_Btranspose_times_C(bs,rtmp_ptr,uik,u); 775 } 776 PetscCall(PetscLogFlops(2.0*bs*bs2*(jmax-jmin))); 777 778 /* ... add i to row list for next nonzero entry */ 779 il[i] = jmin; /* update il(i) in column k+1, ... mbs-1 */ 780 j = bj[jmin]; 781 jl[i] = jl[j]; jl[j] = i; /* update jl */ 782 } 783 i = nexti; 784 } 785 786 /* save nonzero entries in k-th row of U ... */ 787 788 /* invert diagonal block */ 789 diag = ba+k*bs2; 790 PetscCall(PetscArraycpy(diag,dk,bs2)); 791 792 PetscCall(PetscKernel_A_gets_inverse_A(bs,diag,pivots,work,allowzeropivot,&zeropivotdetected)); 793 if (zeropivotdetected) C->factorerrortype = MAT_FACTOR_NUMERIC_ZEROPIVOT; 794 795 jmin = bi[k]; jmax = bi[k+1]; 796 if (jmin < jmax) { 797 for (j=jmin; j<jmax; j++) { 798 vj = bj[j]; /* block col. index of U */ 799 u = ba + j*bs2; 800 rtmp_ptr = rtmp + vj*bs2; 801 for (k1=0; k1<bs2; k1++) { 802 *u++ = *rtmp_ptr; 803 *rtmp_ptr++ = 0.0; 804 } 805 } 806 807 /* ... add k to row list for first nonzero entry in k-th row */ 808 il[k] = jmin; 809 i = bj[jmin]; 810 jl[k] = jl[i]; jl[i] = k; 811 } 812 } 813 814 PetscCall(PetscFree(rtmp)); 815 PetscCall(PetscFree2(il,jl)); 816 PetscCall(PetscFree3(dk,uik,work)); 817 PetscCall(PetscFree(pivots)); 818 819 C->ops->solve = MatSolve_SeqSBAIJ_N_NaturalOrdering_inplace; 820 C->ops->solvetranspose = MatSolve_SeqSBAIJ_N_NaturalOrdering_inplace; 821 C->ops->forwardsolve = MatForwardSolve_SeqSBAIJ_N_NaturalOrdering_inplace; 822 C->ops->backwardsolve = MatBackwardSolve_SeqSBAIJ_N_NaturalOrdering_inplace; 823 C->assembled = PETSC_TRUE; 824 C->preallocated = PETSC_TRUE; 825 826 PetscCall(PetscLogFlops(1.3333*bs*bs2*b->mbs)); /* from inverting diagonal blocks */ 827 PetscFunctionReturn(0); 828 } 829 830 /* 831 Numeric U^T*D*U factorization for SBAIJ format. Modified from SNF of YSMP. 832 Version for blocks 2 by 2. 833 */ 834 PetscErrorCode MatCholeskyFactorNumeric_SeqSBAIJ_2(Mat C,Mat A,const MatFactorInfo *info) 835 { 836 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)A->data,*b = (Mat_SeqSBAIJ*)C->data; 837 IS perm = b->row; 838 const PetscInt *ai,*aj,*perm_ptr; 839 PetscInt i,j,mbs=a->mbs,*bi=b->i,*bj=b->j; 840 PetscInt *a2anew,k,k1,jmin,jmax,*jl,*il,vj,nexti,ili; 841 MatScalar *ba = b->a,*aa,*ap; 842 MatScalar *u,*diag,*rtmp,*rtmp_ptr,dk[4],uik[4]; 843 PetscReal shift = info->shiftamount; 844 PetscBool allowzeropivot,zeropivotdetected; 845 846 PetscFunctionBegin; 847 allowzeropivot = PetscNot(A->erroriffailure); 848 849 /* initialization */ 850 /* il and jl record the first nonzero element in each row of the accessing 851 window U(0:k, k:mbs-1). 852 jl: list of rows to be added to uneliminated rows 853 i>= k: jl(i) is the first row to be added to row i 854 i< k: jl(i) is the row following row i in some list of rows 855 jl(i) = mbs indicates the end of a list 856 il(i): points to the first nonzero element in columns k,...,mbs-1 of 857 row i of U */ 858 PetscCall(PetscCalloc1(4*mbs,&rtmp)); 859 PetscCall(PetscMalloc2(mbs,&il,mbs,&jl)); 860 il[0] = 0; 861 for (i=0; i<mbs; i++) jl[i] = mbs; 862 863 PetscCall(ISGetIndices(perm,&perm_ptr)); 864 865 /* check permutation */ 866 if (!a->permute) { 867 ai = a->i; aj = a->j; aa = a->a; 868 } else { 869 ai = a->inew; aj = a->jnew; 870 PetscCall(PetscMalloc1(4*ai[mbs],&aa)); 871 PetscCall(PetscArraycpy(aa,a->a,4*ai[mbs])); 872 PetscCall(PetscMalloc1(ai[mbs],&a2anew)); 873 PetscCall(PetscArraycpy(a2anew,a->a2anew,ai[mbs])); 874 875 for (i=0; i<mbs; i++) { 876 jmin = ai[i]; jmax = ai[i+1]; 877 for (j=jmin; j<jmax; j++) { 878 while (a2anew[j] != j) { 879 k = a2anew[j]; a2anew[j] = a2anew[k]; a2anew[k] = k; 880 for (k1=0; k1<4; k1++) { 881 dk[k1] = aa[k*4+k1]; 882 aa[k*4+k1] = aa[j*4+k1]; 883 aa[j*4+k1] = dk[k1]; 884 } 885 } 886 /* transform columnoriented blocks that lie in the lower triangle to roworiented blocks */ 887 if (i > aj[j]) { 888 ap = aa + j*4; /* ptr to the beginning of the block */ 889 dk[1] = ap[1]; /* swap ap[1] and ap[2] */ 890 ap[1] = ap[2]; 891 ap[2] = dk[1]; 892 } 893 } 894 } 895 PetscCall(PetscFree(a2anew)); 896 } 897 898 /* for each row k */ 899 for (k = 0; k<mbs; k++) { 900 901 /*initialize k-th row with elements nonzero in row perm(k) of A */ 902 jmin = ai[perm_ptr[k]]; jmax = ai[perm_ptr[k]+1]; 903 ap = aa + jmin*4; 904 for (j = jmin; j < jmax; j++) { 905 vj = perm_ptr[aj[j]]; /* block col. index */ 906 rtmp_ptr = rtmp + vj*4; 907 for (i=0; i<4; i++) *rtmp_ptr++ = *ap++; 908 } 909 910 /* modify k-th row by adding in those rows i with U(i,k) != 0 */ 911 PetscCall(PetscArraycpy(dk,rtmp+k*4,4)); 912 i = jl[k]; /* first row to be added to k_th row */ 913 914 while (i < k) { 915 nexti = jl[i]; /* next row to be added to k_th row */ 916 917 /* compute multiplier */ 918 ili = il[i]; /* index of first nonzero element in U(i,k:bms-1) */ 919 920 /* uik = -inv(Di)*U_bar(i,k): - ba[ili]*ba[i] */ 921 diag = ba + i*4; 922 u = ba + ili*4; 923 uik[0] = -(diag[0]*u[0] + diag[2]*u[1]); 924 uik[1] = -(diag[1]*u[0] + diag[3]*u[1]); 925 uik[2] = -(diag[0]*u[2] + diag[2]*u[3]); 926 uik[3] = -(diag[1]*u[2] + diag[3]*u[3]); 927 928 /* update D(k) += -U(i,k)^T * U_bar(i,k): dk += uik*ba[ili] */ 929 dk[0] += uik[0]*u[0] + uik[1]*u[1]; 930 dk[1] += uik[2]*u[0] + uik[3]*u[1]; 931 dk[2] += uik[0]*u[2] + uik[1]*u[3]; 932 dk[3] += uik[2]*u[2] + uik[3]*u[3]; 933 934 PetscCall(PetscLogFlops(16.0*2.0)); 935 936 /* update -U(i,k): ba[ili] = uik */ 937 PetscCall(PetscArraycpy(ba+ili*4,uik,4)); 938 939 /* add multiple of row i to k-th row ... */ 940 jmin = ili + 1; jmax = bi[i+1]; 941 if (jmin < jmax) { 942 for (j=jmin; j<jmax; j++) { 943 /* rtmp += -U(i,k)^T * U_bar(i,j): rtmp[bj[j]] += uik*ba[j]; */ 944 rtmp_ptr = rtmp + bj[j]*4; 945 u = ba + j*4; 946 rtmp_ptr[0] += uik[0]*u[0] + uik[1]*u[1]; 947 rtmp_ptr[1] += uik[2]*u[0] + uik[3]*u[1]; 948 rtmp_ptr[2] += uik[0]*u[2] + uik[1]*u[3]; 949 rtmp_ptr[3] += uik[2]*u[2] + uik[3]*u[3]; 950 } 951 PetscCall(PetscLogFlops(16.0*(jmax-jmin))); 952 953 /* ... add i to row list for next nonzero entry */ 954 il[i] = jmin; /* update il(i) in column k+1, ... mbs-1 */ 955 j = bj[jmin]; 956 jl[i] = jl[j]; jl[j] = i; /* update jl */ 957 } 958 i = nexti; 959 } 960 961 /* save nonzero entries in k-th row of U ... */ 962 963 /* invert diagonal block */ 964 diag = ba+k*4; 965 PetscCall(PetscArraycpy(diag,dk,4)); 966 PetscCall(PetscKernel_A_gets_inverse_A_2(diag,shift,allowzeropivot,&zeropivotdetected)); 967 if (zeropivotdetected) C->factorerrortype = MAT_FACTOR_NUMERIC_ZEROPIVOT; 968 969 jmin = bi[k]; jmax = bi[k+1]; 970 if (jmin < jmax) { 971 for (j=jmin; j<jmax; j++) { 972 vj = bj[j]; /* block col. index of U */ 973 u = ba + j*4; 974 rtmp_ptr = rtmp + vj*4; 975 for (k1=0; k1<4; k1++) { 976 *u++ = *rtmp_ptr; 977 *rtmp_ptr++ = 0.0; 978 } 979 } 980 981 /* ... add k to row list for first nonzero entry in k-th row */ 982 il[k] = jmin; 983 i = bj[jmin]; 984 jl[k] = jl[i]; jl[i] = k; 985 } 986 } 987 988 PetscCall(PetscFree(rtmp)); 989 PetscCall(PetscFree2(il,jl)); 990 if (a->permute) PetscCall(PetscFree(aa)); 991 PetscCall(ISRestoreIndices(perm,&perm_ptr)); 992 993 C->ops->solve = MatSolve_SeqSBAIJ_2_inplace; 994 C->ops->solvetranspose = MatSolve_SeqSBAIJ_2_inplace; 995 C->assembled = PETSC_TRUE; 996 C->preallocated = PETSC_TRUE; 997 998 PetscCall(PetscLogFlops(1.3333*8*b->mbs)); /* from inverting diagonal blocks */ 999 PetscFunctionReturn(0); 1000 } 1001 1002 /* 1003 Version for when blocks are 2 by 2 Using natural ordering 1004 */ 1005 PetscErrorCode MatCholeskyFactorNumeric_SeqSBAIJ_2_NaturalOrdering(Mat C,Mat A,const MatFactorInfo *info) 1006 { 1007 Mat_SeqSBAIJ *a = (Mat_SeqSBAIJ*)A->data,*b = (Mat_SeqSBAIJ*)C->data; 1008 PetscInt i,j,mbs=a->mbs,*bi=b->i,*bj=b->j; 1009 PetscInt *ai,*aj,k,k1,jmin,jmax,*jl,*il,vj,nexti,ili; 1010 MatScalar *ba = b->a,*aa,*ap,dk[8],uik[8]; 1011 MatScalar *u,*diag,*rtmp,*rtmp_ptr; 1012 PetscReal shift = info->shiftamount; 1013 PetscBool allowzeropivot,zeropivotdetected; 1014 1015 PetscFunctionBegin; 1016 allowzeropivot = PetscNot(A->erroriffailure); 1017 1018 /* initialization */ 1019 /* il and jl record the first nonzero element in each row of the accessing 1020 window U(0:k, k:mbs-1). 1021 jl: list of rows to be added to uneliminated rows 1022 i>= k: jl(i) is the first row to be added to row i 1023 i< k: jl(i) is the row following row i in some list of rows 1024 jl(i) = mbs indicates the end of a list 1025 il(i): points to the first nonzero element in columns k,...,mbs-1 of 1026 row i of U */ 1027 PetscCall(PetscCalloc1(4*mbs,&rtmp)); 1028 PetscCall(PetscMalloc2(mbs,&il,mbs,&jl)); 1029 il[0] = 0; 1030 for (i=0; i<mbs; i++) jl[i] = mbs; 1031 1032 ai = a->i; aj = a->j; aa = a->a; 1033 1034 /* for each row k */ 1035 for (k = 0; k<mbs; k++) { 1036 1037 /*initialize k-th row with elements nonzero in row k of A */ 1038 jmin = ai[k]; jmax = ai[k+1]; 1039 ap = aa + jmin*4; 1040 for (j = jmin; j < jmax; j++) { 1041 vj = aj[j]; /* block col. index */ 1042 rtmp_ptr = rtmp + vj*4; 1043 for (i=0; i<4; i++) *rtmp_ptr++ = *ap++; 1044 } 1045 1046 /* modify k-th row by adding in those rows i with U(i,k) != 0 */ 1047 PetscCall(PetscArraycpy(dk,rtmp+k*4,4)); 1048 i = jl[k]; /* first row to be added to k_th row */ 1049 1050 while (i < k) { 1051 nexti = jl[i]; /* next row to be added to k_th row */ 1052 1053 /* compute multiplier */ 1054 ili = il[i]; /* index of first nonzero element in U(i,k:bms-1) */ 1055 1056 /* uik = -inv(Di)*U_bar(i,k): - ba[ili]*ba[i] */ 1057 diag = ba + i*4; 1058 u = ba + ili*4; 1059 uik[0] = -(diag[0]*u[0] + diag[2]*u[1]); 1060 uik[1] = -(diag[1]*u[0] + diag[3]*u[1]); 1061 uik[2] = -(diag[0]*u[2] + diag[2]*u[3]); 1062 uik[3] = -(diag[1]*u[2] + diag[3]*u[3]); 1063 1064 /* update D(k) += -U(i,k)^T * U_bar(i,k): dk += uik*ba[ili] */ 1065 dk[0] += uik[0]*u[0] + uik[1]*u[1]; 1066 dk[1] += uik[2]*u[0] + uik[3]*u[1]; 1067 dk[2] += uik[0]*u[2] + uik[1]*u[3]; 1068 dk[3] += uik[2]*u[2] + uik[3]*u[3]; 1069 1070 PetscCall(PetscLogFlops(16.0*2.0)); 1071 1072 /* update -U(i,k): ba[ili] = uik */ 1073 PetscCall(PetscArraycpy(ba+ili*4,uik,4)); 1074 1075 /* add multiple of row i to k-th row ... */ 1076 jmin = ili + 1; jmax = bi[i+1]; 1077 if (jmin < jmax) { 1078 for (j=jmin; j<jmax; j++) { 1079 /* rtmp += -U(i,k)^T * U_bar(i,j): rtmp[bj[j]] += uik*ba[j]; */ 1080 rtmp_ptr = rtmp + bj[j]*4; 1081 u = ba + j*4; 1082 rtmp_ptr[0] += uik[0]*u[0] + uik[1]*u[1]; 1083 rtmp_ptr[1] += uik[2]*u[0] + uik[3]*u[1]; 1084 rtmp_ptr[2] += uik[0]*u[2] + uik[1]*u[3]; 1085 rtmp_ptr[3] += uik[2]*u[2] + uik[3]*u[3]; 1086 } 1087 PetscCall(PetscLogFlops(16.0*(jmax-jmin))); 1088 1089 /* ... add i to row list for next nonzero entry */ 1090 il[i] = jmin; /* update il(i) in column k+1, ... mbs-1 */ 1091 j = bj[jmin]; 1092 jl[i] = jl[j]; jl[j] = i; /* update jl */ 1093 } 1094 i = nexti; 1095 } 1096 1097 /* save nonzero entries in k-th row of U ... */ 1098 1099 /* invert diagonal block */ 1100 diag = ba+k*4; 1101 PetscCall(PetscArraycpy(diag,dk,4)); 1102 PetscCall(PetscKernel_A_gets_inverse_A_2(diag,shift,allowzeropivot,&zeropivotdetected)); 1103 if (zeropivotdetected) C->factorerrortype = MAT_FACTOR_NUMERIC_ZEROPIVOT; 1104 1105 jmin = bi[k]; jmax = bi[k+1]; 1106 if (jmin < jmax) { 1107 for (j=jmin; j<jmax; j++) { 1108 vj = bj[j]; /* block col. index of U */ 1109 u = ba + j*4; 1110 rtmp_ptr = rtmp + vj*4; 1111 for (k1=0; k1<4; k1++) { 1112 *u++ = *rtmp_ptr; 1113 *rtmp_ptr++ = 0.0; 1114 } 1115 } 1116 1117 /* ... add k to row list for first nonzero entry in k-th row */ 1118 il[k] = jmin; 1119 i = bj[jmin]; 1120 jl[k] = jl[i]; jl[i] = k; 1121 } 1122 } 1123 1124 PetscCall(PetscFree(rtmp)); 1125 PetscCall(PetscFree2(il,jl)); 1126 1127 C->ops->solve = MatSolve_SeqSBAIJ_2_NaturalOrdering_inplace; 1128 C->ops->solvetranspose = MatSolve_SeqSBAIJ_2_NaturalOrdering_inplace; 1129 C->ops->forwardsolve = MatForwardSolve_SeqSBAIJ_2_NaturalOrdering_inplace; 1130 C->ops->backwardsolve = MatBackwardSolve_SeqSBAIJ_2_NaturalOrdering_inplace; 1131 C->assembled = PETSC_TRUE; 1132 C->preallocated = PETSC_TRUE; 1133 1134 PetscCall(PetscLogFlops(1.3333*8*b->mbs)); /* from inverting diagonal blocks */ 1135 PetscFunctionReturn(0); 1136 } 1137 1138 /* 1139 Numeric U^T*D*U factorization for SBAIJ format. 1140 Version for blocks are 1 by 1. 1141 */ 1142 PetscErrorCode MatCholeskyFactorNumeric_SeqSBAIJ_1_inplace(Mat C,Mat A,const MatFactorInfo *info) 1143 { 1144 Mat_SeqSBAIJ *a=(Mat_SeqSBAIJ*)A->data,*b=(Mat_SeqSBAIJ*)C->data; 1145 IS ip=b->row; 1146 const PetscInt *ai,*aj,*rip; 1147 PetscInt *a2anew,i,j,mbs=a->mbs,*bi=b->i,*bj=b->j,*bcol; 1148 PetscInt k,jmin,jmax,*jl,*il,col,nexti,ili,nz; 1149 MatScalar *rtmp,*ba=b->a,*bval,*aa,dk,uikdi; 1150 PetscReal rs; 1151 FactorShiftCtx sctx; 1152 1153 PetscFunctionBegin; 1154 /* MatPivotSetUp(): initialize shift context sctx */ 1155 PetscCall(PetscMemzero(&sctx,sizeof(FactorShiftCtx))); 1156 1157 PetscCall(ISGetIndices(ip,&rip)); 1158 if (!a->permute) { 1159 ai = a->i; aj = a->j; aa = a->a; 1160 } else { 1161 ai = a->inew; aj = a->jnew; 1162 nz = ai[mbs]; 1163 PetscCall(PetscMalloc1(nz,&aa)); 1164 a2anew = a->a2anew; 1165 bval = a->a; 1166 for (j=0; j<nz; j++) { 1167 aa[a2anew[j]] = *(bval++); 1168 } 1169 } 1170 1171 /* initialization */ 1172 /* il and jl record the first nonzero element in each row of the accessing 1173 window U(0:k, k:mbs-1). 1174 jl: list of rows to be added to uneliminated rows 1175 i>= k: jl(i) is the first row to be added to row i 1176 i< k: jl(i) is the row following row i in some list of rows 1177 jl(i) = mbs indicates the end of a list 1178 il(i): points to the first nonzero element in columns k,...,mbs-1 of 1179 row i of U */ 1180 PetscCall(PetscMalloc3(mbs,&rtmp,mbs,&il,mbs,&jl)); 1181 1182 do { 1183 sctx.newshift = PETSC_FALSE; 1184 il[0] = 0; 1185 for (i=0; i<mbs; i++) { 1186 rtmp[i] = 0.0; jl[i] = mbs; 1187 } 1188 1189 for (k = 0; k<mbs; k++) { 1190 /*initialize k-th row by the perm[k]-th row of A */ 1191 jmin = ai[rip[k]]; jmax = ai[rip[k]+1]; 1192 bval = ba + bi[k]; 1193 for (j = jmin; j < jmax; j++) { 1194 col = rip[aj[j]]; 1195 rtmp[col] = aa[j]; 1196 *bval++ = 0.0; /* for in-place factorization */ 1197 } 1198 1199 /* shift the diagonal of the matrix */ 1200 if (sctx.nshift) rtmp[k] += sctx.shift_amount; 1201 1202 /* modify k-th row by adding in those rows i with U(i,k)!=0 */ 1203 dk = rtmp[k]; 1204 i = jl[k]; /* first row to be added to k_th row */ 1205 1206 while (i < k) { 1207 nexti = jl[i]; /* next row to be added to k_th row */ 1208 1209 /* compute multiplier, update diag(k) and U(i,k) */ 1210 ili = il[i]; /* index of first nonzero element in U(i,k:bms-1) */ 1211 uikdi = -ba[ili]*ba[bi[i]]; /* diagonal(k) */ 1212 dk += uikdi*ba[ili]; 1213 ba[ili] = uikdi; /* -U(i,k) */ 1214 1215 /* add multiple of row i to k-th row */ 1216 jmin = ili + 1; jmax = bi[i+1]; 1217 if (jmin < jmax) { 1218 for (j=jmin; j<jmax; j++) rtmp[bj[j]] += uikdi*ba[j]; 1219 PetscCall(PetscLogFlops(2.0*(jmax-jmin))); 1220 1221 /* update il and jl for row i */ 1222 il[i] = jmin; 1223 j = bj[jmin]; jl[i] = jl[j]; jl[j] = i; 1224 } 1225 i = nexti; 1226 } 1227 1228 /* shift the diagonals when zero pivot is detected */ 1229 /* compute rs=sum of abs(off-diagonal) */ 1230 rs = 0.0; 1231 jmin = bi[k]+1; 1232 nz = bi[k+1] - jmin; 1233 if (nz) { 1234 bcol = bj + jmin; 1235 while (nz--) { 1236 rs += PetscAbsScalar(rtmp[*bcol]); 1237 bcol++; 1238 } 1239 } 1240 1241 sctx.rs = rs; 1242 sctx.pv = dk; 1243 PetscCall(MatPivotCheck(C,A,info,&sctx,k)); 1244 if (sctx.newshift) break; /* sctx.shift_amount is updated */ 1245 dk = sctx.pv; 1246 1247 /* copy data into U(k,:) */ 1248 ba[bi[k]] = 1.0/dk; /* U(k,k) */ 1249 jmin = bi[k]+1; jmax = bi[k+1]; 1250 if (jmin < jmax) { 1251 for (j=jmin; j<jmax; j++) { 1252 col = bj[j]; ba[j] = rtmp[col]; rtmp[col] = 0.0; 1253 } 1254 /* add the k-th row into il and jl */ 1255 il[k] = jmin; 1256 i = bj[jmin]; jl[k] = jl[i]; jl[i] = k; 1257 } 1258 } 1259 } while (sctx.newshift); 1260 PetscCall(PetscFree3(rtmp,il,jl)); 1261 if (a->permute) PetscCall(PetscFree(aa)); 1262 1263 PetscCall(ISRestoreIndices(ip,&rip)); 1264 1265 C->ops->solve = MatSolve_SeqSBAIJ_1_inplace; 1266 C->ops->solves = MatSolves_SeqSBAIJ_1_inplace; 1267 C->ops->solvetranspose = MatSolve_SeqSBAIJ_1_inplace; 1268 C->ops->forwardsolve = MatForwardSolve_SeqSBAIJ_1_inplace; 1269 C->ops->backwardsolve = MatBackwardSolve_SeqSBAIJ_1_inplace; 1270 C->assembled = PETSC_TRUE; 1271 C->preallocated = PETSC_TRUE; 1272 1273 PetscCall(PetscLogFlops(C->rmap->N)); 1274 if (sctx.nshift) { 1275 if (info->shifttype == (PetscReal)MAT_SHIFT_NONZERO) { 1276 PetscCall(PetscInfo(A,"number of shiftnz tries %" PetscInt_FMT ", shift_amount %g\n",sctx.nshift,(double)sctx.shift_amount)); 1277 } else if (info->shifttype == (PetscReal)MAT_SHIFT_POSITIVE_DEFINITE) { 1278 PetscCall(PetscInfo(A,"number of shiftpd tries %" PetscInt_FMT ", shift_amount %g\n",sctx.nshift,(double)sctx.shift_amount)); 1279 } 1280 } 1281 PetscFunctionReturn(0); 1282 } 1283 1284 /* 1285 Version for when blocks are 1 by 1 Using natural ordering under new datastructure 1286 Modified from MatCholeskyFactorNumeric_SeqAIJ() 1287 */ 1288 PetscErrorCode MatCholeskyFactorNumeric_SeqSBAIJ_1_NaturalOrdering(Mat B,Mat A,const MatFactorInfo *info) 1289 { 1290 Mat_SeqSBAIJ *a=(Mat_SeqSBAIJ*)A->data; 1291 Mat_SeqSBAIJ *b=(Mat_SeqSBAIJ*)B->data; 1292 PetscInt i,j,mbs=A->rmap->n,*bi=b->i,*bj=b->j,*bdiag=b->diag,*bjtmp; 1293 PetscInt *ai=a->i,*aj=a->j,*ajtmp; 1294 PetscInt k,jmin,jmax,*c2r,*il,col,nexti,ili,nz; 1295 MatScalar *rtmp,*ba=b->a,*bval,*aa=a->a,dk,uikdi; 1296 FactorShiftCtx sctx; 1297 PetscReal rs; 1298 MatScalar d,*v; 1299 1300 PetscFunctionBegin; 1301 PetscCall(PetscMalloc3(mbs,&rtmp,mbs,&il,mbs,&c2r)); 1302 1303 /* MatPivotSetUp(): initialize shift context sctx */ 1304 PetscCall(PetscMemzero(&sctx,sizeof(FactorShiftCtx))); 1305 1306 if (info->shifttype == (PetscReal)MAT_SHIFT_POSITIVE_DEFINITE) { /* set sctx.shift_top=max{rs} */ 1307 sctx.shift_top = info->zeropivot; 1308 1309 PetscCall(PetscArrayzero(rtmp,mbs)); 1310 1311 for (i=0; i<mbs; i++) { 1312 /* calculate sum(|aij|)-RealPart(aii), amt of shift needed for this row */ 1313 d = (aa)[a->diag[i]]; 1314 rtmp[i] += -PetscRealPart(d); /* diagonal entry */ 1315 ajtmp = aj + ai[i] + 1; /* exclude diagonal */ 1316 v = aa + ai[i] + 1; 1317 nz = ai[i+1] - ai[i] - 1; 1318 for (j=0; j<nz; j++) { 1319 rtmp[i] += PetscAbsScalar(v[j]); 1320 rtmp[ajtmp[j]] += PetscAbsScalar(v[j]); 1321 } 1322 if (PetscRealPart(rtmp[i]) > sctx.shift_top) sctx.shift_top = PetscRealPart(rtmp[i]); 1323 } 1324 sctx.shift_top *= 1.1; 1325 sctx.nshift_max = 5; 1326 sctx.shift_lo = 0.; 1327 sctx.shift_hi = 1.; 1328 } 1329 1330 /* allocate working arrays 1331 c2r: linked list, keep track of pivot rows for a given column. c2r[col]: head of the list for a given col 1332 il: for active k row, il[i] gives the index of the 1st nonzero entry in U[i,k:n-1] in bj and ba arrays 1333 */ 1334 do { 1335 sctx.newshift = PETSC_FALSE; 1336 1337 for (i=0; i<mbs; i++) c2r[i] = mbs; 1338 if (mbs) il[0] = 0; 1339 1340 for (k = 0; k<mbs; k++) { 1341 /* zero rtmp */ 1342 nz = bi[k+1] - bi[k]; 1343 bjtmp = bj + bi[k]; 1344 for (j=0; j<nz; j++) rtmp[bjtmp[j]] = 0.0; 1345 1346 /* load in initial unfactored row */ 1347 bval = ba + bi[k]; 1348 jmin = ai[k]; jmax = ai[k+1]; 1349 for (j = jmin; j < jmax; j++) { 1350 col = aj[j]; 1351 rtmp[col] = aa[j]; 1352 *bval++ = 0.0; /* for in-place factorization */ 1353 } 1354 /* shift the diagonal of the matrix: ZeropivotApply() */ 1355 rtmp[k] += sctx.shift_amount; /* shift the diagonal of the matrix */ 1356 1357 /* modify k-th row by adding in those rows i with U(i,k)!=0 */ 1358 dk = rtmp[k]; 1359 i = c2r[k]; /* first row to be added to k_th row */ 1360 1361 while (i < k) { 1362 nexti = c2r[i]; /* next row to be added to k_th row */ 1363 1364 /* compute multiplier, update diag(k) and U(i,k) */ 1365 ili = il[i]; /* index of first nonzero element in U(i,k:bms-1) */ 1366 uikdi = -ba[ili]*ba[bdiag[i]]; /* diagonal(k) */ 1367 dk += uikdi*ba[ili]; /* update diag[k] */ 1368 ba[ili] = uikdi; /* -U(i,k) */ 1369 1370 /* add multiple of row i to k-th row */ 1371 jmin = ili + 1; jmax = bi[i+1]; 1372 if (jmin < jmax) { 1373 for (j=jmin; j<jmax; j++) rtmp[bj[j]] += uikdi*ba[j]; 1374 /* update il and c2r for row i */ 1375 il[i] = jmin; 1376 j = bj[jmin]; c2r[i] = c2r[j]; c2r[j] = i; 1377 } 1378 i = nexti; 1379 } 1380 1381 /* copy data into U(k,:) */ 1382 rs = 0.0; 1383 jmin = bi[k]; jmax = bi[k+1]-1; 1384 if (jmin < jmax) { 1385 for (j=jmin; j<jmax; j++) { 1386 col = bj[j]; ba[j] = rtmp[col]; rs += PetscAbsScalar(ba[j]); 1387 } 1388 /* add the k-th row into il and c2r */ 1389 il[k] = jmin; 1390 i = bj[jmin]; c2r[k] = c2r[i]; c2r[i] = k; 1391 } 1392 1393 sctx.rs = rs; 1394 sctx.pv = dk; 1395 PetscCall(MatPivotCheck(B,A,info,&sctx,k)); 1396 if (sctx.newshift) break; 1397 dk = sctx.pv; 1398 1399 ba[bdiag[k]] = 1.0/dk; /* U(k,k) */ 1400 } 1401 } while (sctx.newshift); 1402 1403 PetscCall(PetscFree3(rtmp,il,c2r)); 1404 1405 B->ops->solve = MatSolve_SeqSBAIJ_1_NaturalOrdering; 1406 B->ops->solves = MatSolves_SeqSBAIJ_1; 1407 B->ops->solvetranspose = MatSolve_SeqSBAIJ_1_NaturalOrdering; 1408 B->ops->matsolve = MatMatSolve_SeqSBAIJ_1_NaturalOrdering; 1409 B->ops->forwardsolve = MatForwardSolve_SeqSBAIJ_1_NaturalOrdering; 1410 B->ops->backwardsolve = MatBackwardSolve_SeqSBAIJ_1_NaturalOrdering; 1411 1412 B->assembled = PETSC_TRUE; 1413 B->preallocated = PETSC_TRUE; 1414 1415 PetscCall(PetscLogFlops(B->rmap->n)); 1416 1417 /* MatPivotView() */ 1418 if (sctx.nshift) { 1419 if (info->shifttype == (PetscReal)MAT_SHIFT_POSITIVE_DEFINITE) { 1420 PetscCall(PetscInfo(A,"number of shift_pd tries %" PetscInt_FMT ", shift_amount %g, diagonal shifted up by %e fraction top_value %e\n",sctx.nshift,(double)sctx.shift_amount,(double)sctx.shift_fraction,(double)sctx.shift_top)); 1421 } else if (info->shifttype == (PetscReal)MAT_SHIFT_NONZERO) { 1422 PetscCall(PetscInfo(A,"number of shift_nz tries %" PetscInt_FMT ", shift_amount %g\n",sctx.nshift,(double)sctx.shift_amount)); 1423 } else if (info->shifttype == (PetscReal)MAT_SHIFT_INBLOCKS) { 1424 PetscCall(PetscInfo(A,"number of shift_inblocks applied %" PetscInt_FMT ", each shift_amount %g\n",sctx.nshift,(double)info->shiftamount)); 1425 } 1426 } 1427 PetscFunctionReturn(0); 1428 } 1429 1430 PetscErrorCode MatCholeskyFactorNumeric_SeqSBAIJ_1_NaturalOrdering_inplace(Mat C,Mat A,const MatFactorInfo *info) 1431 { 1432 Mat_SeqSBAIJ *a=(Mat_SeqSBAIJ*)A->data,*b=(Mat_SeqSBAIJ*)C->data; 1433 PetscInt i,j,mbs = a->mbs; 1434 PetscInt *ai=a->i,*aj=a->j,*bi=b->i,*bj=b->j; 1435 PetscInt k,jmin,*jl,*il,nexti,ili,*acol,*bcol,nz; 1436 MatScalar *rtmp,*ba=b->a,*aa=a->a,dk,uikdi,*aval,*bval; 1437 PetscReal rs; 1438 FactorShiftCtx sctx; 1439 1440 PetscFunctionBegin; 1441 /* MatPivotSetUp(): initialize shift context sctx */ 1442 PetscCall(PetscMemzero(&sctx,sizeof(FactorShiftCtx))); 1443 1444 /* initialization */ 1445 /* il and jl record the first nonzero element in each row of the accessing 1446 window U(0:k, k:mbs-1). 1447 jl: list of rows to be added to uneliminated rows 1448 i>= k: jl(i) is the first row to be added to row i 1449 i< k: jl(i) is the row following row i in some list of rows 1450 jl(i) = mbs indicates the end of a list 1451 il(i): points to the first nonzero element in U(i,k:mbs-1) 1452 */ 1453 PetscCall(PetscMalloc1(mbs,&rtmp)); 1454 PetscCall(PetscMalloc2(mbs,&il,mbs,&jl)); 1455 1456 do { 1457 sctx.newshift = PETSC_FALSE; 1458 il[0] = 0; 1459 for (i=0; i<mbs; i++) { 1460 rtmp[i] = 0.0; jl[i] = mbs; 1461 } 1462 1463 for (k = 0; k<mbs; k++) { 1464 /*initialize k-th row with elements nonzero in row perm(k) of A */ 1465 nz = ai[k+1] - ai[k]; 1466 acol = aj + ai[k]; 1467 aval = aa + ai[k]; 1468 bval = ba + bi[k]; 1469 while (nz--) { 1470 rtmp[*acol++] = *aval++; 1471 *bval++ = 0.0; /* for in-place factorization */ 1472 } 1473 1474 /* shift the diagonal of the matrix */ 1475 if (sctx.nshift) rtmp[k] += sctx.shift_amount; 1476 1477 /* modify k-th row by adding in those rows i with U(i,k)!=0 */ 1478 dk = rtmp[k]; 1479 i = jl[k]; /* first row to be added to k_th row */ 1480 1481 while (i < k) { 1482 nexti = jl[i]; /* next row to be added to k_th row */ 1483 /* compute multiplier, update D(k) and U(i,k) */ 1484 ili = il[i]; /* index of first nonzero element in U(i,k:bms-1) */ 1485 uikdi = -ba[ili]*ba[bi[i]]; 1486 dk += uikdi*ba[ili]; 1487 ba[ili] = uikdi; /* -U(i,k) */ 1488 1489 /* add multiple of row i to k-th row ... */ 1490 jmin = ili + 1; 1491 nz = bi[i+1] - jmin; 1492 if (nz > 0) { 1493 bcol = bj + jmin; 1494 bval = ba + jmin; 1495 PetscCall(PetscLogFlops(2.0*nz)); 1496 while (nz--) rtmp[*bcol++] += uikdi*(*bval++); 1497 1498 /* update il and jl for i-th row */ 1499 il[i] = jmin; 1500 j = bj[jmin]; jl[i] = jl[j]; jl[j] = i; 1501 } 1502 i = nexti; 1503 } 1504 1505 /* shift the diagonals when zero pivot is detected */ 1506 /* compute rs=sum of abs(off-diagonal) */ 1507 rs = 0.0; 1508 jmin = bi[k]+1; 1509 nz = bi[k+1] - jmin; 1510 if (nz) { 1511 bcol = bj + jmin; 1512 while (nz--) { 1513 rs += PetscAbsScalar(rtmp[*bcol]); 1514 bcol++; 1515 } 1516 } 1517 1518 sctx.rs = rs; 1519 sctx.pv = dk; 1520 PetscCall(MatPivotCheck(C,A,info,&sctx,k)); 1521 if (sctx.newshift) break; /* sctx.shift_amount is updated */ 1522 dk = sctx.pv; 1523 1524 /* copy data into U(k,:) */ 1525 ba[bi[k]] = 1.0/dk; 1526 jmin = bi[k]+1; 1527 nz = bi[k+1] - jmin; 1528 if (nz) { 1529 bcol = bj + jmin; 1530 bval = ba + jmin; 1531 while (nz--) { 1532 *bval++ = rtmp[*bcol]; 1533 rtmp[*bcol++] = 0.0; 1534 } 1535 /* add k-th row into il and jl */ 1536 il[k] = jmin; 1537 i = bj[jmin]; jl[k] = jl[i]; jl[i] = k; 1538 } 1539 } /* end of for (k = 0; k<mbs; k++) */ 1540 } while (sctx.newshift); 1541 PetscCall(PetscFree(rtmp)); 1542 PetscCall(PetscFree2(il,jl)); 1543 1544 C->ops->solve = MatSolve_SeqSBAIJ_1_NaturalOrdering_inplace; 1545 C->ops->solves = MatSolves_SeqSBAIJ_1_inplace; 1546 C->ops->solvetranspose = MatSolve_SeqSBAIJ_1_NaturalOrdering_inplace; 1547 C->ops->forwardsolve = MatForwardSolve_SeqSBAIJ_1_NaturalOrdering_inplace; 1548 C->ops->backwardsolve = MatBackwardSolve_SeqSBAIJ_1_NaturalOrdering_inplace; 1549 1550 C->assembled = PETSC_TRUE; 1551 C->preallocated = PETSC_TRUE; 1552 1553 PetscCall(PetscLogFlops(C->rmap->N)); 1554 if (sctx.nshift) { 1555 if (info->shifttype == (PetscReal)MAT_SHIFT_NONZERO) { 1556 PetscCall(PetscInfo(A,"number of shiftnz tries %" PetscInt_FMT ", shift_amount %g\n",sctx.nshift,(double)sctx.shift_amount)); 1557 } else if (info->shifttype == (PetscReal)MAT_SHIFT_POSITIVE_DEFINITE) { 1558 PetscCall(PetscInfo(A,"number of shiftpd tries %" PetscInt_FMT ", shift_amount %g\n",sctx.nshift,(double)sctx.shift_amount)); 1559 } 1560 } 1561 PetscFunctionReturn(0); 1562 } 1563 1564 PetscErrorCode MatCholeskyFactor_SeqSBAIJ(Mat A,IS perm,const MatFactorInfo *info) 1565 { 1566 Mat C; 1567 1568 PetscFunctionBegin; 1569 PetscCall(MatGetFactor(A,"petsc",MAT_FACTOR_CHOLESKY,&C)); 1570 PetscCall(MatCholeskyFactorSymbolic(C,A,perm,info)); 1571 PetscCall(MatCholeskyFactorNumeric(C,A,info)); 1572 1573 A->ops->solve = C->ops->solve; 1574 A->ops->solvetranspose = C->ops->solvetranspose; 1575 1576 PetscCall(MatHeaderMerge(A,&C)); 1577 PetscFunctionReturn(0); 1578 } 1579