1 2 /* 3 Defines matrix-matrix product routines for pairs of MPIAIJ matrices 4 C = A * B 5 */ 6 #include <../src/mat/impls/aij/seq/aij.h> /*I "petscmat.h" I*/ 7 #include <../src/mat/utils/freespace.h> 8 #include <../src/mat/impls/aij/mpi/mpiaij.h> 9 #include <petscbt.h> 10 #include <../src/mat/impls/dense/mpi/mpidense.h> 11 #include <petsc/private/vecimpl.h> 12 #include <petsc/private/vecscatterimpl.h> 13 14 #if defined(PETSC_HAVE_HYPRE) 15 PETSC_INTERN PetscErrorCode MatMatMultSymbolic_AIJ_AIJ_wHYPRE(Mat,Mat,PetscReal,Mat); 16 #endif 17 18 PETSC_INTERN PetscErrorCode MatProductSymbolic_AB_MPIAIJ_MPIAIJ(Mat C) 19 { 20 PetscErrorCode ierr; 21 Mat_Product *product = C->product; 22 Mat A=product->A,B=product->B; 23 MatProductAlgorithm alg=product->alg; 24 PetscReal fill=product->fill; 25 PetscBool flg; 26 27 PetscFunctionBegin; 28 /* scalable */ 29 ierr = PetscStrcmp(alg,"scalable",&flg);CHKERRQ(ierr); 30 if (flg) { 31 ierr = MatMatMultSymbolic_MPIAIJ_MPIAIJ(A,B,fill,C);CHKERRQ(ierr); 32 goto next; 33 } 34 35 /* nonscalable */ 36 ierr = PetscStrcmp(alg,"nonscalable",&flg);CHKERRQ(ierr); 37 if (flg) { 38 ierr = MatMatMultSymbolic_MPIAIJ_MPIAIJ_nonscalable(A,B,fill,C);CHKERRQ(ierr); 39 goto next; 40 } 41 42 /* seqmpi */ 43 ierr = PetscStrcmp(alg,"seqmpi",&flg);CHKERRQ(ierr); 44 if (flg) { 45 ierr = MatMatMultSymbolic_MPIAIJ_MPIAIJ_seqMPI(A,B,fill,C);CHKERRQ(ierr); 46 goto next; 47 } 48 49 #if defined(PETSC_HAVE_HYPRE) 50 ierr = PetscStrcmp(alg,"hypre",&flg);CHKERRQ(ierr); 51 if (flg) { 52 ierr = MatMatMultSymbolic_AIJ_AIJ_wHYPRE(A,B,fill,C);CHKERRQ(ierr); 53 PetscFunctionReturn(0); 54 } 55 #endif 56 SETERRQ(PetscObjectComm((PetscObject)C),PETSC_ERR_SUP,"Mat Product Algorithm is not supported"); 57 58 next: 59 { 60 Mat_MPIAIJ *c = (Mat_MPIAIJ*)C->data; 61 Mat_APMPI *ap = c->ap; 62 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatFreeIntermediateDataStructures","Mat");CHKERRQ(ierr); 63 ap->freestruct = PETSC_FALSE; 64 ierr = PetscOptionsBool("-mat_freeintermediatedatastructures","Free intermediate data structures", "MatFreeIntermediateDataStructures",ap->freestruct,&ap->freestruct, NULL);CHKERRQ(ierr); 65 ierr = PetscOptionsEnd();CHKERRQ(ierr); 66 } 67 PetscFunctionReturn(0); 68 } 69 70 PetscErrorCode MatDestroy_MPIAIJ_MatMatMult(Mat A) 71 { 72 PetscErrorCode ierr; 73 Mat_MPIAIJ *a = (Mat_MPIAIJ*)A->data; 74 Mat_APMPI *ptap = a->ap; 75 76 PetscFunctionBegin; 77 ierr = PetscFree2(ptap->startsj_s,ptap->startsj_r);CHKERRQ(ierr); 78 ierr = PetscFree(ptap->bufa);CHKERRQ(ierr); 79 ierr = MatDestroy(&ptap->P_loc);CHKERRQ(ierr); 80 ierr = MatDestroy(&ptap->P_oth);CHKERRQ(ierr); 81 ierr = MatDestroy(&ptap->Pt);CHKERRQ(ierr); 82 ierr = PetscFree(ptap->api);CHKERRQ(ierr); 83 ierr = PetscFree(ptap->apj);CHKERRQ(ierr); 84 ierr = PetscFree(ptap->apa);CHKERRQ(ierr); 85 ierr = ptap->destroy(A);CHKERRQ(ierr); 86 ierr = PetscFree(ptap);CHKERRQ(ierr); 87 PetscFunctionReturn(0); 88 } 89 90 PetscErrorCode MatMatMultNumeric_MPIAIJ_MPIAIJ_nonscalable(Mat A,Mat P,Mat C) 91 { 92 PetscErrorCode ierr; 93 Mat_MPIAIJ *a =(Mat_MPIAIJ*)A->data,*c=(Mat_MPIAIJ*)C->data; 94 Mat_SeqAIJ *ad =(Mat_SeqAIJ*)(a->A)->data,*ao=(Mat_SeqAIJ*)(a->B)->data; 95 Mat_SeqAIJ *cd =(Mat_SeqAIJ*)(c->A)->data,*co=(Mat_SeqAIJ*)(c->B)->data; 96 PetscScalar *cda=cd->a,*coa=co->a; 97 Mat_SeqAIJ *p_loc,*p_oth; 98 PetscScalar *apa,*ca; 99 PetscInt cm =C->rmap->n; 100 Mat_APMPI *ptap=c->ap; 101 PetscInt *api,*apj,*apJ,i,k; 102 PetscInt cstart=C->cmap->rstart; 103 PetscInt cdnz,conz,k0,k1; 104 MPI_Comm comm; 105 PetscMPIInt size; 106 107 PetscFunctionBegin; 108 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 109 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 110 111 if (!ptap->P_oth && size>1) SETERRQ(comm,PETSC_ERR_ARG_WRONGSTATE,"AP cannot be reused. Do not call MatFreeIntermediateDataStructures() or use '-mat_freeintermediatedatastructures'"); 112 113 /* 1) get P_oth = ptap->P_oth and P_loc = ptap->P_loc */ 114 /*-----------------------------------------------------*/ 115 /* update numerical values of P_oth and P_loc */ 116 ierr = MatGetBrowsOfAoCols_MPIAIJ(A,P,MAT_REUSE_MATRIX,&ptap->startsj_s,&ptap->startsj_r,&ptap->bufa,&ptap->P_oth);CHKERRQ(ierr); 117 ierr = MatMPIAIJGetLocalMat(P,MAT_REUSE_MATRIX,&ptap->P_loc);CHKERRQ(ierr); 118 119 /* 2) compute numeric C_loc = A_loc*P = Ad*P_loc + Ao*P_oth */ 120 /*----------------------------------------------------------*/ 121 /* get data from symbolic products */ 122 p_loc = (Mat_SeqAIJ*)(ptap->P_loc)->data; 123 p_oth = NULL; 124 if (size >1) { 125 p_oth = (Mat_SeqAIJ*)(ptap->P_oth)->data; 126 } 127 128 /* get apa for storing dense row A[i,:]*P */ 129 apa = ptap->apa; 130 131 api = ptap->api; 132 apj = ptap->apj; 133 for (i=0; i<cm; i++) { 134 /* compute apa = A[i,:]*P */ 135 AProw_nonscalable(i,ad,ao,p_loc,p_oth,apa); 136 137 /* set values in C */ 138 apJ = apj + api[i]; 139 cdnz = cd->i[i+1] - cd->i[i]; 140 conz = co->i[i+1] - co->i[i]; 141 142 /* 1st off-diagoanl part of C */ 143 ca = coa + co->i[i]; 144 k = 0; 145 for (k0=0; k0<conz; k0++) { 146 if (apJ[k] >= cstart) break; 147 ca[k0] = apa[apJ[k]]; 148 apa[apJ[k++]] = 0.0; 149 } 150 151 /* diagonal part of C */ 152 ca = cda + cd->i[i]; 153 for (k1=0; k1<cdnz; k1++) { 154 ca[k1] = apa[apJ[k]]; 155 apa[apJ[k++]] = 0.0; 156 } 157 158 /* 2nd off-diagoanl part of C */ 159 ca = coa + co->i[i]; 160 for (; k0<conz; k0++) { 161 ca[k0] = apa[apJ[k]]; 162 apa[apJ[k++]] = 0.0; 163 } 164 } 165 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 166 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 167 168 if (ptap->freestruct) { 169 ierr = MatFreeIntermediateDataStructures(C);CHKERRQ(ierr); 170 } 171 PetscFunctionReturn(0); 172 } 173 174 PetscErrorCode MatMatMultSymbolic_MPIAIJ_MPIAIJ_nonscalable(Mat A,Mat P,PetscReal fill,Mat C) 175 { 176 PetscErrorCode ierr; 177 MPI_Comm comm; 178 PetscMPIInt size; 179 Mat_APMPI *ptap; 180 PetscFreeSpaceList free_space=NULL,current_space=NULL; 181 Mat_MPIAIJ *a =(Mat_MPIAIJ*)A->data,*c; 182 Mat_SeqAIJ *ad =(Mat_SeqAIJ*)(a->A)->data,*ao=(Mat_SeqAIJ*)(a->B)->data,*p_loc,*p_oth; 183 PetscInt *pi_loc,*pj_loc,*pi_oth,*pj_oth,*dnz,*onz; 184 PetscInt *adi=ad->i,*adj=ad->j,*aoi=ao->i,*aoj=ao->j,rstart=A->rmap->rstart; 185 PetscInt *lnk,i,pnz,row,*api,*apj,*Jptr,apnz,nspacedouble=0,j,nzi; 186 PetscInt am=A->rmap->n,pN=P->cmap->N,pn=P->cmap->n,pm=P->rmap->n; 187 PetscBT lnkbt; 188 PetscReal afill; 189 MatType mtype; 190 191 PetscFunctionBegin; 192 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 193 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 194 195 /* create struct Mat_APMPI and attached it to C later */ 196 ierr = PetscNew(&ptap);CHKERRQ(ierr); 197 198 /* get P_oth by taking rows of P (= non-zero cols of local A) from other processors */ 199 ierr = MatGetBrowsOfAoCols_MPIAIJ(A,P,MAT_INITIAL_MATRIX,&ptap->startsj_s,&ptap->startsj_r,&ptap->bufa,&ptap->P_oth);CHKERRQ(ierr); 200 201 /* get P_loc by taking all local rows of P */ 202 ierr = MatMPIAIJGetLocalMat(P,MAT_INITIAL_MATRIX,&ptap->P_loc);CHKERRQ(ierr); 203 204 p_loc = (Mat_SeqAIJ*)(ptap->P_loc)->data; 205 pi_loc = p_loc->i; pj_loc = p_loc->j; 206 if (size > 1) { 207 p_oth = (Mat_SeqAIJ*)(ptap->P_oth)->data; 208 pi_oth = p_oth->i; pj_oth = p_oth->j; 209 } else { 210 p_oth = NULL; 211 pi_oth = NULL; pj_oth = NULL; 212 } 213 214 /* first, compute symbolic AP = A_loc*P = A_diag*P_loc + A_off*P_oth */ 215 /*-------------------------------------------------------------------*/ 216 ierr = PetscMalloc1(am+2,&api);CHKERRQ(ierr); 217 ptap->api = api; 218 api[0] = 0; 219 220 /* create and initialize a linked list */ 221 ierr = PetscLLCondensedCreate(pN,pN,&lnk,&lnkbt);CHKERRQ(ierr); 222 223 /* Initial FreeSpace size is fill*(nnz(A)+nnz(P)) */ 224 ierr = PetscFreeSpaceGet(PetscRealIntMultTruncate(fill,PetscIntSumTruncate(adi[am],PetscIntSumTruncate(aoi[am],pi_loc[pm]))),&free_space);CHKERRQ(ierr); 225 current_space = free_space; 226 227 ierr = MatPreallocateInitialize(comm,am,pn,dnz,onz);CHKERRQ(ierr); 228 for (i=0; i<am; i++) { 229 /* diagonal portion of A */ 230 nzi = adi[i+1] - adi[i]; 231 for (j=0; j<nzi; j++) { 232 row = *adj++; 233 pnz = pi_loc[row+1] - pi_loc[row]; 234 Jptr = pj_loc + pi_loc[row]; 235 /* add non-zero cols of P into the sorted linked list lnk */ 236 ierr = PetscLLCondensedAddSorted(pnz,Jptr,lnk,lnkbt);CHKERRQ(ierr); 237 } 238 /* off-diagonal portion of A */ 239 nzi = aoi[i+1] - aoi[i]; 240 for (j=0; j<nzi; j++) { 241 row = *aoj++; 242 pnz = pi_oth[row+1] - pi_oth[row]; 243 Jptr = pj_oth + pi_oth[row]; 244 ierr = PetscLLCondensedAddSorted(pnz,Jptr,lnk,lnkbt);CHKERRQ(ierr); 245 } 246 247 apnz = lnk[0]; 248 api[i+1] = api[i] + apnz; 249 250 /* if free space is not available, double the total space in the list */ 251 if (current_space->local_remaining<apnz) { 252 ierr = PetscFreeSpaceGet(PetscIntSumTruncate(apnz,current_space->total_array_size),¤t_space);CHKERRQ(ierr); 253 nspacedouble++; 254 } 255 256 /* Copy data into free space, then initialize lnk */ 257 ierr = PetscLLCondensedClean(pN,apnz,current_space->array,lnk,lnkbt);CHKERRQ(ierr); 258 ierr = MatPreallocateSet(i+rstart,apnz,current_space->array,dnz,onz);CHKERRQ(ierr); 259 260 current_space->array += apnz; 261 current_space->local_used += apnz; 262 current_space->local_remaining -= apnz; 263 } 264 265 /* Allocate space for apj, initialize apj, and */ 266 /* destroy list of free space and other temporary array(s) */ 267 ierr = PetscMalloc1(api[am]+1,&ptap->apj);CHKERRQ(ierr); 268 apj = ptap->apj; 269 ierr = PetscFreeSpaceContiguous(&free_space,ptap->apj);CHKERRQ(ierr); 270 ierr = PetscLLDestroy(lnk,lnkbt);CHKERRQ(ierr); 271 272 /* malloc apa to store dense row A[i,:]*P */ 273 ierr = PetscCalloc1(pN,&ptap->apa);CHKERRQ(ierr); 274 275 /* set and assemble symbolic parallel matrix C */ 276 /*---------------------------------------------*/ 277 ierr = MatSetSizes(C,am,pn,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 278 ierr = MatSetBlockSizesFromMats(C,A,P);CHKERRQ(ierr); 279 280 ierr = MatGetType(A,&mtype);CHKERRQ(ierr); 281 ierr = MatSetType(C,mtype);CHKERRQ(ierr); 282 ierr = MatMPIAIJSetPreallocation(C,0,dnz,0,onz);CHKERRQ(ierr); 283 284 ierr = MatSetValues_MPIAIJ_CopyFromCSRFormat_Symbolic(C, apj, api);CHKERRQ(ierr); 285 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 286 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 287 ierr = MatPreallocateFinalize(dnz,onz);CHKERRQ(ierr); 288 289 ptap->destroy = C->ops->destroy; 290 ptap->duplicate = C->ops->duplicate; 291 C->ops->matmultnumeric = MatMatMultNumeric_MPIAIJ_MPIAIJ_nonscalable; 292 C->ops->productnumeric = MatProductNumeric_AB; 293 C->ops->destroy = MatDestroy_MPIAIJ_MatMatMult; 294 C->ops->freeintermediatedatastructures = MatFreeIntermediateDataStructures_MPIAIJ_AP; 295 296 /* attach the supporting struct to C for reuse */ 297 c = (Mat_MPIAIJ*)C->data; 298 c->ap = ptap; 299 300 /* set MatInfo */ 301 afill = (PetscReal)api[am]/(adi[am]+aoi[am]+pi_loc[pm]+1) + 1.e-5; 302 if (afill < 1.0) afill = 1.0; 303 C->info.mallocs = nspacedouble; 304 C->info.fill_ratio_given = fill; 305 C->info.fill_ratio_needed = afill; 306 307 #if defined(PETSC_USE_INFO) 308 if (api[am]) { 309 ierr = PetscInfo3(C,"Reallocs %D; Fill ratio: given %g needed %g.\n",nspacedouble,(double)fill,(double)afill);CHKERRQ(ierr); 310 ierr = PetscInfo1(C,"Use MatMatMult(A,B,MatReuse,%g,&C) for best performance.;\n",(double)afill);CHKERRQ(ierr); 311 } else { 312 ierr = PetscInfo(C,"Empty matrix product\n");CHKERRQ(ierr); 313 } 314 #endif 315 PetscFunctionReturn(0); 316 } 317 318 /* ------------------------------------------------------- */ 319 static PetscErrorCode MatProductSetFromOptions_MPIAIJ_MPIDense_AB(Mat C) 320 { 321 Mat_Product *product = C->product; 322 Mat A = product->A,B=product->B; 323 324 PetscFunctionBegin; 325 if (A->cmap->rstart != B->rmap->rstart || A->cmap->rend != B->rmap->rend) 326 SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, (%D, %D) != (%D,%D)",A->cmap->rstart,A->cmap->rend,B->rmap->rstart,B->rmap->rend); 327 328 C->ops->matmultsymbolic = MatMatMultSymbolic_MPIAIJ_MPIDense; 329 C->ops->productsymbolic = MatProductSymbolic_AB; 330 C->ops->productnumeric = MatProductNumeric_AB; 331 PetscFunctionReturn(0); 332 } 333 /* -------------------------------------------------------------------- */ 334 static PetscErrorCode MatProductSetFromOptions_MPIAIJ_MPIDense_AtB(Mat C) 335 { 336 Mat_Product *product = C->product; 337 Mat A = product->A,B=product->B; 338 339 PetscFunctionBegin; 340 if (A->rmap->rstart != B->rmap->rstart || A->rmap->rend != B->rmap->rend) 341 SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, (%D, %D) != (%D,%D)",A->rmap->rstart,A->rmap->rend,B->rmap->rstart,B->rmap->rend); 342 343 C->ops->transposematmultsymbolic = MatTransposeMatMultSymbolic_MPIAIJ_MPIDense; 344 C->ops->productsymbolic = MatProductSymbolic_AtB; 345 PetscFunctionReturn(0); 346 } 347 348 /* --------------------------------------------------------------------- */ 349 PETSC_INTERN PetscErrorCode MatProductSetFromOptions_MPIAIJ_MPIDense(Mat C) 350 { 351 PetscErrorCode ierr; 352 Mat_Product *product = C->product; 353 354 PetscFunctionBegin; 355 switch (product->type) { 356 case MATPRODUCT_AB: 357 ierr = MatProductSetFromOptions_MPIAIJ_MPIDense_AB(C);CHKERRQ(ierr); 358 break; 359 case MATPRODUCT_AtB: 360 ierr = MatProductSetFromOptions_MPIAIJ_MPIDense_AtB(C);CHKERRQ(ierr); 361 break; 362 default: 363 /* Use MatProduct_Basic() if there is no specific implementation */ 364 C->ops->productsymbolic = MatProductSymbolic_Basic; 365 } 366 PetscFunctionReturn(0); 367 } 368 /* ------------------------------------------------------- */ 369 370 typedef struct { 371 Mat workB,Bb,Cb,workB1,Bb1,Cb1; 372 MPI_Request *rwaits,*swaits; 373 PetscInt numBb; /* num of Bb matrices */ 374 PetscInt nsends,nrecvs; 375 MPI_Datatype *stype,*rtype; 376 } MPIAIJ_MPIDense; 377 378 PetscErrorCode MatMPIAIJ_MPIDenseDestroy(void *ctx) 379 { 380 MPIAIJ_MPIDense *contents = (MPIAIJ_MPIDense*)ctx; 381 PetscErrorCode ierr; 382 PetscInt i; 383 384 PetscFunctionBegin; 385 ierr = MatDestroy(&contents->workB);CHKERRQ(ierr); 386 387 if (contents->numBb) { 388 ierr = MatDestroy(&contents->Bb);CHKERRQ(ierr); 389 ierr = MatDestroy(&contents->Cb);CHKERRQ(ierr); 390 391 ierr = MatDestroy(&contents->workB1);CHKERRQ(ierr); 392 ierr = MatDestroy(&contents->Bb1);CHKERRQ(ierr); 393 ierr = MatDestroy(&contents->Cb1);CHKERRQ(ierr); 394 } 395 for (i=0; i<contents->nsends; i++) { 396 ierr = MPI_Type_free(&contents->stype[i]);CHKERRQ(ierr); 397 } 398 for (i=0; i<contents->nrecvs; i++) { 399 ierr = MPI_Type_free(&contents->rtype[i]);CHKERRQ(ierr); 400 } 401 ierr = PetscFree4(contents->stype,contents->rtype,contents->rwaits,contents->swaits);CHKERRQ(ierr); 402 ierr = PetscFree(contents);CHKERRQ(ierr); 403 PetscFunctionReturn(0); 404 } 405 406 /* 407 This is a "dummy function" that handles the case where matrix C was created as a dense matrix 408 directly by the user and passed to MatMatMult() with the MAT_REUSE_MATRIX option 409 410 It is the same as MatMatMultSymbolic_MPIAIJ_MPIDense() except does not create C 411 */ 412 PETSC_INTERN PetscErrorCode MatMatMultNumeric_MPIDense(Mat A,Mat B,Mat C) 413 { 414 PetscBool flg; 415 PetscErrorCode ierr; 416 417 PetscFunctionBegin; 418 ierr = PetscObjectTypeCompare((PetscObject)A,MATNEST,&flg);CHKERRQ(ierr); 419 if (flg) { 420 ierr = PetscLogEventBegin(MAT_MatMultSymbolic,A,B,0,0);CHKERRQ(ierr); 421 ierr = MatMatMultSymbolic_Nest_Dense(A,B,PETSC_DEFAULT,&C);CHKERRQ(ierr); 422 ierr = PetscLogEventEnd(MAT_MatMultSymbolic,A,B,0,0);CHKERRQ(ierr); 423 C->ops->matmultnumeric = MatMatMultNumeric_Nest_Dense; 424 } else { 425 ierr = MatMatMultSymbolic_MPIAIJ_MPIDense(A,B,PETSC_DEFAULT,C);CHKERRQ(ierr); 426 } 427 ierr = (*C->ops->matmultnumeric)(A,B,C);CHKERRQ(ierr); 428 PetscFunctionReturn(0); 429 } 430 431 /* 432 Create Bb, Cb, Bb1 and Cb1 matrices to be used by MatMatMultSymbolic_MPIAIJ_MPIDense(). 433 These matrices are used as wrappers for sub-columns of B and C, thus their own matrix operations are not used. 434 Modified from MatCreateDense(). 435 */ 436 PETSC_STATIC_INLINE PetscErrorCode MatCreateSubMPIDense_private(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt M,PetscInt N,PetscInt rbs,PetscInt cbs,PetscScalar *data,Mat *A) 437 { 438 PetscErrorCode ierr; 439 440 PetscFunctionBegin; 441 ierr = MatCreate(comm,A);CHKERRQ(ierr); 442 ierr = MatSetSizes(*A,m,n,M,N);CHKERRQ(ierr); 443 ierr = MatSetBlockSizes(*A,rbs,cbs);CHKERRQ(ierr); 444 ierr = MatSetType(*A,MATMPIDENSE);CHKERRQ(ierr); 445 ierr = MatMPIDenseSetPreallocation(*A,data);CHKERRQ(ierr); 446 (*A)->assembled = PETSC_TRUE; 447 PetscFunctionReturn(0); 448 } 449 450 PetscErrorCode MatMatMultSymbolic_MPIAIJ_MPIDense(Mat A,Mat B,PetscReal fill,Mat C) 451 { 452 PetscErrorCode ierr; 453 Mat_MPIAIJ *aij=(Mat_MPIAIJ*)A->data; 454 Mat_MPIDense *b=(Mat_MPIDense*)B->data; 455 Mat_SeqDense *bseq=(Mat_SeqDense*)(b->A)->data; 456 PetscInt nz=aij->B->cmap->n,nsends,nrecvs,i,nrows_to,j,lda=bseq->lda; 457 PetscContainer container; 458 MPIAIJ_MPIDense *contents; 459 VecScatter ctx=aij->Mvctx; 460 PetscInt Am=A->rmap->n,Bm=B->rmap->n,BN=B->cmap->N,Bbn,Bbn1,bs,nrows_from; 461 MPI_Comm comm; 462 MPI_Datatype type1,*stype,*rtype; 463 const PetscInt *sindices,*sstarts,*rstarts; 464 PetscMPIInt *disp; 465 466 PetscFunctionBegin; 467 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 468 if (!C->preallocated) { 469 ierr = MatSetSizes(C,Am,B->cmap->n,A->rmap->N,BN);CHKERRQ(ierr); 470 ierr = MatSetBlockSizesFromMats(C,A,B);CHKERRQ(ierr); 471 ierr = MatSetType(C,MATMPIDENSE);CHKERRQ(ierr); 472 ierr = MatMPIDenseSetPreallocation(C,NULL);CHKERRQ(ierr); 473 } 474 475 ierr = PetscNew(&contents);CHKERRQ(ierr); 476 contents->numBb = 0; 477 478 ierr = VecScatterGetRemote_Private(ctx,PETSC_TRUE/*send*/,&nsends,&sstarts,&sindices,NULL,NULL);CHKERRQ(ierr); 479 ierr = VecScatterGetRemoteOrdered_Private(ctx,PETSC_FALSE/*recv*/,&nrecvs,&rstarts,NULL,NULL,NULL);CHKERRQ(ierr); 480 481 /* Create column block of B and C for memory scalability when BN is too large */ 482 /* Estimate Bbn, column size of Bb */ 483 if (nz) { 484 Bbn1 = 2*Am*BN/nz; 485 } else Bbn1 = BN; 486 487 bs = PetscAbs(B->cmap->bs); 488 Bbn1 = Bbn1/bs *bs; /* Bbn1 is a multiple of bs */ 489 if (Bbn1 > BN) Bbn1 = BN; 490 ierr = MPI_Allreduce(&Bbn1,&Bbn,1,MPIU_INT,MPI_MAX,comm);CHKERRQ(ierr); 491 492 /* Enable runtime option for Bbn */ 493 ierr = PetscOptionsBegin(comm,((PetscObject)C)->prefix,"MatMatMult","Mat");CHKERRQ(ierr); 494 ierr = PetscOptionsInt("-matmatmult_Bbn","Number of columns in Bb","MatMatMult",Bbn,&Bbn,NULL);CHKERRQ(ierr); 495 if (Bbn > BN) SETERRQ2(comm,PETSC_ERR_ARG_SIZ,"Bbn=%D cannot be larger than %D, column size of B",Bbn,BN); 496 ierr = PetscOptionsEnd();CHKERRQ(ierr); 497 498 if (Bbn < BN) { 499 contents->numBb = BN/Bbn; 500 Bbn1 = BN - contents->numBb*Bbn; 501 } 502 503 if (contents->numBb) { 504 PetscScalar data[1]; /* fake array for Bb and Cb */ 505 ierr = PetscInfo3(C,"use Bb, BN=%D, Bbn=%D; numBb=%D\n",BN,Bbn,contents->numBb);CHKERRQ(ierr); 506 ierr = MatCreateSubMPIDense_private(comm,B->rmap->n,PETSC_DECIDE,A->rmap->N,Bbn,B->rmap->bs,B->cmap->bs,data,&contents->Bb);CHKERRQ(ierr); 507 ierr = MatCreateSubMPIDense_private(comm,Am,PETSC_DECIDE,A->rmap->N,Bbn,C->rmap->bs,C->cmap->bs,data,&contents->Cb);CHKERRQ(ierr); 508 509 if (Bbn1) { /* Create Bb1 and Cb1 for the remaining columns */ 510 ierr = PetscInfo2(C,"use Bb1, BN=%D, Bbn1=%D\n",BN,Bbn1);CHKERRQ(ierr); 511 ierr = MatCreateSubMPIDense_private(comm,B->rmap->n,PETSC_DECIDE,A->rmap->N,Bbn1,B->rmap->bs,B->cmap->bs,data,&contents->Bb1);CHKERRQ(ierr); 512 ierr = MatCreateSubMPIDense_private(comm,Am,PETSC_DECIDE,A->rmap->N,Bbn1,C->rmap->bs,C->cmap->bs,data,&contents->Cb1);CHKERRQ(ierr); 513 514 /* Create work matrix used to store off processor rows of B needed for local product */ 515 ierr = MatCreateSeqDense(PETSC_COMM_SELF,nz,Bbn1,NULL,&contents->workB1);CHKERRQ(ierr); 516 } 517 } 518 519 /* Create work matrix used to store off processor rows of B needed for local product */ 520 ierr = MatCreateSeqDense(PETSC_COMM_SELF,nz,Bbn,NULL,&contents->workB);CHKERRQ(ierr); 521 522 /* Use MPI derived data type to reduce memory required by the send/recv buffers */ 523 ierr = PetscMalloc4(nsends,&stype,nrecvs,&rtype,nrecvs,&contents->rwaits,nsends,&contents->swaits);CHKERRQ(ierr); 524 contents->stype = stype; 525 contents->nsends = nsends; 526 527 contents->rtype = rtype; 528 contents->nrecvs = nrecvs; 529 530 ierr = PetscMalloc1(Bm+1,&disp);CHKERRQ(ierr); 531 for (i=0; i<nsends; i++) { 532 nrows_to = sstarts[i+1]-sstarts[i]; 533 for (j=0; j<nrows_to; j++){ 534 disp[j] = sindices[sstarts[i]+j]; /* rowB to be sent */ 535 } 536 ierr = MPI_Type_create_indexed_block(nrows_to,1,(const PetscMPIInt *)disp,MPIU_SCALAR,&type1);CHKERRQ(ierr); 537 538 ierr = MPI_Type_create_resized(type1,0,lda*sizeof(PetscScalar),&stype[i]);CHKERRQ(ierr); 539 ierr = MPI_Type_commit(&stype[i]);CHKERRQ(ierr); 540 ierr = MPI_Type_free(&type1);CHKERRQ(ierr); 541 } 542 543 for (i=0; i<nrecvs; i++) { 544 /* received values from a process form a (nrows_from x Bbn) row block in workB (column-wise) */ 545 nrows_from = rstarts[i+1]-rstarts[i]; 546 disp[0] = 0; 547 ierr = MPI_Type_create_indexed_block(1, nrows_from, (const PetscMPIInt *)disp, MPIU_SCALAR, &type1);CHKERRQ(ierr); 548 ierr = MPI_Type_create_resized(type1, 0, nz*sizeof(PetscScalar), &rtype[i]);CHKERRQ(ierr); 549 ierr = MPI_Type_commit(&rtype[i]);CHKERRQ(ierr); 550 ierr = MPI_Type_free(&type1);CHKERRQ(ierr); 551 } 552 553 ierr = PetscFree(disp);CHKERRQ(ierr); 554 ierr = VecScatterRestoreRemote_Private(ctx,PETSC_TRUE/*send*/,&nsends,&sstarts,&sindices,NULL,NULL);CHKERRQ(ierr); 555 ierr = VecScatterRestoreRemoteOrdered_Private(ctx,PETSC_FALSE/*recv*/,&nrecvs,&rstarts,NULL,NULL,NULL);CHKERRQ(ierr); 556 557 ierr = PetscContainerCreate(comm,&container);CHKERRQ(ierr); 558 ierr = PetscContainerSetPointer(container,contents);CHKERRQ(ierr); 559 ierr = PetscContainerSetUserDestroy(container,MatMPIAIJ_MPIDenseDestroy);CHKERRQ(ierr); 560 ierr = PetscObjectCompose((PetscObject)C,"workB",(PetscObject)container);CHKERRQ(ierr); 561 ierr = PetscContainerDestroy(&container);CHKERRQ(ierr); 562 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 563 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 564 C->ops->matmultnumeric = MatMatMultNumeric_MPIAIJ_MPIDense; 565 PetscFunctionReturn(0); 566 } 567 568 extern PetscErrorCode MatMatMultNumericAdd_SeqAIJ_SeqDense(Mat,Mat,Mat); 569 /* 570 Performs an efficient scatter on the rows of B needed by this process; this is 571 a modification of the VecScatterBegin_() routines. 572 573 Input: Bbidx = 0: B = Bb 574 = 1: B = Bb1, see MatMatMultSymbolic_MPIAIJ_MPIDense() 575 */ 576 PetscErrorCode MatMPIDenseScatter(Mat A,Mat B,PetscInt Bbidx,Mat C,Mat *outworkB) 577 { 578 Mat_MPIAIJ *aij = (Mat_MPIAIJ*)A->data; 579 PetscErrorCode ierr; 580 const PetscScalar *b; 581 PetscScalar *rvalues; 582 VecScatter ctx = aij->Mvctx; 583 const PetscInt *sindices,*sstarts,*rstarts; 584 const PetscMPIInt *sprocs,*rprocs; 585 PetscInt i,nsends,nrecvs,nrecvs2; 586 MPI_Request *swaits,*rwaits; 587 MPI_Comm comm; 588 PetscMPIInt tag=((PetscObject)ctx)->tag,ncols=B->cmap->N,nrows=aij->B->cmap->n,imdex,nsends_mpi,nrecvs_mpi; 589 MPI_Status status; 590 MPIAIJ_MPIDense *contents; 591 PetscContainer container; 592 Mat workB; 593 MPI_Datatype *stype,*rtype; 594 595 PetscFunctionBegin; 596 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 597 ierr = PetscObjectQuery((PetscObject)C,"workB",(PetscObject*)&container);CHKERRQ(ierr); 598 if (!container) SETERRQ(comm,PETSC_ERR_PLIB,"Container does not exist"); 599 ierr = PetscContainerGetPointer(container,(void**)&contents);CHKERRQ(ierr); 600 601 ierr = VecScatterGetRemote_Private(ctx,PETSC_TRUE/*send*/,&nsends,&sstarts,&sindices,&sprocs,NULL/*bs*/);CHKERRQ(ierr); 602 ierr = VecScatterGetRemoteOrdered_Private(ctx,PETSC_FALSE/*recv*/,&nrecvs,&rstarts,NULL,&rprocs,NULL/*bs*/);CHKERRQ(ierr); 603 ierr = PetscMPIIntCast(nsends,&nsends_mpi);CHKERRQ(ierr); 604 ierr = PetscMPIIntCast(nrecvs,&nrecvs_mpi);CHKERRQ(ierr); 605 if (Bbidx == 0) { 606 workB = *outworkB = contents->workB; 607 } else { 608 workB = *outworkB = contents->workB1; 609 } 610 if (nrows != workB->rmap->n) SETERRQ2(comm,PETSC_ERR_PLIB,"Number of rows of workB %D not equal to columns of aij->B %D",workB->cmap->n,nrows); 611 swaits = contents->swaits; 612 rwaits = contents->rwaits; 613 614 ierr = MatDenseGetArrayRead(B,&b);CHKERRQ(ierr); 615 ierr = MatDenseGetArray(workB,&rvalues);CHKERRQ(ierr); 616 617 /* Post recv, use MPI derived data type to save memory */ 618 rtype = contents->rtype; 619 for (i=0; i<nrecvs; i++) { 620 ierr = MPI_Irecv(rvalues+(rstarts[i]-rstarts[0]),ncols,rtype[i],rprocs[i],tag,comm,rwaits+i);CHKERRQ(ierr); 621 } 622 623 stype = contents->stype; 624 for (i=0; i<nsends; i++) { 625 ierr = MPI_Isend(b,ncols,stype[i],sprocs[i],tag,comm,swaits+i);CHKERRQ(ierr); 626 } 627 628 nrecvs2 = nrecvs; 629 while (nrecvs2) { 630 ierr = MPI_Waitany(nrecvs_mpi,rwaits,&imdex,&status);CHKERRQ(ierr); 631 nrecvs2--; 632 } 633 if (nsends) {ierr = MPI_Waitall(nsends_mpi,swaits,MPI_STATUSES_IGNORE);CHKERRQ(ierr);} 634 635 ierr = VecScatterRestoreRemote_Private(ctx,PETSC_TRUE/*send*/,&nsends,&sstarts,&sindices,&sprocs,NULL);CHKERRQ(ierr); 636 ierr = VecScatterRestoreRemoteOrdered_Private(ctx,PETSC_FALSE/*recv*/,&nrecvs,&rstarts,NULL,&rprocs,NULL);CHKERRQ(ierr); 637 ierr = MatDenseRestoreArrayRead(B,&b);CHKERRQ(ierr); 638 ierr = MatDenseRestoreArray(workB,&rvalues);CHKERRQ(ierr); 639 ierr = MatAssemblyBegin(workB,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 640 ierr = MatAssemblyEnd(workB,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 641 PetscFunctionReturn(0); 642 } 643 644 /* 645 Compute Cb = A*Bb 646 */ 647 PETSC_STATIC_INLINE PetscErrorCode MatMatMultNumeric_MPIAIJ_MPIDense_private(Mat A,Mat Bb,PetscInt Bbidx,PetscInt start,Mat C,const PetscScalar *barray,PetscScalar *carray,Mat Cb) 648 { 649 PetscErrorCode ierr; 650 PetscInt start1; 651 Mat workB; 652 Mat_MPIAIJ *aij = (Mat_MPIAIJ*)A->data; 653 Mat_MPIDense *cbdense = (Mat_MPIDense*)Cb->data; 654 655 PetscFunctionBegin; 656 /* Place barray to Bb */ 657 start1 = start*Bb->rmap->n; 658 ierr = MatDensePlaceArray(Bb,barray+start1);CHKERRQ(ierr); 659 660 /* get off processor parts of Bb needed to complete Cb=A*Bb */ 661 ierr = MatMPIDenseScatter(A,Bb,Bbidx,C,&workB);CHKERRQ(ierr); 662 ierr = MatDenseResetArray(Bb);CHKERRQ(ierr); 663 664 /* off-diagonal block of A times nonlocal rows of Bb */ 665 /* Place carray to Cb */ 666 start1 = start*Cb->rmap->n; 667 ierr = MatDensePlaceArray(Cb,carray+start1);CHKERRQ(ierr); 668 ierr = MatMatMultNumericAdd_SeqAIJ_SeqDense(aij->B,workB,cbdense->A);CHKERRQ(ierr); 669 ierr = MatDenseResetArray(Cb);CHKERRQ(ierr); 670 PetscFunctionReturn(0); 671 } 672 673 PetscErrorCode MatMatMultNumeric_MPIAIJ_MPIDense(Mat A,Mat B,Mat C) 674 { 675 PetscErrorCode ierr; 676 Mat_MPIAIJ *aij = (Mat_MPIAIJ*)A->data; 677 Mat_MPIDense *bdense = (Mat_MPIDense*)B->data; 678 Mat_MPIDense *cdense = (Mat_MPIDense*)C->data; 679 Mat workB; 680 MPIAIJ_MPIDense *contents; 681 PetscContainer container; 682 MPI_Comm comm; 683 PetscInt numBb; 684 685 PetscFunctionBegin; 686 /* diagonal block of A times all local rows of B*/ 687 ierr = MatMatMultNumeric_SeqAIJ_SeqDense(aij->A,bdense->A,cdense->A);CHKERRQ(ierr); 688 689 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 690 ierr = PetscObjectQuery((PetscObject)C,"workB",(PetscObject*)&container);CHKERRQ(ierr); 691 if (!container) SETERRQ(comm,PETSC_ERR_PLIB,"Container does not exist"); 692 ierr = PetscContainerGetPointer(container,(void**)&contents);CHKERRQ(ierr); 693 numBb = contents->numBb; 694 695 if (!numBb) { 696 /* get off processor parts of B needed to complete C=A*B */ 697 ierr = MatMPIDenseScatter(A,B,0,C,&workB);CHKERRQ(ierr); 698 699 /* off-diagonal block of A times nonlocal rows of B */ 700 ierr = MatMatMultNumericAdd_SeqAIJ_SeqDense(aij->B,workB,cdense->A);CHKERRQ(ierr); 701 702 } else { 703 const PetscScalar *barray; 704 PetscScalar *carray; 705 Mat Bb=contents->Bb,Cb=contents->Cb; 706 PetscInt BbN=Bb->cmap->N,start,i; 707 708 ierr = MatDenseGetArrayRead(B,&barray);CHKERRQ(ierr); 709 ierr = MatDenseGetArray(C,&carray);CHKERRQ(ierr); 710 for (i=0; i<numBb; i++) { 711 start = i*BbN; 712 ierr = MatMatMultNumeric_MPIAIJ_MPIDense_private(A,Bb,0,start,C,barray,carray,Cb);CHKERRQ(ierr); 713 } 714 715 if (contents->Bb1) { 716 Bb = contents->Bb1; Cb = contents->Cb1; 717 start = i*BbN; 718 ierr = MatMatMultNumeric_MPIAIJ_MPIDense_private(A,Bb,1,start,C,barray,carray,Cb);CHKERRQ(ierr); 719 } 720 ierr = MatDenseRestoreArrayRead(B,&barray);CHKERRQ(ierr); 721 ierr = MatDenseRestoreArray(C,&carray);CHKERRQ(ierr); 722 } 723 PetscFunctionReturn(0); 724 } 725 726 PetscErrorCode MatMatMultNumeric_MPIAIJ_MPIAIJ(Mat A,Mat P,Mat C) 727 { 728 PetscErrorCode ierr; 729 Mat_MPIAIJ *a = (Mat_MPIAIJ*)A->data,*c=(Mat_MPIAIJ*)C->data; 730 Mat_SeqAIJ *ad = (Mat_SeqAIJ*)(a->A)->data,*ao=(Mat_SeqAIJ*)(a->B)->data; 731 Mat_SeqAIJ *cd = (Mat_SeqAIJ*)(c->A)->data,*co=(Mat_SeqAIJ*)(c->B)->data; 732 PetscInt *adi = ad->i,*adj,*aoi=ao->i,*aoj; 733 PetscScalar *ada,*aoa,*cda=cd->a,*coa=co->a; 734 Mat_SeqAIJ *p_loc,*p_oth; 735 PetscInt *pi_loc,*pj_loc,*pi_oth,*pj_oth,*pj; 736 PetscScalar *pa_loc,*pa_oth,*pa,valtmp,*ca; 737 PetscInt cm = C->rmap->n,anz,pnz; 738 Mat_APMPI *ptap = c->ap; 739 PetscScalar *apa_sparse; 740 PetscInt *api,*apj,*apJ,i,j,k,row; 741 PetscInt cstart = C->cmap->rstart; 742 PetscInt cdnz,conz,k0,k1,nextp; 743 MPI_Comm comm; 744 PetscMPIInt size; 745 746 PetscFunctionBegin; 747 ierr = PetscObjectGetComm((PetscObject)C,&comm);CHKERRQ(ierr); 748 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 749 750 if (!ptap->P_oth && size>1) { 751 SETERRQ(comm,PETSC_ERR_ARG_WRONGSTATE,"AP cannot be reused. Do not call MatFreeIntermediateDataStructures() or use '-mat_freeintermediatedatastructures'"); 752 } 753 apa_sparse = ptap->apa; 754 755 /* 1) get P_oth = ptap->P_oth and P_loc = ptap->P_loc */ 756 /*-----------------------------------------------------*/ 757 /* update numerical values of P_oth and P_loc */ 758 ierr = MatGetBrowsOfAoCols_MPIAIJ(A,P,MAT_REUSE_MATRIX,&ptap->startsj_s,&ptap->startsj_r,&ptap->bufa,&ptap->P_oth);CHKERRQ(ierr); 759 ierr = MatMPIAIJGetLocalMat(P,MAT_REUSE_MATRIX,&ptap->P_loc);CHKERRQ(ierr); 760 761 /* 2) compute numeric C_loc = A_loc*P = Ad*P_loc + Ao*P_oth */ 762 /*----------------------------------------------------------*/ 763 /* get data from symbolic products */ 764 p_loc = (Mat_SeqAIJ*)(ptap->P_loc)->data; 765 pi_loc = p_loc->i; pj_loc = p_loc->j; pa_loc = p_loc->a; 766 if (size >1) { 767 p_oth = (Mat_SeqAIJ*)(ptap->P_oth)->data; 768 pi_oth = p_oth->i; pj_oth = p_oth->j; pa_oth = p_oth->a; 769 } else { 770 p_oth = NULL; pi_oth = NULL; pj_oth = NULL; pa_oth = NULL; 771 } 772 773 api = ptap->api; 774 apj = ptap->apj; 775 for (i=0; i<cm; i++) { 776 apJ = apj + api[i]; 777 778 /* diagonal portion of A */ 779 anz = adi[i+1] - adi[i]; 780 adj = ad->j + adi[i]; 781 ada = ad->a + adi[i]; 782 for (j=0; j<anz; j++) { 783 row = adj[j]; 784 pnz = pi_loc[row+1] - pi_loc[row]; 785 pj = pj_loc + pi_loc[row]; 786 pa = pa_loc + pi_loc[row]; 787 /* perform sparse axpy */ 788 valtmp = ada[j]; 789 nextp = 0; 790 for (k=0; nextp<pnz; k++) { 791 if (apJ[k] == pj[nextp]) { /* column of AP == column of P */ 792 apa_sparse[k] += valtmp*pa[nextp++]; 793 } 794 } 795 ierr = PetscLogFlops(2.0*pnz);CHKERRQ(ierr); 796 } 797 798 /* off-diagonal portion of A */ 799 anz = aoi[i+1] - aoi[i]; 800 aoj = ao->j + aoi[i]; 801 aoa = ao->a + aoi[i]; 802 for (j=0; j<anz; j++) { 803 row = aoj[j]; 804 pnz = pi_oth[row+1] - pi_oth[row]; 805 pj = pj_oth + pi_oth[row]; 806 pa = pa_oth + pi_oth[row]; 807 /* perform sparse axpy */ 808 valtmp = aoa[j]; 809 nextp = 0; 810 for (k=0; nextp<pnz; k++) { 811 if (apJ[k] == pj[nextp]) { /* column of AP == column of P */ 812 apa_sparse[k] += valtmp*pa[nextp++]; 813 } 814 } 815 ierr = PetscLogFlops(2.0*pnz);CHKERRQ(ierr); 816 } 817 818 /* set values in C */ 819 cdnz = cd->i[i+1] - cd->i[i]; 820 conz = co->i[i+1] - co->i[i]; 821 822 /* 1st off-diagoanl part of C */ 823 ca = coa + co->i[i]; 824 k = 0; 825 for (k0=0; k0<conz; k0++) { 826 if (apJ[k] >= cstart) break; 827 ca[k0] = apa_sparse[k]; 828 apa_sparse[k] = 0.0; 829 k++; 830 } 831 832 /* diagonal part of C */ 833 ca = cda + cd->i[i]; 834 for (k1=0; k1<cdnz; k1++) { 835 ca[k1] = apa_sparse[k]; 836 apa_sparse[k] = 0.0; 837 k++; 838 } 839 840 /* 2nd off-diagoanl part of C */ 841 ca = coa + co->i[i]; 842 for (; k0<conz; k0++) { 843 ca[k0] = apa_sparse[k]; 844 apa_sparse[k] = 0.0; 845 k++; 846 } 847 } 848 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 849 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 850 851 if (ptap->freestruct) { 852 ierr = MatFreeIntermediateDataStructures(C);CHKERRQ(ierr); 853 } 854 PetscFunctionReturn(0); 855 } 856 857 /* same as MatMatMultSymbolic_MPIAIJ_MPIAIJ_nonscalable(), except using LLCondensed to avoid O(BN) memory requirement */ 858 PetscErrorCode MatMatMultSymbolic_MPIAIJ_MPIAIJ(Mat A,Mat P,PetscReal fill,Mat C) 859 { 860 PetscErrorCode ierr; 861 MPI_Comm comm; 862 PetscMPIInt size; 863 Mat_APMPI *ptap; 864 PetscFreeSpaceList free_space = NULL,current_space=NULL; 865 Mat_MPIAIJ *a = (Mat_MPIAIJ*)A->data,*c; 866 Mat_SeqAIJ *ad = (Mat_SeqAIJ*)(a->A)->data,*ao=(Mat_SeqAIJ*)(a->B)->data,*p_loc,*p_oth; 867 PetscInt *pi_loc,*pj_loc,*pi_oth,*pj_oth,*dnz,*onz; 868 PetscInt *adi=ad->i,*adj=ad->j,*aoi=ao->i,*aoj=ao->j,rstart=A->rmap->rstart; 869 PetscInt i,pnz,row,*api,*apj,*Jptr,apnz,nspacedouble=0,j,nzi,*lnk,apnz_max=0; 870 PetscInt am=A->rmap->n,pn=P->cmap->n,pm=P->rmap->n,lsize=pn+20; 871 PetscReal afill; 872 MatType mtype; 873 874 PetscFunctionBegin; 875 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 876 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 877 878 /* create struct Mat_APMPI and attached it to C later */ 879 ierr = PetscNew(&ptap);CHKERRQ(ierr); 880 881 /* get P_oth by taking rows of P (= non-zero cols of local A) from other processors */ 882 ierr = MatGetBrowsOfAoCols_MPIAIJ(A,P,MAT_INITIAL_MATRIX,&ptap->startsj_s,&ptap->startsj_r,&ptap->bufa,&ptap->P_oth);CHKERRQ(ierr); 883 884 /* get P_loc by taking all local rows of P */ 885 ierr = MatMPIAIJGetLocalMat(P,MAT_INITIAL_MATRIX,&ptap->P_loc);CHKERRQ(ierr); 886 887 p_loc = (Mat_SeqAIJ*)(ptap->P_loc)->data; 888 pi_loc = p_loc->i; pj_loc = p_loc->j; 889 if (size > 1) { 890 p_oth = (Mat_SeqAIJ*)(ptap->P_oth)->data; 891 pi_oth = p_oth->i; pj_oth = p_oth->j; 892 } else { 893 p_oth = NULL; 894 pi_oth = NULL; pj_oth = NULL; 895 } 896 897 /* first, compute symbolic AP = A_loc*P = A_diag*P_loc + A_off*P_oth */ 898 /*-------------------------------------------------------------------*/ 899 ierr = PetscMalloc1(am+2,&api);CHKERRQ(ierr); 900 ptap->api = api; 901 api[0] = 0; 902 903 ierr = PetscLLCondensedCreate_Scalable(lsize,&lnk);CHKERRQ(ierr); 904 905 /* Initial FreeSpace size is fill*(nnz(A)+nnz(P)) */ 906 ierr = PetscFreeSpaceGet(PetscRealIntMultTruncate(fill,PetscIntSumTruncate(adi[am],PetscIntSumTruncate(aoi[am],pi_loc[pm]))),&free_space);CHKERRQ(ierr); 907 current_space = free_space; 908 ierr = MatPreallocateInitialize(comm,am,pn,dnz,onz);CHKERRQ(ierr); 909 for (i=0; i<am; i++) { 910 /* diagonal portion of A */ 911 nzi = adi[i+1] - adi[i]; 912 for (j=0; j<nzi; j++) { 913 row = *adj++; 914 pnz = pi_loc[row+1] - pi_loc[row]; 915 Jptr = pj_loc + pi_loc[row]; 916 /* Expand list if it is not long enough */ 917 if (pnz+apnz_max > lsize) { 918 lsize = pnz+apnz_max; 919 ierr = PetscLLCondensedExpand_Scalable(lsize, &lnk);CHKERRQ(ierr); 920 } 921 /* add non-zero cols of P into the sorted linked list lnk */ 922 ierr = PetscLLCondensedAddSorted_Scalable(pnz,Jptr,lnk);CHKERRQ(ierr); 923 apnz = *lnk; /* The first element in the list is the number of items in the list */ 924 api[i+1] = api[i] + apnz; 925 if (apnz > apnz_max) apnz_max = apnz; 926 } 927 /* off-diagonal portion of A */ 928 nzi = aoi[i+1] - aoi[i]; 929 for (j=0; j<nzi; j++) { 930 row = *aoj++; 931 pnz = pi_oth[row+1] - pi_oth[row]; 932 Jptr = pj_oth + pi_oth[row]; 933 /* Expand list if it is not long enough */ 934 if (pnz+apnz_max > lsize) { 935 lsize = pnz + apnz_max; 936 ierr = PetscLLCondensedExpand_Scalable(lsize, &lnk);CHKERRQ(ierr); 937 } 938 /* add non-zero cols of P into the sorted linked list lnk */ 939 ierr = PetscLLCondensedAddSorted_Scalable(pnz,Jptr,lnk);CHKERRQ(ierr); 940 apnz = *lnk; /* The first element in the list is the number of items in the list */ 941 api[i+1] = api[i] + apnz; 942 if (apnz > apnz_max) apnz_max = apnz; 943 } 944 apnz = *lnk; 945 api[i+1] = api[i] + apnz; 946 if (apnz > apnz_max) apnz_max = apnz; 947 948 /* if free space is not available, double the total space in the list */ 949 if (current_space->local_remaining<apnz) { 950 ierr = PetscFreeSpaceGet(PetscIntSumTruncate(apnz,current_space->total_array_size),¤t_space);CHKERRQ(ierr); 951 nspacedouble++; 952 } 953 954 /* Copy data into free space, then initialize lnk */ 955 ierr = PetscLLCondensedClean_Scalable(apnz,current_space->array,lnk);CHKERRQ(ierr); 956 ierr = MatPreallocateSet(i+rstart,apnz,current_space->array,dnz,onz);CHKERRQ(ierr); 957 958 current_space->array += apnz; 959 current_space->local_used += apnz; 960 current_space->local_remaining -= apnz; 961 } 962 963 /* Allocate space for apj, initialize apj, and */ 964 /* destroy list of free space and other temporary array(s) */ 965 ierr = PetscMalloc1(api[am]+1,&ptap->apj);CHKERRQ(ierr); 966 apj = ptap->apj; 967 ierr = PetscFreeSpaceContiguous(&free_space,ptap->apj);CHKERRQ(ierr); 968 ierr = PetscLLCondensedDestroy_Scalable(lnk);CHKERRQ(ierr); 969 970 /* create and assemble symbolic parallel matrix C */ 971 /*----------------------------------------------------*/ 972 ierr = MatSetSizes(C,am,pn,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 973 ierr = MatSetBlockSizesFromMats(C,A,P);CHKERRQ(ierr); 974 ierr = MatGetType(A,&mtype);CHKERRQ(ierr); 975 ierr = MatSetType(C,mtype);CHKERRQ(ierr); 976 ierr = MatMPIAIJSetPreallocation(C,0,dnz,0,onz);CHKERRQ(ierr); 977 978 /* malloc apa for assembly C */ 979 ierr = PetscCalloc1(apnz_max,&ptap->apa);CHKERRQ(ierr); 980 981 ierr = MatSetValues_MPIAIJ_CopyFromCSRFormat_Symbolic(C, apj, api);CHKERRQ(ierr); 982 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 983 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 984 ierr = MatPreallocateFinalize(dnz,onz);CHKERRQ(ierr); 985 986 ptap->destroy = C->ops->destroy; 987 ptap->duplicate = C->ops->duplicate; 988 C->ops->matmultnumeric = MatMatMultNumeric_MPIAIJ_MPIAIJ; 989 C->ops->productnumeric = MatProductNumeric_AB; 990 C->ops->destroy = MatDestroy_MPIAIJ_MatMatMult; 991 C->ops->freeintermediatedatastructures = MatFreeIntermediateDataStructures_MPIAIJ_AP; 992 993 /* attach the supporting struct to C for reuse */ 994 c = (Mat_MPIAIJ*)C->data; 995 c->ap = ptap; 996 997 /* set MatInfo */ 998 afill = (PetscReal)api[am]/(adi[am]+aoi[am]+pi_loc[pm]+1) + 1.e-5; 999 if (afill < 1.0) afill = 1.0; 1000 C->info.mallocs = nspacedouble; 1001 C->info.fill_ratio_given = fill; 1002 C->info.fill_ratio_needed = afill; 1003 1004 #if defined(PETSC_USE_INFO) 1005 if (api[am]) { 1006 ierr = PetscInfo3(C,"Reallocs %D; Fill ratio: given %g needed %g.\n",nspacedouble,(double)fill,(double)afill);CHKERRQ(ierr); 1007 ierr = PetscInfo1(C,"Use MatMatMult(A,B,MatReuse,%g,&C) for best performance.;\n",(double)afill);CHKERRQ(ierr); 1008 } else { 1009 ierr = PetscInfo(C,"Empty matrix product\n");CHKERRQ(ierr); 1010 } 1011 #endif 1012 PetscFunctionReturn(0); 1013 } 1014 1015 /* This function is needed for the seqMPI matrix-matrix multiplication. */ 1016 /* Three input arrays are merged to one output array. The size of the */ 1017 /* output array is also output. Duplicate entries only show up once. */ 1018 static void Merge3SortedArrays(PetscInt size1, PetscInt *in1, 1019 PetscInt size2, PetscInt *in2, 1020 PetscInt size3, PetscInt *in3, 1021 PetscInt *size4, PetscInt *out) 1022 { 1023 int i = 0, j = 0, k = 0, l = 0; 1024 1025 /* Traverse all three arrays */ 1026 while (i<size1 && j<size2 && k<size3) { 1027 if (in1[i] < in2[j] && in1[i] < in3[k]) { 1028 out[l++] = in1[i++]; 1029 } 1030 else if(in2[j] < in1[i] && in2[j] < in3[k]) { 1031 out[l++] = in2[j++]; 1032 } 1033 else if(in3[k] < in1[i] && in3[k] < in2[j]) { 1034 out[l++] = in3[k++]; 1035 } 1036 else if(in1[i] == in2[j] && in1[i] < in3[k]) { 1037 out[l++] = in1[i]; 1038 i++, j++; 1039 } 1040 else if(in1[i] == in3[k] && in1[i] < in2[j]) { 1041 out[l++] = in1[i]; 1042 i++, k++; 1043 } 1044 else if(in3[k] == in2[j] && in2[j] < in1[i]) { 1045 out[l++] = in2[j]; 1046 k++, j++; 1047 } 1048 else if(in1[i] == in2[j] && in1[i] == in3[k]) { 1049 out[l++] = in1[i]; 1050 i++, j++, k++; 1051 } 1052 } 1053 1054 /* Traverse two remaining arrays */ 1055 while (i<size1 && j<size2) { 1056 if (in1[i] < in2[j]) { 1057 out[l++] = in1[i++]; 1058 } 1059 else if(in1[i] > in2[j]) { 1060 out[l++] = in2[j++]; 1061 } 1062 else { 1063 out[l++] = in1[i]; 1064 i++, j++; 1065 } 1066 } 1067 1068 while (i<size1 && k<size3) { 1069 if (in1[i] < in3[k]) { 1070 out[l++] = in1[i++]; 1071 } 1072 else if(in1[i] > in3[k]) { 1073 out[l++] = in3[k++]; 1074 } 1075 else { 1076 out[l++] = in1[i]; 1077 i++, k++; 1078 } 1079 } 1080 1081 while (k<size3 && j<size2) { 1082 if (in3[k] < in2[j]) { 1083 out[l++] = in3[k++]; 1084 } 1085 else if(in3[k] > in2[j]) { 1086 out[l++] = in2[j++]; 1087 } 1088 else { 1089 out[l++] = in3[k]; 1090 k++, j++; 1091 } 1092 } 1093 1094 /* Traverse one remaining array */ 1095 while (i<size1) out[l++] = in1[i++]; 1096 while (j<size2) out[l++] = in2[j++]; 1097 while (k<size3) out[l++] = in3[k++]; 1098 1099 *size4 = l; 1100 } 1101 1102 /* This matrix-matrix multiplication algorithm divides the multiplication into three multiplications and */ 1103 /* adds up the products. Two of these three multiplications are performed with existing (sequential) */ 1104 /* matrix-matrix multiplications. */ 1105 PetscErrorCode MatMatMultSymbolic_MPIAIJ_MPIAIJ_seqMPI(Mat A, Mat P, PetscReal fill, Mat C) 1106 { 1107 PetscErrorCode ierr; 1108 MPI_Comm comm; 1109 PetscMPIInt size; 1110 Mat_APMPI *ptap; 1111 PetscFreeSpaceList free_space_diag=NULL, current_space=NULL; 1112 Mat_MPIAIJ *a =(Mat_MPIAIJ*)A->data; 1113 Mat_SeqAIJ *ad =(Mat_SeqAIJ*)(a->A)->data,*ao=(Mat_SeqAIJ*)(a->B)->data,*p_loc; 1114 Mat_MPIAIJ *p =(Mat_MPIAIJ*)P->data; 1115 Mat_MPIAIJ *c; 1116 Mat_SeqAIJ *adpd_seq, *p_off, *aopoth_seq; 1117 PetscInt adponz, adpdnz; 1118 PetscInt *pi_loc,*dnz,*onz; 1119 PetscInt *adi=ad->i,*adj=ad->j,*aoi=ao->i,rstart=A->rmap->rstart; 1120 PetscInt *lnk,i, i1=0,pnz,row,*adpoi,*adpoj, *api, *adpoJ, *aopJ, *apJ,*Jptr, aopnz, nspacedouble=0,j,nzi, 1121 *apj,apnz, *adpdi, *adpdj, *adpdJ, *poff_i, *poff_j, *j_temp, *aopothi, *aopothj; 1122 PetscInt am=A->rmap->n,pN=P->cmap->N,pn=P->cmap->n,pm=P->rmap->n, p_colstart, p_colend; 1123 PetscBT lnkbt; 1124 PetscReal afill; 1125 PetscMPIInt rank; 1126 Mat adpd, aopoth; 1127 MatType mtype; 1128 const char *prefix; 1129 1130 PetscFunctionBegin; 1131 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 1132 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 1133 ierr = MPI_Comm_rank(comm, &rank);CHKERRQ(ierr); 1134 ierr = MatGetOwnershipRangeColumn(P, &p_colstart, &p_colend); CHKERRQ(ierr); 1135 1136 /* create struct Mat_APMPI and attached it to C later */ 1137 ierr = PetscNew(&ptap);CHKERRQ(ierr); 1138 1139 /* get P_oth by taking rows of P (= non-zero cols of local A) from other processors */ 1140 ierr = MatGetBrowsOfAoCols_MPIAIJ(A,P,MAT_INITIAL_MATRIX,&ptap->startsj_s,&ptap->startsj_r,&ptap->bufa,&ptap->P_oth);CHKERRQ(ierr); 1141 1142 /* get P_loc by taking all local rows of P */ 1143 ierr = MatMPIAIJGetLocalMat(P,MAT_INITIAL_MATRIX,&ptap->P_loc);CHKERRQ(ierr); 1144 1145 1146 p_loc = (Mat_SeqAIJ*)(ptap->P_loc)->data; 1147 pi_loc = p_loc->i; 1148 1149 /* Allocate memory for the i arrays of the matrices A*P, A_diag*P_off and A_offd * P */ 1150 ierr = PetscMalloc1(am+2,&api);CHKERRQ(ierr); 1151 ierr = PetscMalloc1(am+2,&adpoi);CHKERRQ(ierr); 1152 1153 adpoi[0] = 0; 1154 ptap->api = api; 1155 api[0] = 0; 1156 1157 /* create and initialize a linked list, will be used for both A_diag * P_loc_off and A_offd * P_oth */ 1158 ierr = PetscLLCondensedCreate(pN,pN,&lnk,&lnkbt);CHKERRQ(ierr); 1159 ierr = MatPreallocateInitialize(comm,am,pn,dnz,onz);CHKERRQ(ierr); 1160 1161 /* Symbolic calc of A_loc_diag * P_loc_diag */ 1162 ierr = MatGetOptionsPrefix(A,&prefix);CHKERRQ(ierr); 1163 ierr = MatProductCreate(a->A,p->A,NULL,&adpd);CHKERRQ(ierr); 1164 ierr = MatGetOptionsPrefix(A,&prefix);CHKERRQ(ierr); 1165 ierr = MatSetOptionsPrefix(adpd,prefix);CHKERRQ(ierr); 1166 ierr = MatAppendOptionsPrefix(adpd,"inner_diag_");CHKERRQ(ierr); 1167 1168 ierr = MatProductSetType(adpd,MATPRODUCT_AB);CHKERRQ(ierr); 1169 ierr = MatProductSetAlgorithm(adpd,"sorted");CHKERRQ(ierr); 1170 ierr = MatProductSetFill(adpd,fill);CHKERRQ(ierr); 1171 ierr = MatProductSetFromOptions(adpd);CHKERRQ(ierr); 1172 ierr = MatProductSymbolic(adpd);CHKERRQ(ierr); 1173 1174 adpd_seq = (Mat_SeqAIJ*)((adpd)->data); 1175 adpdi = adpd_seq->i; adpdj = adpd_seq->j; 1176 p_off = (Mat_SeqAIJ*)((p->B)->data); 1177 poff_i = p_off->i; poff_j = p_off->j; 1178 1179 /* j_temp stores indices of a result row before they are added to the linked list */ 1180 ierr = PetscMalloc1(pN+2,&j_temp);CHKERRQ(ierr); 1181 1182 1183 /* Symbolic calc of the A_diag * p_loc_off */ 1184 /* Initial FreeSpace size is fill*(nnz(A)+nnz(P)) */ 1185 ierr = PetscFreeSpaceGet(PetscRealIntMultTruncate(fill,PetscIntSumTruncate(adi[am],PetscIntSumTruncate(aoi[am],pi_loc[pm]))),&free_space_diag);CHKERRQ(ierr); 1186 current_space = free_space_diag; 1187 1188 for (i=0; i<am; i++) { 1189 /* A_diag * P_loc_off */ 1190 nzi = adi[i+1] - adi[i]; 1191 for (j=0; j<nzi; j++) { 1192 row = *adj++; 1193 pnz = poff_i[row+1] - poff_i[row]; 1194 Jptr = poff_j + poff_i[row]; 1195 for(i1 = 0; i1 < pnz; i1++) { 1196 j_temp[i1] = p->garray[Jptr[i1]]; 1197 } 1198 /* add non-zero cols of P into the sorted linked list lnk */ 1199 ierr = PetscLLCondensedAddSorted(pnz,j_temp,lnk,lnkbt);CHKERRQ(ierr); 1200 } 1201 1202 adponz = lnk[0]; 1203 adpoi[i+1] = adpoi[i] + adponz; 1204 1205 /* if free space is not available, double the total space in the list */ 1206 if (current_space->local_remaining<adponz) { 1207 ierr = PetscFreeSpaceGet(PetscIntSumTruncate(adponz,current_space->total_array_size),¤t_space);CHKERRQ(ierr); 1208 nspacedouble++; 1209 } 1210 1211 /* Copy data into free space, then initialize lnk */ 1212 ierr = PetscLLCondensedClean(pN,adponz,current_space->array,lnk,lnkbt);CHKERRQ(ierr); 1213 1214 current_space->array += adponz; 1215 current_space->local_used += adponz; 1216 current_space->local_remaining -= adponz; 1217 } 1218 1219 /* Symbolic calc of A_off * P_oth */ 1220 ierr = MatSetOptionsPrefix(a->B,prefix);CHKERRQ(ierr); 1221 ierr = MatAppendOptionsPrefix(a->B,"inner_offdiag_");CHKERRQ(ierr); 1222 ierr = MatCreate(PETSC_COMM_SELF,&aopoth);CHKERRQ(ierr); 1223 ierr = MatMatMultSymbolic_SeqAIJ_SeqAIJ(a->B, ptap->P_oth, fill, aopoth);CHKERRQ(ierr); 1224 aopoth_seq = (Mat_SeqAIJ*)((aopoth)->data); 1225 aopothi = aopoth_seq->i; aopothj = aopoth_seq->j; 1226 1227 /* Allocate space for apj, adpj, aopj, ... */ 1228 /* destroy lists of free space and other temporary array(s) */ 1229 1230 ierr = PetscMalloc1(aopothi[am] + adpoi[am] + adpdi[am]+2, &ptap->apj);CHKERRQ(ierr); 1231 ierr = PetscMalloc1(adpoi[am]+2, &adpoj);CHKERRQ(ierr); 1232 1233 /* Copy from linked list to j-array */ 1234 ierr = PetscFreeSpaceContiguous(&free_space_diag,adpoj);CHKERRQ(ierr); 1235 ierr = PetscLLDestroy(lnk,lnkbt);CHKERRQ(ierr); 1236 1237 adpoJ = adpoj; 1238 adpdJ = adpdj; 1239 aopJ = aopothj; 1240 apj = ptap->apj; 1241 apJ = apj; /* still empty */ 1242 1243 /* Merge j-arrays of A_off * P, A_diag * P_loc_off, and */ 1244 /* A_diag * P_loc_diag to get A*P */ 1245 for (i = 0; i < am; i++) { 1246 aopnz = aopothi[i+1] - aopothi[i]; 1247 adponz = adpoi[i+1] - adpoi[i]; 1248 adpdnz = adpdi[i+1] - adpdi[i]; 1249 1250 /* Correct indices from A_diag*P_diag */ 1251 for(i1 = 0; i1 < adpdnz; i1++) { 1252 adpdJ[i1] += p_colstart; 1253 } 1254 /* Merge j-arrays of A_diag * P_loc_off and A_diag * P_loc_diag and A_off * P_oth */ 1255 Merge3SortedArrays(adponz, adpoJ, adpdnz, adpdJ, aopnz, aopJ, &apnz, apJ); 1256 ierr = MatPreallocateSet(i+rstart, apnz, apJ, dnz, onz); CHKERRQ(ierr); 1257 1258 aopJ += aopnz; 1259 adpoJ += adponz; 1260 adpdJ += adpdnz; 1261 apJ += apnz; 1262 api[i+1] = api[i] + apnz; 1263 } 1264 1265 /* malloc apa to store dense row A[i,:]*P */ 1266 ierr = PetscCalloc1(pN+2,&ptap->apa);CHKERRQ(ierr); 1267 1268 /* create and assemble symbolic parallel matrix C */ 1269 ierr = MatSetSizes(C,am,pn,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 1270 ierr = MatSetBlockSizesFromMats(C,A,P);CHKERRQ(ierr); 1271 ierr = MatGetType(A,&mtype);CHKERRQ(ierr); 1272 ierr = MatSetType(C,mtype);CHKERRQ(ierr); 1273 ierr = MatMPIAIJSetPreallocation(C,0,dnz,0,onz);CHKERRQ(ierr); 1274 1275 1276 ierr = MatSetValues_MPIAIJ_CopyFromCSRFormat_Symbolic(C, apj, api);CHKERRQ(ierr); 1277 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1278 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1279 ierr = MatPreallocateFinalize(dnz,onz);CHKERRQ(ierr); 1280 1281 1282 ptap->destroy = C->ops->destroy; 1283 ptap->duplicate = C->ops->duplicate; 1284 C->ops->matmultnumeric = MatMatMultNumeric_MPIAIJ_MPIAIJ_nonscalable; 1285 C->ops->productnumeric = MatProductNumeric_AB; 1286 C->ops->destroy = MatDestroy_MPIAIJ_MatMatMult; 1287 1288 /* attach the supporting struct to C for reuse */ 1289 c = (Mat_MPIAIJ*)C->data; 1290 c->ap = ptap; 1291 1292 /* set MatInfo */ 1293 afill = (PetscReal)api[am]/(adi[am]+aoi[am]+pi_loc[pm]+1) + 1.e-5; 1294 if (afill < 1.0) afill = 1.0; 1295 C->info.mallocs = nspacedouble; 1296 C->info.fill_ratio_given = fill; 1297 C->info.fill_ratio_needed = afill; 1298 1299 #if defined(PETSC_USE_INFO) 1300 if (api[am]) { 1301 ierr = PetscInfo3(C,"Reallocs %D; Fill ratio: given %g needed %g.\n",nspacedouble,(double)fill,(double)afill);CHKERRQ(ierr); 1302 ierr = PetscInfo1(C,"Use MatMatMult(A,B,MatReuse,%g,&C) for best performance.;\n",(double)afill);CHKERRQ(ierr); 1303 } else { 1304 ierr = PetscInfo(C,"Empty matrix product\n");CHKERRQ(ierr); 1305 } 1306 #endif 1307 1308 ierr = MatDestroy(&aopoth);CHKERRQ(ierr); 1309 ierr = MatDestroy(&adpd);CHKERRQ(ierr); 1310 ierr = PetscFree(j_temp);CHKERRQ(ierr); 1311 ierr = PetscFree(adpoj);CHKERRQ(ierr); 1312 ierr = PetscFree(adpoi);CHKERRQ(ierr); 1313 PetscFunctionReturn(0); 1314 } 1315 1316 /*-------------------------------------------------------------------------*/ 1317 /* This routine only works when scall=MAT_REUSE_MATRIX! */ 1318 PetscErrorCode MatTransposeMatMultNumeric_MPIAIJ_MPIAIJ_matmatmult(Mat P,Mat A,Mat C) 1319 { 1320 PetscErrorCode ierr; 1321 Mat_MPIAIJ *c=(Mat_MPIAIJ*)C->data; 1322 Mat_APMPI *ptap= c->ap; 1323 Mat Pt; 1324 1325 PetscFunctionBegin; 1326 if (!ptap->Pt) { 1327 MPI_Comm comm; 1328 ierr = PetscObjectGetComm((PetscObject)C,&comm);CHKERRQ(ierr); 1329 SETERRQ(comm,PETSC_ERR_ARG_WRONGSTATE,"PtA cannot be reused. Do not call MatFreeIntermediateDataStructures() or use '-mat_freeintermediatedatastructures'"); 1330 } 1331 1332 Pt = ptap->Pt; 1333 ierr = MatTranspose(P,MAT_REUSE_MATRIX,&Pt);CHKERRQ(ierr); 1334 ierr = MatMatMultNumeric_MPIAIJ_MPIAIJ(Pt,A,C);CHKERRQ(ierr); 1335 1336 /* supporting struct ptap consumes almost same amount of memory as C=PtAP, release it if C will not be updated by A and P */ 1337 if (ptap->freestruct) { 1338 ierr = MatFreeIntermediateDataStructures(C);CHKERRQ(ierr); 1339 } 1340 PetscFunctionReturn(0); 1341 } 1342 1343 /* This routine is modified from MatPtAPSymbolic_MPIAIJ_MPIAIJ() */ 1344 PetscErrorCode MatTransposeMatMultSymbolic_MPIAIJ_MPIAIJ_nonscalable(Mat P,Mat A,PetscReal fill,Mat C) 1345 { 1346 PetscErrorCode ierr; 1347 Mat_APMPI *ptap; 1348 Mat_MPIAIJ *p=(Mat_MPIAIJ*)P->data,*c; 1349 MPI_Comm comm; 1350 PetscMPIInt size,rank; 1351 PetscFreeSpaceList free_space=NULL,current_space=NULL; 1352 PetscInt pn=P->cmap->n,aN=A->cmap->N,an=A->cmap->n; 1353 PetscInt *lnk,i,k,nsend; 1354 PetscBT lnkbt; 1355 PetscMPIInt tagi,tagj,*len_si,*len_s,*len_ri,icompleted=0,nrecv; 1356 PetscInt **buf_rj,**buf_ri,**buf_ri_k; 1357 PetscInt len,proc,*dnz,*onz,*owners,nzi; 1358 PetscInt nrows,*buf_s,*buf_si,*buf_si_i,**nextrow,**nextci; 1359 MPI_Request *swaits,*rwaits; 1360 MPI_Status *sstatus,rstatus; 1361 PetscLayout rowmap; 1362 PetscInt *owners_co,*coi,*coj; /* i and j array of (p->B)^T*A*P - used in the communication */ 1363 PetscMPIInt *len_r,*id_r; /* array of length of comm->size, store send/recv matrix values */ 1364 PetscInt *Jptr,*prmap=p->garray,con,j,Crmax; 1365 Mat_SeqAIJ *a_loc,*c_loc,*c_oth; 1366 PetscTable ta; 1367 MatType mtype; 1368 const char *prefix; 1369 1370 PetscFunctionBegin; 1371 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 1372 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 1373 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 1374 1375 /* create symbolic parallel matrix C */ 1376 ierr = MatGetType(A,&mtype);CHKERRQ(ierr); 1377 ierr = MatSetType(C,mtype);CHKERRQ(ierr); 1378 1379 C->ops->transposematmultnumeric = MatTransposeMatMultNumeric_MPIAIJ_MPIAIJ_nonscalable; 1380 1381 /* create struct Mat_APMPI and attached it to C later */ 1382 ierr = PetscNew(&ptap);CHKERRQ(ierr); 1383 ptap->reuse = MAT_INITIAL_MATRIX; 1384 1385 /* (0) compute Rd = Pd^T, Ro = Po^T */ 1386 /* --------------------------------- */ 1387 ierr = MatTranspose_SeqAIJ(p->A,MAT_INITIAL_MATRIX,&ptap->Rd);CHKERRQ(ierr); 1388 ierr = MatTranspose_SeqAIJ(p->B,MAT_INITIAL_MATRIX,&ptap->Ro);CHKERRQ(ierr); 1389 1390 /* (1) compute symbolic A_loc */ 1391 /* ---------------------------*/ 1392 ierr = MatMPIAIJGetLocalMat(A,MAT_INITIAL_MATRIX,&ptap->A_loc);CHKERRQ(ierr); 1393 1394 /* (2-1) compute symbolic C_oth = Ro*A_loc */ 1395 /* ------------------------------------ */ 1396 ierr = MatGetOptionsPrefix(A,&prefix);CHKERRQ(ierr); 1397 ierr = MatSetOptionsPrefix(ptap->Ro,prefix);CHKERRQ(ierr); 1398 ierr = MatAppendOptionsPrefix(ptap->Ro,"inner_offdiag_");CHKERRQ(ierr); 1399 ierr = MatCreate(PETSC_COMM_SELF,&ptap->C_oth);CHKERRQ(ierr); 1400 ierr = MatMatMultSymbolic_SeqAIJ_SeqAIJ(ptap->Ro,ptap->A_loc,fill,ptap->C_oth);CHKERRQ(ierr); 1401 1402 /* (3) send coj of C_oth to other processors */ 1403 /* ------------------------------------------ */ 1404 /* determine row ownership */ 1405 ierr = PetscLayoutCreate(comm,&rowmap);CHKERRQ(ierr); 1406 rowmap->n = pn; 1407 rowmap->bs = 1; 1408 ierr = PetscLayoutSetUp(rowmap);CHKERRQ(ierr); 1409 owners = rowmap->range; 1410 1411 /* determine the number of messages to send, their lengths */ 1412 ierr = PetscMalloc4(size,&len_s,size,&len_si,size,&sstatus,size+2,&owners_co);CHKERRQ(ierr); 1413 ierr = PetscArrayzero(len_s,size);CHKERRQ(ierr); 1414 ierr = PetscArrayzero(len_si,size);CHKERRQ(ierr); 1415 1416 c_oth = (Mat_SeqAIJ*)ptap->C_oth->data; 1417 coi = c_oth->i; coj = c_oth->j; 1418 con = ptap->C_oth->rmap->n; 1419 proc = 0; 1420 for (i=0; i<con; i++) { 1421 while (prmap[i] >= owners[proc+1]) proc++; 1422 len_si[proc]++; /* num of rows in Co(=Pt*A) to be sent to [proc] */ 1423 len_s[proc] += coi[i+1] - coi[i]; /* num of nonzeros in Co to be sent to [proc] */ 1424 } 1425 1426 len = 0; /* max length of buf_si[], see (4) */ 1427 owners_co[0] = 0; 1428 nsend = 0; 1429 for (proc=0; proc<size; proc++) { 1430 owners_co[proc+1] = owners_co[proc] + len_si[proc]; 1431 if (len_s[proc]) { 1432 nsend++; 1433 len_si[proc] = 2*(len_si[proc] + 1); /* length of buf_si to be sent to [proc] */ 1434 len += len_si[proc]; 1435 } 1436 } 1437 1438 /* determine the number and length of messages to receive for coi and coj */ 1439 ierr = PetscGatherNumberOfMessages(comm,NULL,len_s,&nrecv);CHKERRQ(ierr); 1440 ierr = PetscGatherMessageLengths2(comm,nsend,nrecv,len_s,len_si,&id_r,&len_r,&len_ri);CHKERRQ(ierr); 1441 1442 /* post the Irecv and Isend of coj */ 1443 ierr = PetscCommGetNewTag(comm,&tagj);CHKERRQ(ierr); 1444 ierr = PetscPostIrecvInt(comm,tagj,nrecv,id_r,len_r,&buf_rj,&rwaits);CHKERRQ(ierr); 1445 ierr = PetscMalloc1(nsend+1,&swaits);CHKERRQ(ierr); 1446 for (proc=0, k=0; proc<size; proc++) { 1447 if (!len_s[proc]) continue; 1448 i = owners_co[proc]; 1449 ierr = MPI_Isend(coj+coi[i],len_s[proc],MPIU_INT,proc,tagj,comm,swaits+k);CHKERRQ(ierr); 1450 k++; 1451 } 1452 1453 /* (2-2) compute symbolic C_loc = Rd*A_loc */ 1454 /* ---------------------------------------- */ 1455 ierr = MatSetOptionsPrefix(ptap->Rd,prefix);CHKERRQ(ierr); 1456 ierr = MatAppendOptionsPrefix(ptap->Rd,"inner_diag_");CHKERRQ(ierr); 1457 ierr = MatCreate(PETSC_COMM_SELF,&ptap->C_loc);CHKERRQ(ierr); 1458 ierr = MatMatMultSymbolic_SeqAIJ_SeqAIJ(ptap->Rd,ptap->A_loc,fill,ptap->C_loc);CHKERRQ(ierr); 1459 c_loc = (Mat_SeqAIJ*)ptap->C_loc->data; 1460 1461 /* receives coj are complete */ 1462 for (i=0; i<nrecv; i++) { 1463 ierr = MPI_Waitany(nrecv,rwaits,&icompleted,&rstatus);CHKERRQ(ierr); 1464 } 1465 ierr = PetscFree(rwaits);CHKERRQ(ierr); 1466 if (nsend) {ierr = MPI_Waitall(nsend,swaits,sstatus);CHKERRQ(ierr);} 1467 1468 /* add received column indices into ta to update Crmax */ 1469 a_loc = (Mat_SeqAIJ*)(ptap->A_loc)->data; 1470 1471 /* create and initialize a linked list */ 1472 ierr = PetscTableCreate(an,aN,&ta);CHKERRQ(ierr); /* for compute Crmax */ 1473 MatRowMergeMax_SeqAIJ(a_loc,ptap->A_loc->rmap->N,ta); 1474 1475 for (k=0; k<nrecv; k++) {/* k-th received message */ 1476 Jptr = buf_rj[k]; 1477 for (j=0; j<len_r[k]; j++) { 1478 ierr = PetscTableAdd(ta,*(Jptr+j)+1,1,INSERT_VALUES);CHKERRQ(ierr); 1479 } 1480 } 1481 ierr = PetscTableGetCount(ta,&Crmax);CHKERRQ(ierr); 1482 ierr = PetscTableDestroy(&ta);CHKERRQ(ierr); 1483 1484 /* (4) send and recv coi */ 1485 /*-----------------------*/ 1486 ierr = PetscCommGetNewTag(comm,&tagi);CHKERRQ(ierr); 1487 ierr = PetscPostIrecvInt(comm,tagi,nrecv,id_r,len_ri,&buf_ri,&rwaits);CHKERRQ(ierr); 1488 ierr = PetscMalloc1(len+1,&buf_s);CHKERRQ(ierr); 1489 buf_si = buf_s; /* points to the beginning of k-th msg to be sent */ 1490 for (proc=0,k=0; proc<size; proc++) { 1491 if (!len_s[proc]) continue; 1492 /* form outgoing message for i-structure: 1493 buf_si[0]: nrows to be sent 1494 [1:nrows]: row index (global) 1495 [nrows+1:2*nrows+1]: i-structure index 1496 */ 1497 /*-------------------------------------------*/ 1498 nrows = len_si[proc]/2 - 1; /* num of rows in Co to be sent to [proc] */ 1499 buf_si_i = buf_si + nrows+1; 1500 buf_si[0] = nrows; 1501 buf_si_i[0] = 0; 1502 nrows = 0; 1503 for (i=owners_co[proc]; i<owners_co[proc+1]; i++) { 1504 nzi = coi[i+1] - coi[i]; 1505 buf_si_i[nrows+1] = buf_si_i[nrows] + nzi; /* i-structure */ 1506 buf_si[nrows+1] = prmap[i] -owners[proc]; /* local row index */ 1507 nrows++; 1508 } 1509 ierr = MPI_Isend(buf_si,len_si[proc],MPIU_INT,proc,tagi,comm,swaits+k);CHKERRQ(ierr); 1510 k++; 1511 buf_si += len_si[proc]; 1512 } 1513 for (i=0; i<nrecv; i++) { 1514 ierr = MPI_Waitany(nrecv,rwaits,&icompleted,&rstatus);CHKERRQ(ierr); 1515 } 1516 ierr = PetscFree(rwaits);CHKERRQ(ierr); 1517 if (nsend) {ierr = MPI_Waitall(nsend,swaits,sstatus);CHKERRQ(ierr);} 1518 1519 ierr = PetscFree4(len_s,len_si,sstatus,owners_co);CHKERRQ(ierr); 1520 ierr = PetscFree(len_ri);CHKERRQ(ierr); 1521 ierr = PetscFree(swaits);CHKERRQ(ierr); 1522 ierr = PetscFree(buf_s);CHKERRQ(ierr); 1523 1524 /* (5) compute the local portion of C */ 1525 /* ------------------------------------------ */ 1526 /* set initial free space to be Crmax, sufficient for holding nozeros in each row of C */ 1527 ierr = PetscFreeSpaceGet(Crmax,&free_space);CHKERRQ(ierr); 1528 current_space = free_space; 1529 1530 ierr = PetscMalloc3(nrecv,&buf_ri_k,nrecv,&nextrow,nrecv,&nextci);CHKERRQ(ierr); 1531 for (k=0; k<nrecv; k++) { 1532 buf_ri_k[k] = buf_ri[k]; /* beginning of k-th recved i-structure */ 1533 nrows = *buf_ri_k[k]; 1534 nextrow[k] = buf_ri_k[k] + 1; /* next row number of k-th recved i-structure */ 1535 nextci[k] = buf_ri_k[k] + (nrows + 1); /* poins to the next i-structure of k-th recved i-structure */ 1536 } 1537 1538 ierr = MatPreallocateInitialize(comm,pn,an,dnz,onz);CHKERRQ(ierr); 1539 ierr = PetscLLCondensedCreate(Crmax,aN,&lnk,&lnkbt);CHKERRQ(ierr); 1540 for (i=0; i<pn; i++) { 1541 /* add C_loc into C */ 1542 nzi = c_loc->i[i+1] - c_loc->i[i]; 1543 Jptr = c_loc->j + c_loc->i[i]; 1544 ierr = PetscLLCondensedAddSorted(nzi,Jptr,lnk,lnkbt);CHKERRQ(ierr); 1545 1546 /* add received col data into lnk */ 1547 for (k=0; k<nrecv; k++) { /* k-th received message */ 1548 if (i == *nextrow[k]) { /* i-th row */ 1549 nzi = *(nextci[k]+1) - *nextci[k]; 1550 Jptr = buf_rj[k] + *nextci[k]; 1551 ierr = PetscLLCondensedAddSorted(nzi,Jptr,lnk,lnkbt);CHKERRQ(ierr); 1552 nextrow[k]++; nextci[k]++; 1553 } 1554 } 1555 nzi = lnk[0]; 1556 1557 /* copy data into free space, then initialize lnk */ 1558 ierr = PetscLLCondensedClean(aN,nzi,current_space->array,lnk,lnkbt);CHKERRQ(ierr); 1559 ierr = MatPreallocateSet(i+owners[rank],nzi,current_space->array,dnz,onz);CHKERRQ(ierr); 1560 } 1561 ierr = PetscFree3(buf_ri_k,nextrow,nextci);CHKERRQ(ierr); 1562 ierr = PetscLLDestroy(lnk,lnkbt);CHKERRQ(ierr); 1563 ierr = PetscFreeSpaceDestroy(free_space);CHKERRQ(ierr); 1564 1565 /* local sizes and preallocation */ 1566 ierr = MatSetSizes(C,pn,an,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 1567 if (P->cmap->bs > 0) {ierr = PetscLayoutSetBlockSize(C->rmap,P->cmap->bs);CHKERRQ(ierr);} 1568 if (A->cmap->bs > 0) {ierr = PetscLayoutSetBlockSize(C->cmap,A->cmap->bs);CHKERRQ(ierr);} 1569 ierr = MatMPIAIJSetPreallocation(C,0,dnz,0,onz);CHKERRQ(ierr); 1570 ierr = MatPreallocateFinalize(dnz,onz);CHKERRQ(ierr); 1571 1572 /* members in merge */ 1573 ierr = PetscFree(id_r);CHKERRQ(ierr); 1574 ierr = PetscFree(len_r);CHKERRQ(ierr); 1575 ierr = PetscFree(buf_ri[0]);CHKERRQ(ierr); 1576 ierr = PetscFree(buf_ri);CHKERRQ(ierr); 1577 ierr = PetscFree(buf_rj[0]);CHKERRQ(ierr); 1578 ierr = PetscFree(buf_rj);CHKERRQ(ierr); 1579 ierr = PetscLayoutDestroy(&rowmap);CHKERRQ(ierr); 1580 1581 /* attach the supporting struct to C for reuse */ 1582 c = (Mat_MPIAIJ*)C->data; 1583 c->ap = ptap; 1584 ptap->destroy = C->ops->destroy; 1585 1586 /* C is not ready for use - assembly will be done by MatPtAPNumeric() */ 1587 C->assembled = PETSC_FALSE; 1588 C->ops->destroy = MatDestroy_MPIAIJ_PtAP; 1589 C->ops->freeintermediatedatastructures = MatFreeIntermediateDataStructures_MPIAIJ_AP; 1590 PetscFunctionReturn(0); 1591 } 1592 1593 PetscErrorCode MatTransposeMatMultNumeric_MPIAIJ_MPIAIJ_nonscalable(Mat P,Mat A,Mat C) 1594 { 1595 PetscErrorCode ierr; 1596 Mat_MPIAIJ *p=(Mat_MPIAIJ*)P->data,*c=(Mat_MPIAIJ*)C->data; 1597 Mat_SeqAIJ *c_seq; 1598 Mat_APMPI *ptap = c->ap; 1599 Mat A_loc,C_loc,C_oth; 1600 PetscInt i,rstart,rend,cm,ncols,row; 1601 const PetscInt *cols; 1602 const PetscScalar *vals; 1603 1604 PetscFunctionBegin; 1605 if (!ptap->A_loc) { 1606 MPI_Comm comm; 1607 ierr = PetscObjectGetComm((PetscObject)C,&comm);CHKERRQ(ierr); 1608 SETERRQ(comm,PETSC_ERR_ARG_WRONGSTATE,"PtA cannot be reused. Do not call MatFreeIntermediateDataStructures() or use '-mat_freeintermediatedatastructures'"); 1609 } 1610 1611 ierr = MatZeroEntries(C);CHKERRQ(ierr); 1612 1613 if (ptap->reuse == MAT_REUSE_MATRIX) { 1614 /* These matrices are obtained in MatTransposeMatMultSymbolic() */ 1615 /* 1) get R = Pd^T, Ro = Po^T */ 1616 /*----------------------------*/ 1617 ierr = MatTranspose_SeqAIJ(p->A,MAT_REUSE_MATRIX,&ptap->Rd);CHKERRQ(ierr); 1618 ierr = MatTranspose_SeqAIJ(p->B,MAT_REUSE_MATRIX,&ptap->Ro);CHKERRQ(ierr); 1619 1620 /* 2) compute numeric A_loc */ 1621 /*--------------------------*/ 1622 ierr = MatMPIAIJGetLocalMat(A,MAT_REUSE_MATRIX,&ptap->A_loc);CHKERRQ(ierr); 1623 } 1624 1625 /* 3) C_loc = Rd*A_loc, C_oth = Ro*A_loc */ 1626 A_loc = ptap->A_loc; 1627 ierr = ((ptap->C_loc)->ops->matmultnumeric)(ptap->Rd,A_loc,ptap->C_loc);CHKERRQ(ierr); 1628 ierr = ((ptap->C_oth)->ops->matmultnumeric)(ptap->Ro,A_loc,ptap->C_oth);CHKERRQ(ierr); 1629 C_loc = ptap->C_loc; 1630 C_oth = ptap->C_oth; 1631 1632 /* add C_loc and Co to to C */ 1633 ierr = MatGetOwnershipRange(C,&rstart,&rend);CHKERRQ(ierr); 1634 1635 /* C_loc -> C */ 1636 cm = C_loc->rmap->N; 1637 c_seq = (Mat_SeqAIJ*)C_loc->data; 1638 cols = c_seq->j; 1639 vals = c_seq->a; 1640 for (i=0; i<cm; i++) { 1641 ncols = c_seq->i[i+1] - c_seq->i[i]; 1642 row = rstart + i; 1643 ierr = MatSetValues(C,1,&row,ncols,cols,vals,ADD_VALUES);CHKERRQ(ierr); 1644 cols += ncols; vals += ncols; 1645 } 1646 1647 /* Co -> C, off-processor part */ 1648 cm = C_oth->rmap->N; 1649 c_seq = (Mat_SeqAIJ*)C_oth->data; 1650 cols = c_seq->j; 1651 vals = c_seq->a; 1652 for (i=0; i<cm; i++) { 1653 ncols = c_seq->i[i+1] - c_seq->i[i]; 1654 row = p->garray[i]; 1655 ierr = MatSetValues(C,1,&row,ncols,cols,vals,ADD_VALUES);CHKERRQ(ierr); 1656 cols += ncols; vals += ncols; 1657 } 1658 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1659 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1660 1661 ptap->reuse = MAT_REUSE_MATRIX; 1662 1663 /* supporting struct ptap consumes almost same amount of memory as C=PtAP, release it if C will not be updated by A and P */ 1664 if (ptap->freestruct) { 1665 ierr = MatFreeIntermediateDataStructures(C);CHKERRQ(ierr); 1666 } 1667 PetscFunctionReturn(0); 1668 } 1669 1670 PetscErrorCode MatTransposeMatMultNumeric_MPIAIJ_MPIAIJ(Mat P,Mat A,Mat C) 1671 { 1672 PetscErrorCode ierr; 1673 Mat_Merge_SeqsToMPI *merge; 1674 Mat_MPIAIJ *p =(Mat_MPIAIJ*)P->data,*c=(Mat_MPIAIJ*)C->data; 1675 Mat_SeqAIJ *pd=(Mat_SeqAIJ*)(p->A)->data,*po=(Mat_SeqAIJ*)(p->B)->data; 1676 Mat_APMPI *ptap; 1677 PetscInt *adj; 1678 PetscInt i,j,k,anz,pnz,row,*cj,nexta; 1679 MatScalar *ada,*ca,valtmp; 1680 PetscInt am =A->rmap->n,cm=C->rmap->n,pon=(p->B)->cmap->n; 1681 MPI_Comm comm; 1682 PetscMPIInt size,rank,taga,*len_s; 1683 PetscInt *owners,proc,nrows,**buf_ri_k,**nextrow,**nextci; 1684 PetscInt **buf_ri,**buf_rj; 1685 PetscInt cnz=0,*bj_i,*bi,*bj,bnz,nextcj; /* bi,bj,ba: local array of C(mpi mat) */ 1686 MPI_Request *s_waits,*r_waits; 1687 MPI_Status *status; 1688 MatScalar **abuf_r,*ba_i,*pA,*coa,*ba; 1689 PetscInt *ai,*aj,*coi,*coj,*poJ,*pdJ; 1690 Mat A_loc; 1691 Mat_SeqAIJ *a_loc; 1692 1693 PetscFunctionBegin; 1694 ierr = PetscObjectGetComm((PetscObject)C,&comm);CHKERRQ(ierr); 1695 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 1696 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 1697 1698 ptap = c->ap; 1699 if (!ptap->A_loc) SETERRQ(comm,PETSC_ERR_ARG_WRONGSTATE,"PtA cannot be reused. Do not call MatFreeIntermediateDataStructures() or use '-mat_freeintermediatedatastructures'"); 1700 merge = ptap->merge; 1701 1702 /* 2) compute numeric C_seq = P_loc^T*A_loc */ 1703 /*------------------------------------------*/ 1704 /* get data from symbolic products */ 1705 coi = merge->coi; coj = merge->coj; 1706 ierr = PetscCalloc1(coi[pon]+1,&coa);CHKERRQ(ierr); 1707 bi = merge->bi; bj = merge->bj; 1708 owners = merge->rowmap->range; 1709 ierr = PetscCalloc1(bi[cm]+1,&ba);CHKERRQ(ierr); 1710 1711 /* get A_loc by taking all local rows of A */ 1712 A_loc = ptap->A_loc; 1713 ierr = MatMPIAIJGetLocalMat(A,MAT_REUSE_MATRIX,&A_loc);CHKERRQ(ierr); 1714 a_loc = (Mat_SeqAIJ*)(A_loc)->data; 1715 ai = a_loc->i; 1716 aj = a_loc->j; 1717 1718 for (i=0; i<am; i++) { 1719 anz = ai[i+1] - ai[i]; 1720 adj = aj + ai[i]; 1721 ada = a_loc->a + ai[i]; 1722 1723 /* 2-b) Compute Cseq = P_loc[i,:]^T*A[i,:] using outer product */ 1724 /*-------------------------------------------------------------*/ 1725 /* put the value into Co=(p->B)^T*A (off-diagonal part, send to others) */ 1726 pnz = po->i[i+1] - po->i[i]; 1727 poJ = po->j + po->i[i]; 1728 pA = po->a + po->i[i]; 1729 for (j=0; j<pnz; j++) { 1730 row = poJ[j]; 1731 cj = coj + coi[row]; 1732 ca = coa + coi[row]; 1733 /* perform sparse axpy */ 1734 nexta = 0; 1735 valtmp = pA[j]; 1736 for (k=0; nexta<anz; k++) { 1737 if (cj[k] == adj[nexta]) { 1738 ca[k] += valtmp*ada[nexta]; 1739 nexta++; 1740 } 1741 } 1742 ierr = PetscLogFlops(2.0*anz);CHKERRQ(ierr); 1743 } 1744 1745 /* put the value into Cd (diagonal part) */ 1746 pnz = pd->i[i+1] - pd->i[i]; 1747 pdJ = pd->j + pd->i[i]; 1748 pA = pd->a + pd->i[i]; 1749 for (j=0; j<pnz; j++) { 1750 row = pdJ[j]; 1751 cj = bj + bi[row]; 1752 ca = ba + bi[row]; 1753 /* perform sparse axpy */ 1754 nexta = 0; 1755 valtmp = pA[j]; 1756 for (k=0; nexta<anz; k++) { 1757 if (cj[k] == adj[nexta]) { 1758 ca[k] += valtmp*ada[nexta]; 1759 nexta++; 1760 } 1761 } 1762 ierr = PetscLogFlops(2.0*anz);CHKERRQ(ierr); 1763 } 1764 } 1765 1766 /* 3) send and recv matrix values coa */ 1767 /*------------------------------------*/ 1768 buf_ri = merge->buf_ri; 1769 buf_rj = merge->buf_rj; 1770 len_s = merge->len_s; 1771 ierr = PetscCommGetNewTag(comm,&taga);CHKERRQ(ierr); 1772 ierr = PetscPostIrecvScalar(comm,taga,merge->nrecv,merge->id_r,merge->len_r,&abuf_r,&r_waits);CHKERRQ(ierr); 1773 1774 ierr = PetscMalloc2(merge->nsend+1,&s_waits,size,&status);CHKERRQ(ierr); 1775 for (proc=0,k=0; proc<size; proc++) { 1776 if (!len_s[proc]) continue; 1777 i = merge->owners_co[proc]; 1778 ierr = MPI_Isend(coa+coi[i],len_s[proc],MPIU_MATSCALAR,proc,taga,comm,s_waits+k);CHKERRQ(ierr); 1779 k++; 1780 } 1781 if (merge->nrecv) {ierr = MPI_Waitall(merge->nrecv,r_waits,status);CHKERRQ(ierr);} 1782 if (merge->nsend) {ierr = MPI_Waitall(merge->nsend,s_waits,status);CHKERRQ(ierr);} 1783 1784 ierr = PetscFree2(s_waits,status);CHKERRQ(ierr); 1785 ierr = PetscFree(r_waits);CHKERRQ(ierr); 1786 ierr = PetscFree(coa);CHKERRQ(ierr); 1787 1788 /* 4) insert local Cseq and received values into Cmpi */ 1789 /*----------------------------------------------------*/ 1790 ierr = PetscMalloc3(merge->nrecv,&buf_ri_k,merge->nrecv,&nextrow,merge->nrecv,&nextci);CHKERRQ(ierr); 1791 for (k=0; k<merge->nrecv; k++) { 1792 buf_ri_k[k] = buf_ri[k]; /* beginning of k-th recved i-structure */ 1793 nrows = *(buf_ri_k[k]); 1794 nextrow[k] = buf_ri_k[k]+1; /* next row number of k-th recved i-structure */ 1795 nextci[k] = buf_ri_k[k] + (nrows + 1); /* poins to the next i-structure of k-th recved i-structure */ 1796 } 1797 1798 for (i=0; i<cm; i++) { 1799 row = owners[rank] + i; /* global row index of C_seq */ 1800 bj_i = bj + bi[i]; /* col indices of the i-th row of C */ 1801 ba_i = ba + bi[i]; 1802 bnz = bi[i+1] - bi[i]; 1803 /* add received vals into ba */ 1804 for (k=0; k<merge->nrecv; k++) { /* k-th received message */ 1805 /* i-th row */ 1806 if (i == *nextrow[k]) { 1807 cnz = *(nextci[k]+1) - *nextci[k]; 1808 cj = buf_rj[k] + *(nextci[k]); 1809 ca = abuf_r[k] + *(nextci[k]); 1810 nextcj = 0; 1811 for (j=0; nextcj<cnz; j++) { 1812 if (bj_i[j] == cj[nextcj]) { /* bcol == ccol */ 1813 ba_i[j] += ca[nextcj++]; 1814 } 1815 } 1816 nextrow[k]++; nextci[k]++; 1817 ierr = PetscLogFlops(2.0*cnz);CHKERRQ(ierr); 1818 } 1819 } 1820 ierr = MatSetValues(C,1,&row,bnz,bj_i,ba_i,INSERT_VALUES);CHKERRQ(ierr); 1821 } 1822 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1823 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 1824 1825 ierr = PetscFree(ba);CHKERRQ(ierr); 1826 ierr = PetscFree(abuf_r[0]);CHKERRQ(ierr); 1827 ierr = PetscFree(abuf_r);CHKERRQ(ierr); 1828 ierr = PetscFree3(buf_ri_k,nextrow,nextci);CHKERRQ(ierr); 1829 1830 if (ptap->freestruct) { 1831 ierr = MatFreeIntermediateDataStructures(C);CHKERRQ(ierr); 1832 } 1833 PetscFunctionReturn(0); 1834 } 1835 1836 PetscErrorCode MatTransposeMatMultSymbolic_MPIAIJ_MPIAIJ(Mat P,Mat A,PetscReal fill,Mat C) 1837 { 1838 PetscErrorCode ierr; 1839 Mat A_loc,POt,PDt; 1840 Mat_APMPI *ptap; 1841 PetscFreeSpaceList free_space=NULL,current_space=NULL; 1842 Mat_MPIAIJ *p=(Mat_MPIAIJ*)P->data,*a=(Mat_MPIAIJ*)A->data,*c; 1843 PetscInt *pdti,*pdtj,*poti,*potj,*ptJ; 1844 PetscInt nnz; 1845 PetscInt *lnk,*owners_co,*coi,*coj,i,k,pnz,row; 1846 PetscInt am =A->rmap->n,pn=P->cmap->n; 1847 MPI_Comm comm; 1848 PetscMPIInt size,rank,tagi,tagj,*len_si,*len_s,*len_ri; 1849 PetscInt **buf_rj,**buf_ri,**buf_ri_k; 1850 PetscInt len,proc,*dnz,*onz,*owners; 1851 PetscInt nzi,*bi,*bj; 1852 PetscInt nrows,*buf_s,*buf_si,*buf_si_i,**nextrow,**nextci; 1853 MPI_Request *swaits,*rwaits; 1854 MPI_Status *sstatus,rstatus; 1855 Mat_Merge_SeqsToMPI *merge; 1856 PetscInt *ai,*aj,*Jptr,anz,*prmap=p->garray,pon,nspacedouble=0,j; 1857 PetscReal afill =1.0,afill_tmp; 1858 PetscInt rstart = P->cmap->rstart,rmax,aN=A->cmap->N,Armax; 1859 Mat_SeqAIJ *a_loc,*pdt,*pot; 1860 PetscTable ta; 1861 MatType mtype; 1862 1863 PetscFunctionBegin; 1864 ierr = PetscObjectGetComm((PetscObject)A,&comm);CHKERRQ(ierr); 1865 /* check if matrix local sizes are compatible */ 1866 if (A->rmap->rstart != P->rmap->rstart || A->rmap->rend != P->rmap->rend) SETERRQ4(comm,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, A (%D, %D) != P (%D,%D)",A->rmap->rstart,A->rmap->rend,P->rmap->rstart,P->rmap->rend); 1867 1868 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 1869 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 1870 1871 /* create struct Mat_APMPI and attached it to C later */ 1872 ierr = PetscNew(&ptap);CHKERRQ(ierr); 1873 1874 /* get A_loc by taking all local rows of A */ 1875 ierr = MatMPIAIJGetLocalMat(A,MAT_INITIAL_MATRIX,&A_loc);CHKERRQ(ierr); 1876 1877 ptap->A_loc = A_loc; 1878 a_loc = (Mat_SeqAIJ*)(A_loc)->data; 1879 ai = a_loc->i; 1880 aj = a_loc->j; 1881 1882 /* determine symbolic Co=(p->B)^T*A - send to others */ 1883 /*----------------------------------------------------*/ 1884 ierr = MatTransposeSymbolic_SeqAIJ(p->A,&PDt);CHKERRQ(ierr); 1885 pdt = (Mat_SeqAIJ*)PDt->data; 1886 pdti = pdt->i; pdtj = pdt->j; 1887 1888 ierr = MatTransposeSymbolic_SeqAIJ(p->B,&POt);CHKERRQ(ierr); 1889 pot = (Mat_SeqAIJ*)POt->data; 1890 poti = pot->i; potj = pot->j; 1891 1892 /* then, compute symbolic Co = (p->B)^T*A */ 1893 pon = (p->B)->cmap->n; /* total num of rows to be sent to other processors 1894 >= (num of nonzero rows of C_seq) - pn */ 1895 ierr = PetscMalloc1(pon+1,&coi);CHKERRQ(ierr); 1896 coi[0] = 0; 1897 1898 /* set initial free space to be fill*(nnz(p->B) + nnz(A)) */ 1899 nnz = PetscRealIntMultTruncate(fill,PetscIntSumTruncate(poti[pon],ai[am])); 1900 ierr = PetscFreeSpaceGet(nnz,&free_space);CHKERRQ(ierr); 1901 current_space = free_space; 1902 1903 /* create and initialize a linked list */ 1904 ierr = PetscTableCreate(A->cmap->n + a->B->cmap->N,aN,&ta);CHKERRQ(ierr); 1905 MatRowMergeMax_SeqAIJ(a_loc,am,ta); 1906 ierr = PetscTableGetCount(ta,&Armax);CHKERRQ(ierr); 1907 1908 ierr = PetscLLCondensedCreate_Scalable(Armax,&lnk);CHKERRQ(ierr); 1909 1910 for (i=0; i<pon; i++) { 1911 pnz = poti[i+1] - poti[i]; 1912 ptJ = potj + poti[i]; 1913 for (j=0; j<pnz; j++) { 1914 row = ptJ[j]; /* row of A_loc == col of Pot */ 1915 anz = ai[row+1] - ai[row]; 1916 Jptr = aj + ai[row]; 1917 /* add non-zero cols of AP into the sorted linked list lnk */ 1918 ierr = PetscLLCondensedAddSorted_Scalable(anz,Jptr,lnk);CHKERRQ(ierr); 1919 } 1920 nnz = lnk[0]; 1921 1922 /* If free space is not available, double the total space in the list */ 1923 if (current_space->local_remaining<nnz) { 1924 ierr = PetscFreeSpaceGet(PetscIntSumTruncate(nnz,current_space->total_array_size),¤t_space);CHKERRQ(ierr); 1925 nspacedouble++; 1926 } 1927 1928 /* Copy data into free space, and zero out denserows */ 1929 ierr = PetscLLCondensedClean_Scalable(nnz,current_space->array,lnk);CHKERRQ(ierr); 1930 1931 current_space->array += nnz; 1932 current_space->local_used += nnz; 1933 current_space->local_remaining -= nnz; 1934 1935 coi[i+1] = coi[i] + nnz; 1936 } 1937 1938 ierr = PetscMalloc1(coi[pon]+1,&coj);CHKERRQ(ierr); 1939 ierr = PetscFreeSpaceContiguous(&free_space,coj);CHKERRQ(ierr); 1940 ierr = PetscLLCondensedDestroy_Scalable(lnk);CHKERRQ(ierr); /* must destroy to get a new one for C */ 1941 1942 afill_tmp = (PetscReal)coi[pon]/(poti[pon] + ai[am]+1); 1943 if (afill_tmp > afill) afill = afill_tmp; 1944 1945 /* send j-array (coj) of Co to other processors */ 1946 /*----------------------------------------------*/ 1947 /* determine row ownership */ 1948 ierr = PetscNew(&merge);CHKERRQ(ierr); 1949 ierr = PetscLayoutCreate(comm,&merge->rowmap);CHKERRQ(ierr); 1950 1951 merge->rowmap->n = pn; 1952 merge->rowmap->bs = 1; 1953 1954 ierr = PetscLayoutSetUp(merge->rowmap);CHKERRQ(ierr); 1955 owners = merge->rowmap->range; 1956 1957 /* determine the number of messages to send, their lengths */ 1958 ierr = PetscCalloc1(size,&len_si);CHKERRQ(ierr); 1959 ierr = PetscCalloc1(size,&merge->len_s);CHKERRQ(ierr); 1960 1961 len_s = merge->len_s; 1962 merge->nsend = 0; 1963 1964 ierr = PetscMalloc1(size+2,&owners_co);CHKERRQ(ierr); 1965 1966 proc = 0; 1967 for (i=0; i<pon; i++) { 1968 while (prmap[i] >= owners[proc+1]) proc++; 1969 len_si[proc]++; /* num of rows in Co to be sent to [proc] */ 1970 len_s[proc] += coi[i+1] - coi[i]; 1971 } 1972 1973 len = 0; /* max length of buf_si[] */ 1974 owners_co[0] = 0; 1975 for (proc=0; proc<size; proc++) { 1976 owners_co[proc+1] = owners_co[proc] + len_si[proc]; 1977 if (len_si[proc]) { 1978 merge->nsend++; 1979 len_si[proc] = 2*(len_si[proc] + 1); 1980 len += len_si[proc]; 1981 } 1982 } 1983 1984 /* determine the number and length of messages to receive for coi and coj */ 1985 ierr = PetscGatherNumberOfMessages(comm,NULL,len_s,&merge->nrecv);CHKERRQ(ierr); 1986 ierr = PetscGatherMessageLengths2(comm,merge->nsend,merge->nrecv,len_s,len_si,&merge->id_r,&merge->len_r,&len_ri);CHKERRQ(ierr); 1987 1988 /* post the Irecv and Isend of coj */ 1989 ierr = PetscCommGetNewTag(comm,&tagj);CHKERRQ(ierr); 1990 ierr = PetscPostIrecvInt(comm,tagj,merge->nrecv,merge->id_r,merge->len_r,&buf_rj,&rwaits);CHKERRQ(ierr); 1991 ierr = PetscMalloc1(merge->nsend+1,&swaits);CHKERRQ(ierr); 1992 for (proc=0, k=0; proc<size; proc++) { 1993 if (!len_s[proc]) continue; 1994 i = owners_co[proc]; 1995 ierr = MPI_Isend(coj+coi[i],len_s[proc],MPIU_INT,proc,tagj,comm,swaits+k);CHKERRQ(ierr); 1996 k++; 1997 } 1998 1999 /* receives and sends of coj are complete */ 2000 ierr = PetscMalloc1(size,&sstatus);CHKERRQ(ierr); 2001 for (i=0; i<merge->nrecv; i++) { 2002 PetscMPIInt icompleted; 2003 ierr = MPI_Waitany(merge->nrecv,rwaits,&icompleted,&rstatus);CHKERRQ(ierr); 2004 } 2005 ierr = PetscFree(rwaits);CHKERRQ(ierr); 2006 if (merge->nsend) {ierr = MPI_Waitall(merge->nsend,swaits,sstatus);CHKERRQ(ierr);} 2007 2008 /* add received column indices into table to update Armax */ 2009 /* Armax can be as large as aN if a P[row,:] is dense, see src/ksp/ksp/tutorials/ex56.c! */ 2010 for (k=0; k<merge->nrecv; k++) {/* k-th received message */ 2011 Jptr = buf_rj[k]; 2012 for (j=0; j<merge->len_r[k]; j++) { 2013 ierr = PetscTableAdd(ta,*(Jptr+j)+1,1,INSERT_VALUES);CHKERRQ(ierr); 2014 } 2015 } 2016 ierr = PetscTableGetCount(ta,&Armax);CHKERRQ(ierr); 2017 /* printf("Armax %d, an %d + Bn %d = %d, aN %d\n",Armax,A->cmap->n,a->B->cmap->N,A->cmap->n+a->B->cmap->N,aN); */ 2018 2019 /* send and recv coi */ 2020 /*-------------------*/ 2021 ierr = PetscCommGetNewTag(comm,&tagi);CHKERRQ(ierr); 2022 ierr = PetscPostIrecvInt(comm,tagi,merge->nrecv,merge->id_r,len_ri,&buf_ri,&rwaits);CHKERRQ(ierr); 2023 ierr = PetscMalloc1(len+1,&buf_s);CHKERRQ(ierr); 2024 buf_si = buf_s; /* points to the beginning of k-th msg to be sent */ 2025 for (proc=0,k=0; proc<size; proc++) { 2026 if (!len_s[proc]) continue; 2027 /* form outgoing message for i-structure: 2028 buf_si[0]: nrows to be sent 2029 [1:nrows]: row index (global) 2030 [nrows+1:2*nrows+1]: i-structure index 2031 */ 2032 /*-------------------------------------------*/ 2033 nrows = len_si[proc]/2 - 1; 2034 buf_si_i = buf_si + nrows+1; 2035 buf_si[0] = nrows; 2036 buf_si_i[0] = 0; 2037 nrows = 0; 2038 for (i=owners_co[proc]; i<owners_co[proc+1]; i++) { 2039 nzi = coi[i+1] - coi[i]; 2040 buf_si_i[nrows+1] = buf_si_i[nrows] + nzi; /* i-structure */ 2041 buf_si[nrows+1] = prmap[i] -owners[proc]; /* local row index */ 2042 nrows++; 2043 } 2044 ierr = MPI_Isend(buf_si,len_si[proc],MPIU_INT,proc,tagi,comm,swaits+k);CHKERRQ(ierr); 2045 k++; 2046 buf_si += len_si[proc]; 2047 } 2048 i = merge->nrecv; 2049 while (i--) { 2050 PetscMPIInt icompleted; 2051 ierr = MPI_Waitany(merge->nrecv,rwaits,&icompleted,&rstatus);CHKERRQ(ierr); 2052 } 2053 ierr = PetscFree(rwaits);CHKERRQ(ierr); 2054 if (merge->nsend) {ierr = MPI_Waitall(merge->nsend,swaits,sstatus);CHKERRQ(ierr);} 2055 ierr = PetscFree(len_si);CHKERRQ(ierr); 2056 ierr = PetscFree(len_ri);CHKERRQ(ierr); 2057 ierr = PetscFree(swaits);CHKERRQ(ierr); 2058 ierr = PetscFree(sstatus);CHKERRQ(ierr); 2059 ierr = PetscFree(buf_s);CHKERRQ(ierr); 2060 2061 /* compute the local portion of C (mpi mat) */ 2062 /*------------------------------------------*/ 2063 /* allocate bi array and free space for accumulating nonzero column info */ 2064 ierr = PetscMalloc1(pn+1,&bi);CHKERRQ(ierr); 2065 bi[0] = 0; 2066 2067 /* set initial free space to be fill*(nnz(P) + nnz(AP)) */ 2068 nnz = PetscRealIntMultTruncate(fill,PetscIntSumTruncate(pdti[pn],PetscIntSumTruncate(poti[pon],ai[am]))); 2069 ierr = PetscFreeSpaceGet(nnz,&free_space);CHKERRQ(ierr); 2070 current_space = free_space; 2071 2072 ierr = PetscMalloc3(merge->nrecv,&buf_ri_k,merge->nrecv,&nextrow,merge->nrecv,&nextci);CHKERRQ(ierr); 2073 for (k=0; k<merge->nrecv; k++) { 2074 buf_ri_k[k] = buf_ri[k]; /* beginning of k-th recved i-structure */ 2075 nrows = *buf_ri_k[k]; 2076 nextrow[k] = buf_ri_k[k] + 1; /* next row number of k-th recved i-structure */ 2077 nextci[k] = buf_ri_k[k] + (nrows + 1); /* points to the next i-structure of k-th recieved i-structure */ 2078 } 2079 2080 ierr = PetscLLCondensedCreate_Scalable(Armax,&lnk);CHKERRQ(ierr); 2081 ierr = MatPreallocateInitialize(comm,pn,A->cmap->n,dnz,onz);CHKERRQ(ierr); 2082 rmax = 0; 2083 for (i=0; i<pn; i++) { 2084 /* add pdt[i,:]*AP into lnk */ 2085 pnz = pdti[i+1] - pdti[i]; 2086 ptJ = pdtj + pdti[i]; 2087 for (j=0; j<pnz; j++) { 2088 row = ptJ[j]; /* row of AP == col of Pt */ 2089 anz = ai[row+1] - ai[row]; 2090 Jptr = aj + ai[row]; 2091 /* add non-zero cols of AP into the sorted linked list lnk */ 2092 ierr = PetscLLCondensedAddSorted_Scalable(anz,Jptr,lnk);CHKERRQ(ierr); 2093 } 2094 2095 /* add received col data into lnk */ 2096 for (k=0; k<merge->nrecv; k++) { /* k-th received message */ 2097 if (i == *nextrow[k]) { /* i-th row */ 2098 nzi = *(nextci[k]+1) - *nextci[k]; 2099 Jptr = buf_rj[k] + *nextci[k]; 2100 ierr = PetscLLCondensedAddSorted_Scalable(nzi,Jptr,lnk);CHKERRQ(ierr); 2101 nextrow[k]++; nextci[k]++; 2102 } 2103 } 2104 nnz = lnk[0]; 2105 2106 /* if free space is not available, make more free space */ 2107 if (current_space->local_remaining<nnz) { 2108 ierr = PetscFreeSpaceGet(PetscIntSumTruncate(nnz,current_space->total_array_size),¤t_space);CHKERRQ(ierr); 2109 nspacedouble++; 2110 } 2111 /* copy data into free space, then initialize lnk */ 2112 ierr = PetscLLCondensedClean_Scalable(nnz,current_space->array,lnk);CHKERRQ(ierr); 2113 ierr = MatPreallocateSet(i+owners[rank],nnz,current_space->array,dnz,onz);CHKERRQ(ierr); 2114 2115 current_space->array += nnz; 2116 current_space->local_used += nnz; 2117 current_space->local_remaining -= nnz; 2118 2119 bi[i+1] = bi[i] + nnz; 2120 if (nnz > rmax) rmax = nnz; 2121 } 2122 ierr = PetscFree3(buf_ri_k,nextrow,nextci);CHKERRQ(ierr); 2123 2124 ierr = PetscMalloc1(bi[pn]+1,&bj);CHKERRQ(ierr); 2125 ierr = PetscFreeSpaceContiguous(&free_space,bj);CHKERRQ(ierr); 2126 afill_tmp = (PetscReal)bi[pn]/(pdti[pn] + poti[pon] + ai[am]+1); 2127 if (afill_tmp > afill) afill = afill_tmp; 2128 ierr = PetscLLCondensedDestroy_Scalable(lnk);CHKERRQ(ierr); 2129 ierr = PetscTableDestroy(&ta);CHKERRQ(ierr); 2130 2131 ierr = MatDestroy(&POt);CHKERRQ(ierr); 2132 ierr = MatDestroy(&PDt);CHKERRQ(ierr); 2133 2134 /* create symbolic parallel matrix C - why cannot be assembled in Numeric part */ 2135 /*-------------------------------------------------------------------------------*/ 2136 ierr = MatSetSizes(C,pn,A->cmap->n,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 2137 ierr = MatSetBlockSizes(C,PetscAbs(P->cmap->bs),PetscAbs(A->cmap->bs));CHKERRQ(ierr); 2138 ierr = MatGetType(A,&mtype);CHKERRQ(ierr); 2139 ierr = MatSetType(C,mtype);CHKERRQ(ierr); 2140 ierr = MatMPIAIJSetPreallocation(C,0,dnz,0,onz);CHKERRQ(ierr); 2141 ierr = MatPreallocateFinalize(dnz,onz);CHKERRQ(ierr); 2142 ierr = MatSetBlockSize(C,1);CHKERRQ(ierr); 2143 for (i=0; i<pn; i++) { 2144 row = i + rstart; 2145 nnz = bi[i+1] - bi[i]; 2146 Jptr = bj + bi[i]; 2147 ierr = MatSetValues(C,1,&row,nnz,Jptr,NULL,INSERT_VALUES);CHKERRQ(ierr); 2148 } 2149 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2150 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 2151 merge->bi = bi; 2152 merge->bj = bj; 2153 merge->coi = coi; 2154 merge->coj = coj; 2155 merge->buf_ri = buf_ri; 2156 merge->buf_rj = buf_rj; 2157 merge->owners_co = owners_co; 2158 2159 /* attach the supporting struct to C for reuse */ 2160 c = (Mat_MPIAIJ*)C->data; 2161 2162 c->ap = ptap; 2163 ptap->api = NULL; 2164 ptap->apj = NULL; 2165 ptap->merge = merge; 2166 ptap->apa = NULL; 2167 ptap->destroy = C->ops->destroy; 2168 ptap->duplicate = C->ops->duplicate; 2169 2170 C->ops->mattransposemultnumeric = MatTransposeMatMultNumeric_MPIAIJ_MPIAIJ; 2171 C->ops->destroy = MatDestroy_MPIAIJ_PtAP; 2172 C->ops->freeintermediatedatastructures = MatFreeIntermediateDataStructures_MPIAIJ_AP; 2173 2174 #if defined(PETSC_USE_INFO) 2175 if (bi[pn] != 0) { 2176 ierr = PetscInfo3(C,"Reallocs %D; Fill ratio: given %g needed %g.\n",nspacedouble,(double)fill,(double)afill);CHKERRQ(ierr); 2177 ierr = PetscInfo1(C,"Use MatTransposeMatMult(A,B,MatReuse,%g,&C) for best performance.\n",(double)afill);CHKERRQ(ierr); 2178 } else { 2179 ierr = PetscInfo(C,"Empty matrix product\n");CHKERRQ(ierr); 2180 } 2181 #endif 2182 PetscFunctionReturn(0); 2183 } 2184 2185 /* ---------------------------------------------------------------- */ 2186 static PetscErrorCode MatProductSymbolic_AtB_MPIAIJ_MPIAIJ(Mat C) 2187 { 2188 PetscErrorCode ierr; 2189 Mat_Product *product = C->product; 2190 Mat A=product->A,B=product->B; 2191 PetscReal fill=product->fill; 2192 PetscBool flg; 2193 2194 PetscFunctionBegin; 2195 /* scalable */ 2196 ierr = PetscStrcmp(product->alg,"scalable",&flg);CHKERRQ(ierr); 2197 if (flg) { 2198 ierr = MatTransposeMatMultSymbolic_MPIAIJ_MPIAIJ(A,B,fill,C);CHKERRQ(ierr); 2199 goto next; 2200 } 2201 2202 /* nonscalable */ 2203 ierr = PetscStrcmp(product->alg,"nonscalable",&flg);CHKERRQ(ierr); 2204 if (flg) { 2205 ierr = MatTransposeMatMultSymbolic_MPIAIJ_MPIAIJ_nonscalable(A,B,fill,C);CHKERRQ(ierr); 2206 goto next; 2207 } 2208 2209 /* matmatmult */ 2210 ierr = PetscStrcmp(product->alg,"at*b",&flg);CHKERRQ(ierr); 2211 if (flg) { 2212 Mat At; 2213 Mat_APMPI *ptap; 2214 Mat_MPIAIJ *c; 2215 ierr = MatTranspose(A,MAT_INITIAL_MATRIX,&At);CHKERRQ(ierr); 2216 2217 ierr = MatMatMultSymbolic_MPIAIJ_MPIAIJ(At,B,fill,C);CHKERRQ(ierr); 2218 c = (Mat_MPIAIJ*)C->data; 2219 ptap = c->ap; 2220 if (ptap) { 2221 ptap->Pt = At; 2222 C->ops->freeintermediatedatastructures = MatFreeIntermediateDataStructures_MPIAIJ_AP; 2223 } 2224 C->ops->transposematmultnumeric = MatTransposeMatMultNumeric_MPIAIJ_MPIAIJ_matmatmult; 2225 goto next; 2226 } 2227 2228 SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"MatProduct type is not supported"); 2229 2230 next: 2231 C->ops->productnumeric = MatProductNumeric_AtB; 2232 2233 { 2234 Mat_MPIAIJ *c = (Mat_MPIAIJ*)C->data; 2235 Mat_APMPI *ap = c->ap; 2236 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatFreeIntermediateDataStructures","Mat");CHKERRQ(ierr); 2237 ap->freestruct = PETSC_FALSE; 2238 ierr = PetscOptionsBool("-mat_freeintermediatedatastructures","Free intermediate data structures", "MatFreeIntermediateDataStructures",ap->freestruct,&ap->freestruct, NULL);CHKERRQ(ierr); 2239 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2240 } 2241 PetscFunctionReturn(0); 2242 } 2243 2244 /* ---------------------------------------------------------------- */ 2245 /* Set options for MatMatMultxxx_MPIAIJ_MPIAIJ */ 2246 static PetscErrorCode MatProductSetFromOptions_MPIAIJ_AB(Mat C) 2247 { 2248 PetscErrorCode ierr; 2249 Mat_Product *product = C->product; 2250 Mat A=product->A,B=product->B; 2251 #if defined(PETSC_HAVE_HYPRE) 2252 const char *algTypes[4] = {"scalable","nonscalable","seqmpi","hypre"}; 2253 PetscInt nalg = 4; 2254 #else 2255 const char *algTypes[3] = {"scalable","nonscalable","seqmpi"}; 2256 PetscInt nalg = 3; 2257 #endif 2258 PetscInt alg = 1; /* set nonscalable algorithm as default */ 2259 PetscBool flg; 2260 MPI_Comm comm; 2261 2262 PetscFunctionBegin; 2263 /* Check matrix local sizes */ 2264 ierr = PetscObjectGetComm((PetscObject)C,&comm);CHKERRQ(ierr); 2265 if (A->cmap->rstart != B->rmap->rstart || A->cmap->rend != B->rmap->rend) SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, (%D, %D) != (%D,%D)",A->cmap->rstart,A->cmap->rend,B->rmap->rstart,B->rmap->rend); 2266 2267 /* Set "nonscalable" as default algorithm */ 2268 ierr = PetscStrcmp(C->product->alg,"default",&flg);CHKERRQ(ierr); 2269 if (flg) { 2270 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2271 2272 /* Set "scalable" as default if BN and local nonzeros of A and B are large */ 2273 if (B->cmap->N > 100000) { /* may switch to scalable algorithm as default */ 2274 MatInfo Ainfo,Binfo; 2275 PetscInt nz_local; 2276 PetscBool alg_scalable_loc=PETSC_FALSE,alg_scalable; 2277 2278 ierr = MatGetInfo(A,MAT_LOCAL,&Ainfo);CHKERRQ(ierr); 2279 ierr = MatGetInfo(B,MAT_LOCAL,&Binfo);CHKERRQ(ierr); 2280 nz_local = (PetscInt)(Ainfo.nz_allocated + Binfo.nz_allocated); 2281 2282 if (B->cmap->N > product->fill*nz_local) alg_scalable_loc = PETSC_TRUE; 2283 ierr = MPIU_Allreduce(&alg_scalable_loc,&alg_scalable,1,MPIU_BOOL,MPI_LOR,comm);CHKERRQ(ierr); 2284 2285 if (alg_scalable) { 2286 alg = 0; /* scalable algorithm would 50% slower than nonscalable algorithm */ 2287 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2288 ierr = PetscInfo2(B,"Use scalable algorithm, BN %D, fill*nz_allocated %g\n",B->cmap->N,product->fill*nz_local);CHKERRQ(ierr); 2289 } 2290 } 2291 } 2292 2293 /* Get runtime option */ 2294 if (product->api_user) { 2295 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatMatMult","Mat");CHKERRQ(ierr); 2296 ierr = PetscOptionsEList("-matmatmult_via","Algorithmic approach","MatMatMult",algTypes,nalg,algTypes[alg],&alg,&flg);CHKERRQ(ierr); 2297 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2298 } else { 2299 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatProduct_AB","Mat");CHKERRQ(ierr); 2300 ierr = PetscOptionsEList("-matproduct_ab_via","Algorithmic approach","MatMatMult",algTypes,nalg,algTypes[alg],&alg,&flg);CHKERRQ(ierr); 2301 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2302 } 2303 if (flg) { 2304 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2305 } 2306 2307 C->ops->productsymbolic = MatProductSymbolic_AB_MPIAIJ_MPIAIJ; 2308 PetscFunctionReturn(0); 2309 } 2310 2311 /* Set options for MatTransposeMatMultXXX_MPIAIJ_MPIAIJ */ 2312 static PetscErrorCode MatProductSetFromOptions_MPIAIJ_AtB(Mat C) 2313 { 2314 PetscErrorCode ierr; 2315 Mat_Product *product = C->product; 2316 Mat A=product->A,B=product->B; 2317 const char *algTypes[3] = {"scalable","nonscalable","at*b"}; 2318 PetscInt nalg = 3; 2319 PetscInt alg = 1; /* set default algorithm */ 2320 PetscBool flg; 2321 MPI_Comm comm; 2322 2323 PetscFunctionBegin; 2324 /* Check matrix local sizes */ 2325 ierr = PetscObjectGetComm((PetscObject)C,&comm);CHKERRQ(ierr); 2326 if (A->rmap->rstart != B->rmap->rstart || A->rmap->rend != B->rmap->rend) SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, A (%D, %D) != B (%D,%D)",A->rmap->rstart,A->rmap->rend,B->rmap->rstart,B->rmap->rend); 2327 2328 /* Set default algorithm */ 2329 ierr = PetscStrcmp(C->product->alg,"default",&flg);CHKERRQ(ierr); 2330 if (flg) { 2331 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2332 } 2333 2334 /* Set "scalable" as default if BN and local nonzeros of A and B are large */ 2335 if (alg && B->cmap->N > 100000) { /* may switch to scalable algorithm as default */ 2336 MatInfo Ainfo,Binfo; 2337 PetscInt nz_local; 2338 PetscBool alg_scalable_loc=PETSC_FALSE,alg_scalable; 2339 2340 ierr = MatGetInfo(A,MAT_LOCAL,&Ainfo);CHKERRQ(ierr); 2341 ierr = MatGetInfo(B,MAT_LOCAL,&Binfo);CHKERRQ(ierr); 2342 nz_local = (PetscInt)(Ainfo.nz_allocated + Binfo.nz_allocated); 2343 2344 if (B->cmap->N > product->fill*nz_local) alg_scalable_loc = PETSC_TRUE; 2345 ierr = MPIU_Allreduce(&alg_scalable_loc,&alg_scalable,1,MPIU_BOOL,MPI_LOR,comm);CHKERRQ(ierr); 2346 2347 if (alg_scalable) { 2348 alg = 0; /* scalable algorithm would 50% slower than nonscalable algorithm */ 2349 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2350 ierr = PetscInfo2(B,"Use scalable algorithm, BN %D, fill*nz_allocated %g\n",B->cmap->N,product->fill*nz_local);CHKERRQ(ierr); 2351 } 2352 } 2353 2354 /* Get runtime option */ 2355 if (product->api_user) { 2356 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatTransposeMatMult","Mat");CHKERRQ(ierr); 2357 ierr = PetscOptionsEList("-mattransposematmult_via","Algorithmic approach","MatTransposeMatMult",algTypes,nalg,algTypes[alg],&alg,&flg);CHKERRQ(ierr); 2358 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2359 } else { 2360 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatProduct_AtB","Mat");CHKERRQ(ierr); 2361 ierr = PetscOptionsEList("-matproduct_atb_via","Algorithmic approach","MatTransposeMatMult",algTypes,nalg,algTypes[alg],&alg,&flg);CHKERRQ(ierr); 2362 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2363 } 2364 if (flg) { 2365 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2366 } 2367 2368 C->ops->productsymbolic = MatProductSymbolic_AtB_MPIAIJ_MPIAIJ; 2369 PetscFunctionReturn(0); 2370 } 2371 2372 static PetscErrorCode MatProductSetFromOptions_MPIAIJ_PtAP(Mat C) 2373 { 2374 PetscErrorCode ierr; 2375 Mat_Product *product = C->product; 2376 Mat A=product->A,P=product->B; 2377 MPI_Comm comm; 2378 PetscBool flg; 2379 PetscInt alg=1; /* set default algorithm */ 2380 #if !defined(PETSC_HAVE_HYPRE) 2381 const char *algTypes[4] = {"scalable","nonscalable","allatonce","allatonce_merged"}; 2382 PetscInt nalg=4; 2383 #else 2384 const char *algTypes[5] = {"scalable","nonscalable","allatonce","allatonce_merged","hypre"}; 2385 PetscInt nalg=5; 2386 #endif 2387 PetscInt pN=P->cmap->N; 2388 2389 PetscFunctionBegin; 2390 /* Check matrix local sizes */ 2391 ierr = PetscObjectGetComm((PetscObject)C,&comm);CHKERRQ(ierr); 2392 if (A->rmap->rstart != P->rmap->rstart || A->rmap->rend != P->rmap->rend) SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, Arow (%D, %D) != Prow (%D,%D)",A->rmap->rstart,A->rmap->rend,P->rmap->rstart,P->rmap->rend); 2393 if (A->cmap->rstart != P->rmap->rstart || A->cmap->rend != P->rmap->rend) SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, Acol (%D, %D) != Prow (%D,%D)",A->cmap->rstart,A->cmap->rend,P->rmap->rstart,P->rmap->rend); 2394 2395 /* Set "nonscalable" as default algorithm */ 2396 ierr = PetscStrcmp(C->product->alg,"default",&flg);CHKERRQ(ierr); 2397 if (flg) { 2398 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2399 2400 /* Set "scalable" as default if BN and local nonzeros of A and B are large */ 2401 if (pN > 100000) { 2402 MatInfo Ainfo,Pinfo; 2403 PetscInt nz_local; 2404 PetscBool alg_scalable_loc=PETSC_FALSE,alg_scalable; 2405 2406 ierr = MatGetInfo(A,MAT_LOCAL,&Ainfo);CHKERRQ(ierr); 2407 ierr = MatGetInfo(P,MAT_LOCAL,&Pinfo);CHKERRQ(ierr); 2408 nz_local = (PetscInt)(Ainfo.nz_allocated + Pinfo.nz_allocated); 2409 2410 if (pN > product->fill*nz_local) alg_scalable_loc = PETSC_TRUE; 2411 ierr = MPIU_Allreduce(&alg_scalable_loc,&alg_scalable,1,MPIU_BOOL,MPI_LOR,comm);CHKERRQ(ierr); 2412 2413 if (alg_scalable) { 2414 alg = 0; /* scalable algorithm would 50% slower than nonscalable algorithm */ 2415 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2416 } 2417 } 2418 } 2419 2420 /* Get runtime option */ 2421 if (product->api_user) { 2422 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatPtAP","Mat");CHKERRQ(ierr); 2423 ierr = PetscOptionsEList("-matptap_via","Algorithmic approach","MatPtAP",algTypes,nalg,algTypes[alg],&alg,&flg);CHKERRQ(ierr); 2424 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2425 } else { 2426 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatProduct_PtAP","Mat");CHKERRQ(ierr); 2427 ierr = PetscOptionsEList("-matproduct_ptap_via","Algorithmic approach","MatPtAP",algTypes,nalg,algTypes[alg],&alg,&flg);CHKERRQ(ierr); 2428 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2429 } 2430 if (flg) { 2431 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2432 } 2433 2434 C->ops->productsymbolic = MatProductSymbolic_PtAP_MPIAIJ_MPIAIJ; 2435 PetscFunctionReturn(0); 2436 } 2437 2438 static PetscErrorCode MatProductSetFromOptions_MPIAIJ_RARt(Mat C) 2439 { 2440 Mat_Product *product = C->product; 2441 Mat A = product->A,R=product->B; 2442 2443 PetscFunctionBegin; 2444 /* Check matrix local sizes */ 2445 if (A->cmap->n != R->cmap->n || A->rmap->n != R->cmap->n) SETERRQ4(PETSC_COMM_SELF,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, A local (%D, %D), R local (%D,%D)",A->rmap->n,A->rmap->n,R->rmap->n,R->cmap->n); 2446 2447 C->ops->productsymbolic = MatProductSymbolic_RARt_MPIAIJ_MPIAIJ; 2448 PetscFunctionReturn(0); 2449 } 2450 2451 /* 2452 Set options for ABC = A*B*C = A*(B*C); ABC's algorithm must be chosen from AB's algorithm 2453 */ 2454 static PetscErrorCode MatProductSetFromOptions_MPIAIJ_ABC(Mat C) 2455 { 2456 PetscErrorCode ierr; 2457 Mat_Product *product = C->product; 2458 PetscBool flg = PETSC_FALSE; 2459 PetscInt alg = 1; /* default algorithm */ 2460 const char *algTypes[3] = {"scalable","nonscalable","seqmpi"}; 2461 PetscInt nalg = 3; 2462 2463 PetscFunctionBegin; 2464 /* Set default algorithm */ 2465 ierr = PetscStrcmp(C->product->alg,"default",&flg);CHKERRQ(ierr); 2466 if (flg) { 2467 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2468 } 2469 2470 /* Get runtime option */ 2471 if (product->api_user) { 2472 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatMatMatMult","Mat");CHKERRQ(ierr); 2473 ierr = PetscOptionsEList("-matmatmatmult_via","Algorithmic approach","MatMatMatMult",algTypes,nalg,algTypes[alg],&alg,&flg);CHKERRQ(ierr); 2474 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2475 } else { 2476 ierr = PetscOptionsBegin(PetscObjectComm((PetscObject)C),((PetscObject)C)->prefix,"MatProduct_ABC","Mat");CHKERRQ(ierr); 2477 ierr = PetscOptionsEList("-matproduct_abc_via","Algorithmic approach","MatProduct_ABC",algTypes,nalg,algTypes[alg],&alg,&flg);CHKERRQ(ierr); 2478 ierr = PetscOptionsEnd();CHKERRQ(ierr); 2479 } 2480 if (flg) { 2481 ierr = MatProductSetAlgorithm(C,(MatProductAlgorithm)algTypes[alg]);CHKERRQ(ierr); 2482 } 2483 2484 C->ops->matmatmultsymbolic = MatMatMatMultSymbolic_MPIAIJ_MPIAIJ_MPIAIJ; 2485 C->ops->productsymbolic = MatProductSymbolic_ABC; 2486 PetscFunctionReturn(0); 2487 } 2488 2489 PETSC_INTERN PetscErrorCode MatProductSetFromOptions_MPIAIJ(Mat C) 2490 { 2491 PetscErrorCode ierr; 2492 Mat_Product *product = C->product; 2493 2494 PetscFunctionBegin; 2495 switch (product->type) { 2496 case MATPRODUCT_AB: 2497 ierr = MatProductSetFromOptions_MPIAIJ_AB(C);CHKERRQ(ierr); 2498 break; 2499 case MATPRODUCT_AtB: 2500 ierr = MatProductSetFromOptions_MPIAIJ_AtB(C);CHKERRQ(ierr); 2501 break; 2502 case MATPRODUCT_PtAP: 2503 ierr = MatProductSetFromOptions_MPIAIJ_PtAP(C);CHKERRQ(ierr); 2504 break; 2505 case MATPRODUCT_RARt: 2506 ierr = MatProductSetFromOptions_MPIAIJ_RARt(C);CHKERRQ(ierr); 2507 break; 2508 case MATPRODUCT_ABC: 2509 ierr = MatProductSetFromOptions_MPIAIJ_ABC(C);CHKERRQ(ierr); 2510 break; 2511 default: SETERRQ(PetscObjectComm((PetscObject)C),PETSC_ERR_SUP,"MatProduct type is not supported"); 2512 } 2513 PetscFunctionReturn(0); 2514 } 2515