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