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