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