1 2 /* 3 Defines projective product routines where A is a MPIAIJ matrix 4 C = P^T * A * P 5 */ 6 7 #include <../src/mat/impls/aij/seq/aij.h> /*I "petscmat.h" I*/ 8 #include <../src/mat/utils/freespace.h> 9 #include <../src/mat/impls/aij/mpi/mpiaij.h> 10 #include <petscbt.h> 11 12 /* #define PTAP_PROFILE */ 13 14 extern PetscErrorCode MatDestroy_MPIAIJ(Mat); 15 #undef __FUNCT__ 16 #define __FUNCT__ "MatDestroy_MPIAIJ_PtAP" 17 PetscErrorCode MatDestroy_MPIAIJ_PtAP(Mat A) 18 { 19 PetscErrorCode ierr; 20 Mat_MPIAIJ *a=(Mat_MPIAIJ*)A->data; 21 Mat_PtAPMPI *ptap=a->ptap; 22 23 PetscFunctionBegin; 24 if (ptap){ 25 Mat_Merge_SeqsToMPI *merge=ptap->merge; 26 ierr = PetscFree2(ptap->startsj_s,ptap->startsj_r);CHKERRQ(ierr); 27 ierr = PetscFree(ptap->bufa);CHKERRQ(ierr); 28 ierr = MatDestroy(&ptap->P_loc);CHKERRQ(ierr); 29 ierr = MatDestroy(&ptap->P_oth);CHKERRQ(ierr); 30 ierr = MatDestroy(&ptap->A_loc);CHKERRQ(ierr); /* used by MatTransposeMatMult() */ 31 if (ptap->api){ierr = PetscFree(ptap->api);CHKERRQ(ierr);} 32 if (ptap->apj){ierr = PetscFree(ptap->apj);CHKERRQ(ierr);} 33 if (ptap->apa){ierr = PetscFree(ptap->apa);CHKERRQ(ierr);} 34 if (merge) { 35 ierr = PetscFree(merge->id_r);CHKERRQ(ierr); 36 ierr = PetscFree(merge->len_s);CHKERRQ(ierr); 37 ierr = PetscFree(merge->len_r);CHKERRQ(ierr); 38 ierr = PetscFree(merge->bi);CHKERRQ(ierr); 39 ierr = PetscFree(merge->bj);CHKERRQ(ierr); 40 ierr = PetscFree(merge->buf_ri[0]);CHKERRQ(ierr); 41 ierr = PetscFree(merge->buf_ri);CHKERRQ(ierr); 42 ierr = PetscFree(merge->buf_rj[0]);CHKERRQ(ierr); 43 ierr = PetscFree(merge->buf_rj);CHKERRQ(ierr); 44 ierr = PetscFree(merge->coi);CHKERRQ(ierr); 45 ierr = PetscFree(merge->coj);CHKERRQ(ierr); 46 ierr = PetscFree(merge->owners_co);CHKERRQ(ierr); 47 ierr = PetscLayoutDestroy(&merge->rowmap);CHKERRQ(ierr); 48 ierr = merge->destroy(A);CHKERRQ(ierr); 49 ierr = PetscFree(ptap->merge);CHKERRQ(ierr); 50 } 51 ierr = PetscFree(ptap);CHKERRQ(ierr); 52 } 53 PetscFunctionReturn(0); 54 } 55 56 #undef __FUNCT__ 57 #define __FUNCT__ "MatDuplicate_MPIAIJ_MatPtAP" 58 PetscErrorCode MatDuplicate_MPIAIJ_MatPtAP(Mat A, MatDuplicateOption op, Mat *M) 59 { 60 PetscErrorCode ierr; 61 Mat_MPIAIJ *a=(Mat_MPIAIJ*)A->data; 62 Mat_PtAPMPI *ptap = a->ptap; 63 Mat_Merge_SeqsToMPI *merge = ptap->merge; 64 65 PetscFunctionBegin; 66 ierr = (*merge->duplicate)(A,op,M);CHKERRQ(ierr); 67 (*M)->ops->destroy = merge->destroy; 68 (*M)->ops->duplicate = merge->duplicate; 69 PetscFunctionReturn(0); 70 } 71 72 #undef __FUNCT__ 73 #define __FUNCT__ "MatPtAP_MPIAIJ_MPIAIJ" 74 PetscErrorCode MatPtAP_MPIAIJ_MPIAIJ(Mat A,Mat P,MatReuse scall,PetscReal fill,Mat *C) 75 { 76 PetscErrorCode ierr; 77 78 PetscFunctionBegin; 79 if (scall == MAT_INITIAL_MATRIX){ 80 ierr = PetscLogEventBegin(MAT_PtAPSymbolic,A,P,0,0);CHKERRQ(ierr); 81 ierr = MatPtAPSymbolic_MPIAIJ_MPIAIJ(A,P,fill,C);CHKERRQ(ierr); 82 ierr = PetscLogEventEnd(MAT_PtAPSymbolic,A,P,0,0);CHKERRQ(ierr); 83 } 84 ierr = PetscLogEventBegin(MAT_PtAPNumeric,A,P,0,0);CHKERRQ(ierr); 85 ierr = MatPtAPNumeric_MPIAIJ_MPIAIJ(A,P,*C);CHKERRQ(ierr); 86 ierr = PetscLogEventEnd(MAT_PtAPNumeric,A,P,0,0);CHKERRQ(ierr); 87 PetscFunctionReturn(0); 88 } 89 90 #undef __FUNCT__ 91 #define __FUNCT__ "MatPtAPSymbolic_MPIAIJ_MPIAIJ" 92 PetscErrorCode MatPtAPSymbolic_MPIAIJ_MPIAIJ(Mat A,Mat P,PetscReal fill,Mat *C) 93 { 94 PetscErrorCode ierr; 95 Mat Cmpi; 96 Mat_PtAPMPI *ptap; 97 PetscFreeSpaceList free_space=PETSC_NULL,current_space=PETSC_NULL; 98 Mat_MPIAIJ *a=(Mat_MPIAIJ*)A->data,*p=(Mat_MPIAIJ*)P->data,*c; 99 Mat_SeqAIJ *ad=(Mat_SeqAIJ*)(a->A)->data,*ao=(Mat_SeqAIJ*)(a->B)->data; 100 Mat_SeqAIJ *p_loc,*p_oth; 101 PetscInt *pi_loc,*pj_loc,*pi_oth,*pj_oth,*pdti,*pdtj,*poti,*potj,*ptJ; 102 PetscInt *adi=ad->i,*aj,*aoi=ao->i,nnz; 103 PetscInt *lnk,*owners_co,*coi,*coj,i,k,pnz,row; 104 PetscInt am=A->rmap->n,pN=P->cmap->N,pm=P->rmap->n,pn=P->cmap->n; 105 PetscBT lnkbt; 106 MPI_Comm comm=((PetscObject)A)->comm; 107 PetscMPIInt size,rank,tagi,tagj,*len_si,*len_s,*len_ri,icompleted=0; 108 PetscInt **buf_rj,**buf_ri,**buf_ri_k; 109 PetscInt len,proc,*dnz,*onz,*owners; 110 PetscInt nzi,*pti,*ptj; 111 PetscInt nrows,*buf_s,*buf_si,*buf_si_i,**nextrow,**nextci; 112 MPI_Request *swaits,*rwaits; 113 MPI_Status *sstatus,rstatus; 114 Mat_Merge_SeqsToMPI *merge; 115 PetscInt *api,*apj,*Jptr,apnz,*prmap=p->garray,pon,nspacedouble=0,j,ap_rmax=0; 116 PetscReal afill=1.0,afill_tmp; 117 PetscInt rmax; 118 #if defined(PTAP_PROFILE) 119 PetscLogDouble t0,t1,t2,t3; 120 #endif 121 122 PetscFunctionBegin; 123 #if defined(PTAP_PROFILE) 124 ierr = PetscGetTime(&t0);CHKERRQ(ierr); 125 #endif 126 127 /* check if matrix local sizes are compatible */ 128 if (A->rmap->rstart != P->rmap->rstart || A->rmap->rend != P->rmap->rend){ 129 SETERRQ4(comm,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, Arow (%D, %D) != Prow (%D,%D)",A->rmap->rstart,A->rmap->rend,P->rmap->rstart,P->rmap->rend); 130 } 131 if (A->cmap->rstart != P->rmap->rstart || A->cmap->rend != P->rmap->rend){ 132 SETERRQ4(comm,PETSC_ERR_ARG_SIZ,"Matrix local dimensions are incompatible, Acol (%D, %D) != Prow (%D,%D)",A->cmap->rstart,A->cmap->rend,P->rmap->rstart,P->rmap->rend); 133 } 134 135 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 136 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 137 138 /* create struct Mat_PtAPMPI and attached it to C later */ 139 ierr = PetscNew(Mat_PtAPMPI,&ptap);CHKERRQ(ierr); 140 ptap->reuse = MAT_INITIAL_MATRIX; 141 142 /* get P_oth by taking rows of P (= non-zero cols of local A) from other processors */ 143 ierr = MatGetBrowsOfAoCols_MPIAIJ(A,P,MAT_INITIAL_MATRIX,&ptap->startsj_s,&ptap->startsj_r,&ptap->bufa,&ptap->P_oth);CHKERRQ(ierr); 144 145 /* get P_loc by taking all local rows of P */ 146 ierr = MatMPIAIJGetLocalMat(P,MAT_INITIAL_MATRIX,&ptap->P_loc);CHKERRQ(ierr); 147 148 p_loc = (Mat_SeqAIJ*)(ptap->P_loc)->data; 149 p_oth = (Mat_SeqAIJ*)(ptap->P_oth)->data; 150 pi_loc = p_loc->i; pj_loc = p_loc->j; 151 pi_oth = p_oth->i; pj_oth = p_oth->j; 152 #if defined(PTAP_PROFILE) 153 ierr = PetscGetTime(&t1);CHKERRQ(ierr); 154 #endif 155 156 /* first, compute symbolic AP = A_loc*P = A_diag*P_loc + A_off*P_oth */ 157 /*-------------------------------------------------------------------*/ 158 ierr = PetscMalloc((am+1)*sizeof(PetscInt),&api);CHKERRQ(ierr); 159 api[0] = 0; 160 161 /* create and initialize a linked list */ 162 ierr = PetscLLCondensedCreate(pN,pN,&lnk,&lnkbt);CHKERRQ(ierr); 163 164 /* Initial FreeSpace size is fill*(nnz(A) + nnz(P)) -OOM for ex56, np=8k on Intrepid! */ 165 ierr = PetscFreeSpaceGet((PetscInt)(fill*(adi[am]+aoi[am]+pi_loc[pm])),&free_space);CHKERRQ(ierr); 166 current_space = free_space; 167 168 for (i=0; i<am; i++) { 169 /* diagonal portion of A */ 170 nzi = adi[i+1] - adi[i]; 171 aj = ad->j + adi[i]; 172 for (j=0; j<nzi; j++){ 173 row = aj[j]; 174 pnz = pi_loc[row+1] - pi_loc[row]; 175 Jptr = pj_loc + pi_loc[row]; 176 /* add non-zero cols of P into the sorted linked list lnk */ 177 ierr = PetscLLCondensedAddSorted(pnz,Jptr,lnk,lnkbt);CHKERRQ(ierr); 178 } 179 /* off-diagonal portion of A */ 180 nzi = aoi[i+1] - aoi[i]; 181 aj = ao->j + aoi[i]; 182 for (j=0; j<nzi; j++){ 183 row = aj[j]; 184 pnz = pi_oth[row+1] - pi_oth[row]; 185 Jptr = pj_oth + pi_oth[row]; 186 ierr = PetscLLCondensedAddSorted(pnz,Jptr,lnk,lnkbt);CHKERRQ(ierr); 187 } 188 apnz = lnk[0]; 189 api[i+1] = api[i] + apnz; 190 if (ap_rmax < apnz) ap_rmax = apnz; 191 192 /* if free space is not available, double the total space in the list */ 193 if (current_space->local_remaining<apnz) { 194 ierr = PetscFreeSpaceGet(apnz+current_space->total_array_size,¤t_space);CHKERRQ(ierr); 195 nspacedouble++; 196 } 197 198 /* Copy data into free space, then initialize lnk */ 199 ierr = PetscLLCondensedClean(pN,apnz,current_space->array,lnk,lnkbt);CHKERRQ(ierr); 200 current_space->array += apnz; 201 current_space->local_used += apnz; 202 current_space->local_remaining -= apnz; 203 } 204 205 /* Allocate space for apj, initialize apj, and */ 206 /* destroy list of free space and other temporary array(s) */ 207 ierr = PetscMalloc((api[am]+1)*sizeof(PetscInt),&apj);CHKERRQ(ierr); 208 ierr = PetscFreeSpaceContiguous(&free_space,apj);CHKERRQ(ierr); 209 afill_tmp = (PetscReal)api[am]/(adi[am]+aoi[am]+pi_loc[pm]+1); 210 if (afill_tmp > afill) afill = afill_tmp; 211 212 #if defined(PTAP_PROFILE) 213 ierr = PetscGetTime(&t2);CHKERRQ(ierr); 214 #endif 215 216 /* determine symbolic Co=(p->B)^T*AP - send to others */ 217 /*----------------------------------------------------*/ 218 ierr = MatGetSymbolicTranspose_SeqAIJ(p->B,&poti,&potj);CHKERRQ(ierr); 219 220 /* then, compute symbolic Co = (p->B)^T*AP */ 221 pon = (p->B)->cmap->n; /* total num of rows to be sent to other processors 222 >= (num of nonzero rows of C_seq) - pn */ 223 ierr = PetscMalloc((pon+1)*sizeof(PetscInt),&coi);CHKERRQ(ierr); 224 coi[0] = 0; 225 226 /* set initial free space to be fill*(nnz(p->B) + nnz(AP)) */ 227 nnz = fill*(poti[pon] + api[am]); 228 ierr = PetscFreeSpaceGet(nnz,&free_space); 229 current_space = free_space; 230 231 for (i=0; i<pon; i++) { 232 pnz = poti[i+1] - poti[i]; 233 ptJ = potj + poti[i]; 234 for (j=0; j<pnz; j++){ 235 row = ptJ[j]; /* row of AP == col of Pot */ 236 apnz = api[row+1] - api[row]; 237 Jptr = apj + api[row]; 238 /* add non-zero cols of AP into the sorted linked list lnk */ 239 ierr = PetscLLCondensedAddSorted(apnz,Jptr,lnk,lnkbt);CHKERRQ(ierr); 240 } 241 nnz = lnk[0]; 242 243 /* If free space is not available, double the total space in the list */ 244 if (current_space->local_remaining<nnz) { 245 ierr = PetscFreeSpaceGet(nnz+current_space->total_array_size,¤t_space);CHKERRQ(ierr); 246 nspacedouble++; 247 } 248 249 /* Copy data into free space, and zero out denserows */ 250 ierr = PetscLLCondensedClean(pN,nnz,current_space->array,lnk,lnkbt);CHKERRQ(ierr); 251 current_space->array += nnz; 252 current_space->local_used += nnz; 253 current_space->local_remaining -= nnz; 254 coi[i+1] = coi[i] + nnz; 255 } 256 ierr = PetscMalloc((coi[pon]+1)*sizeof(PetscInt),&coj);CHKERRQ(ierr); 257 ierr = PetscFreeSpaceContiguous(&free_space,coj);CHKERRQ(ierr); 258 afill_tmp = (PetscReal)coi[pon]/(poti[pon] + api[am]+1); 259 if (afill_tmp > afill) afill = afill_tmp; 260 ierr = MatRestoreSymbolicTranspose_SeqAIJ(p->B,&poti,&potj);CHKERRQ(ierr); 261 262 /* send j-array (coj) of Co to other processors */ 263 /*----------------------------------------------*/ 264 /* determine row ownership */ 265 ierr = PetscNew(Mat_Merge_SeqsToMPI,&merge);CHKERRQ(ierr); 266 ierr = PetscLayoutCreate(comm,&merge->rowmap);CHKERRQ(ierr); 267 merge->rowmap->n = pn; 268 merge->rowmap->bs = 1; 269 ierr = PetscLayoutSetUp(merge->rowmap);CHKERRQ(ierr); 270 owners = merge->rowmap->range; 271 272 /* determine the number of messages to send, their lengths */ 273 ierr = PetscMalloc2(size,PetscMPIInt,&len_si,size,MPI_Status,&sstatus);CHKERRQ(ierr); 274 ierr = PetscMemzero(len_si,size*sizeof(PetscMPIInt));CHKERRQ(ierr); 275 ierr = PetscMalloc(size*sizeof(PetscMPIInt),&merge->len_s);CHKERRQ(ierr); 276 len_s = merge->len_s; 277 merge->nsend = 0; 278 279 ierr = PetscMalloc((size+2)*sizeof(PetscInt),&owners_co);CHKERRQ(ierr); 280 ierr = PetscMemzero(len_s,size*sizeof(PetscMPIInt));CHKERRQ(ierr); 281 282 proc = 0; 283 for (i=0; i<pon; i++){ 284 while (prmap[i] >= owners[proc+1]) proc++; 285 len_si[proc]++; /* num of rows in Co to be sent to [proc] */ 286 len_s[proc] += coi[i+1] - coi[i]; 287 } 288 289 len = 0; /* max length of buf_si[] */ 290 owners_co[0] = 0; 291 for (proc=0; proc<size; proc++){ 292 owners_co[proc+1] = owners_co[proc] + len_si[proc]; 293 if (len_si[proc]){ 294 merge->nsend++; 295 len_si[proc] = 2*(len_si[proc] + 1); 296 len += len_si[proc]; 297 } 298 } 299 300 /* determine the number and length of messages to receive for coi and coj */ 301 ierr = PetscGatherNumberOfMessages(comm,PETSC_NULL,len_s,&merge->nrecv);CHKERRQ(ierr); 302 ierr = PetscGatherMessageLengths2(comm,merge->nsend,merge->nrecv,len_s,len_si,&merge->id_r,&merge->len_r,&len_ri);CHKERRQ(ierr); 303 304 /* post the Irecv and Isend of coj */ 305 ierr = PetscCommGetNewTag(comm,&tagj);CHKERRQ(ierr); 306 ierr = PetscPostIrecvInt(comm,tagj,merge->nrecv,merge->id_r,merge->len_r,&buf_rj,&rwaits);CHKERRQ(ierr); 307 ierr = PetscMalloc((merge->nsend+1)*sizeof(MPI_Request),&swaits);CHKERRQ(ierr); 308 for (proc=0, k=0; proc<size; proc++){ 309 if (!len_s[proc]) continue; 310 i = owners_co[proc]; 311 ierr = MPI_Isend(coj+coi[i],len_s[proc],MPIU_INT,proc,tagj,comm,swaits+k);CHKERRQ(ierr); 312 k++; 313 } 314 315 /* receives and sends of coj are complete */ 316 for (i=0; i<merge->nrecv; i++){ 317 ierr = MPI_Waitany(merge->nrecv,rwaits,&icompleted,&rstatus);CHKERRQ(ierr); 318 } 319 ierr = PetscFree(rwaits);CHKERRQ(ierr); 320 if (merge->nsend) {ierr = MPI_Waitall(merge->nsend,swaits,sstatus);CHKERRQ(ierr);} 321 322 /* send and recv coi */ 323 /*-------------------*/ 324 ierr = PetscCommGetNewTag(comm,&tagi);CHKERRQ(ierr); 325 ierr = PetscPostIrecvInt(comm,tagi,merge->nrecv,merge->id_r,len_ri,&buf_ri,&rwaits);CHKERRQ(ierr); 326 ierr = PetscMalloc((len+1)*sizeof(PetscInt),&buf_s);CHKERRQ(ierr); 327 buf_si = buf_s; /* points to the beginning of k-th msg to be sent */ 328 for (proc=0,k=0; proc<size; proc++){ 329 if (!len_s[proc]) continue; 330 /* form outgoing message for i-structure: 331 buf_si[0]: nrows to be sent 332 [1:nrows]: row index (global) 333 [nrows+1:2*nrows+1]: i-structure index 334 */ 335 /*-------------------------------------------*/ 336 nrows = len_si[proc]/2 - 1; 337 buf_si_i = buf_si + nrows+1; 338 buf_si[0] = nrows; 339 buf_si_i[0] = 0; 340 nrows = 0; 341 for (i=owners_co[proc]; i<owners_co[proc+1]; i++){ 342 nzi = coi[i+1] - coi[i]; 343 buf_si_i[nrows+1] = buf_si_i[nrows] + nzi; /* i-structure */ 344 buf_si[nrows+1] =prmap[i] -owners[proc]; /* local row index */ 345 nrows++; 346 } 347 ierr = MPI_Isend(buf_si,len_si[proc],MPIU_INT,proc,tagi,comm,swaits+k);CHKERRQ(ierr); 348 k++; 349 buf_si += len_si[proc]; 350 } 351 i = merge->nrecv; 352 while (i--) { 353 ierr = MPI_Waitany(merge->nrecv,rwaits,&icompleted,&rstatus);CHKERRQ(ierr); 354 } 355 ierr = PetscFree(rwaits);CHKERRQ(ierr); 356 if (merge->nsend) {ierr = MPI_Waitall(merge->nsend,swaits,sstatus);CHKERRQ(ierr);} 357 358 ierr = PetscFree2(len_si,sstatus);CHKERRQ(ierr); 359 ierr = PetscFree(len_ri);CHKERRQ(ierr); 360 ierr = PetscFree(swaits);CHKERRQ(ierr); 361 ierr = PetscFree(buf_s);CHKERRQ(ierr); 362 363 /* compute the local portion of C (mpi mat) */ 364 /*------------------------------------------*/ 365 ierr = MatGetSymbolicTranspose_SeqAIJ(p->A,&pdti,&pdtj);CHKERRQ(ierr); 366 367 /* allocate pti array and free space for accumulating nonzero column info */ 368 ierr = PetscMalloc((pn+1)*sizeof(PetscInt),&pti);CHKERRQ(ierr); 369 pti[0] = 0; 370 371 /* set initial free space to be fill*(nnz(P) + nnz(AP)) */ 372 nnz = fill*(pi_loc[pm] + api[am]); 373 ierr = PetscFreeSpaceGet(nnz,&free_space); 374 current_space = free_space; 375 376 ierr = PetscMalloc3(merge->nrecv,PetscInt**,&buf_ri_k,merge->nrecv,PetscInt*,&nextrow,merge->nrecv,PetscInt*,&nextci);CHKERRQ(ierr); 377 for (k=0; k<merge->nrecv; k++){ 378 buf_ri_k[k] = buf_ri[k]; /* beginning of k-th recved i-structure */ 379 nrows = *buf_ri_k[k]; 380 nextrow[k] = buf_ri_k[k] + 1; /* next row number of k-th recved i-structure */ 381 nextci[k] = buf_ri_k[k] + (nrows + 1);/* poins to the next i-structure of k-th recved i-structure */ 382 } 383 ierr = MatPreallocateInitialize(comm,pn,pn,dnz,onz);CHKERRQ(ierr); 384 rmax = 0; 385 for (i=0; i<pn; i++) { 386 /* add pdt[i,:]*AP into lnk */ 387 pnz = pdti[i+1] - pdti[i]; 388 ptJ = pdtj + pdti[i]; 389 for (j=0; j<pnz; j++){ 390 row = ptJ[j]; /* row of AP == col of Pt */ 391 apnz = api[row+1] - api[row]; 392 Jptr = apj + api[row]; 393 /* add non-zero cols of AP into the sorted linked list lnk */ 394 ierr = PetscLLCondensedAddSorted(apnz,Jptr,lnk,lnkbt);CHKERRQ(ierr); 395 } 396 397 /* add received col data into lnk */ 398 for (k=0; k<merge->nrecv; k++){ /* k-th received message */ 399 if (i == *nextrow[k]) { /* i-th row */ 400 nzi = *(nextci[k]+1) - *nextci[k]; 401 Jptr = buf_rj[k] + *nextci[k]; 402 ierr = PetscLLCondensedAddSorted(nzi,Jptr,lnk,lnkbt);CHKERRQ(ierr); 403 nextrow[k]++; nextci[k]++; 404 } 405 } 406 nnz = lnk[0]; 407 408 /* if free space is not available, make more free space */ 409 if (current_space->local_remaining<nnz) { 410 ierr = PetscFreeSpaceGet(nnz+current_space->total_array_size,¤t_space);CHKERRQ(ierr); 411 nspacedouble++; 412 } 413 /* copy data into free space, then initialize lnk */ 414 ierr = PetscLLCondensedClean(pN,nnz,current_space->array,lnk,lnkbt);CHKERRQ(ierr); 415 ierr = MatPreallocateSet(i+owners[rank],nnz,current_space->array,dnz,onz);CHKERRQ(ierr); 416 current_space->array += nnz; 417 current_space->local_used += nnz; 418 current_space->local_remaining -= nnz; 419 pti[i+1] = pti[i] + nnz; 420 if (nnz > rmax) rmax = nnz; 421 } 422 ierr = MatRestoreSymbolicTranspose_SeqAIJ(p->A,&pdti,&pdtj);CHKERRQ(ierr); 423 ierr = PetscFree3(buf_ri_k,nextrow,nextci);CHKERRQ(ierr); 424 425 ierr = PetscMalloc((pti[pn]+1)*sizeof(PetscInt),&ptj);CHKERRQ(ierr); 426 ierr = PetscFreeSpaceContiguous(&free_space,ptj);CHKERRQ(ierr); 427 afill_tmp = (PetscReal)pti[pn]/(pi_loc[pm] + api[am]+1); 428 if (afill_tmp > afill) afill = afill_tmp; 429 ierr = PetscLLDestroy(lnk,lnkbt);CHKERRQ(ierr); 430 431 /* create symbolic parallel matrix Cmpi */ 432 /*--------------------------------------*/ 433 ierr = MatCreate(comm,&Cmpi);CHKERRQ(ierr); 434 ierr = MatSetSizes(Cmpi,pn,pn,PETSC_DETERMINE,PETSC_DETERMINE);CHKERRQ(ierr); 435 ierr = MatSetBlockSizes(Cmpi,P->cmap->bs,P->cmap->bs);CHKERRQ(ierr); 436 ierr = MatSetType(Cmpi,MATMPIAIJ);CHKERRQ(ierr); 437 ierr = MatMPIAIJSetPreallocation(Cmpi,0,dnz,0,onz);CHKERRQ(ierr); 438 ierr = MatPreallocateFinalize(dnz,onz);CHKERRQ(ierr); 439 440 merge->bi = pti; 441 merge->bj = ptj; 442 merge->coi = coi; 443 merge->coj = coj; 444 merge->buf_ri = buf_ri; 445 merge->buf_rj = buf_rj; 446 merge->owners_co = owners_co; 447 merge->destroy = Cmpi->ops->destroy; 448 merge->duplicate = Cmpi->ops->duplicate; 449 450 /* Cmpi is not ready for use - assembly will be done by MatPtAPNumeric() */ 451 Cmpi->assembled = PETSC_FALSE; 452 Cmpi->ops->destroy = MatDestroy_MPIAIJ_PtAP; 453 Cmpi->ops->duplicate = MatDuplicate_MPIAIJ_MatPtAP; 454 455 /* attach the supporting struct to Cmpi for reuse */ 456 c = (Mat_MPIAIJ*)Cmpi->data; 457 c->ptap = ptap; 458 ptap->api = api; 459 ptap->apj = apj; 460 ptap->merge = merge; 461 ptap->rmax = ap_rmax; 462 *C = Cmpi; 463 464 /* flag 'scalable' determines which implementations to be used: 465 0: do dense axpy in MatPtAPNumeric() - fast, but requires storage of a nonscalable dense array apa; 466 1: do sparse axpy in MatPtAPNumeric() - might slow, uses a sparse array apa */ 467 /* set default scalable */ 468 ptap->scalable = PETSC_TRUE; 469 ierr = PetscOptionsGetBool(((PetscObject)Cmpi)->prefix,"-matptap_scalable",&ptap->scalable,PETSC_NULL);CHKERRQ(ierr); 470 if (!ptap->scalable){ /* Do dense axpy */ 471 ierr = PetscMalloc(pN*sizeof(PetscScalar),&ptap->apa);CHKERRQ(ierr); 472 } else { 473 ierr = PetscMalloc((ap_rmax+1)*sizeof(PetscScalar),&ptap->apa);CHKERRQ(ierr); 474 } 475 476 #if defined(PTAP_PROFILE) 477 ierr = PetscGetTime(&t3);CHKERRQ(ierr); 478 if (!rank) PetscPrintf(MPI_COMM_SELF," [%d] PtAPSymbolic %g (Ploc)+ %g (A*P) + %g (Pt*AP) = %g\n",rank,t1-t0,t2-t1,t3-t2,t3-t0);CHKERRQ(ierr); 479 #endif 480 481 #if defined(PETSC_USE_INFO) 482 if (pti[pn] != 0) { 483 ierr = PetscInfo3(Cmpi,"Reallocs %D; Fill ratio: given %G needed %G.\n",nspacedouble,fill,afill);CHKERRQ(ierr); 484 ierr = PetscInfo1(Cmpi,"Use MatPtAP(A,P,MatReuse,%G,&C) for best performance.\n",afill);CHKERRQ(ierr); 485 } else { 486 ierr = PetscInfo(Cmpi,"Empty matrix product\n");CHKERRQ(ierr); 487 } 488 #endif 489 PetscFunctionReturn(0); 490 } 491 492 #undef __FUNCT__ 493 #define __FUNCT__ "MatPtAPNumeric_MPIAIJ_MPIAIJ" 494 PetscErrorCode MatPtAPNumeric_MPIAIJ_MPIAIJ(Mat A,Mat P,Mat C) 495 { 496 PetscErrorCode ierr; 497 Mat_Merge_SeqsToMPI *merge; 498 Mat_MPIAIJ *a=(Mat_MPIAIJ*)A->data,*p=(Mat_MPIAIJ*)P->data,*c=(Mat_MPIAIJ*)C->data; 499 Mat_SeqAIJ *ad=(Mat_SeqAIJ*)(a->A)->data,*ao=(Mat_SeqAIJ*)(a->B)->data; 500 Mat_SeqAIJ *pd=(Mat_SeqAIJ*)(p->A)->data,*po=(Mat_SeqAIJ*)(p->B)->data; 501 Mat_SeqAIJ *p_loc,*p_oth; 502 Mat_PtAPMPI *ptap; 503 PetscInt *adi=ad->i,*aoi=ao->i,*adj,*aoj,*apJ,nextp; 504 PetscInt *pi_loc,*pj_loc,*pi_oth,*pj_oth,*pJ,*pj; 505 PetscInt i,j,k,anz,pnz,apnz,nextap,row,*cj; 506 MatScalar *ada,*aoa,*apa,*pa,*ca,*pa_loc,*pa_oth,valtmp; 507 PetscInt am=A->rmap->n,cm=C->rmap->n,pon=(p->B)->cmap->n; 508 MPI_Comm comm=((PetscObject)C)->comm; 509 PetscMPIInt size,rank,taga,*len_s; 510 PetscInt *owners,proc,nrows,**buf_ri_k,**nextrow,**nextci; 511 PetscInt **buf_ri,**buf_rj; 512 PetscInt cnz=0,*bj_i,*bi,*bj,bnz,nextcj; /* bi,bj,ba: local array of C(mpi mat) */ 513 MPI_Request *s_waits,*r_waits; 514 MPI_Status *status; 515 MatScalar **abuf_r,*ba_i,*pA,*coa,*ba; 516 PetscInt *api,*apj,*coi,*coj; 517 PetscInt *poJ=po->j,*pdJ=pd->j,pcstart=P->cmap->rstart,pcend=P->cmap->rend; 518 PetscBool scalable; 519 520 PetscFunctionBegin; 521 ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); 522 ierr = MPI_Comm_rank(comm,&rank);CHKERRQ(ierr); 523 524 ptap = c->ptap; 525 if (!ptap) SETERRQ(((PetscObject)C)->comm,PETSC_ERR_ARG_INCOMP,"MatPtAP() has not been called to create matrix C yet, cannot use MAT_REUSE_MATRIX"); 526 merge = ptap->merge; 527 apa = ptap->apa; 528 scalable = ptap->scalable; 529 530 /* 1) get P_oth = ptap->P_oth and P_loc = ptap->P_loc */ 531 /*--------------------------------------------------*/ 532 if (ptap->reuse == MAT_INITIAL_MATRIX){ 533 ptap->reuse = MAT_REUSE_MATRIX; 534 } else { /* update numerical values of P_oth and P_loc */ 535 ierr = MatGetBrowsOfAoCols_MPIAIJ(A,P,MAT_REUSE_MATRIX,&ptap->startsj_s,&ptap->startsj_r,&ptap->bufa,&ptap->P_oth);CHKERRQ(ierr); 536 ierr = MatMPIAIJGetLocalMat(P,MAT_REUSE_MATRIX,&ptap->P_loc);CHKERRQ(ierr); 537 } 538 539 /* 2) compute numeric C_seq = P_loc^T*A_loc*P - dominating part */ 540 /*--------------------------------------------------------------*/ 541 /* get data from symbolic products */ 542 p_loc = (Mat_SeqAIJ*)(ptap->P_loc)->data; 543 p_oth = (Mat_SeqAIJ*)(ptap->P_oth)->data; 544 pi_loc=p_loc->i; pj_loc=p_loc->j; pJ=pj_loc; pa_loc=p_loc->a; 545 pi_oth=p_oth->i; pj_oth=p_oth->j; pa_oth=p_oth->a; 546 547 coi = merge->coi; coj = merge->coj; 548 ierr = PetscMalloc((coi[pon]+1)*sizeof(MatScalar),&coa);CHKERRQ(ierr); 549 ierr = PetscMemzero(coa,coi[pon]*sizeof(MatScalar));CHKERRQ(ierr); 550 551 bi = merge->bi; bj = merge->bj; 552 owners = merge->rowmap->range; 553 ierr = PetscMalloc((bi[cm]+1)*sizeof(MatScalar),&ba);CHKERRQ(ierr); 554 ierr = PetscMemzero(ba,bi[cm]*sizeof(MatScalar));CHKERRQ(ierr); 555 556 api = ptap->api; apj = ptap->apj; 557 558 if (!scalable){ /* Do dense axpy */ 559 /*--------------------------------------------------*/ 560 /* apa (length of pN) stores dense row A[i,:]*P - nonscalable! */ 561 ierr = PetscMemzero(apa,P->cmap->N*sizeof(PetscScalar));CHKERRQ(ierr); 562 563 for (i=0; i<am; i++) { 564 /* 2-a) form i-th sparse row of A_loc*P = Ad*P_loc + Ao*P_oth */ 565 /*------------------------------------------------------------*/ 566 apJ = apj + api[i]; 567 568 /* diagonal portion of A */ 569 anz = adi[i+1] - adi[i]; 570 adj = ad->j + adi[i]; 571 ada = ad->a + adi[i]; 572 for (j=0; j<anz; j++) { 573 row = adj[j]; 574 pnz = pi_loc[row+1] - pi_loc[row]; 575 pj = pj_loc + pi_loc[row]; 576 pa = pa_loc + pi_loc[row]; 577 578 /* perform dense axpy */ 579 valtmp = ada[j]; 580 for (k=0; k<pnz; k++){ 581 apa[pj[k]] += valtmp*pa[k]; 582 } 583 ierr = PetscLogFlops(2.0*pnz);CHKERRQ(ierr); 584 } 585 586 /* off-diagonal portion of A */ 587 anz = aoi[i+1] - aoi[i]; 588 aoj = ao->j + aoi[i]; 589 aoa = ao->a + aoi[i]; 590 for (j=0; j<anz; j++) { 591 row = aoj[j]; 592 pnz = pi_oth[row+1] - pi_oth[row]; 593 pj = pj_oth + pi_oth[row]; 594 pa = pa_oth + pi_oth[row]; 595 596 /* perform dense axpy */ 597 valtmp = aoa[j]; 598 for (k=0; k<pnz; k++){ 599 apa[pj[k]] += valtmp*pa[k]; 600 } 601 ierr = PetscLogFlops(2.0*pnz);CHKERRQ(ierr); 602 } 603 604 /* 2-b) Compute Cseq = P_loc[i,:]^T*AP[i,:] using outer product */ 605 /*--------------------------------------------------------------*/ 606 apnz = api[i+1] - api[i]; 607 /* put the value into Co=(p->B)^T*AP (off-diagonal part, send to others) */ 608 pnz = po->i[i+1] - po->i[i]; 609 poJ = po->j + po->i[i]; 610 pA = po->a + po->i[i]; 611 for (j=0; j<pnz; j++){ 612 row = poJ[j]; 613 cnz = coi[row+1] - coi[row]; 614 cj = coj + coi[row]; 615 ca = coa + coi[row]; 616 /* perform dense axpy */ 617 valtmp = pA[j]; 618 for (k=0; k<cnz; k++) { 619 ca[k] += valtmp*apa[cj[k]]; 620 } 621 ierr = PetscLogFlops(2.0*cnz);CHKERRQ(ierr); 622 } 623 624 /* put the value into Cd (diagonal part) */ 625 pnz = pd->i[i+1] - pd->i[i]; 626 pdJ = pd->j + pd->i[i]; 627 pA = pd->a + pd->i[i]; 628 for (j=0; j<pnz; j++){ 629 row = pdJ[j]; 630 cnz = bi[row+1] - bi[row]; 631 cj = bj + bi[row]; 632 ca = ba + bi[row]; 633 /* perform dense axpy */ 634 valtmp = pA[j]; 635 for (k=0; k<cnz; k++) { 636 ca[k] += valtmp*apa[cj[k]]; 637 } 638 ierr = PetscLogFlops(2.0*cnz);CHKERRQ(ierr); 639 } 640 641 /* zero the current row of A*P */ 642 for (k=0; k<apnz; k++) apa[apJ[k]] = 0.0; 643 } 644 } else {/* Perform sparse axpy */ 645 /*----------------------------------------------------*/ 646 /* apa (length ap_rmax) stores sparse row A[i,:]*P */ 647 ierr = PetscMemzero(apa,ptap->rmax*sizeof(MatScalar));CHKERRQ(ierr); 648 649 pA=pa_loc; 650 for (i=0; i<am; i++) { 651 /* form i-th sparse row of A*P */ 652 apnz = api[i+1] - api[i]; 653 apJ = apj + api[i]; 654 /* diagonal portion of A */ 655 anz = adi[i+1] - adi[i]; 656 adj = ad->j + adi[i]; 657 ada = ad->a + adi[i]; 658 for (j=0; j<anz; j++) { 659 row = adj[j]; 660 pnz = pi_loc[row+1] - pi_loc[row]; 661 pj = pj_loc + pi_loc[row]; 662 pa = pa_loc + pi_loc[row]; 663 valtmp = ada[j]; 664 nextp = 0; 665 for (k=0; nextp<pnz; k++) { 666 if (apJ[k] == pj[nextp]) { /* col of AP == col of P */ 667 apa[k] += valtmp*pa[nextp++]; 668 } 669 } 670 ierr = PetscLogFlops(2.0*pnz);CHKERRQ(ierr); 671 } 672 /* off-diagonal portion of A */ 673 anz = aoi[i+1] - aoi[i]; 674 aoj = ao->j + aoi[i]; 675 aoa = ao->a + aoi[i]; 676 for (j=0; j<anz; j++) { 677 row = aoj[j]; 678 pnz = pi_oth[row+1] - pi_oth[row]; 679 pj = pj_oth + pi_oth[row]; 680 pa = pa_oth + pi_oth[row]; 681 valtmp = aoa[j]; 682 nextp = 0; 683 for (k=0; nextp<pnz; k++) { 684 if (apJ[k] == pj[nextp]) { /* col of AP == col of P */ 685 apa[k] += valtmp*pa[nextp++]; 686 } 687 } 688 ierr = PetscLogFlops(2.0*pnz);CHKERRQ(ierr); 689 } 690 691 /* 2-b) Compute Cseq = P_loc[i,:]^T*AP[i,:] using outer product */ 692 /*--------------------------------------------------------------*/ 693 pnz = pi_loc[i+1] - pi_loc[i]; 694 pJ = pj_loc + pi_loc[i]; 695 for (j=0; j<pnz; j++) { 696 nextap = 0; 697 row = pJ[j]; /* global index */ 698 if (row < pcstart || row >=pcend) { /* put the value into Co */ 699 row = *poJ; 700 cj = coj + coi[row]; 701 ca = coa + coi[row]; poJ++; 702 } else { /* put the value into Cd */ 703 row = *pdJ; 704 cj = bj + bi[row]; 705 ca = ba + bi[row]; pdJ++; 706 } 707 valtmp = pA[j]; 708 for (k=0; nextap<apnz; k++) { 709 if (cj[k]==apJ[nextap]) ca[k] += valtmp*apa[nextap++]; 710 } 711 ierr = PetscLogFlops(2.0*apnz);CHKERRQ(ierr); 712 } 713 pA += pnz; 714 /* zero the current row info for A*P */ 715 ierr = PetscMemzero(apa,apnz*sizeof(MatScalar));CHKERRQ(ierr); 716 } 717 } 718 719 /* 3) send and recv matrix values coa */ 720 /*------------------------------------*/ 721 buf_ri = merge->buf_ri; 722 buf_rj = merge->buf_rj; 723 len_s = merge->len_s; 724 ierr = PetscCommGetNewTag(comm,&taga);CHKERRQ(ierr); 725 ierr = PetscPostIrecvScalar(comm,taga,merge->nrecv,merge->id_r,merge->len_r,&abuf_r,&r_waits);CHKERRQ(ierr); 726 727 ierr = PetscMalloc2(merge->nsend+1,MPI_Request,&s_waits,size,MPI_Status,&status);CHKERRQ(ierr); 728 for (proc=0,k=0; proc<size; proc++){ 729 if (!len_s[proc]) continue; 730 i = merge->owners_co[proc]; 731 ierr = MPI_Isend(coa+coi[i],len_s[proc],MPIU_MATSCALAR,proc,taga,comm,s_waits+k);CHKERRQ(ierr); 732 k++; 733 } 734 if (merge->nrecv) {ierr = MPI_Waitall(merge->nrecv,r_waits,status);CHKERRQ(ierr);} 735 if (merge->nsend) {ierr = MPI_Waitall(merge->nsend,s_waits,status);CHKERRQ(ierr);} 736 737 ierr = PetscFree2(s_waits,status);CHKERRQ(ierr); 738 ierr = PetscFree(r_waits);CHKERRQ(ierr); 739 ierr = PetscFree(coa);CHKERRQ(ierr); 740 741 /* 4) insert local Cseq and received values into Cmpi */ 742 /*------------------------------------------------------*/ 743 ierr = PetscMalloc3(merge->nrecv,PetscInt**,&buf_ri_k,merge->nrecv,PetscInt*,&nextrow,merge->nrecv,PetscInt*,&nextci);CHKERRQ(ierr); 744 for (k=0; k<merge->nrecv; k++){ 745 buf_ri_k[k] = buf_ri[k]; /* beginning of k-th recved i-structure */ 746 nrows = *(buf_ri_k[k]); 747 nextrow[k] = buf_ri_k[k]+1; /* next row number of k-th recved i-structure */ 748 nextci[k] = buf_ri_k[k] + (nrows + 1);/* poins to the next i-structure of k-th recved i-structure */ 749 } 750 751 for (i=0; i<cm; i++) { 752 row = owners[rank] + i; /* global row index of C_seq */ 753 bj_i = bj + bi[i]; /* col indices of the i-th row of C */ 754 ba_i = ba + bi[i]; 755 bnz = bi[i+1] - bi[i]; 756 /* add received vals into ba */ 757 for (k=0; k<merge->nrecv; k++){ /* k-th received message */ 758 /* i-th row */ 759 if (i == *nextrow[k]) { 760 cnz = *(nextci[k]+1) - *nextci[k]; 761 cj = buf_rj[k] + *(nextci[k]); 762 ca = abuf_r[k] + *(nextci[k]); 763 nextcj = 0; 764 for (j=0; nextcj<cnz; j++){ 765 if (bj_i[j] == cj[nextcj]){ /* bcol == ccol */ 766 ba_i[j] += ca[nextcj++]; 767 } 768 } 769 nextrow[k]++; nextci[k]++; 770 ierr = PetscLogFlops(2.0*cnz);CHKERRQ(ierr); 771 } 772 } 773 ierr = MatSetValues(C,1,&row,bnz,bj_i,ba_i,INSERT_VALUES);CHKERRQ(ierr); 774 } 775 ierr = MatAssemblyBegin(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 776 ierr = MatAssemblyEnd(C,MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 777 778 ierr = PetscFree(ba);CHKERRQ(ierr); 779 ierr = PetscFree(abuf_r[0]);CHKERRQ(ierr); 780 ierr = PetscFree(abuf_r);CHKERRQ(ierr); 781 ierr = PetscFree3(buf_ri_k,nextrow,nextci);CHKERRQ(ierr); 782 #if defined(PETSC_USE_INFO) 783 if (scalable){ 784 ierr = PetscInfo(C,"Use scalable sparse axpy\n");CHKERRQ(ierr); 785 } else { 786 ierr = PetscInfo(C,"Use non-scalable dense axpy\n");CHKERRQ(ierr); 787 } 788 #endif 789 PetscFunctionReturn(0); 790 } 791