1 #define PETSCMAT_DLL 2 3 /* 4 Support for the parallel dense matrix vector multiply 5 */ 6 #include "../src/mat/impls/dense/mpi/mpidense.h" 7 8 #undef __FUNCT__ 9 #define __FUNCT__ "MatSetUpMultiply_MPIDense" 10 PetscErrorCode MatSetUpMultiply_MPIDense(Mat mat) 11 { 12 Mat_MPIDense *mdn = (Mat_MPIDense*)mat->data; 13 PetscErrorCode ierr; 14 IS from,to; 15 Vec gvec; 16 17 PetscFunctionBegin; 18 /* Create local vector that is used to scatter into */ 19 ierr = VecCreateSeq(PETSC_COMM_SELF,mat->cmap->N,&mdn->lvec);CHKERRQ(ierr); 20 21 /* Create temporary index set for building scatter gather */ 22 ierr = ISCreateStride(((PetscObject)mat)->comm,mat->cmap->N,0,1,&from);CHKERRQ(ierr); 23 ierr = ISCreateStride(PETSC_COMM_SELF,mat->cmap->N,0,1,&to);CHKERRQ(ierr); 24 25 /* Create temporary global vector to generate scatter context */ 26 /* n = mdn->cowners[mdn->rank+1] - mdn->cowners[mdn->rank]; */ 27 28 ierr = VecCreateMPIWithArray(((PetscObject)mat)->comm,mdn->nvec,mat->cmap->N,PETSC_NULL,&gvec);CHKERRQ(ierr); 29 30 /* Generate the scatter context */ 31 ierr = VecScatterCreate(gvec,from,mdn->lvec,to,&mdn->Mvctx);CHKERRQ(ierr); 32 ierr = PetscLogObjectParent(mat,mdn->Mvctx);CHKERRQ(ierr); 33 ierr = PetscLogObjectParent(mat,mdn->lvec);CHKERRQ(ierr); 34 ierr = PetscLogObjectParent(mat,from);CHKERRQ(ierr); 35 ierr = PetscLogObjectParent(mat,to);CHKERRQ(ierr); 36 ierr = PetscLogObjectParent(mat,gvec);CHKERRQ(ierr); 37 38 ierr = ISDestroy(to);CHKERRQ(ierr); 39 ierr = ISDestroy(from);CHKERRQ(ierr); 40 ierr = VecDestroy(gvec);CHKERRQ(ierr); 41 PetscFunctionReturn(0); 42 } 43 44 EXTERN PetscErrorCode MatGetSubMatrices_MPIDense_Local(Mat,PetscInt,const IS[],const IS[],MatReuse,Mat*); 45 #undef __FUNCT__ 46 #define __FUNCT__ "MatGetSubMatrices_MPIDense" 47 PetscErrorCode MatGetSubMatrices_MPIDense(Mat C,PetscInt ismax,const IS isrow[],const IS iscol[],MatReuse scall,Mat *submat[]) 48 { 49 PetscErrorCode ierr; 50 PetscInt nmax,nstages_local,nstages,i,pos,max_no; 51 52 PetscFunctionBegin; 53 /* Allocate memory to hold all the submatrices */ 54 if (scall != MAT_REUSE_MATRIX) { 55 ierr = PetscMalloc((ismax+1)*sizeof(Mat),submat);CHKERRQ(ierr); 56 } 57 /* Determine the number of stages through which submatrices are done */ 58 nmax = 20*1000000 / (C->cmap->N * sizeof(PetscInt)); 59 if (!nmax) nmax = 1; 60 nstages_local = ismax/nmax + ((ismax % nmax)?1:0); 61 62 /* Make sure every processor loops through the nstages */ 63 ierr = MPI_Allreduce(&nstages_local,&nstages,1,MPIU_INT,MPI_MAX,((PetscObject)C)->comm);CHKERRQ(ierr); 64 65 66 for (i=0,pos=0; i<nstages; i++) { 67 if (pos+nmax <= ismax) max_no = nmax; 68 else if (pos == ismax) max_no = 0; 69 else max_no = ismax-pos; 70 ierr = MatGetSubMatrices_MPIDense_Local(C,max_no,isrow+pos,iscol+pos,scall,*submat+pos);CHKERRQ(ierr); 71 pos += max_no; 72 } 73 PetscFunctionReturn(0); 74 } 75 /* -------------------------------------------------------------------------*/ 76 #undef __FUNCT__ 77 #define __FUNCT__ "MatGetSubMatrices_MPIDense_Local" 78 PetscErrorCode MatGetSubMatrices_MPIDense_Local(Mat C,PetscInt ismax,const IS isrow[],const IS iscol[],MatReuse scall,Mat *submats) 79 { 80 Mat_MPIDense *c = (Mat_MPIDense*)C->data; 81 Mat A = c->A; 82 Mat_SeqDense *a = (Mat_SeqDense*)A->data,*mat; 83 PetscErrorCode ierr; 84 PetscMPIInt rank,size,tag0,tag1,idex,end,i; 85 PetscInt N = C->cmap->N,rstart = C->rmap->rstart,count; 86 const PetscInt **irow,**icol,*irow_i; 87 PetscInt *nrow,*ncol,*w1,*w3,*w4,*rtable,start; 88 PetscInt **sbuf1,m,j,k,l,ct1,**rbuf1,row,proc; 89 PetscInt nrqs,msz,**ptr,*ctr,*pa,*tmp,bsz,nrqr; 90 PetscInt is_no,jmax,**rmap,*rmap_i; 91 PetscInt ctr_j,*sbuf1_j,*rbuf1_i; 92 MPI_Request *s_waits1,*r_waits1,*s_waits2,*r_waits2; 93 MPI_Status *r_status1,*r_status2,*s_status1,*s_status2; 94 MPI_Comm comm; 95 PetscScalar **rbuf2,**sbuf2; 96 PetscTruth sorted; 97 98 PetscFunctionBegin; 99 comm = ((PetscObject)C)->comm; 100 tag0 = ((PetscObject)C)->tag; 101 size = c->size; 102 rank = c->rank; 103 m = C->rmap->N; 104 105 /* Get some new tags to keep the communication clean */ 106 ierr = PetscObjectGetNewTag((PetscObject)C,&tag1);CHKERRQ(ierr); 107 108 /* Check if the col indices are sorted */ 109 for (i=0; i<ismax; i++) { 110 ierr = ISSorted(isrow[i],&sorted);CHKERRQ(ierr); 111 if (!sorted) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"ISrow is not sorted"); 112 ierr = ISSorted(iscol[i],&sorted);CHKERRQ(ierr); 113 if (!sorted) SETERRQ(PETSC_ERR_ARG_WRONGSTATE,"IScol is not sorted"); 114 } 115 116 ierr = PetscMalloc(2*ismax*(sizeof(PetscInt*)+sizeof(PetscInt)) + (m+1)*sizeof(PetscInt),&irow);CHKERRQ(ierr); 117 icol = irow + ismax; 118 nrow = (PetscInt*)(icol + ismax); 119 ncol = nrow + ismax; 120 rtable = ncol + ismax; 121 122 for (i=0; i<ismax; i++) { 123 ierr = ISGetIndices(isrow[i],&irow[i]);CHKERRQ(ierr); 124 ierr = ISGetIndices(iscol[i],&icol[i]);CHKERRQ(ierr); 125 ierr = ISGetLocalSize(isrow[i],&nrow[i]);CHKERRQ(ierr); 126 ierr = ISGetLocalSize(iscol[i],&ncol[i]);CHKERRQ(ierr); 127 } 128 129 /* Create hash table for the mapping :row -> proc*/ 130 for (i=0,j=0; i<size; i++) { 131 jmax = C->rmap->range[i+1]; 132 for (; j<jmax; j++) { 133 rtable[j] = i; 134 } 135 } 136 137 /* evaluate communication - mesg to who,length of mesg, and buffer space 138 required. Based on this, buffers are allocated, and data copied into them*/ 139 ierr = PetscMalloc(size*4*sizeof(PetscInt),&w1);CHKERRQ(ierr); /* mesg size */ 140 w3 = w1 + 2*size; /* no of IS that needs to be sent to proc i */ 141 w4 = w3 + size; /* temp work space used in determining w1, w3 */ 142 ierr = PetscMemzero(w1,size*3*sizeof(PetscInt));CHKERRQ(ierr); /* initialize work vector*/ 143 for (i=0; i<ismax; i++) { 144 ierr = PetscMemzero(w4,size*sizeof(PetscInt));CHKERRQ(ierr); /* initialize work vector*/ 145 jmax = nrow[i]; 146 irow_i = irow[i]; 147 for (j=0; j<jmax; j++) { 148 row = irow_i[j]; 149 proc = rtable[row]; 150 w4[proc]++; 151 } 152 for (j=0; j<size; j++) { 153 if (w4[j]) { w1[2*j] += w4[j]; w3[j]++;} 154 } 155 } 156 157 nrqs = 0; /* no of outgoing messages */ 158 msz = 0; /* total mesg length (for all procs) */ 159 w1[2*rank] = 0; /* no mesg sent to self */ 160 w3[rank] = 0; 161 for (i=0; i<size; i++) { 162 if (w1[2*i]) { w1[2*i+1] = 1; nrqs++;} /* there exists a message to proc i */ 163 } 164 ierr = PetscMalloc((nrqs+1)*sizeof(PetscInt),&pa);CHKERRQ(ierr); /*(proc -array)*/ 165 for (i=0,j=0; i<size; i++) { 166 if (w1[2*i]) { pa[j] = i; j++; } 167 } 168 169 /* Each message would have a header = 1 + 2*(no of IS) + data */ 170 for (i=0; i<nrqs; i++) { 171 j = pa[i]; 172 w1[2*j] += w1[2*j+1] + 2* w3[j]; 173 msz += w1[2*j]; 174 } 175 /* Do a global reduction to determine how many messages to expect*/ 176 ierr = PetscMaxSum(comm,w1,&bsz,&nrqr);CHKERRQ(ierr); 177 178 /* Allocate memory for recv buffers . Prob none if nrqr = 0 ???? */ 179 ierr = PetscMalloc((nrqr+1)*sizeof(PetscInt*) + nrqr*bsz*sizeof(PetscInt),&rbuf1);CHKERRQ(ierr); 180 rbuf1[0] = (PetscInt*)(rbuf1 + nrqr); 181 for (i=1; i<nrqr; ++i) rbuf1[i] = rbuf1[i-1] + bsz; 182 183 /* Post the receives */ 184 ierr = PetscMalloc((nrqr+1)*sizeof(MPI_Request),&r_waits1);CHKERRQ(ierr); 185 for (i=0; i<nrqr; ++i) { 186 ierr = MPI_Irecv(rbuf1[i],bsz,MPIU_INT,MPI_ANY_SOURCE,tag0,comm,r_waits1+i);CHKERRQ(ierr); 187 } 188 189 /* Allocate Memory for outgoing messages */ 190 ierr = PetscMalloc(2*size*sizeof(PetscInt*) + 2*msz*sizeof(PetscInt)+ size*sizeof(PetscInt),&sbuf1);CHKERRQ(ierr); 191 ptr = sbuf1 + size; /* Pointers to the data in outgoing buffers */ 192 ierr = PetscMemzero(sbuf1,2*size*sizeof(PetscInt*));CHKERRQ(ierr); 193 /* allocate memory for outgoing data + buf to receive the first reply */ 194 tmp = (PetscInt*)(ptr + size); 195 ctr = tmp + 2*msz; 196 197 { 198 PetscInt *iptr = tmp,ict = 0; 199 for (i=0; i<nrqs; i++) { 200 j = pa[i]; 201 iptr += ict; 202 sbuf1[j] = iptr; 203 ict = w1[2*j]; 204 } 205 } 206 207 /* Form the outgoing messages */ 208 /* Initialize the header space */ 209 for (i=0; i<nrqs; i++) { 210 j = pa[i]; 211 sbuf1[j][0] = 0; 212 ierr = PetscMemzero(sbuf1[j]+1,2*w3[j]*sizeof(PetscInt));CHKERRQ(ierr); 213 ptr[j] = sbuf1[j] + 2*w3[j] + 1; 214 } 215 216 /* Parse the isrow and copy data into outbuf */ 217 for (i=0; i<ismax; i++) { 218 ierr = PetscMemzero(ctr,size*sizeof(PetscInt));CHKERRQ(ierr); 219 irow_i = irow[i]; 220 jmax = nrow[i]; 221 for (j=0; j<jmax; j++) { /* parse the indices of each IS */ 222 row = irow_i[j]; 223 proc = rtable[row]; 224 if (proc != rank) { /* copy to the outgoing buf*/ 225 ctr[proc]++; 226 *ptr[proc] = row; 227 ptr[proc]++; 228 } 229 } 230 /* Update the headers for the current IS */ 231 for (j=0; j<size; j++) { /* Can Optimise this loop too */ 232 if ((ctr_j = ctr[j])) { 233 sbuf1_j = sbuf1[j]; 234 k = ++sbuf1_j[0]; 235 sbuf1_j[2*k] = ctr_j; 236 sbuf1_j[2*k-1] = i; 237 } 238 } 239 } 240 241 /* Now post the sends */ 242 ierr = PetscMalloc((nrqs+1)*sizeof(MPI_Request),&s_waits1);CHKERRQ(ierr); 243 for (i=0; i<nrqs; ++i) { 244 j = pa[i]; 245 ierr = MPI_Isend(sbuf1[j],w1[2*j],MPIU_INT,j,tag0,comm,s_waits1+i);CHKERRQ(ierr); 246 } 247 248 /* Post recieves to capture the row_data from other procs */ 249 ierr = PetscMalloc((nrqs+1)*sizeof(MPI_Request),&r_waits2);CHKERRQ(ierr); 250 ierr = PetscMalloc((nrqs+1)*sizeof(PetscScalar*),&rbuf2);CHKERRQ(ierr); 251 for (i=0; i<nrqs; i++) { 252 j = pa[i]; 253 count = (w1[2*j] - (2*sbuf1[j][0] + 1))*N; 254 ierr = PetscMalloc((count+1)*sizeof(PetscScalar),&rbuf2[i]);CHKERRQ(ierr); 255 ierr = MPI_Irecv(rbuf2[i],count,MPIU_SCALAR,j,tag1,comm,r_waits2+i);CHKERRQ(ierr); 256 } 257 258 /* Receive messages(row_nos) and then, pack and send off the rowvalues 259 to the correct processors */ 260 261 ierr = PetscMalloc((nrqr+1)*sizeof(MPI_Request),&s_waits2);CHKERRQ(ierr); 262 ierr = PetscMalloc((nrqr+1)*sizeof(MPI_Status),&r_status1);CHKERRQ(ierr); 263 ierr = PetscMalloc((nrqr+1)*sizeof(PetscScalar*),&sbuf2);CHKERRQ(ierr); 264 265 { 266 PetscScalar *sbuf2_i,*v_start; 267 PetscInt s_proc; 268 for (i=0; i<nrqr; ++i) { 269 ierr = MPI_Waitany(nrqr,r_waits1,&idex,r_status1+i);CHKERRQ(ierr); 270 s_proc = r_status1[i].MPI_SOURCE; /* send processor */ 271 rbuf1_i = rbuf1[idex]; /* Actual message from s_proc */ 272 /* no of rows = end - start; since start is array idex[], 0idex, whel end 273 is length of the buffer - which is 1idex */ 274 start = 2*rbuf1_i[0] + 1; 275 ierr = MPI_Get_count(r_status1+i,MPIU_INT,&end);CHKERRQ(ierr); 276 /* allocate memory sufficinet to hold all the row values */ 277 ierr = PetscMalloc((end-start)*N*sizeof(PetscScalar),&sbuf2[idex]);CHKERRQ(ierr); 278 sbuf2_i = sbuf2[idex]; 279 /* Now pack the data */ 280 for (j=start; j<end; j++) { 281 row = rbuf1_i[j] - rstart; 282 v_start = a->v + row; 283 for (k=0; k<N; k++) { 284 sbuf2_i[0] = v_start[0]; 285 sbuf2_i++; v_start += C->rmap->n; 286 } 287 } 288 /* Now send off the data */ 289 ierr = MPI_Isend(sbuf2[idex],(end-start)*N,MPIU_SCALAR,s_proc,tag1,comm,s_waits2+i);CHKERRQ(ierr); 290 } 291 } 292 /* End Send-Recv of IS + row_numbers */ 293 ierr = PetscFree(r_status1);CHKERRQ(ierr); 294 ierr = PetscFree(r_waits1);CHKERRQ(ierr); 295 ierr = PetscMalloc((nrqs+1)*sizeof(MPI_Status),&s_status1);CHKERRQ(ierr); 296 if (nrqs) {ierr = MPI_Waitall(nrqs,s_waits1,s_status1);CHKERRQ(ierr);} 297 ierr = PetscFree(s_status1);CHKERRQ(ierr); 298 ierr = PetscFree(s_waits1);CHKERRQ(ierr); 299 300 /* Create the submatrices */ 301 if (scall == MAT_REUSE_MATRIX) { 302 for (i=0; i<ismax; i++) { 303 mat = (Mat_SeqDense *)(submats[i]->data); 304 if ((submats[i]->rmap->n != nrow[i]) || (submats[i]->cmap->n != ncol[i])) { 305 SETERRQ(PETSC_ERR_ARG_SIZ,"Cannot reuse matrix. wrong size"); 306 } 307 ierr = PetscMemzero(mat->v,submats[i]->rmap->n*submats[i]->cmap->n*sizeof(PetscScalar));CHKERRQ(ierr); 308 submats[i]->factor = C->factor; 309 } 310 } else { 311 for (i=0; i<ismax; i++) { 312 ierr = MatCreate(PETSC_COMM_SELF,submats+i);CHKERRQ(ierr); 313 ierr = MatSetSizes(submats[i],nrow[i],ncol[i],nrow[i],ncol[i]);CHKERRQ(ierr); 314 ierr = MatSetType(submats[i],((PetscObject)A)->type_name);CHKERRQ(ierr); 315 ierr = MatSeqDenseSetPreallocation(submats[i],PETSC_NULL);CHKERRQ(ierr); 316 } 317 } 318 319 /* Assemble the matrices */ 320 { 321 PetscInt col; 322 PetscScalar *imat_v,*mat_v,*imat_vi,*mat_vi; 323 324 for (i=0; i<ismax; i++) { 325 mat = (Mat_SeqDense*)submats[i]->data; 326 mat_v = a->v; 327 imat_v = mat->v; 328 irow_i = irow[i]; 329 m = nrow[i]; 330 for (j=0; j<m; j++) { 331 row = irow_i[j] ; 332 proc = rtable[row]; 333 if (proc == rank) { 334 row = row - rstart; 335 mat_vi = mat_v + row; 336 imat_vi = imat_v + j; 337 for (k=0; k<ncol[i]; k++) { 338 col = icol[i][k]; 339 imat_vi[k*m] = mat_vi[col*C->rmap->n]; 340 } 341 } 342 } 343 } 344 } 345 346 /* Create row map-> This maps c->row to submat->row for each submat*/ 347 /* this is a very expensive operation wrt memory usage */ 348 ierr = PetscMalloc((1+ismax)*sizeof(PetscInt*)+ ismax*C->rmap->N*sizeof(PetscInt),&rmap);CHKERRQ(ierr); 349 rmap[0] = (PetscInt*)(rmap + ismax); 350 ierr = PetscMemzero(rmap[0],ismax*C->rmap->N*sizeof(PetscInt));CHKERRQ(ierr); 351 for (i=1; i<ismax; i++) { rmap[i] = rmap[i-1] + C->rmap->N;} 352 for (i=0; i<ismax; i++) { 353 rmap_i = rmap[i]; 354 irow_i = irow[i]; 355 jmax = nrow[i]; 356 for (j=0; j<jmax; j++) { 357 rmap_i[irow_i[j]] = j; 358 } 359 } 360 361 /* Now Receive the row_values and assemble the rest of the matrix */ 362 ierr = PetscMalloc((nrqs+1)*sizeof(MPI_Status),&r_status2);CHKERRQ(ierr); 363 364 { 365 PetscInt is_max,tmp1,col,*sbuf1_i,is_sz; 366 PetscScalar *rbuf2_i,*imat_v,*imat_vi; 367 368 for (tmp1=0; tmp1<nrqs; tmp1++) { /* For each message */ 369 ierr = MPI_Waitany(nrqs,r_waits2,&i,r_status2+tmp1);CHKERRQ(ierr); 370 /* Now dig out the corresponding sbuf1, which contains the IS data_structure */ 371 sbuf1_i = sbuf1[pa[i]]; 372 is_max = sbuf1_i[0]; 373 ct1 = 2*is_max+1; 374 rbuf2_i = rbuf2[i]; 375 for (j=1; j<=is_max; j++) { /* For each IS belonging to the message */ 376 is_no = sbuf1_i[2*j-1]; 377 is_sz = sbuf1_i[2*j]; 378 mat = (Mat_SeqDense*)submats[is_no]->data; 379 imat_v = mat->v; 380 rmap_i = rmap[is_no]; 381 m = nrow[is_no]; 382 for (k=0; k<is_sz; k++,rbuf2_i+=N) { /* For each row */ 383 row = sbuf1_i[ct1]; ct1++; 384 row = rmap_i[row]; 385 imat_vi = imat_v + row; 386 for (l=0; l<ncol[is_no]; l++) { /* For each col */ 387 col = icol[is_no][l]; 388 imat_vi[l*m] = rbuf2_i[col]; 389 } 390 } 391 } 392 } 393 } 394 /* End Send-Recv of row_values */ 395 ierr = PetscFree(r_status2);CHKERRQ(ierr); 396 ierr = PetscFree(r_waits2);CHKERRQ(ierr); 397 ierr = PetscMalloc((nrqr+1)*sizeof(MPI_Status),&s_status2);CHKERRQ(ierr); 398 if (nrqr) {ierr = MPI_Waitall(nrqr,s_waits2,s_status2);CHKERRQ(ierr);} 399 ierr = PetscFree(s_status2);CHKERRQ(ierr); 400 ierr = PetscFree(s_waits2);CHKERRQ(ierr); 401 402 /* Restore the indices */ 403 for (i=0; i<ismax; i++) { 404 ierr = ISRestoreIndices(isrow[i],irow+i);CHKERRQ(ierr); 405 ierr = ISRestoreIndices(iscol[i],icol+i);CHKERRQ(ierr); 406 } 407 408 /* Destroy allocated memory */ 409 ierr = PetscFree(irow);CHKERRQ(ierr); 410 ierr = PetscFree(w1);CHKERRQ(ierr); 411 ierr = PetscFree(pa);CHKERRQ(ierr); 412 413 414 for (i=0; i<nrqs; ++i) { 415 ierr = PetscFree(rbuf2[i]);CHKERRQ(ierr); 416 } 417 ierr = PetscFree(rbuf2);CHKERRQ(ierr); 418 ierr = PetscFree(sbuf1);CHKERRQ(ierr); 419 ierr = PetscFree(rbuf1);CHKERRQ(ierr); 420 421 for (i=0; i<nrqr; ++i) { 422 ierr = PetscFree(sbuf2[i]);CHKERRQ(ierr); 423 } 424 425 ierr = PetscFree(sbuf2);CHKERRQ(ierr); 426 ierr = PetscFree(rmap);CHKERRQ(ierr); 427 428 for (i=0; i<ismax; i++) { 429 ierr = MatAssemblyBegin(submats[i],MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 430 ierr = MatAssemblyEnd(submats[i],MAT_FINAL_ASSEMBLY);CHKERRQ(ierr); 431 } 432 433 PetscFunctionReturn(0); 434 } 435 436 #include "petscblaslapack.h" 437 #undef __FUNCT__ 438 #define __FUNCT__ "MatScale_MPIDense" 439 PetscErrorCode MatScale_MPIDense(Mat inA,PetscScalar alpha) 440 { 441 Mat_MPIDense *A = (Mat_MPIDense*)inA->data; 442 Mat_SeqDense *a = (Mat_SeqDense*)A->A->data; 443 PetscScalar oalpha = alpha; 444 PetscErrorCode ierr; 445 PetscBLASInt one = 1,nz = PetscBLASIntCast(inA->rmap->n*inA->cmap->N); 446 447 PetscFunctionBegin; 448 BLASscal_(&nz,&oalpha,a->v,&one); 449 ierr = PetscLogFlops(nz);CHKERRQ(ierr); 450 PetscFunctionReturn(0); 451 } 452