1 /* 2 Defines the basic matrix operations for the AIJ (compressed row) 3 matrix storage format using the CUSPARSE library, 4 */ 5 #define PETSC_SKIP_SPINLOCK 6 7 #include <petscconf.h> 8 #include <../src/mat/impls/aij/seq/aij.h> /*I "petscmat.h" I*/ 9 #include <../src/mat/impls/sbaij/seq/sbaij.h> 10 #include <../src/vec/vec/impls/dvecimpl.h> 11 #include <petsc/private/vecimpl.h> 12 #undef VecType 13 #include <../src/mat/impls/aij/seq/seqcusparse/cusparsematimpl.h> 14 15 const char *const MatCUSPARSEStorageFormats[] = {"CSR","ELL","HYB","MatCUSPARSEStorageFormat","MAT_CUSPARSE_",0}; 16 17 static PetscErrorCode MatICCFactorSymbolic_SeqAIJCUSPARSE(Mat,Mat,IS,const MatFactorInfo*); 18 static PetscErrorCode MatCholeskyFactorSymbolic_SeqAIJCUSPARSE(Mat,Mat,IS,const MatFactorInfo*); 19 static PetscErrorCode MatCholeskyFactorNumeric_SeqAIJCUSPARSE(Mat,Mat,const MatFactorInfo*); 20 21 static PetscErrorCode MatILUFactorSymbolic_SeqAIJCUSPARSE(Mat,Mat,IS,IS,const MatFactorInfo*); 22 static PetscErrorCode MatLUFactorSymbolic_SeqAIJCUSPARSE(Mat,Mat,IS,IS,const MatFactorInfo*); 23 static PetscErrorCode MatLUFactorNumeric_SeqAIJCUSPARSE(Mat,Mat,const MatFactorInfo*); 24 25 static PetscErrorCode MatSolve_SeqAIJCUSPARSE(Mat,Vec,Vec); 26 static PetscErrorCode MatSolve_SeqAIJCUSPARSE_NaturalOrdering(Mat,Vec,Vec); 27 static PetscErrorCode MatSolveTranspose_SeqAIJCUSPARSE(Mat,Vec,Vec); 28 static PetscErrorCode MatSolveTranspose_SeqAIJCUSPARSE_NaturalOrdering(Mat,Vec,Vec); 29 static PetscErrorCode MatSetFromOptions_SeqAIJCUSPARSE(PetscOptionItems *PetscOptionsObject,Mat); 30 static PetscErrorCode MatMult_SeqAIJCUSPARSE(Mat,Vec,Vec); 31 static PetscErrorCode MatMultAdd_SeqAIJCUSPARSE(Mat,Vec,Vec,Vec); 32 static PetscErrorCode MatMultTranspose_SeqAIJCUSPARSE(Mat,Vec,Vec); 33 static PetscErrorCode MatMultTransposeAdd_SeqAIJCUSPARSE(Mat,Vec,Vec,Vec); 34 35 static PetscErrorCode CsrMatrix_Destroy(CsrMatrix**); 36 static PetscErrorCode MatSeqAIJCUSPARSEMultStruct_Destroy(Mat_SeqAIJCUSPARSETriFactorStruct**); 37 static PetscErrorCode MatSeqAIJCUSPARSEMultStruct_Destroy(Mat_SeqAIJCUSPARSEMultStruct**,MatCUSPARSEStorageFormat); 38 static PetscErrorCode MatSeqAIJCUSPARSETriFactors_Destroy(Mat_SeqAIJCUSPARSETriFactors**); 39 static PetscErrorCode MatSeqAIJCUSPARSE_Destroy(Mat_SeqAIJCUSPARSE**); 40 41 PetscErrorCode MatCUSPARSESetStream(Mat A,const cudaStream_t stream) 42 { 43 cusparseStatus_t stat; 44 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 45 46 PetscFunctionBegin; 47 cusparsestruct->stream = stream; 48 stat = cusparseSetStream(cusparsestruct->handle,cusparsestruct->stream);CHKERRCUDA(stat); 49 PetscFunctionReturn(0); 50 } 51 52 PetscErrorCode MatCUSPARSESetHandle(Mat A,const cusparseHandle_t handle) 53 { 54 cusparseStatus_t stat; 55 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 56 57 PetscFunctionBegin; 58 if (cusparsestruct->handle != handle) { 59 if (cusparsestruct->handle) { 60 stat = cusparseDestroy(cusparsestruct->handle);CHKERRCUDA(stat); 61 } 62 cusparsestruct->handle = handle; 63 } 64 stat = cusparseSetPointerMode(cusparsestruct->handle, CUSPARSE_POINTER_MODE_DEVICE);CHKERRCUDA(stat); 65 PetscFunctionReturn(0); 66 } 67 68 PetscErrorCode MatCUSPARSEClearHandle(Mat A) 69 { 70 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 71 PetscFunctionBegin; 72 if (cusparsestruct->handle) 73 cusparsestruct->handle = 0; 74 PetscFunctionReturn(0); 75 } 76 77 PetscErrorCode MatFactorGetSolverType_seqaij_cusparse(Mat A,MatSolverType *type) 78 { 79 PetscFunctionBegin; 80 *type = MATSOLVERCUSPARSE; 81 PetscFunctionReturn(0); 82 } 83 84 /*MC 85 MATSOLVERCUSPARSE = "cusparse" - A matrix type providing triangular solvers for seq matrices 86 on a single GPU of type, seqaijcusparse, aijcusparse, or seqaijcusp, aijcusp. Currently supported 87 algorithms are ILU(k) and ICC(k). Typically, deeper factorizations (larger k) results in poorer 88 performance in the triangular solves. Full LU, and Cholesky decompositions can be solved through the 89 CUSPARSE triangular solve algorithm. However, the performance can be quite poor and thus these 90 algorithms are not recommended. This class does NOT support direct solver operations. 91 92 Level: beginner 93 94 .seealso: PCFactorSetMatSolverType(), MatSolverType, MatCreateSeqAIJCUSPARSE(), MATAIJCUSPARSE, MatCreateAIJCUSPARSE(), MatCUSPARSESetFormat(), MatCUSPARSEStorageFormat, MatCUSPARSEFormatOperation 95 M*/ 96 97 PETSC_EXTERN PetscErrorCode MatGetFactor_seqaijcusparse_cusparse(Mat A,MatFactorType ftype,Mat *B) 98 { 99 PetscErrorCode ierr; 100 PetscInt n = A->rmap->n; 101 102 PetscFunctionBegin; 103 ierr = MatCreate(PetscObjectComm((PetscObject)A),B);CHKERRQ(ierr); 104 (*B)->factortype = ftype; 105 ierr = MatSetSizes(*B,n,n,n,n);CHKERRQ(ierr); 106 ierr = MatSetType(*B,MATSEQAIJCUSPARSE);CHKERRQ(ierr); 107 108 if (ftype == MAT_FACTOR_LU || ftype == MAT_FACTOR_ILU || ftype == MAT_FACTOR_ILUDT) { 109 ierr = MatSetBlockSizesFromMats(*B,A,A);CHKERRQ(ierr); 110 (*B)->ops->ilufactorsymbolic = MatILUFactorSymbolic_SeqAIJCUSPARSE; 111 (*B)->ops->lufactorsymbolic = MatLUFactorSymbolic_SeqAIJCUSPARSE; 112 } else if (ftype == MAT_FACTOR_CHOLESKY || ftype == MAT_FACTOR_ICC) { 113 (*B)->ops->iccfactorsymbolic = MatICCFactorSymbolic_SeqAIJCUSPARSE; 114 (*B)->ops->choleskyfactorsymbolic = MatCholeskyFactorSymbolic_SeqAIJCUSPARSE; 115 } else SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"Factor type not supported for CUSPARSE Matrix Types"); 116 117 ierr = MatSeqAIJSetPreallocation(*B,MAT_SKIP_ALLOCATION,NULL);CHKERRQ(ierr); 118 ierr = PetscObjectComposeFunction((PetscObject)(*B),"MatFactorGetSolverType_C",MatFactorGetSolverType_seqaij_cusparse);CHKERRQ(ierr); 119 PetscFunctionReturn(0); 120 } 121 122 PETSC_INTERN PetscErrorCode MatCUSPARSESetFormat_SeqAIJCUSPARSE(Mat A,MatCUSPARSEFormatOperation op,MatCUSPARSEStorageFormat format) 123 { 124 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 125 126 PetscFunctionBegin; 127 #if CUDA_VERSION>=4020 128 switch (op) { 129 case MAT_CUSPARSE_MULT: 130 cusparsestruct->format = format; 131 break; 132 case MAT_CUSPARSE_ALL: 133 cusparsestruct->format = format; 134 break; 135 default: 136 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_SUP,"unsupported operation %d for MatCUSPARSEFormatOperation. MAT_CUSPARSE_MULT and MAT_CUSPARSE_ALL are currently supported.",op); 137 } 138 #else 139 if (format==MAT_CUSPARSE_ELL || format==MAT_CUSPARSE_HYB) SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"ELL (Ellpack) and HYB (Hybrid) storage format require CUDA 4.2 or later."); 140 #endif 141 PetscFunctionReturn(0); 142 } 143 144 /*@ 145 MatCUSPARSESetFormat - Sets the storage format of CUSPARSE matrices for a particular 146 operation. Only the MatMult operation can use different GPU storage formats 147 for MPIAIJCUSPARSE matrices. 148 Not Collective 149 150 Input Parameters: 151 + A - Matrix of type SEQAIJCUSPARSE 152 . op - MatCUSPARSEFormatOperation. SEQAIJCUSPARSE matrices support MAT_CUSPARSE_MULT and MAT_CUSPARSE_ALL. MPIAIJCUSPARSE matrices support MAT_CUSPARSE_MULT_DIAG, MAT_CUSPARSE_MULT_OFFDIAG, and MAT_CUSPARSE_ALL. 153 - format - MatCUSPARSEStorageFormat (one of MAT_CUSPARSE_CSR, MAT_CUSPARSE_ELL, MAT_CUSPARSE_HYB. The latter two require CUDA 4.2) 154 155 Output Parameter: 156 157 Level: intermediate 158 159 .seealso: MatCUSPARSEStorageFormat, MatCUSPARSEFormatOperation 160 @*/ 161 PetscErrorCode MatCUSPARSESetFormat(Mat A,MatCUSPARSEFormatOperation op,MatCUSPARSEStorageFormat format) 162 { 163 PetscErrorCode ierr; 164 165 PetscFunctionBegin; 166 PetscValidHeaderSpecific(A, MAT_CLASSID,1); 167 ierr = PetscTryMethod(A, "MatCUSPARSESetFormat_C",(Mat,MatCUSPARSEFormatOperation,MatCUSPARSEStorageFormat),(A,op,format));CHKERRQ(ierr); 168 PetscFunctionReturn(0); 169 } 170 171 static PetscErrorCode MatSetFromOptions_SeqAIJCUSPARSE(PetscOptionItems *PetscOptionsObject,Mat A) 172 { 173 PetscErrorCode ierr; 174 MatCUSPARSEStorageFormat format; 175 PetscBool flg; 176 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 177 178 PetscFunctionBegin; 179 ierr = PetscOptionsHead(PetscOptionsObject,"SeqAIJCUSPARSE options");CHKERRQ(ierr); 180 if (A->factortype==MAT_FACTOR_NONE) { 181 ierr = PetscOptionsEnum("-mat_cusparse_mult_storage_format","sets storage format of (seq)aijcusparse gpu matrices for SpMV", 182 "MatCUSPARSESetFormat",MatCUSPARSEStorageFormats,(PetscEnum)cusparsestruct->format,(PetscEnum*)&format,&flg);CHKERRQ(ierr); 183 if (flg) { 184 ierr = MatCUSPARSESetFormat(A,MAT_CUSPARSE_MULT,format);CHKERRQ(ierr); 185 } 186 } 187 ierr = PetscOptionsEnum("-mat_cusparse_storage_format","sets storage format of (seq)aijcusparse gpu matrices for SpMV and TriSolve", 188 "MatCUSPARSESetFormat",MatCUSPARSEStorageFormats,(PetscEnum)cusparsestruct->format,(PetscEnum*)&format,&flg);CHKERRQ(ierr); 189 if (flg) { 190 ierr = MatCUSPARSESetFormat(A,MAT_CUSPARSE_ALL,format);CHKERRQ(ierr); 191 } 192 ierr = PetscOptionsTail();CHKERRQ(ierr); 193 PetscFunctionReturn(0); 194 195 } 196 197 static PetscErrorCode MatILUFactorSymbolic_SeqAIJCUSPARSE(Mat B,Mat A,IS isrow,IS iscol,const MatFactorInfo *info) 198 { 199 PetscErrorCode ierr; 200 201 PetscFunctionBegin; 202 ierr = MatILUFactorSymbolic_SeqAIJ(B,A,isrow,iscol,info);CHKERRQ(ierr); 203 B->ops->lufactornumeric = MatLUFactorNumeric_SeqAIJCUSPARSE; 204 PetscFunctionReturn(0); 205 } 206 207 static PetscErrorCode MatLUFactorSymbolic_SeqAIJCUSPARSE(Mat B,Mat A,IS isrow,IS iscol,const MatFactorInfo *info) 208 { 209 PetscErrorCode ierr; 210 211 PetscFunctionBegin; 212 ierr = MatLUFactorSymbolic_SeqAIJ(B,A,isrow,iscol,info);CHKERRQ(ierr); 213 B->ops->lufactornumeric = MatLUFactorNumeric_SeqAIJCUSPARSE; 214 PetscFunctionReturn(0); 215 } 216 217 static PetscErrorCode MatICCFactorSymbolic_SeqAIJCUSPARSE(Mat B,Mat A,IS perm,const MatFactorInfo *info) 218 { 219 PetscErrorCode ierr; 220 221 PetscFunctionBegin; 222 ierr = MatICCFactorSymbolic_SeqAIJ(B,A,perm,info);CHKERRQ(ierr); 223 B->ops->choleskyfactornumeric = MatCholeskyFactorNumeric_SeqAIJCUSPARSE; 224 PetscFunctionReturn(0); 225 } 226 227 static PetscErrorCode MatCholeskyFactorSymbolic_SeqAIJCUSPARSE(Mat B,Mat A,IS perm,const MatFactorInfo *info) 228 { 229 PetscErrorCode ierr; 230 231 PetscFunctionBegin; 232 ierr = MatCholeskyFactorSymbolic_SeqAIJ(B,A,perm,info);CHKERRQ(ierr); 233 B->ops->choleskyfactornumeric = MatCholeskyFactorNumeric_SeqAIJCUSPARSE; 234 PetscFunctionReturn(0); 235 } 236 237 static PetscErrorCode MatSeqAIJCUSPARSEBuildILULowerTriMatrix(Mat A) 238 { 239 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 240 PetscInt n = A->rmap->n; 241 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 242 Mat_SeqAIJCUSPARSETriFactorStruct *loTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtr; 243 cusparseStatus_t stat; 244 const PetscInt *ai = a->i,*aj = a->j,*vi; 245 const MatScalar *aa = a->a,*v; 246 PetscInt *AiLo, *AjLo; 247 PetscScalar *AALo; 248 PetscInt i,nz, nzLower, offset, rowOffset; 249 PetscErrorCode ierr; 250 251 PetscFunctionBegin; 252 if (A->valid_GPU_matrix == PETSC_OFFLOAD_UNALLOCATED || A->valid_GPU_matrix == PETSC_OFFLOAD_CPU) { 253 try { 254 /* first figure out the number of nonzeros in the lower triangular matrix including 1's on the diagonal. */ 255 nzLower=n+ai[n]-ai[1]; 256 257 /* Allocate Space for the lower triangular matrix */ 258 ierr = cudaMallocHost((void**) &AiLo, (n+1)*sizeof(PetscInt));CHKERRCUDA(ierr); 259 ierr = cudaMallocHost((void**) &AjLo, nzLower*sizeof(PetscInt));CHKERRCUDA(ierr); 260 ierr = cudaMallocHost((void**) &AALo, nzLower*sizeof(PetscScalar));CHKERRCUDA(ierr); 261 262 /* Fill the lower triangular matrix */ 263 AiLo[0] = (PetscInt) 0; 264 AiLo[n] = nzLower; 265 AjLo[0] = (PetscInt) 0; 266 AALo[0] = (MatScalar) 1.0; 267 v = aa; 268 vi = aj; 269 offset = 1; 270 rowOffset= 1; 271 for (i=1; i<n; i++) { 272 nz = ai[i+1] - ai[i]; 273 /* additional 1 for the term on the diagonal */ 274 AiLo[i] = rowOffset; 275 rowOffset += nz+1; 276 277 ierr = PetscArraycpy(&(AjLo[offset]), vi, nz);CHKERRQ(ierr); 278 ierr = PetscArraycpy(&(AALo[offset]), v, nz);CHKERRQ(ierr); 279 280 offset += nz; 281 AjLo[offset] = (PetscInt) i; 282 AALo[offset] = (MatScalar) 1.0; 283 offset += 1; 284 285 v += nz; 286 vi += nz; 287 } 288 289 /* allocate space for the triangular factor information */ 290 loTriFactor = new Mat_SeqAIJCUSPARSETriFactorStruct; 291 292 /* Create the matrix description */ 293 stat = cusparseCreateMatDescr(&loTriFactor->descr);CHKERRCUDA(stat); 294 stat = cusparseSetMatIndexBase(loTriFactor->descr, CUSPARSE_INDEX_BASE_ZERO);CHKERRCUDA(stat); 295 stat = cusparseSetMatType(loTriFactor->descr, CUSPARSE_MATRIX_TYPE_TRIANGULAR);CHKERRCUDA(stat); 296 stat = cusparseSetMatFillMode(loTriFactor->descr, CUSPARSE_FILL_MODE_LOWER);CHKERRCUDA(stat); 297 stat = cusparseSetMatDiagType(loTriFactor->descr, CUSPARSE_DIAG_TYPE_UNIT);CHKERRCUDA(stat); 298 299 /* Create the solve analysis information */ 300 stat = cusparseCreateSolveAnalysisInfo(&loTriFactor->solveInfo);CHKERRCUDA(stat); 301 302 /* set the operation */ 303 loTriFactor->solveOp = CUSPARSE_OPERATION_NON_TRANSPOSE; 304 305 /* set the matrix */ 306 loTriFactor->csrMat = new CsrMatrix; 307 loTriFactor->csrMat->num_rows = n; 308 loTriFactor->csrMat->num_cols = n; 309 loTriFactor->csrMat->num_entries = nzLower; 310 311 loTriFactor->csrMat->row_offsets = new THRUSTINTARRAY32(n+1); 312 loTriFactor->csrMat->row_offsets->assign(AiLo, AiLo+n+1); 313 314 loTriFactor->csrMat->column_indices = new THRUSTINTARRAY32(nzLower); 315 loTriFactor->csrMat->column_indices->assign(AjLo, AjLo+nzLower); 316 317 loTriFactor->csrMat->values = new THRUSTARRAY(nzLower); 318 loTriFactor->csrMat->values->assign(AALo, AALo+nzLower); 319 320 /* perform the solve analysis */ 321 stat = cusparse_analysis(cusparseTriFactors->handle, loTriFactor->solveOp, 322 loTriFactor->csrMat->num_rows, loTriFactor->csrMat->num_entries, loTriFactor->descr, 323 loTriFactor->csrMat->values->data().get(), loTriFactor->csrMat->row_offsets->data().get(), 324 loTriFactor->csrMat->column_indices->data().get(), loTriFactor->solveInfo);CHKERRCUDA(stat); 325 326 /* assign the pointer. Is this really necessary? */ 327 ((Mat_SeqAIJCUSPARSETriFactors*)A->spptr)->loTriFactorPtr = loTriFactor; 328 329 ierr = cudaFreeHost(AiLo);CHKERRCUDA(ierr); 330 ierr = cudaFreeHost(AjLo);CHKERRCUDA(ierr); 331 ierr = cudaFreeHost(AALo);CHKERRCUDA(ierr); 332 ierr = PetscLogCpuToGpu((n+1+nzLower)*sizeof(int)+nzLower*sizeof(PetscScalar));CHKERRQ(ierr); 333 } catch(char *ex) { 334 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"CUSPARSE error: %s", ex); 335 } 336 } 337 PetscFunctionReturn(0); 338 } 339 340 static PetscErrorCode MatSeqAIJCUSPARSEBuildILUUpperTriMatrix(Mat A) 341 { 342 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 343 PetscInt n = A->rmap->n; 344 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 345 Mat_SeqAIJCUSPARSETriFactorStruct *upTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtr; 346 cusparseStatus_t stat; 347 const PetscInt *aj = a->j,*adiag = a->diag,*vi; 348 const MatScalar *aa = a->a,*v; 349 PetscInt *AiUp, *AjUp; 350 PetscScalar *AAUp; 351 PetscInt i,nz, nzUpper, offset; 352 PetscErrorCode ierr; 353 354 PetscFunctionBegin; 355 if (A->valid_GPU_matrix == PETSC_OFFLOAD_UNALLOCATED || A->valid_GPU_matrix == PETSC_OFFLOAD_CPU) { 356 try { 357 /* next, figure out the number of nonzeros in the upper triangular matrix. */ 358 nzUpper = adiag[0]-adiag[n]; 359 360 /* Allocate Space for the upper triangular matrix */ 361 ierr = cudaMallocHost((void**) &AiUp, (n+1)*sizeof(PetscInt));CHKERRCUDA(ierr); 362 ierr = cudaMallocHost((void**) &AjUp, nzUpper*sizeof(PetscInt));CHKERRCUDA(ierr); 363 ierr = cudaMallocHost((void**) &AAUp, nzUpper*sizeof(PetscScalar));CHKERRCUDA(ierr); 364 365 /* Fill the upper triangular matrix */ 366 AiUp[0]=(PetscInt) 0; 367 AiUp[n]=nzUpper; 368 offset = nzUpper; 369 for (i=n-1; i>=0; i--) { 370 v = aa + adiag[i+1] + 1; 371 vi = aj + adiag[i+1] + 1; 372 373 /* number of elements NOT on the diagonal */ 374 nz = adiag[i] - adiag[i+1]-1; 375 376 /* decrement the offset */ 377 offset -= (nz+1); 378 379 /* first, set the diagonal elements */ 380 AjUp[offset] = (PetscInt) i; 381 AAUp[offset] = (MatScalar)1./v[nz]; 382 AiUp[i] = AiUp[i+1] - (nz+1); 383 384 ierr = PetscArraycpy(&(AjUp[offset+1]), vi, nz);CHKERRQ(ierr); 385 ierr = PetscArraycpy(&(AAUp[offset+1]), v, nz);CHKERRQ(ierr); 386 } 387 388 /* allocate space for the triangular factor information */ 389 upTriFactor = new Mat_SeqAIJCUSPARSETriFactorStruct; 390 391 /* Create the matrix description */ 392 stat = cusparseCreateMatDescr(&upTriFactor->descr);CHKERRCUDA(stat); 393 stat = cusparseSetMatIndexBase(upTriFactor->descr, CUSPARSE_INDEX_BASE_ZERO);CHKERRCUDA(stat); 394 stat = cusparseSetMatType(upTriFactor->descr, CUSPARSE_MATRIX_TYPE_TRIANGULAR);CHKERRCUDA(stat); 395 stat = cusparseSetMatFillMode(upTriFactor->descr, CUSPARSE_FILL_MODE_UPPER);CHKERRCUDA(stat); 396 stat = cusparseSetMatDiagType(upTriFactor->descr, CUSPARSE_DIAG_TYPE_NON_UNIT);CHKERRCUDA(stat); 397 398 /* Create the solve analysis information */ 399 stat = cusparseCreateSolveAnalysisInfo(&upTriFactor->solveInfo);CHKERRCUDA(stat); 400 401 /* set the operation */ 402 upTriFactor->solveOp = CUSPARSE_OPERATION_NON_TRANSPOSE; 403 404 /* set the matrix */ 405 upTriFactor->csrMat = new CsrMatrix; 406 upTriFactor->csrMat->num_rows = n; 407 upTriFactor->csrMat->num_cols = n; 408 upTriFactor->csrMat->num_entries = nzUpper; 409 410 upTriFactor->csrMat->row_offsets = new THRUSTINTARRAY32(n+1); 411 upTriFactor->csrMat->row_offsets->assign(AiUp, AiUp+n+1); 412 413 upTriFactor->csrMat->column_indices = new THRUSTINTARRAY32(nzUpper); 414 upTriFactor->csrMat->column_indices->assign(AjUp, AjUp+nzUpper); 415 416 upTriFactor->csrMat->values = new THRUSTARRAY(nzUpper); 417 upTriFactor->csrMat->values->assign(AAUp, AAUp+nzUpper); 418 419 /* perform the solve analysis */ 420 stat = cusparse_analysis(cusparseTriFactors->handle, upTriFactor->solveOp, 421 upTriFactor->csrMat->num_rows, upTriFactor->csrMat->num_entries, upTriFactor->descr, 422 upTriFactor->csrMat->values->data().get(), upTriFactor->csrMat->row_offsets->data().get(), 423 upTriFactor->csrMat->column_indices->data().get(), upTriFactor->solveInfo);CHKERRCUDA(stat); 424 425 /* assign the pointer. Is this really necessary? */ 426 ((Mat_SeqAIJCUSPARSETriFactors*)A->spptr)->upTriFactorPtr = upTriFactor; 427 428 ierr = cudaFreeHost(AiUp);CHKERRCUDA(ierr); 429 ierr = cudaFreeHost(AjUp);CHKERRCUDA(ierr); 430 ierr = cudaFreeHost(AAUp);CHKERRCUDA(ierr); 431 ierr = PetscLogCpuToGpu((n+1+nzUpper)*sizeof(int)+nzUpper*sizeof(PetscScalar));CHKERRQ(ierr); 432 } catch(char *ex) { 433 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"CUSPARSE error: %s", ex); 434 } 435 } 436 PetscFunctionReturn(0); 437 } 438 439 static PetscErrorCode MatSeqAIJCUSPARSEILUAnalysisAndCopyToGPU(Mat A) 440 { 441 PetscErrorCode ierr; 442 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 443 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 444 IS isrow = a->row,iscol = a->icol; 445 PetscBool row_identity,col_identity; 446 const PetscInt *r,*c; 447 PetscInt n = A->rmap->n; 448 449 PetscFunctionBegin; 450 ierr = MatSeqAIJCUSPARSEBuildILULowerTriMatrix(A);CHKERRQ(ierr); 451 ierr = MatSeqAIJCUSPARSEBuildILUUpperTriMatrix(A);CHKERRQ(ierr); 452 453 cusparseTriFactors->workVector = new THRUSTARRAY(n); 454 cusparseTriFactors->nnz=a->nz; 455 456 A->valid_GPU_matrix = PETSC_OFFLOAD_BOTH; 457 /*lower triangular indices */ 458 ierr = ISGetIndices(isrow,&r);CHKERRQ(ierr); 459 ierr = ISIdentity(isrow,&row_identity);CHKERRQ(ierr); 460 if (!row_identity) { 461 cusparseTriFactors->rpermIndices = new THRUSTINTARRAY(n); 462 cusparseTriFactors->rpermIndices->assign(r, r+n); 463 } 464 ierr = ISRestoreIndices(isrow,&r);CHKERRQ(ierr); 465 466 /*upper triangular indices */ 467 ierr = ISGetIndices(iscol,&c);CHKERRQ(ierr); 468 ierr = ISIdentity(iscol,&col_identity);CHKERRQ(ierr); 469 if (!col_identity) { 470 cusparseTriFactors->cpermIndices = new THRUSTINTARRAY(n); 471 cusparseTriFactors->cpermIndices->assign(c, c+n); 472 } 473 474 if(!row_identity && !col_identity) { 475 ierr = PetscLogCpuToGpu(2*n*sizeof(PetscInt));CHKERRQ(ierr); 476 } else if(!row_identity) { 477 ierr = PetscLogCpuToGpu(n*sizeof(PetscInt));CHKERRQ(ierr); 478 } else if(!col_identity) { 479 ierr = PetscLogCpuToGpu(n*sizeof(PetscInt));CHKERRQ(ierr); 480 } 481 482 ierr = ISRestoreIndices(iscol,&c);CHKERRQ(ierr); 483 PetscFunctionReturn(0); 484 } 485 486 static PetscErrorCode MatSeqAIJCUSPARSEBuildICCTriMatrices(Mat A) 487 { 488 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 489 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 490 Mat_SeqAIJCUSPARSETriFactorStruct *loTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtr; 491 Mat_SeqAIJCUSPARSETriFactorStruct *upTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtr; 492 cusparseStatus_t stat; 493 PetscErrorCode ierr; 494 PetscInt *AiUp, *AjUp; 495 PetscScalar *AAUp; 496 PetscScalar *AALo; 497 PetscInt nzUpper = a->nz,n = A->rmap->n,i,offset,nz,j; 498 Mat_SeqSBAIJ *b = (Mat_SeqSBAIJ*)A->data; 499 const PetscInt *ai = b->i,*aj = b->j,*vj; 500 const MatScalar *aa = b->a,*v; 501 502 PetscFunctionBegin; 503 if (A->valid_GPU_matrix == PETSC_OFFLOAD_UNALLOCATED || A->valid_GPU_matrix == PETSC_OFFLOAD_CPU) { 504 try { 505 /* Allocate Space for the upper triangular matrix */ 506 ierr = cudaMallocHost((void**) &AiUp, (n+1)*sizeof(PetscInt));CHKERRCUDA(ierr); 507 ierr = cudaMallocHost((void**) &AjUp, nzUpper*sizeof(PetscInt));CHKERRCUDA(ierr); 508 ierr = cudaMallocHost((void**) &AAUp, nzUpper*sizeof(PetscScalar));CHKERRCUDA(ierr); 509 ierr = cudaMallocHost((void**) &AALo, nzUpper*sizeof(PetscScalar));CHKERRCUDA(ierr); 510 511 /* Fill the upper triangular matrix */ 512 AiUp[0]=(PetscInt) 0; 513 AiUp[n]=nzUpper; 514 offset = 0; 515 for (i=0; i<n; i++) { 516 /* set the pointers */ 517 v = aa + ai[i]; 518 vj = aj + ai[i]; 519 nz = ai[i+1] - ai[i] - 1; /* exclude diag[i] */ 520 521 /* first, set the diagonal elements */ 522 AjUp[offset] = (PetscInt) i; 523 AAUp[offset] = (MatScalar)1.0/v[nz]; 524 AiUp[i] = offset; 525 AALo[offset] = (MatScalar)1.0/v[nz]; 526 527 offset+=1; 528 if (nz>0) { 529 ierr = PetscArraycpy(&(AjUp[offset]), vj, nz);CHKERRQ(ierr); 530 ierr = PetscArraycpy(&(AAUp[offset]), v, nz);CHKERRQ(ierr); 531 for (j=offset; j<offset+nz; j++) { 532 AAUp[j] = -AAUp[j]; 533 AALo[j] = AAUp[j]/v[nz]; 534 } 535 offset+=nz; 536 } 537 } 538 539 /* allocate space for the triangular factor information */ 540 upTriFactor = new Mat_SeqAIJCUSPARSETriFactorStruct; 541 542 /* Create the matrix description */ 543 stat = cusparseCreateMatDescr(&upTriFactor->descr);CHKERRCUDA(stat); 544 stat = cusparseSetMatIndexBase(upTriFactor->descr, CUSPARSE_INDEX_BASE_ZERO);CHKERRCUDA(stat); 545 stat = cusparseSetMatType(upTriFactor->descr, CUSPARSE_MATRIX_TYPE_TRIANGULAR);CHKERRCUDA(stat); 546 stat = cusparseSetMatFillMode(upTriFactor->descr, CUSPARSE_FILL_MODE_UPPER);CHKERRCUDA(stat); 547 stat = cusparseSetMatDiagType(upTriFactor->descr, CUSPARSE_DIAG_TYPE_UNIT);CHKERRCUDA(stat); 548 549 /* Create the solve analysis information */ 550 stat = cusparseCreateSolveAnalysisInfo(&upTriFactor->solveInfo);CHKERRCUDA(stat); 551 552 /* set the operation */ 553 upTriFactor->solveOp = CUSPARSE_OPERATION_NON_TRANSPOSE; 554 555 /* set the matrix */ 556 upTriFactor->csrMat = new CsrMatrix; 557 upTriFactor->csrMat->num_rows = A->rmap->n; 558 upTriFactor->csrMat->num_cols = A->cmap->n; 559 upTriFactor->csrMat->num_entries = a->nz; 560 561 upTriFactor->csrMat->row_offsets = new THRUSTINTARRAY32(A->rmap->n+1); 562 upTriFactor->csrMat->row_offsets->assign(AiUp, AiUp+A->rmap->n+1); 563 564 upTriFactor->csrMat->column_indices = new THRUSTINTARRAY32(a->nz); 565 upTriFactor->csrMat->column_indices->assign(AjUp, AjUp+a->nz); 566 567 upTriFactor->csrMat->values = new THRUSTARRAY(a->nz); 568 upTriFactor->csrMat->values->assign(AAUp, AAUp+a->nz); 569 570 /* perform the solve analysis */ 571 stat = cusparse_analysis(cusparseTriFactors->handle, upTriFactor->solveOp, 572 upTriFactor->csrMat->num_rows, upTriFactor->csrMat->num_entries, upTriFactor->descr, 573 upTriFactor->csrMat->values->data().get(), upTriFactor->csrMat->row_offsets->data().get(), 574 upTriFactor->csrMat->column_indices->data().get(), upTriFactor->solveInfo);CHKERRCUDA(stat); 575 576 /* assign the pointer. Is this really necessary? */ 577 ((Mat_SeqAIJCUSPARSETriFactors*)A->spptr)->upTriFactorPtr = upTriFactor; 578 579 /* allocate space for the triangular factor information */ 580 loTriFactor = new Mat_SeqAIJCUSPARSETriFactorStruct; 581 582 /* Create the matrix description */ 583 stat = cusparseCreateMatDescr(&loTriFactor->descr);CHKERRCUDA(stat); 584 stat = cusparseSetMatIndexBase(loTriFactor->descr, CUSPARSE_INDEX_BASE_ZERO);CHKERRCUDA(stat); 585 stat = cusparseSetMatType(loTriFactor->descr, CUSPARSE_MATRIX_TYPE_TRIANGULAR);CHKERRCUDA(stat); 586 stat = cusparseSetMatFillMode(loTriFactor->descr, CUSPARSE_FILL_MODE_UPPER);CHKERRCUDA(stat); 587 stat = cusparseSetMatDiagType(loTriFactor->descr, CUSPARSE_DIAG_TYPE_NON_UNIT);CHKERRCUDA(stat); 588 589 /* Create the solve analysis information */ 590 stat = cusparseCreateSolveAnalysisInfo(&loTriFactor->solveInfo);CHKERRCUDA(stat); 591 592 /* set the operation */ 593 loTriFactor->solveOp = CUSPARSE_OPERATION_TRANSPOSE; 594 595 /* set the matrix */ 596 loTriFactor->csrMat = new CsrMatrix; 597 loTriFactor->csrMat->num_rows = A->rmap->n; 598 loTriFactor->csrMat->num_cols = A->cmap->n; 599 loTriFactor->csrMat->num_entries = a->nz; 600 601 loTriFactor->csrMat->row_offsets = new THRUSTINTARRAY32(A->rmap->n+1); 602 loTriFactor->csrMat->row_offsets->assign(AiUp, AiUp+A->rmap->n+1); 603 604 loTriFactor->csrMat->column_indices = new THRUSTINTARRAY32(a->nz); 605 loTriFactor->csrMat->column_indices->assign(AjUp, AjUp+a->nz); 606 607 loTriFactor->csrMat->values = new THRUSTARRAY(a->nz); 608 loTriFactor->csrMat->values->assign(AALo, AALo+a->nz); 609 ierr = PetscLogCpuToGpu(2*(((A->rmap->n+1)+(a->nz))*sizeof(int)+(a->nz)*sizeof(PetscScalar)));CHKERRQ(ierr); 610 611 /* perform the solve analysis */ 612 stat = cusparse_analysis(cusparseTriFactors->handle, loTriFactor->solveOp, 613 loTriFactor->csrMat->num_rows, loTriFactor->csrMat->num_entries, loTriFactor->descr, 614 loTriFactor->csrMat->values->data().get(), loTriFactor->csrMat->row_offsets->data().get(), 615 loTriFactor->csrMat->column_indices->data().get(), loTriFactor->solveInfo);CHKERRCUDA(stat); 616 617 /* assign the pointer. Is this really necessary? */ 618 ((Mat_SeqAIJCUSPARSETriFactors*)A->spptr)->loTriFactorPtr = loTriFactor; 619 620 A->valid_GPU_matrix = PETSC_OFFLOAD_BOTH; 621 ierr = cudaFreeHost(AiUp);CHKERRCUDA(ierr); 622 ierr = cudaFreeHost(AjUp);CHKERRCUDA(ierr); 623 ierr = cudaFreeHost(AAUp);CHKERRCUDA(ierr); 624 ierr = cudaFreeHost(AALo);CHKERRCUDA(ierr); 625 } catch(char *ex) { 626 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"CUSPARSE error: %s", ex); 627 } 628 } 629 PetscFunctionReturn(0); 630 } 631 632 static PetscErrorCode MatSeqAIJCUSPARSEICCAnalysisAndCopyToGPU(Mat A) 633 { 634 PetscErrorCode ierr; 635 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 636 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 637 IS ip = a->row; 638 const PetscInt *rip; 639 PetscBool perm_identity; 640 PetscInt n = A->rmap->n; 641 642 PetscFunctionBegin; 643 ierr = MatSeqAIJCUSPARSEBuildICCTriMatrices(A);CHKERRQ(ierr); 644 cusparseTriFactors->workVector = new THRUSTARRAY(n); 645 cusparseTriFactors->nnz=(a->nz-n)*2 + n; 646 647 /*lower triangular indices */ 648 ierr = ISGetIndices(ip,&rip);CHKERRQ(ierr); 649 ierr = ISIdentity(ip,&perm_identity);CHKERRQ(ierr); 650 if (!perm_identity) { 651 cusparseTriFactors->rpermIndices = new THRUSTINTARRAY(n); 652 cusparseTriFactors->rpermIndices->assign(rip, rip+n); 653 cusparseTriFactors->cpermIndices = new THRUSTINTARRAY(n); 654 cusparseTriFactors->cpermIndices->assign(rip, rip+n); 655 ierr = PetscLogCpuToGpu(2*n*sizeof(PetscInt));CHKERRQ(ierr); 656 } 657 ierr = ISRestoreIndices(ip,&rip);CHKERRQ(ierr); 658 PetscFunctionReturn(0); 659 } 660 661 static PetscErrorCode MatLUFactorNumeric_SeqAIJCUSPARSE(Mat B,Mat A,const MatFactorInfo *info) 662 { 663 Mat_SeqAIJ *b = (Mat_SeqAIJ*)B->data; 664 IS isrow = b->row,iscol = b->col; 665 PetscBool row_identity,col_identity; 666 PetscErrorCode ierr; 667 668 PetscFunctionBegin; 669 ierr = MatLUFactorNumeric_SeqAIJ(B,A,info);CHKERRQ(ierr); 670 /* determine which version of MatSolve needs to be used. */ 671 ierr = ISIdentity(isrow,&row_identity);CHKERRQ(ierr); 672 ierr = ISIdentity(iscol,&col_identity);CHKERRQ(ierr); 673 if (row_identity && col_identity) { 674 B->ops->solve = MatSolve_SeqAIJCUSPARSE_NaturalOrdering; 675 B->ops->solvetranspose = MatSolveTranspose_SeqAIJCUSPARSE_NaturalOrdering; 676 } else { 677 B->ops->solve = MatSolve_SeqAIJCUSPARSE; 678 B->ops->solvetranspose = MatSolveTranspose_SeqAIJCUSPARSE; 679 } 680 681 /* get the triangular factors */ 682 ierr = MatSeqAIJCUSPARSEILUAnalysisAndCopyToGPU(B);CHKERRQ(ierr); 683 PetscFunctionReturn(0); 684 } 685 686 static PetscErrorCode MatCholeskyFactorNumeric_SeqAIJCUSPARSE(Mat B,Mat A,const MatFactorInfo *info) 687 { 688 Mat_SeqAIJ *b = (Mat_SeqAIJ*)B->data; 689 IS ip = b->row; 690 PetscBool perm_identity; 691 PetscErrorCode ierr; 692 693 PetscFunctionBegin; 694 ierr = MatCholeskyFactorNumeric_SeqAIJ(B,A,info);CHKERRQ(ierr); 695 696 /* determine which version of MatSolve needs to be used. */ 697 ierr = ISIdentity(ip,&perm_identity);CHKERRQ(ierr); 698 if (perm_identity) { 699 B->ops->solve = MatSolve_SeqAIJCUSPARSE_NaturalOrdering; 700 B->ops->solvetranspose = MatSolveTranspose_SeqAIJCUSPARSE_NaturalOrdering; 701 } else { 702 B->ops->solve = MatSolve_SeqAIJCUSPARSE; 703 B->ops->solvetranspose = MatSolveTranspose_SeqAIJCUSPARSE; 704 } 705 706 /* get the triangular factors */ 707 ierr = MatSeqAIJCUSPARSEICCAnalysisAndCopyToGPU(B);CHKERRQ(ierr); 708 PetscFunctionReturn(0); 709 } 710 711 static PetscErrorCode MatSeqAIJCUSPARSEAnalyzeTransposeForSolve(Mat A) 712 { 713 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 714 Mat_SeqAIJCUSPARSETriFactorStruct *loTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtr; 715 Mat_SeqAIJCUSPARSETriFactorStruct *upTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtr; 716 Mat_SeqAIJCUSPARSETriFactorStruct *loTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtrTranspose; 717 Mat_SeqAIJCUSPARSETriFactorStruct *upTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtrTranspose; 718 cusparseStatus_t stat; 719 cusparseIndexBase_t indexBase; 720 cusparseMatrixType_t matrixType; 721 cusparseFillMode_t fillMode; 722 cusparseDiagType_t diagType; 723 724 PetscFunctionBegin; 725 726 /*********************************************/ 727 /* Now the Transpose of the Lower Tri Factor */ 728 /*********************************************/ 729 730 /* allocate space for the transpose of the lower triangular factor */ 731 loTriFactorT = new Mat_SeqAIJCUSPARSETriFactorStruct; 732 733 /* set the matrix descriptors of the lower triangular factor */ 734 matrixType = cusparseGetMatType(loTriFactor->descr); 735 indexBase = cusparseGetMatIndexBase(loTriFactor->descr); 736 fillMode = cusparseGetMatFillMode(loTriFactor->descr)==CUSPARSE_FILL_MODE_UPPER ? 737 CUSPARSE_FILL_MODE_LOWER : CUSPARSE_FILL_MODE_UPPER; 738 diagType = cusparseGetMatDiagType(loTriFactor->descr); 739 740 /* Create the matrix description */ 741 stat = cusparseCreateMatDescr(&loTriFactorT->descr);CHKERRCUDA(stat); 742 stat = cusparseSetMatIndexBase(loTriFactorT->descr, indexBase);CHKERRCUDA(stat); 743 stat = cusparseSetMatType(loTriFactorT->descr, matrixType);CHKERRCUDA(stat); 744 stat = cusparseSetMatFillMode(loTriFactorT->descr, fillMode);CHKERRCUDA(stat); 745 stat = cusparseSetMatDiagType(loTriFactorT->descr, diagType);CHKERRCUDA(stat); 746 747 /* Create the solve analysis information */ 748 stat = cusparseCreateSolveAnalysisInfo(&loTriFactorT->solveInfo);CHKERRCUDA(stat); 749 750 /* set the operation */ 751 loTriFactorT->solveOp = CUSPARSE_OPERATION_NON_TRANSPOSE; 752 753 /* allocate GPU space for the CSC of the lower triangular factor*/ 754 loTriFactorT->csrMat = new CsrMatrix; 755 loTriFactorT->csrMat->num_rows = loTriFactor->csrMat->num_rows; 756 loTriFactorT->csrMat->num_cols = loTriFactor->csrMat->num_cols; 757 loTriFactorT->csrMat->num_entries = loTriFactor->csrMat->num_entries; 758 loTriFactorT->csrMat->row_offsets = new THRUSTINTARRAY32(loTriFactor->csrMat->num_rows+1); 759 loTriFactorT->csrMat->column_indices = new THRUSTINTARRAY32(loTriFactor->csrMat->num_entries); 760 loTriFactorT->csrMat->values = new THRUSTARRAY(loTriFactor->csrMat->num_entries); 761 762 /* compute the transpose of the lower triangular factor, i.e. the CSC */ 763 stat = cusparse_csr2csc(cusparseTriFactors->handle, loTriFactor->csrMat->num_rows, 764 loTriFactor->csrMat->num_cols, loTriFactor->csrMat->num_entries, 765 loTriFactor->csrMat->values->data().get(), 766 loTriFactor->csrMat->row_offsets->data().get(), 767 loTriFactor->csrMat->column_indices->data().get(), 768 loTriFactorT->csrMat->values->data().get(), 769 loTriFactorT->csrMat->column_indices->data().get(), 770 loTriFactorT->csrMat->row_offsets->data().get(), 771 CUSPARSE_ACTION_NUMERIC, indexBase);CHKERRCUDA(stat); 772 773 /* perform the solve analysis on the transposed matrix */ 774 stat = cusparse_analysis(cusparseTriFactors->handle, loTriFactorT->solveOp, 775 loTriFactorT->csrMat->num_rows, loTriFactorT->csrMat->num_entries, 776 loTriFactorT->descr, loTriFactorT->csrMat->values->data().get(), 777 loTriFactorT->csrMat->row_offsets->data().get(), loTriFactorT->csrMat->column_indices->data().get(), 778 loTriFactorT->solveInfo);CHKERRCUDA(stat); 779 780 /* assign the pointer. Is this really necessary? */ 781 ((Mat_SeqAIJCUSPARSETriFactors*)A->spptr)->loTriFactorPtrTranspose = loTriFactorT; 782 783 /*********************************************/ 784 /* Now the Transpose of the Upper Tri Factor */ 785 /*********************************************/ 786 787 /* allocate space for the transpose of the upper triangular factor */ 788 upTriFactorT = new Mat_SeqAIJCUSPARSETriFactorStruct; 789 790 /* set the matrix descriptors of the upper triangular factor */ 791 matrixType = cusparseGetMatType(upTriFactor->descr); 792 indexBase = cusparseGetMatIndexBase(upTriFactor->descr); 793 fillMode = cusparseGetMatFillMode(upTriFactor->descr)==CUSPARSE_FILL_MODE_UPPER ? 794 CUSPARSE_FILL_MODE_LOWER : CUSPARSE_FILL_MODE_UPPER; 795 diagType = cusparseGetMatDiagType(upTriFactor->descr); 796 797 /* Create the matrix description */ 798 stat = cusparseCreateMatDescr(&upTriFactorT->descr);CHKERRCUDA(stat); 799 stat = cusparseSetMatIndexBase(upTriFactorT->descr, indexBase);CHKERRCUDA(stat); 800 stat = cusparseSetMatType(upTriFactorT->descr, matrixType);CHKERRCUDA(stat); 801 stat = cusparseSetMatFillMode(upTriFactorT->descr, fillMode);CHKERRCUDA(stat); 802 stat = cusparseSetMatDiagType(upTriFactorT->descr, diagType);CHKERRCUDA(stat); 803 804 /* Create the solve analysis information */ 805 stat = cusparseCreateSolveAnalysisInfo(&upTriFactorT->solveInfo);CHKERRCUDA(stat); 806 807 /* set the operation */ 808 upTriFactorT->solveOp = CUSPARSE_OPERATION_NON_TRANSPOSE; 809 810 /* allocate GPU space for the CSC of the upper triangular factor*/ 811 upTriFactorT->csrMat = new CsrMatrix; 812 upTriFactorT->csrMat->num_rows = upTriFactor->csrMat->num_rows; 813 upTriFactorT->csrMat->num_cols = upTriFactor->csrMat->num_cols; 814 upTriFactorT->csrMat->num_entries = upTriFactor->csrMat->num_entries; 815 upTriFactorT->csrMat->row_offsets = new THRUSTINTARRAY32(upTriFactor->csrMat->num_rows+1); 816 upTriFactorT->csrMat->column_indices = new THRUSTINTARRAY32(upTriFactor->csrMat->num_entries); 817 upTriFactorT->csrMat->values = new THRUSTARRAY(upTriFactor->csrMat->num_entries); 818 819 /* compute the transpose of the upper triangular factor, i.e. the CSC */ 820 stat = cusparse_csr2csc(cusparseTriFactors->handle, upTriFactor->csrMat->num_rows, 821 upTriFactor->csrMat->num_cols, upTriFactor->csrMat->num_entries, 822 upTriFactor->csrMat->values->data().get(), 823 upTriFactor->csrMat->row_offsets->data().get(), 824 upTriFactor->csrMat->column_indices->data().get(), 825 upTriFactorT->csrMat->values->data().get(), 826 upTriFactorT->csrMat->column_indices->data().get(), 827 upTriFactorT->csrMat->row_offsets->data().get(), 828 CUSPARSE_ACTION_NUMERIC, indexBase);CHKERRCUDA(stat); 829 830 /* perform the solve analysis on the transposed matrix */ 831 stat = cusparse_analysis(cusparseTriFactors->handle, upTriFactorT->solveOp, 832 upTriFactorT->csrMat->num_rows, upTriFactorT->csrMat->num_entries, 833 upTriFactorT->descr, upTriFactorT->csrMat->values->data().get(), 834 upTriFactorT->csrMat->row_offsets->data().get(), upTriFactorT->csrMat->column_indices->data().get(), 835 upTriFactorT->solveInfo);CHKERRCUDA(stat); 836 837 /* assign the pointer. Is this really necessary? */ 838 ((Mat_SeqAIJCUSPARSETriFactors*)A->spptr)->upTriFactorPtrTranspose = upTriFactorT; 839 PetscFunctionReturn(0); 840 } 841 842 static PetscErrorCode MatSeqAIJCUSPARSEGenerateTransposeForMult(Mat A) 843 { 844 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 845 Mat_SeqAIJCUSPARSEMultStruct *matstruct = (Mat_SeqAIJCUSPARSEMultStruct*)cusparsestruct->mat; 846 Mat_SeqAIJCUSPARSEMultStruct *matstructT = (Mat_SeqAIJCUSPARSEMultStruct*)cusparsestruct->matTranspose; 847 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 848 cusparseStatus_t stat; 849 cusparseIndexBase_t indexBase; 850 cudaError_t err; 851 PetscErrorCode ierr; 852 853 PetscFunctionBegin; 854 855 /* allocate space for the triangular factor information */ 856 matstructT = new Mat_SeqAIJCUSPARSEMultStruct; 857 stat = cusparseCreateMatDescr(&matstructT->descr);CHKERRCUDA(stat); 858 indexBase = cusparseGetMatIndexBase(matstruct->descr); 859 stat = cusparseSetMatIndexBase(matstructT->descr, indexBase);CHKERRCUDA(stat); 860 stat = cusparseSetMatType(matstructT->descr, CUSPARSE_MATRIX_TYPE_GENERAL);CHKERRCUDA(stat); 861 862 /* set alpha and beta */ 863 err = cudaMalloc((void **)&(matstructT->alpha), sizeof(PetscScalar));CHKERRCUDA(err); 864 err = cudaMalloc((void **)&(matstructT->beta_zero),sizeof(PetscScalar));CHKERRCUDA(err); 865 err = cudaMalloc((void **)&(matstructT->beta_one), sizeof(PetscScalar));CHKERRCUDA(err); 866 err = cudaMemcpy(matstructT->alpha, &PETSC_CUSPARSE_ONE, sizeof(PetscScalar),cudaMemcpyHostToDevice);CHKERRCUDA(err); 867 err = cudaMemcpy(matstructT->beta_zero,&PETSC_CUSPARSE_ZERO,sizeof(PetscScalar),cudaMemcpyHostToDevice);CHKERRCUDA(err); 868 err = cudaMemcpy(matstructT->beta_one, &PETSC_CUSPARSE_ONE, sizeof(PetscScalar),cudaMemcpyHostToDevice);CHKERRCUDA(err); 869 stat = cusparseSetPointerMode(cusparsestruct->handle, CUSPARSE_POINTER_MODE_DEVICE);CHKERRCUDA(stat); 870 871 if (cusparsestruct->format==MAT_CUSPARSE_CSR) { 872 CsrMatrix *matrix = (CsrMatrix*)matstruct->mat; 873 CsrMatrix *matrixT= new CsrMatrix; 874 matrixT->num_rows = A->cmap->n; 875 matrixT->num_cols = A->rmap->n; 876 matrixT->num_entries = a->nz; 877 matrixT->row_offsets = new THRUSTINTARRAY32(A->rmap->n+1); 878 matrixT->column_indices = new THRUSTINTARRAY32(a->nz); 879 matrixT->values = new THRUSTARRAY(a->nz); 880 881 /* compute the transpose of the upper triangular factor, i.e. the CSC */ 882 indexBase = cusparseGetMatIndexBase(matstruct->descr); 883 stat = cusparse_csr2csc(cusparsestruct->handle, matrix->num_rows, 884 matrix->num_cols, matrix->num_entries, 885 matrix->values->data().get(), 886 matrix->row_offsets->data().get(), 887 matrix->column_indices->data().get(), 888 matrixT->values->data().get(), 889 matrixT->column_indices->data().get(), 890 matrixT->row_offsets->data().get(), 891 CUSPARSE_ACTION_NUMERIC, indexBase);CHKERRCUDA(stat); 892 893 /* assign the pointer */ 894 matstructT->mat = matrixT; 895 ierr = PetscLogCpuToGpu(((A->rmap->n+1)+(a->nz))*sizeof(int)+(3+a->nz)*sizeof(PetscScalar));CHKERRQ(ierr); 896 } else if (cusparsestruct->format==MAT_CUSPARSE_ELL || cusparsestruct->format==MAT_CUSPARSE_HYB) { 897 #if CUDA_VERSION>=5000 898 /* First convert HYB to CSR */ 899 CsrMatrix *temp= new CsrMatrix; 900 temp->num_rows = A->rmap->n; 901 temp->num_cols = A->cmap->n; 902 temp->num_entries = a->nz; 903 temp->row_offsets = new THRUSTINTARRAY32(A->rmap->n+1); 904 temp->column_indices = new THRUSTINTARRAY32(a->nz); 905 temp->values = new THRUSTARRAY(a->nz); 906 907 908 stat = cusparse_hyb2csr(cusparsestruct->handle, 909 matstruct->descr, (cusparseHybMat_t)matstruct->mat, 910 temp->values->data().get(), 911 temp->row_offsets->data().get(), 912 temp->column_indices->data().get());CHKERRCUDA(stat); 913 914 /* Next, convert CSR to CSC (i.e. the matrix transpose) */ 915 CsrMatrix *tempT= new CsrMatrix; 916 tempT->num_rows = A->rmap->n; 917 tempT->num_cols = A->cmap->n; 918 tempT->num_entries = a->nz; 919 tempT->row_offsets = new THRUSTINTARRAY32(A->rmap->n+1); 920 tempT->column_indices = new THRUSTINTARRAY32(a->nz); 921 tempT->values = new THRUSTARRAY(a->nz); 922 923 stat = cusparse_csr2csc(cusparsestruct->handle, temp->num_rows, 924 temp->num_cols, temp->num_entries, 925 temp->values->data().get(), 926 temp->row_offsets->data().get(), 927 temp->column_indices->data().get(), 928 tempT->values->data().get(), 929 tempT->column_indices->data().get(), 930 tempT->row_offsets->data().get(), 931 CUSPARSE_ACTION_NUMERIC, indexBase);CHKERRCUDA(stat); 932 933 /* Last, convert CSC to HYB */ 934 cusparseHybMat_t hybMat; 935 stat = cusparseCreateHybMat(&hybMat);CHKERRCUDA(stat); 936 cusparseHybPartition_t partition = cusparsestruct->format==MAT_CUSPARSE_ELL ? 937 CUSPARSE_HYB_PARTITION_MAX : CUSPARSE_HYB_PARTITION_AUTO; 938 stat = cusparse_csr2hyb(cusparsestruct->handle, A->rmap->n, A->cmap->n, 939 matstructT->descr, tempT->values->data().get(), 940 tempT->row_offsets->data().get(), 941 tempT->column_indices->data().get(), 942 hybMat, 0, partition);CHKERRCUDA(stat); 943 944 /* assign the pointer */ 945 matstructT->mat = hybMat; 946 ierr = PetscLogCpuToGpu((2*(((A->rmap->n+1)+(a->nz))*sizeof(int)+(a->nz)*sizeof(PetscScalar)))+3*sizeof(PetscScalar));CHKERRQ(ierr); 947 948 /* delete temporaries */ 949 if (tempT) { 950 if (tempT->values) delete (THRUSTARRAY*) tempT->values; 951 if (tempT->column_indices) delete (THRUSTINTARRAY32*) tempT->column_indices; 952 if (tempT->row_offsets) delete (THRUSTINTARRAY32*) tempT->row_offsets; 953 delete (CsrMatrix*) tempT; 954 } 955 if (temp) { 956 if (temp->values) delete (THRUSTARRAY*) temp->values; 957 if (temp->column_indices) delete (THRUSTINTARRAY32*) temp->column_indices; 958 if (temp->row_offsets) delete (THRUSTINTARRAY32*) temp->row_offsets; 959 delete (CsrMatrix*) temp; 960 } 961 #else 962 SETERRQ(PETSC_COMM_SELF,PETSC_ERR_SUP,"ELL (Ellpack) and HYB (Hybrid) storage format for the Matrix Transpose (in MatMultTranspose) require CUDA 5.0 or later."); 963 #endif 964 } 965 /* assign the compressed row indices */ 966 matstructT->cprowIndices = new THRUSTINTARRAY; 967 matstructT->cprowIndices->resize(A->cmap->n); 968 thrust::sequence(matstructT->cprowIndices->begin(), matstructT->cprowIndices->end()); 969 /* assign the pointer */ 970 ((Mat_SeqAIJCUSPARSE*)A->spptr)->matTranspose = matstructT; 971 PetscFunctionReturn(0); 972 } 973 974 static PetscErrorCode MatSolveTranspose_SeqAIJCUSPARSE(Mat A,Vec bb,Vec xx) 975 { 976 PetscInt n = xx->map->n; 977 const PetscScalar *barray; 978 PetscScalar *xarray; 979 thrust::device_ptr<const PetscScalar> bGPU; 980 thrust::device_ptr<PetscScalar> xGPU; 981 cusparseStatus_t stat; 982 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 983 Mat_SeqAIJCUSPARSETriFactorStruct *loTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtrTranspose; 984 Mat_SeqAIJCUSPARSETriFactorStruct *upTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtrTranspose; 985 THRUSTARRAY *tempGPU = (THRUSTARRAY*)cusparseTriFactors->workVector; 986 PetscErrorCode ierr; 987 988 PetscFunctionBegin; 989 /* Analyze the matrix and create the transpose ... on the fly */ 990 if (!loTriFactorT && !upTriFactorT) { 991 ierr = MatSeqAIJCUSPARSEAnalyzeTransposeForSolve(A);CHKERRQ(ierr); 992 loTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtrTranspose; 993 upTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtrTranspose; 994 } 995 996 /* Get the GPU pointers */ 997 ierr = VecCUDAGetArrayWrite(xx,&xarray);CHKERRQ(ierr); 998 ierr = VecCUDAGetArrayRead(bb,&barray);CHKERRQ(ierr); 999 xGPU = thrust::device_pointer_cast(xarray); 1000 bGPU = thrust::device_pointer_cast(barray); 1001 1002 /* First, reorder with the row permutation */ 1003 thrust::copy(thrust::make_permutation_iterator(bGPU, cusparseTriFactors->rpermIndices->begin()), 1004 thrust::make_permutation_iterator(bGPU+n, cusparseTriFactors->rpermIndices->end()), 1005 xGPU); 1006 1007 /* First, solve U */ 1008 stat = cusparse_solve(cusparseTriFactors->handle, upTriFactorT->solveOp, 1009 upTriFactorT->csrMat->num_rows, &PETSC_CUSPARSE_ONE, upTriFactorT->descr, 1010 upTriFactorT->csrMat->values->data().get(), 1011 upTriFactorT->csrMat->row_offsets->data().get(), 1012 upTriFactorT->csrMat->column_indices->data().get(), 1013 upTriFactorT->solveInfo, 1014 xarray, tempGPU->data().get());CHKERRCUDA(stat); 1015 1016 /* Then, solve L */ 1017 stat = cusparse_solve(cusparseTriFactors->handle, loTriFactorT->solveOp, 1018 loTriFactorT->csrMat->num_rows, &PETSC_CUSPARSE_ONE, loTriFactorT->descr, 1019 loTriFactorT->csrMat->values->data().get(), 1020 loTriFactorT->csrMat->row_offsets->data().get(), 1021 loTriFactorT->csrMat->column_indices->data().get(), 1022 loTriFactorT->solveInfo, 1023 tempGPU->data().get(), xarray);CHKERRCUDA(stat); 1024 1025 /* Last, copy the solution, xGPU, into a temporary with the column permutation ... can't be done in place. */ 1026 thrust::copy(thrust::make_permutation_iterator(xGPU, cusparseTriFactors->cpermIndices->begin()), 1027 thrust::make_permutation_iterator(xGPU+n, cusparseTriFactors->cpermIndices->end()), 1028 tempGPU->begin()); 1029 1030 /* Copy the temporary to the full solution. */ 1031 thrust::copy(tempGPU->begin(), tempGPU->end(), xGPU); 1032 1033 /* restore */ 1034 ierr = VecCUDARestoreArrayRead(bb,&barray);CHKERRQ(ierr); 1035 ierr = VecCUDARestoreArrayWrite(xx,&xarray);CHKERRQ(ierr); 1036 ierr = WaitForGPU();CHKERRCUDA(ierr); 1037 1038 ierr = PetscLogFlops(2.0*cusparseTriFactors->nnz - A->cmap->n);CHKERRQ(ierr); 1039 PetscFunctionReturn(0); 1040 } 1041 1042 static PetscErrorCode MatSolveTranspose_SeqAIJCUSPARSE_NaturalOrdering(Mat A,Vec bb,Vec xx) 1043 { 1044 const PetscScalar *barray; 1045 PetscScalar *xarray; 1046 cusparseStatus_t stat; 1047 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 1048 Mat_SeqAIJCUSPARSETriFactorStruct *loTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtrTranspose; 1049 Mat_SeqAIJCUSPARSETriFactorStruct *upTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtrTranspose; 1050 THRUSTARRAY *tempGPU = (THRUSTARRAY*)cusparseTriFactors->workVector; 1051 PetscErrorCode ierr; 1052 1053 PetscFunctionBegin; 1054 /* Analyze the matrix and create the transpose ... on the fly */ 1055 if (!loTriFactorT && !upTriFactorT) { 1056 ierr = MatSeqAIJCUSPARSEAnalyzeTransposeForSolve(A);CHKERRQ(ierr); 1057 loTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtrTranspose; 1058 upTriFactorT = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtrTranspose; 1059 } 1060 1061 /* Get the GPU pointers */ 1062 ierr = VecCUDAGetArrayWrite(xx,&xarray);CHKERRQ(ierr); 1063 ierr = VecCUDAGetArrayRead(bb,&barray);CHKERRQ(ierr); 1064 1065 /* First, solve U */ 1066 stat = cusparse_solve(cusparseTriFactors->handle, upTriFactorT->solveOp, 1067 upTriFactorT->csrMat->num_rows, &PETSC_CUSPARSE_ONE, upTriFactorT->descr, 1068 upTriFactorT->csrMat->values->data().get(), 1069 upTriFactorT->csrMat->row_offsets->data().get(), 1070 upTriFactorT->csrMat->column_indices->data().get(), 1071 upTriFactorT->solveInfo, 1072 barray, tempGPU->data().get());CHKERRCUDA(stat); 1073 1074 /* Then, solve L */ 1075 stat = cusparse_solve(cusparseTriFactors->handle, loTriFactorT->solveOp, 1076 loTriFactorT->csrMat->num_rows, &PETSC_CUSPARSE_ONE, loTriFactorT->descr, 1077 loTriFactorT->csrMat->values->data().get(), 1078 loTriFactorT->csrMat->row_offsets->data().get(), 1079 loTriFactorT->csrMat->column_indices->data().get(), 1080 loTriFactorT->solveInfo, 1081 tempGPU->data().get(), xarray);CHKERRCUDA(stat); 1082 1083 /* restore */ 1084 ierr = VecCUDARestoreArrayRead(bb,&barray);CHKERRQ(ierr); 1085 ierr = VecCUDARestoreArrayWrite(xx,&xarray);CHKERRQ(ierr); 1086 ierr = WaitForGPU();CHKERRCUDA(ierr); 1087 ierr = PetscLogFlops(2.0*cusparseTriFactors->nnz - A->cmap->n);CHKERRQ(ierr); 1088 PetscFunctionReturn(0); 1089 } 1090 1091 static PetscErrorCode MatSolve_SeqAIJCUSPARSE(Mat A,Vec bb,Vec xx) 1092 { 1093 const PetscScalar *barray; 1094 PetscScalar *xarray; 1095 thrust::device_ptr<const PetscScalar> bGPU; 1096 thrust::device_ptr<PetscScalar> xGPU; 1097 cusparseStatus_t stat; 1098 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 1099 Mat_SeqAIJCUSPARSETriFactorStruct *loTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtr; 1100 Mat_SeqAIJCUSPARSETriFactorStruct *upTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtr; 1101 THRUSTARRAY *tempGPU = (THRUSTARRAY*)cusparseTriFactors->workVector; 1102 PetscErrorCode ierr; 1103 1104 PetscFunctionBegin; 1105 1106 /* Get the GPU pointers */ 1107 ierr = VecCUDAGetArrayWrite(xx,&xarray);CHKERRQ(ierr); 1108 ierr = VecCUDAGetArrayRead(bb,&barray);CHKERRQ(ierr); 1109 xGPU = thrust::device_pointer_cast(xarray); 1110 bGPU = thrust::device_pointer_cast(barray); 1111 1112 /* First, reorder with the row permutation */ 1113 thrust::copy(thrust::make_permutation_iterator(bGPU, cusparseTriFactors->rpermIndices->begin()), 1114 thrust::make_permutation_iterator(bGPU, cusparseTriFactors->rpermIndices->end()), 1115 xGPU); 1116 1117 /* Next, solve L */ 1118 stat = cusparse_solve(cusparseTriFactors->handle, loTriFactor->solveOp, 1119 loTriFactor->csrMat->num_rows, &PETSC_CUSPARSE_ONE, loTriFactor->descr, 1120 loTriFactor->csrMat->values->data().get(), 1121 loTriFactor->csrMat->row_offsets->data().get(), 1122 loTriFactor->csrMat->column_indices->data().get(), 1123 loTriFactor->solveInfo, 1124 xarray, tempGPU->data().get());CHKERRCUDA(stat); 1125 1126 /* Then, solve U */ 1127 stat = cusparse_solve(cusparseTriFactors->handle, upTriFactor->solveOp, 1128 upTriFactor->csrMat->num_rows, &PETSC_CUSPARSE_ONE, upTriFactor->descr, 1129 upTriFactor->csrMat->values->data().get(), 1130 upTriFactor->csrMat->row_offsets->data().get(), 1131 upTriFactor->csrMat->column_indices->data().get(), 1132 upTriFactor->solveInfo, 1133 tempGPU->data().get(), xarray);CHKERRCUDA(stat); 1134 1135 /* Last, copy the solution, xGPU, into a temporary with the column permutation ... can't be done in place. */ 1136 thrust::copy(thrust::make_permutation_iterator(xGPU, cusparseTriFactors->cpermIndices->begin()), 1137 thrust::make_permutation_iterator(xGPU, cusparseTriFactors->cpermIndices->end()), 1138 tempGPU->begin()); 1139 1140 /* Copy the temporary to the full solution. */ 1141 thrust::copy(tempGPU->begin(), tempGPU->end(), xGPU); 1142 1143 ierr = VecCUDARestoreArrayRead(bb,&barray);CHKERRQ(ierr); 1144 ierr = VecCUDARestoreArrayWrite(xx,&xarray);CHKERRQ(ierr); 1145 ierr = WaitForGPU();CHKERRCUDA(ierr); 1146 ierr = PetscLogFlops(2.0*cusparseTriFactors->nnz - A->cmap->n);CHKERRQ(ierr); 1147 PetscFunctionReturn(0); 1148 } 1149 1150 static PetscErrorCode MatSolve_SeqAIJCUSPARSE_NaturalOrdering(Mat A,Vec bb,Vec xx) 1151 { 1152 const PetscScalar *barray; 1153 PetscScalar *xarray; 1154 cusparseStatus_t stat; 1155 Mat_SeqAIJCUSPARSETriFactors *cusparseTriFactors = (Mat_SeqAIJCUSPARSETriFactors*)A->spptr; 1156 Mat_SeqAIJCUSPARSETriFactorStruct *loTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->loTriFactorPtr; 1157 Mat_SeqAIJCUSPARSETriFactorStruct *upTriFactor = (Mat_SeqAIJCUSPARSETriFactorStruct*)cusparseTriFactors->upTriFactorPtr; 1158 THRUSTARRAY *tempGPU = (THRUSTARRAY*)cusparseTriFactors->workVector; 1159 PetscErrorCode ierr; 1160 1161 PetscFunctionBegin; 1162 /* Get the GPU pointers */ 1163 ierr = VecCUDAGetArrayWrite(xx,&xarray);CHKERRQ(ierr); 1164 ierr = VecCUDAGetArrayRead(bb,&barray);CHKERRQ(ierr); 1165 1166 /* First, solve L */ 1167 stat = cusparse_solve(cusparseTriFactors->handle, loTriFactor->solveOp, 1168 loTriFactor->csrMat->num_rows, &PETSC_CUSPARSE_ONE, loTriFactor->descr, 1169 loTriFactor->csrMat->values->data().get(), 1170 loTriFactor->csrMat->row_offsets->data().get(), 1171 loTriFactor->csrMat->column_indices->data().get(), 1172 loTriFactor->solveInfo, 1173 barray, tempGPU->data().get());CHKERRCUDA(stat); 1174 1175 /* Next, solve U */ 1176 stat = cusparse_solve(cusparseTriFactors->handle, upTriFactor->solveOp, 1177 upTriFactor->csrMat->num_rows, &PETSC_CUSPARSE_ONE, upTriFactor->descr, 1178 upTriFactor->csrMat->values->data().get(), 1179 upTriFactor->csrMat->row_offsets->data().get(), 1180 upTriFactor->csrMat->column_indices->data().get(), 1181 upTriFactor->solveInfo, 1182 tempGPU->data().get(), xarray);CHKERRCUDA(stat); 1183 1184 ierr = VecCUDARestoreArrayRead(bb,&barray);CHKERRQ(ierr); 1185 ierr = VecCUDARestoreArrayWrite(xx,&xarray);CHKERRQ(ierr); 1186 ierr = WaitForGPU();CHKERRCUDA(ierr); 1187 ierr = PetscLogFlops(2.0*cusparseTriFactors->nnz - A->cmap->n);CHKERRQ(ierr); 1188 PetscFunctionReturn(0); 1189 } 1190 1191 static PetscErrorCode MatSeqAIJCUSPARSECopyToGPU(Mat A) 1192 { 1193 1194 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 1195 Mat_SeqAIJCUSPARSEMultStruct *matstruct = (Mat_SeqAIJCUSPARSEMultStruct*)cusparsestruct->mat; 1196 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 1197 PetscInt m = A->rmap->n,*ii,*ridx; 1198 PetscErrorCode ierr; 1199 cusparseStatus_t stat; 1200 cudaError_t err; 1201 1202 PetscFunctionBegin; 1203 if (A->pinnedtocpu) PetscFunctionReturn(0); 1204 if (A->valid_GPU_matrix == PETSC_OFFLOAD_UNALLOCATED || A->valid_GPU_matrix == PETSC_OFFLOAD_CPU) { 1205 ierr = PetscLogEventBegin(MAT_CUSPARSECopyToGPU,A,0,0,0);CHKERRQ(ierr); 1206 if (A->assembled && A->nonzerostate == cusparsestruct->nonzerostate && cusparsestruct->format == MAT_CUSPARSE_CSR) { 1207 CsrMatrix *matrix = (CsrMatrix*)matstruct->mat; 1208 /* copy values only */ 1209 matrix->values->assign(a->a, a->a+a->nz); 1210 ierr = PetscLogCpuToGpu((a->nz)*sizeof(PetscScalar));CHKERRQ(ierr); 1211 } else { 1212 MatSeqAIJCUSPARSEMultStruct_Destroy(&matstruct,cusparsestruct->format); 1213 try { 1214 cusparsestruct->nonzerorow=0; 1215 for (int j = 0; j<m; j++) cusparsestruct->nonzerorow += ((a->i[j+1]-a->i[j])>0); 1216 1217 if (a->compressedrow.use) { 1218 m = a->compressedrow.nrows; 1219 ii = a->compressedrow.i; 1220 ridx = a->compressedrow.rindex; 1221 } else { 1222 /* Forcing compressed row on the GPU */ 1223 int k=0; 1224 ierr = PetscMalloc1(cusparsestruct->nonzerorow+1, &ii);CHKERRQ(ierr); 1225 ierr = PetscMalloc1(cusparsestruct->nonzerorow, &ridx);CHKERRQ(ierr); 1226 ii[0]=0; 1227 for (int j = 0; j<m; j++) { 1228 if ((a->i[j+1]-a->i[j])>0) { 1229 ii[k] = a->i[j]; 1230 ridx[k]= j; 1231 k++; 1232 } 1233 } 1234 ii[cusparsestruct->nonzerorow] = a->nz; 1235 m = cusparsestruct->nonzerorow; 1236 } 1237 1238 /* allocate space for the triangular factor information */ 1239 matstruct = new Mat_SeqAIJCUSPARSEMultStruct; 1240 stat = cusparseCreateMatDescr(&matstruct->descr);CHKERRCUDA(stat); 1241 stat = cusparseSetMatIndexBase(matstruct->descr, CUSPARSE_INDEX_BASE_ZERO);CHKERRCUDA(stat); 1242 stat = cusparseSetMatType(matstruct->descr, CUSPARSE_MATRIX_TYPE_GENERAL);CHKERRCUDA(stat); 1243 1244 err = cudaMalloc((void **)&(matstruct->alpha), sizeof(PetscScalar));CHKERRCUDA(err); 1245 err = cudaMalloc((void **)&(matstruct->beta_zero),sizeof(PetscScalar));CHKERRCUDA(err); 1246 err = cudaMalloc((void **)&(matstruct->beta_one), sizeof(PetscScalar));CHKERRCUDA(err); 1247 err = cudaMemcpy(matstruct->alpha, &PETSC_CUSPARSE_ONE, sizeof(PetscScalar),cudaMemcpyHostToDevice);CHKERRCUDA(err); 1248 err = cudaMemcpy(matstruct->beta_zero,&PETSC_CUSPARSE_ZERO,sizeof(PetscScalar),cudaMemcpyHostToDevice);CHKERRCUDA(err); 1249 err = cudaMemcpy(matstruct->beta_one, &PETSC_CUSPARSE_ONE, sizeof(PetscScalar),cudaMemcpyHostToDevice);CHKERRCUDA(err); 1250 stat = cusparseSetPointerMode(cusparsestruct->handle, CUSPARSE_POINTER_MODE_DEVICE);CHKERRCUDA(stat); 1251 1252 /* Build a hybrid/ellpack matrix if this option is chosen for the storage */ 1253 if (cusparsestruct->format==MAT_CUSPARSE_CSR) { 1254 /* set the matrix */ 1255 CsrMatrix *matrix= new CsrMatrix; 1256 matrix->num_rows = m; 1257 matrix->num_cols = A->cmap->n; 1258 matrix->num_entries = a->nz; 1259 matrix->row_offsets = new THRUSTINTARRAY32(m+1); 1260 matrix->row_offsets->assign(ii, ii + m+1); 1261 1262 matrix->column_indices = new THRUSTINTARRAY32(a->nz); 1263 matrix->column_indices->assign(a->j, a->j+a->nz); 1264 1265 matrix->values = new THRUSTARRAY(a->nz); 1266 matrix->values->assign(a->a, a->a+a->nz); 1267 1268 /* assign the pointer */ 1269 matstruct->mat = matrix; 1270 1271 } else if (cusparsestruct->format==MAT_CUSPARSE_ELL || cusparsestruct->format==MAT_CUSPARSE_HYB) { 1272 #if CUDA_VERSION>=4020 1273 CsrMatrix *matrix= new CsrMatrix; 1274 matrix->num_rows = m; 1275 matrix->num_cols = A->cmap->n; 1276 matrix->num_entries = a->nz; 1277 matrix->row_offsets = new THRUSTINTARRAY32(m+1); 1278 matrix->row_offsets->assign(ii, ii + m+1); 1279 1280 matrix->column_indices = new THRUSTINTARRAY32(a->nz); 1281 matrix->column_indices->assign(a->j, a->j+a->nz); 1282 1283 matrix->values = new THRUSTARRAY(a->nz); 1284 matrix->values->assign(a->a, a->a+a->nz); 1285 1286 cusparseHybMat_t hybMat; 1287 stat = cusparseCreateHybMat(&hybMat);CHKERRCUDA(stat); 1288 cusparseHybPartition_t partition = cusparsestruct->format==MAT_CUSPARSE_ELL ? 1289 CUSPARSE_HYB_PARTITION_MAX : CUSPARSE_HYB_PARTITION_AUTO; 1290 stat = cusparse_csr2hyb(cusparsestruct->handle, matrix->num_rows, matrix->num_cols, 1291 matstruct->descr, matrix->values->data().get(), 1292 matrix->row_offsets->data().get(), 1293 matrix->column_indices->data().get(), 1294 hybMat, 0, partition);CHKERRCUDA(stat); 1295 /* assign the pointer */ 1296 matstruct->mat = hybMat; 1297 1298 if (matrix) { 1299 if (matrix->values) delete (THRUSTARRAY*)matrix->values; 1300 if (matrix->column_indices) delete (THRUSTINTARRAY32*)matrix->column_indices; 1301 if (matrix->row_offsets) delete (THRUSTINTARRAY32*)matrix->row_offsets; 1302 delete (CsrMatrix*)matrix; 1303 } 1304 #endif 1305 } 1306 1307 /* assign the compressed row indices */ 1308 matstruct->cprowIndices = new THRUSTINTARRAY(m); 1309 matstruct->cprowIndices->assign(ridx,ridx+m); 1310 ierr = PetscLogCpuToGpu(((m+1)+(a->nz))*sizeof(int)+m*sizeof(PetscInt)+(3+(a->nz))*sizeof(PetscScalar));CHKERRQ(ierr); 1311 1312 /* assign the pointer */ 1313 cusparsestruct->mat = matstruct; 1314 1315 if (!a->compressedrow.use) { 1316 ierr = PetscFree(ii);CHKERRQ(ierr); 1317 ierr = PetscFree(ridx);CHKERRQ(ierr); 1318 } 1319 cusparsestruct->workVector = new THRUSTARRAY(m); 1320 } catch(char *ex) { 1321 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"CUSPARSE error: %s", ex); 1322 } 1323 cusparsestruct->nonzerostate = A->nonzerostate; 1324 } 1325 ierr = WaitForGPU();CHKERRCUDA(ierr); 1326 A->valid_GPU_matrix = PETSC_OFFLOAD_BOTH; 1327 ierr = PetscLogEventEnd(MAT_CUSPARSECopyToGPU,A,0,0,0);CHKERRQ(ierr); 1328 } 1329 PetscFunctionReturn(0); 1330 } 1331 1332 struct VecCUDAPlusEquals 1333 { 1334 template <typename Tuple> 1335 __host__ __device__ 1336 void operator()(Tuple t) 1337 { 1338 thrust::get<1>(t) = thrust::get<1>(t) + thrust::get<0>(t); 1339 } 1340 }; 1341 1342 static PetscErrorCode MatMult_SeqAIJCUSPARSE(Mat A,Vec xx,Vec yy) 1343 { 1344 PetscErrorCode ierr; 1345 1346 PetscFunctionBegin; 1347 ierr = MatMultAdd_SeqAIJCUSPARSE(A,xx,NULL,yy);CHKERRQ(ierr); 1348 PetscFunctionReturn(0); 1349 } 1350 1351 static PetscErrorCode MatMultTranspose_SeqAIJCUSPARSE(Mat A,Vec xx,Vec yy) 1352 { 1353 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 1354 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 1355 Mat_SeqAIJCUSPARSEMultStruct *matstructT; 1356 const PetscScalar *xarray; 1357 PetscScalar *yarray; 1358 PetscErrorCode ierr; 1359 cusparseStatus_t stat; 1360 1361 PetscFunctionBegin; 1362 /* The line below is necessary due to the operations that modify the matrix on the CPU (axpy, scale, etc) */ 1363 ierr = MatSeqAIJCUSPARSECopyToGPU(A);CHKERRQ(ierr); 1364 matstructT = (Mat_SeqAIJCUSPARSEMultStruct*)cusparsestruct->matTranspose; 1365 if (!matstructT) { 1366 ierr = MatSeqAIJCUSPARSEGenerateTransposeForMult(A);CHKERRQ(ierr); 1367 matstructT = (Mat_SeqAIJCUSPARSEMultStruct*)cusparsestruct->matTranspose; 1368 } 1369 ierr = VecCUDAGetArrayRead(xx,&xarray);CHKERRQ(ierr); 1370 ierr = VecSet(yy,0);CHKERRQ(ierr); 1371 ierr = VecCUDAGetArrayWrite(yy,&yarray);CHKERRQ(ierr); 1372 1373 if (cusparsestruct->format==MAT_CUSPARSE_CSR) { 1374 CsrMatrix *mat = (CsrMatrix*)matstructT->mat; 1375 stat = cusparse_csr_spmv(cusparsestruct->handle, CUSPARSE_OPERATION_NON_TRANSPOSE, 1376 mat->num_rows, mat->num_cols, 1377 mat->num_entries, matstructT->alpha, matstructT->descr, 1378 mat->values->data().get(), mat->row_offsets->data().get(), 1379 mat->column_indices->data().get(), xarray, matstructT->beta_zero, 1380 yarray);CHKERRCUDA(stat); 1381 } else { 1382 #if CUDA_VERSION>=4020 1383 cusparseHybMat_t hybMat = (cusparseHybMat_t)matstructT->mat; 1384 stat = cusparse_hyb_spmv(cusparsestruct->handle, CUSPARSE_OPERATION_NON_TRANSPOSE, 1385 matstructT->alpha, matstructT->descr, hybMat, 1386 xarray, matstructT->beta_zero, 1387 yarray);CHKERRCUDA(stat); 1388 #endif 1389 } 1390 ierr = VecCUDARestoreArrayRead(xx,&xarray);CHKERRQ(ierr); 1391 ierr = VecCUDARestoreArrayWrite(yy,&yarray);CHKERRQ(ierr); 1392 if (!cusparsestruct->stream) { 1393 ierr = WaitForGPU();CHKERRCUDA(ierr); 1394 } 1395 ierr = PetscLogFlops(2.0*a->nz - cusparsestruct->nonzerorow);CHKERRQ(ierr); 1396 PetscFunctionReturn(0); 1397 } 1398 1399 1400 static PetscErrorCode MatMultAdd_SeqAIJCUSPARSE(Mat A,Vec xx,Vec yy,Vec zz) 1401 { 1402 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 1403 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 1404 Mat_SeqAIJCUSPARSEMultStruct *matstruct; 1405 const PetscScalar *xarray; 1406 PetscScalar *zarray,*dptr,*beta; 1407 PetscErrorCode ierr; 1408 cusparseStatus_t stat; 1409 1410 PetscFunctionBegin; 1411 /* The line below is necessary due to the operations that modify the matrix on the CPU (axpy, scale, etc) */ 1412 ierr = MatSeqAIJCUSPARSECopyToGPU(A);CHKERRQ(ierr); 1413 matstruct = (Mat_SeqAIJCUSPARSEMultStruct*)cusparsestruct->mat; 1414 try { 1415 ierr = VecCUDAGetArrayRead(xx,&xarray);CHKERRQ(ierr); 1416 ierr = VecCUDAGetArray(zz,&zarray);CHKERRQ(ierr); 1417 dptr = cusparsestruct->workVector->size() == (thrust::detail::vector_base<PetscScalar, thrust::device_malloc_allocator<PetscScalar> >::size_type)(A->rmap->n) ? zarray : cusparsestruct->workVector->data().get(); 1418 beta = (yy == zz && dptr == zarray) ? matstruct->beta_one : matstruct->beta_zero; 1419 1420 /* csr_spmv is multiply add */ 1421 if (cusparsestruct->format == MAT_CUSPARSE_CSR) { 1422 /* here we need to be careful to set the number of rows in the multiply to the 1423 number of compressed rows in the matrix ... which is equivalent to the 1424 size of the workVector */ 1425 CsrMatrix *mat = (CsrMatrix*)matstruct->mat; 1426 stat = cusparse_csr_spmv(cusparsestruct->handle, CUSPARSE_OPERATION_NON_TRANSPOSE, 1427 mat->num_rows, mat->num_cols, 1428 mat->num_entries, matstruct->alpha, matstruct->descr, 1429 mat->values->data().get(), mat->row_offsets->data().get(), 1430 mat->column_indices->data().get(), xarray, beta, 1431 dptr);CHKERRCUDA(stat); 1432 } else { 1433 #if CUDA_VERSION>=4020 1434 cusparseHybMat_t hybMat = (cusparseHybMat_t)matstruct->mat; 1435 if (cusparsestruct->workVector->size()) { 1436 stat = cusparse_hyb_spmv(cusparsestruct->handle, CUSPARSE_OPERATION_NON_TRANSPOSE, 1437 matstruct->alpha, matstruct->descr, hybMat, 1438 xarray, beta, 1439 dptr);CHKERRCUDA(stat); 1440 } 1441 #endif 1442 } 1443 1444 if (yy) { 1445 if (dptr != zarray) { 1446 ierr = VecCopy_SeqCUDA(yy,zz);CHKERRQ(ierr); 1447 } else if (zz != yy) { 1448 ierr = VecAXPY_SeqCUDA(zz,1.0,yy);CHKERRQ(ierr); 1449 } 1450 } else if (dptr != zarray) { 1451 ierr = VecSet(zz,0);CHKERRQ(ierr); 1452 } 1453 /* scatter the data from the temporary into the full vector with a += operation */ 1454 if (dptr != zarray) { 1455 thrust::device_ptr<PetscScalar> zptr; 1456 1457 zptr = thrust::device_pointer_cast(zarray); 1458 thrust::for_each(thrust::make_zip_iterator(thrust::make_tuple(cusparsestruct->workVector->begin(), thrust::make_permutation_iterator(zptr, matstruct->cprowIndices->begin()))), 1459 thrust::make_zip_iterator(thrust::make_tuple(cusparsestruct->workVector->begin(), thrust::make_permutation_iterator(zptr, matstruct->cprowIndices->begin()))) + cusparsestruct->workVector->size(), 1460 VecCUDAPlusEquals()); 1461 } 1462 ierr = VecCUDARestoreArrayRead(xx,&xarray);CHKERRQ(ierr); 1463 ierr = VecCUDARestoreArray(zz,&zarray);CHKERRQ(ierr); 1464 } catch(char *ex) { 1465 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"CUSPARSE error: %s", ex); 1466 } 1467 if (!yy) { /* MatMult */ 1468 if (!cusparsestruct->stream) { 1469 ierr = WaitForGPU();CHKERRCUDA(ierr); 1470 } 1471 } 1472 ierr = PetscLogFlops(2.0*a->nz);CHKERRQ(ierr); 1473 PetscFunctionReturn(0); 1474 } 1475 1476 static PetscErrorCode MatMultTransposeAdd_SeqAIJCUSPARSE(Mat A,Vec xx,Vec yy,Vec zz) 1477 { 1478 Mat_SeqAIJ *a = (Mat_SeqAIJ*)A->data; 1479 Mat_SeqAIJCUSPARSE *cusparsestruct = (Mat_SeqAIJCUSPARSE*)A->spptr; 1480 Mat_SeqAIJCUSPARSEMultStruct *matstructT; 1481 thrust::device_ptr<PetscScalar> zptr; 1482 const PetscScalar *xarray; 1483 PetscScalar *zarray; 1484 PetscErrorCode ierr; 1485 cusparseStatus_t stat; 1486 1487 PetscFunctionBegin; 1488 /* The line below is necessary due to the operations that modify the matrix on the CPU (axpy, scale, etc) */ 1489 ierr = MatSeqAIJCUSPARSECopyToGPU(A);CHKERRQ(ierr); 1490 matstructT = (Mat_SeqAIJCUSPARSEMultStruct*)cusparsestruct->matTranspose; 1491 if (!matstructT) { 1492 ierr = MatSeqAIJCUSPARSEGenerateTransposeForMult(A);CHKERRQ(ierr); 1493 matstructT = (Mat_SeqAIJCUSPARSEMultStruct*)cusparsestruct->matTranspose; 1494 } 1495 1496 try { 1497 ierr = VecCopy_SeqCUDA(yy,zz);CHKERRQ(ierr); 1498 ierr = VecCUDAGetArrayRead(xx,&xarray);CHKERRQ(ierr); 1499 ierr = VecCUDAGetArray(zz,&zarray);CHKERRQ(ierr); 1500 zptr = thrust::device_pointer_cast(zarray); 1501 1502 /* multiply add with matrix transpose */ 1503 if (cusparsestruct->format==MAT_CUSPARSE_CSR) { 1504 CsrMatrix *mat = (CsrMatrix*)matstructT->mat; 1505 /* here we need to be careful to set the number of rows in the multiply to the 1506 number of compressed rows in the matrix ... which is equivalent to the 1507 size of the workVector */ 1508 stat = cusparse_csr_spmv(cusparsestruct->handle, CUSPARSE_OPERATION_NON_TRANSPOSE, 1509 mat->num_rows, mat->num_cols, 1510 mat->num_entries, matstructT->alpha, matstructT->descr, 1511 mat->values->data().get(), mat->row_offsets->data().get(), 1512 mat->column_indices->data().get(), xarray, matstructT->beta_zero, 1513 cusparsestruct->workVector->data().get());CHKERRCUDA(stat); 1514 } else { 1515 #if CUDA_VERSION>=4020 1516 cusparseHybMat_t hybMat = (cusparseHybMat_t)matstructT->mat; 1517 if (cusparsestruct->workVector->size()) { 1518 stat = cusparse_hyb_spmv(cusparsestruct->handle, CUSPARSE_OPERATION_NON_TRANSPOSE, 1519 matstructT->alpha, matstructT->descr, hybMat, 1520 xarray, matstructT->beta_zero, 1521 cusparsestruct->workVector->data().get());CHKERRCUDA(stat); 1522 } 1523 #endif 1524 } 1525 1526 /* scatter the data from the temporary into the full vector with a += operation */ 1527 thrust::for_each(thrust::make_zip_iterator(thrust::make_tuple(cusparsestruct->workVector->begin(), thrust::make_permutation_iterator(zptr, matstructT->cprowIndices->begin()))), 1528 thrust::make_zip_iterator(thrust::make_tuple(cusparsestruct->workVector->begin(), thrust::make_permutation_iterator(zptr, matstructT->cprowIndices->begin()))) + A->cmap->n, 1529 VecCUDAPlusEquals()); 1530 1531 ierr = VecCUDARestoreArrayRead(xx,&xarray);CHKERRQ(ierr); 1532 ierr = VecCUDARestoreArray(zz,&zarray);CHKERRQ(ierr); 1533 1534 } catch(char *ex) { 1535 SETERRQ1(PETSC_COMM_SELF,PETSC_ERR_LIB,"CUSPARSE error: %s", ex); 1536 } 1537 ierr = WaitForGPU();CHKERRCUDA(ierr); 1538 ierr = PetscLogFlops(2.0*a->nz);CHKERRQ(ierr); 1539 PetscFunctionReturn(0); 1540 } 1541 1542 static PetscErrorCode MatAssemblyEnd_SeqAIJCUSPARSE(Mat A,MatAssemblyType mode) 1543 { 1544 PetscErrorCode ierr; 1545 1546 PetscFunctionBegin; 1547 ierr = MatAssemblyEnd_SeqAIJ(A,mode);CHKERRQ(ierr); 1548 if (A->factortype==MAT_FACTOR_NONE) { 1549 ierr = MatSeqAIJCUSPARSECopyToGPU(A);CHKERRQ(ierr); 1550 } 1551 if (mode == MAT_FLUSH_ASSEMBLY) PetscFunctionReturn(0); 1552 A->ops->mult = MatMult_SeqAIJCUSPARSE; 1553 A->ops->multadd = MatMultAdd_SeqAIJCUSPARSE; 1554 A->ops->multtranspose = MatMultTranspose_SeqAIJCUSPARSE; 1555 A->ops->multtransposeadd = MatMultTransposeAdd_SeqAIJCUSPARSE; 1556 PetscFunctionReturn(0); 1557 } 1558 1559 /* --------------------------------------------------------------------------------*/ 1560 /*@ 1561 MatCreateSeqAIJCUSPARSE - Creates a sparse matrix in AIJ (compressed row) format 1562 (the default parallel PETSc format). This matrix will ultimately pushed down 1563 to NVidia GPUs and use the CUSPARSE library for calculations. For good matrix 1564 assembly performance the user should preallocate the matrix storage by setting 1565 the parameter nz (or the array nnz). By setting these parameters accurately, 1566 performance during matrix assembly can be increased by more than a factor of 50. 1567 1568 Collective 1569 1570 Input Parameters: 1571 + comm - MPI communicator, set to PETSC_COMM_SELF 1572 . m - number of rows 1573 . n - number of columns 1574 . nz - number of nonzeros per row (same for all rows) 1575 - nnz - array containing the number of nonzeros in the various rows 1576 (possibly different for each row) or NULL 1577 1578 Output Parameter: 1579 . A - the matrix 1580 1581 It is recommended that one use the MatCreate(), MatSetType() and/or MatSetFromOptions(), 1582 MatXXXXSetPreallocation() paradgm instead of this routine directly. 1583 [MatXXXXSetPreallocation() is, for example, MatSeqAIJSetPreallocation] 1584 1585 Notes: 1586 If nnz is given then nz is ignored 1587 1588 The AIJ format (also called the Yale sparse matrix format or 1589 compressed row storage), is fully compatible with standard Fortran 77 1590 storage. That is, the stored row and column indices can begin at 1591 either one (as in Fortran) or zero. See the users' manual for details. 1592 1593 Specify the preallocated storage with either nz or nnz (not both). 1594 Set nz=PETSC_DEFAULT and nnz=NULL for PETSc to control dynamic memory 1595 allocation. For large problems you MUST preallocate memory or you 1596 will get TERRIBLE performance, see the users' manual chapter on matrices. 1597 1598 By default, this format uses inodes (identical nodes) when possible, to 1599 improve numerical efficiency of matrix-vector products and solves. We 1600 search for consecutive rows with the same nonzero structure, thereby 1601 reusing matrix information to achieve increased efficiency. 1602 1603 Level: intermediate 1604 1605 .seealso: MatCreate(), MatCreateAIJ(), MatSetValues(), MatSeqAIJSetColumnIndices(), MatCreateSeqAIJWithArrays(), MatCreateAIJ(), MATSEQAIJCUSPARSE, MATAIJCUSPARSE 1606 @*/ 1607 PetscErrorCode MatCreateSeqAIJCUSPARSE(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt nz,const PetscInt nnz[],Mat *A) 1608 { 1609 PetscErrorCode ierr; 1610 1611 PetscFunctionBegin; 1612 ierr = MatCreate(comm,A);CHKERRQ(ierr); 1613 ierr = MatSetSizes(*A,m,n,m,n);CHKERRQ(ierr); 1614 ierr = MatSetType(*A,MATSEQAIJCUSPARSE);CHKERRQ(ierr); 1615 ierr = MatSeqAIJSetPreallocation_SeqAIJ(*A,nz,(PetscInt*)nnz);CHKERRQ(ierr); 1616 PetscFunctionReturn(0); 1617 } 1618 1619 static PetscErrorCode MatDestroy_SeqAIJCUSPARSE(Mat A) 1620 { 1621 PetscErrorCode ierr; 1622 1623 PetscFunctionBegin; 1624 if (A->factortype==MAT_FACTOR_NONE) { 1625 if (A->valid_GPU_matrix != PETSC_OFFLOAD_UNALLOCATED) { 1626 ierr = MatSeqAIJCUSPARSE_Destroy((Mat_SeqAIJCUSPARSE**)&A->spptr);CHKERRQ(ierr); 1627 } 1628 } else { 1629 ierr = MatSeqAIJCUSPARSETriFactors_Destroy((Mat_SeqAIJCUSPARSETriFactors**)&A->spptr);CHKERRQ(ierr); 1630 } 1631 ierr = MatDestroy_SeqAIJ(A);CHKERRQ(ierr); 1632 PetscFunctionReturn(0); 1633 } 1634 1635 static PetscErrorCode MatDuplicate_SeqAIJCUSPARSE(Mat A,MatDuplicateOption cpvalues,Mat *B) 1636 { 1637 PetscErrorCode ierr; 1638 Mat C; 1639 cusparseStatus_t stat; 1640 cusparseHandle_t handle=0; 1641 1642 PetscFunctionBegin; 1643 ierr = MatDuplicate_SeqAIJ(A,cpvalues,B);CHKERRQ(ierr); 1644 C = *B; 1645 ierr = PetscFree(C->defaultvectype);CHKERRQ(ierr); 1646 ierr = PetscStrallocpy(VECCUDA,&C->defaultvectype);CHKERRQ(ierr); 1647 1648 /* inject CUSPARSE-specific stuff */ 1649 if (C->factortype==MAT_FACTOR_NONE) { 1650 /* you cannot check the inode.use flag here since the matrix was just created. 1651 now build a GPU matrix data structure */ 1652 C->spptr = new Mat_SeqAIJCUSPARSE; 1653 ((Mat_SeqAIJCUSPARSE*)C->spptr)->mat = 0; 1654 ((Mat_SeqAIJCUSPARSE*)C->spptr)->matTranspose = 0; 1655 ((Mat_SeqAIJCUSPARSE*)C->spptr)->workVector = 0; 1656 ((Mat_SeqAIJCUSPARSE*)C->spptr)->format = MAT_CUSPARSE_CSR; 1657 ((Mat_SeqAIJCUSPARSE*)C->spptr)->stream = 0; 1658 stat = cusparseCreate(&handle);CHKERRCUDA(stat); 1659 ((Mat_SeqAIJCUSPARSE*)C->spptr)->handle = handle; 1660 ((Mat_SeqAIJCUSPARSE*)C->spptr)->nonzerostate = 0; 1661 } else { 1662 /* NEXT, set the pointers to the triangular factors */ 1663 C->spptr = new Mat_SeqAIJCUSPARSETriFactors; 1664 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->loTriFactorPtr = 0; 1665 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->upTriFactorPtr = 0; 1666 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->loTriFactorPtrTranspose = 0; 1667 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->upTriFactorPtrTranspose = 0; 1668 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->rpermIndices = 0; 1669 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->cpermIndices = 0; 1670 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->workVector = 0; 1671 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->handle = 0; 1672 stat = cusparseCreate(&handle);CHKERRCUDA(stat); 1673 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->handle = handle; 1674 ((Mat_SeqAIJCUSPARSETriFactors*)C->spptr)->nnz = 0; 1675 } 1676 1677 C->ops->assemblyend = MatAssemblyEnd_SeqAIJCUSPARSE; 1678 C->ops->destroy = MatDestroy_SeqAIJCUSPARSE; 1679 C->ops->setfromoptions = MatSetFromOptions_SeqAIJCUSPARSE; 1680 C->ops->mult = MatMult_SeqAIJCUSPARSE; 1681 C->ops->multadd = MatMultAdd_SeqAIJCUSPARSE; 1682 C->ops->multtranspose = MatMultTranspose_SeqAIJCUSPARSE; 1683 C->ops->multtransposeadd = MatMultTransposeAdd_SeqAIJCUSPARSE; 1684 C->ops->duplicate = MatDuplicate_SeqAIJCUSPARSE; 1685 1686 ierr = PetscObjectChangeTypeName((PetscObject)C,MATSEQAIJCUSPARSE);CHKERRQ(ierr); 1687 1688 C->valid_GPU_matrix = PETSC_OFFLOAD_UNALLOCATED; 1689 1690 ierr = PetscObjectComposeFunction((PetscObject)C, "MatCUSPARSESetFormat_C", MatCUSPARSESetFormat_SeqAIJCUSPARSE);CHKERRQ(ierr); 1691 PetscFunctionReturn(0); 1692 } 1693 1694 PETSC_EXTERN PetscErrorCode MatConvert_SeqAIJ_SeqAIJCUSPARSE(Mat B) 1695 { 1696 PetscErrorCode ierr; 1697 cusparseStatus_t stat; 1698 cusparseHandle_t handle=0; 1699 1700 PetscFunctionBegin; 1701 ierr = PetscFree(B->defaultvectype);CHKERRQ(ierr); 1702 ierr = PetscStrallocpy(VECCUDA,&B->defaultvectype);CHKERRQ(ierr); 1703 1704 if (B->factortype==MAT_FACTOR_NONE) { 1705 /* you cannot check the inode.use flag here since the matrix was just created. 1706 now build a GPU matrix data structure */ 1707 B->spptr = new Mat_SeqAIJCUSPARSE; 1708 ((Mat_SeqAIJCUSPARSE*)B->spptr)->mat = 0; 1709 ((Mat_SeqAIJCUSPARSE*)B->spptr)->matTranspose = 0; 1710 ((Mat_SeqAIJCUSPARSE*)B->spptr)->workVector = 0; 1711 ((Mat_SeqAIJCUSPARSE*)B->spptr)->format = MAT_CUSPARSE_CSR; 1712 ((Mat_SeqAIJCUSPARSE*)B->spptr)->stream = 0; 1713 stat = cusparseCreate(&handle);CHKERRCUDA(stat); 1714 ((Mat_SeqAIJCUSPARSE*)B->spptr)->handle = handle; 1715 ((Mat_SeqAIJCUSPARSE*)B->spptr)->nonzerostate = 0; 1716 } else { 1717 /* NEXT, set the pointers to the triangular factors */ 1718 B->spptr = new Mat_SeqAIJCUSPARSETriFactors; 1719 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->loTriFactorPtr = 0; 1720 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->upTriFactorPtr = 0; 1721 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->loTriFactorPtrTranspose = 0; 1722 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->upTriFactorPtrTranspose = 0; 1723 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->rpermIndices = 0; 1724 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->cpermIndices = 0; 1725 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->workVector = 0; 1726 stat = cusparseCreate(&handle);CHKERRCUDA(stat); 1727 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->handle = handle; 1728 ((Mat_SeqAIJCUSPARSETriFactors*)B->spptr)->nnz = 0; 1729 } 1730 1731 B->ops->assemblyend = MatAssemblyEnd_SeqAIJCUSPARSE; 1732 B->ops->destroy = MatDestroy_SeqAIJCUSPARSE; 1733 B->ops->setfromoptions = MatSetFromOptions_SeqAIJCUSPARSE; 1734 B->ops->mult = MatMult_SeqAIJCUSPARSE; 1735 B->ops->multadd = MatMultAdd_SeqAIJCUSPARSE; 1736 B->ops->multtranspose = MatMultTranspose_SeqAIJCUSPARSE; 1737 B->ops->multtransposeadd = MatMultTransposeAdd_SeqAIJCUSPARSE; 1738 B->ops->duplicate = MatDuplicate_SeqAIJCUSPARSE; 1739 1740 ierr = PetscObjectChangeTypeName((PetscObject)B,MATSEQAIJCUSPARSE);CHKERRQ(ierr); 1741 1742 B->valid_GPU_matrix = PETSC_OFFLOAD_UNALLOCATED; 1743 1744 ierr = PetscObjectComposeFunction((PetscObject)B, "MatCUSPARSESetFormat_C", MatCUSPARSESetFormat_SeqAIJCUSPARSE);CHKERRQ(ierr); 1745 PetscFunctionReturn(0); 1746 } 1747 1748 PETSC_EXTERN PetscErrorCode MatCreate_SeqAIJCUSPARSE(Mat B) 1749 { 1750 PetscErrorCode ierr; 1751 1752 PetscFunctionBegin; 1753 ierr = MatCreate_SeqAIJ(B);CHKERRQ(ierr); 1754 ierr = MatConvert_SeqAIJ_SeqAIJCUSPARSE(B); 1755 PetscFunctionReturn(0); 1756 } 1757 1758 /*MC 1759 MATSEQAIJCUSPARSE - MATAIJCUSPARSE = "(seq)aijcusparse" - A matrix type to be used for sparse matrices. 1760 1761 A matrix type type whose data resides on Nvidia GPUs. These matrices can be in either 1762 CSR, ELL, or Hybrid format. The ELL and HYB formats require CUDA 4.2 or later. 1763 All matrix calculations are performed on Nvidia GPUs using the CUSPARSE library. 1764 1765 Options Database Keys: 1766 + -mat_type aijcusparse - sets the matrix type to "seqaijcusparse" during a call to MatSetFromOptions() 1767 . -mat_cusparse_storage_format csr - sets the storage format of matrices (for MatMult and factors in MatSolve) during a call to MatSetFromOptions(). Other options include ell (ellpack) or hyb (hybrid). 1768 - -mat_cusparse_mult_storage_format csr - sets the storage format of matrices (for MatMult) during a call to MatSetFromOptions(). Other options include ell (ellpack) or hyb (hybrid). 1769 1770 Level: beginner 1771 1772 .seealso: MatCreateSeqAIJCUSPARSE(), MATAIJCUSPARSE, MatCreateAIJCUSPARSE(), MatCUSPARSESetFormat(), MatCUSPARSEStorageFormat, MatCUSPARSEFormatOperation 1773 M*/ 1774 1775 PETSC_EXTERN PetscErrorCode MatGetFactor_seqaijcusparse_cusparse(Mat,MatFactorType,Mat*); 1776 1777 1778 PETSC_EXTERN PetscErrorCode MatSolverTypeRegister_CUSPARSE(void) 1779 { 1780 PetscErrorCode ierr; 1781 1782 PetscFunctionBegin; 1783 ierr = MatSolverTypeRegister(MATSOLVERCUSPARSE,MATSEQAIJCUSPARSE,MAT_FACTOR_LU,MatGetFactor_seqaijcusparse_cusparse);CHKERRQ(ierr); 1784 ierr = MatSolverTypeRegister(MATSOLVERCUSPARSE,MATSEQAIJCUSPARSE,MAT_FACTOR_CHOLESKY,MatGetFactor_seqaijcusparse_cusparse);CHKERRQ(ierr); 1785 ierr = MatSolverTypeRegister(MATSOLVERCUSPARSE,MATSEQAIJCUSPARSE,MAT_FACTOR_ILU,MatGetFactor_seqaijcusparse_cusparse);CHKERRQ(ierr); 1786 ierr = MatSolverTypeRegister(MATSOLVERCUSPARSE,MATSEQAIJCUSPARSE,MAT_FACTOR_ICC,MatGetFactor_seqaijcusparse_cusparse);CHKERRQ(ierr); 1787 PetscFunctionReturn(0); 1788 } 1789 1790 1791 static PetscErrorCode MatSeqAIJCUSPARSE_Destroy(Mat_SeqAIJCUSPARSE **cusparsestruct) 1792 { 1793 cusparseStatus_t stat; 1794 cusparseHandle_t handle; 1795 1796 PetscFunctionBegin; 1797 if (*cusparsestruct) { 1798 MatSeqAIJCUSPARSEMultStruct_Destroy(&(*cusparsestruct)->mat,(*cusparsestruct)->format); 1799 MatSeqAIJCUSPARSEMultStruct_Destroy(&(*cusparsestruct)->matTranspose,(*cusparsestruct)->format); 1800 delete (*cusparsestruct)->workVector; 1801 if (handle = (*cusparsestruct)->handle) { 1802 stat = cusparseDestroy(handle);CHKERRCUDA(stat); 1803 } 1804 delete *cusparsestruct; 1805 *cusparsestruct = 0; 1806 } 1807 PetscFunctionReturn(0); 1808 } 1809 1810 static PetscErrorCode CsrMatrix_Destroy(CsrMatrix **mat) 1811 { 1812 PetscFunctionBegin; 1813 if (*mat) { 1814 delete (*mat)->values; 1815 delete (*mat)->column_indices; 1816 delete (*mat)->row_offsets; 1817 delete *mat; 1818 *mat = 0; 1819 } 1820 PetscFunctionReturn(0); 1821 } 1822 1823 static PetscErrorCode MatSeqAIJCUSPARSEMultStruct_Destroy(Mat_SeqAIJCUSPARSETriFactorStruct **trifactor) 1824 { 1825 cusparseStatus_t stat; 1826 PetscErrorCode ierr; 1827 1828 PetscFunctionBegin; 1829 if (*trifactor) { 1830 if ((*trifactor)->descr) { stat = cusparseDestroyMatDescr((*trifactor)->descr);CHKERRCUDA(stat); } 1831 if ((*trifactor)->solveInfo) { stat = cusparseDestroySolveAnalysisInfo((*trifactor)->solveInfo);CHKERRCUDA(stat); } 1832 ierr = CsrMatrix_Destroy(&(*trifactor)->csrMat);CHKERRQ(ierr); 1833 delete *trifactor; 1834 *trifactor = 0; 1835 } 1836 PetscFunctionReturn(0); 1837 } 1838 1839 static PetscErrorCode MatSeqAIJCUSPARSEMultStruct_Destroy(Mat_SeqAIJCUSPARSEMultStruct **matstruct,MatCUSPARSEStorageFormat format) 1840 { 1841 CsrMatrix *mat; 1842 cusparseStatus_t stat; 1843 cudaError_t err; 1844 1845 PetscFunctionBegin; 1846 if (*matstruct) { 1847 if ((*matstruct)->mat) { 1848 if (format==MAT_CUSPARSE_ELL || format==MAT_CUSPARSE_HYB) { 1849 cusparseHybMat_t hybMat = (cusparseHybMat_t)(*matstruct)->mat; 1850 stat = cusparseDestroyHybMat(hybMat);CHKERRCUDA(stat); 1851 } else { 1852 mat = (CsrMatrix*)(*matstruct)->mat; 1853 CsrMatrix_Destroy(&mat); 1854 } 1855 } 1856 if ((*matstruct)->descr) { stat = cusparseDestroyMatDescr((*matstruct)->descr);CHKERRCUDA(stat); } 1857 delete (*matstruct)->cprowIndices; 1858 if ((*matstruct)->alpha) { err=cudaFree((*matstruct)->alpha);CHKERRCUDA(err); } 1859 if ((*matstruct)->beta_zero) { err=cudaFree((*matstruct)->beta_zero);CHKERRCUDA(err); } 1860 if ((*matstruct)->beta_one) { err=cudaFree((*matstruct)->beta_one);CHKERRCUDA(err); } 1861 delete *matstruct; 1862 *matstruct = 0; 1863 } 1864 PetscFunctionReturn(0); 1865 } 1866 1867 static PetscErrorCode MatSeqAIJCUSPARSETriFactors_Destroy(Mat_SeqAIJCUSPARSETriFactors** trifactors) 1868 { 1869 cusparseHandle_t handle; 1870 cusparseStatus_t stat; 1871 1872 PetscFunctionBegin; 1873 if (*trifactors) { 1874 MatSeqAIJCUSPARSEMultStruct_Destroy(&(*trifactors)->loTriFactorPtr); 1875 MatSeqAIJCUSPARSEMultStruct_Destroy(&(*trifactors)->upTriFactorPtr); 1876 MatSeqAIJCUSPARSEMultStruct_Destroy(&(*trifactors)->loTriFactorPtrTranspose); 1877 MatSeqAIJCUSPARSEMultStruct_Destroy(&(*trifactors)->upTriFactorPtrTranspose); 1878 delete (*trifactors)->rpermIndices; 1879 delete (*trifactors)->cpermIndices; 1880 delete (*trifactors)->workVector; 1881 if (handle = (*trifactors)->handle) { 1882 stat = cusparseDestroy(handle);CHKERRCUDA(stat); 1883 } 1884 delete *trifactors; 1885 *trifactors = 0; 1886 } 1887 PetscFunctionReturn(0); 1888 } 1889 1890