1 #define PETSC_SKIP_IMMINTRIN_H_CUDAWORKAROUND 1 2 3 #include <petscconf.h> 4 #include <../src/mat/impls/aij/mpi/mpiaij.h> /*I "petscmat.h" I*/ 5 #include <../src/mat/impls/aij/seq/seqviennacl/viennaclmatimpl.h> 6 7 PetscErrorCode MatMPIAIJSetPreallocation_MPIAIJViennaCL(Mat B,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[]) 8 { 9 Mat_MPIAIJ *b = (Mat_MPIAIJ*)B->data; 10 PetscErrorCode ierr; 11 12 PetscFunctionBegin; 13 ierr = PetscLayoutSetUp(B->rmap);CHKERRQ(ierr); 14 ierr = PetscLayoutSetUp(B->cmap);CHKERRQ(ierr); 15 if (!B->preallocated) { 16 /* Explicitly create the two MATSEQAIJVIENNACL matrices. */ 17 ierr = MatCreate(PETSC_COMM_SELF,&b->A);CHKERRQ(ierr); 18 ierr = MatSetSizes(b->A,B->rmap->n,B->cmap->n,B->rmap->n,B->cmap->n);CHKERRQ(ierr); 19 ierr = MatSetType(b->A,MATSEQAIJVIENNACL);CHKERRQ(ierr); 20 ierr = PetscLogObjectParent((PetscObject)B,(PetscObject)b->A);CHKERRQ(ierr); 21 ierr = MatCreate(PETSC_COMM_SELF,&b->B);CHKERRQ(ierr); 22 ierr = MatSetSizes(b->B,B->rmap->n,B->cmap->N,B->rmap->n,B->cmap->N);CHKERRQ(ierr); 23 ierr = MatSetType(b->B,MATSEQAIJVIENNACL);CHKERRQ(ierr); 24 ierr = PetscLogObjectParent((PetscObject)B,(PetscObject)b->B);CHKERRQ(ierr); 25 } 26 ierr = MatSeqAIJSetPreallocation(b->A,d_nz,d_nnz);CHKERRQ(ierr); 27 ierr = MatSeqAIJSetPreallocation(b->B,o_nz,o_nnz);CHKERRQ(ierr); 28 B->preallocated = PETSC_TRUE; 29 PetscFunctionReturn(0); 30 } 31 32 PetscErrorCode MatAssemblyEnd_MPIAIJViennaCL(Mat A,MatAssemblyType mode) 33 { 34 Mat_MPIAIJ *b = (Mat_MPIAIJ*)A->data; 35 PetscErrorCode ierr; 36 PetscBool v; 37 38 PetscFunctionBegin; 39 ierr = MatAssemblyEnd_MPIAIJ(A,mode);CHKERRQ(ierr); 40 ierr = PetscObjectTypeCompare((PetscObject)b->lvec,VECSEQVIENNACL,&v);CHKERRQ(ierr); 41 if (!v) { 42 PetscInt m; 43 ierr = VecGetSize(b->lvec,&m);CHKERRQ(ierr); 44 ierr = VecDestroy(&b->lvec);CHKERRQ(ierr); 45 ierr = VecCreateSeqViennaCL(PETSC_COMM_SELF,m,&b->lvec);CHKERRQ(ierr); 46 } 47 PetscFunctionReturn(0); 48 } 49 50 PetscErrorCode MatDestroy_MPIAIJViennaCL(Mat A) 51 { 52 PetscErrorCode ierr; 53 54 PetscFunctionBegin; 55 ierr = MatDestroy_MPIAIJ(A);CHKERRQ(ierr); 56 PetscFunctionReturn(0); 57 } 58 59 PETSC_EXTERN PetscErrorCode MatCreate_MPIAIJViennaCL(Mat A) 60 { 61 PetscErrorCode ierr; 62 63 PetscFunctionBegin; 64 ierr = MatCreate_MPIAIJ(A);CHKERRQ(ierr); 65 A->boundtocpu = PETSC_FALSE; 66 ierr = PetscFree(A->defaultvectype);CHKERRQ(ierr); 67 ierr = PetscStrallocpy(VECVIENNACL,&A->defaultvectype);CHKERRQ(ierr); 68 ierr = PetscObjectComposeFunction((PetscObject)A,"MatMPIAIJSetPreallocation_C",MatMPIAIJSetPreallocation_MPIAIJViennaCL);CHKERRQ(ierr); 69 A->ops->assemblyend = MatAssemblyEnd_MPIAIJViennaCL; 70 ierr = PetscObjectChangeTypeName((PetscObject)A,MATMPIAIJVIENNACL);CHKERRQ(ierr); 71 PetscFunctionReturn(0); 72 } 73 74 75 /*@C 76 MatCreateAIJViennaCL - Creates a sparse matrix in AIJ (compressed row) format 77 (the default parallel PETSc format). This matrix will ultimately be pushed down 78 to GPUs and use the ViennaCL library for calculations. For good matrix 79 assembly performance the user should preallocate the matrix storage by setting 80 the parameter nz (or the array nnz). By setting these parameters accurately, 81 performance during matrix assembly can be increased substantially. 82 83 84 Collective 85 86 Input Parameters: 87 + comm - MPI communicator, set to PETSC_COMM_SELF 88 . m - number of rows 89 . n - number of columns 90 . nz - number of nonzeros per row (same for all rows) 91 - nnz - array containing the number of nonzeros in the various rows 92 (possibly different for each row) or NULL 93 94 Output Parameter: 95 . A - the matrix 96 97 It is recommended that one use the MatCreate(), MatSetType() and/or MatSetFromOptions(), 98 MatXXXXSetPreallocation() paradigm instead of this routine directly. 99 [MatXXXXSetPreallocation() is, for example, MatSeqAIJSetPreallocation] 100 101 Notes: 102 If nnz is given then nz is ignored 103 104 The AIJ format (also called the Yale sparse matrix format or 105 compressed row storage), is fully compatible with standard Fortran 77 106 storage. That is, the stored row and column indices can begin at 107 either one (as in Fortran) or zero. See the users' manual for details. 108 109 Specify the preallocated storage with either nz or nnz (not both). 110 Set nz=PETSC_DEFAULT and nnz=NULL for PETSc to control dynamic memory 111 allocation. For large problems you MUST preallocate memory or you 112 will get TERRIBLE performance, see the users' manual chapter on matrices. 113 114 Level: intermediate 115 116 .seealso: MatCreate(), MatCreateAIJ(), MatCreateAIJCUSPARSE(), MatSetValues(), MatSeqAIJSetColumnIndices(), MatCreateSeqAIJWithArrays(), MatCreateAIJ(), MATMPIAIJVIENNACL, MATAIJVIENNACL 117 @*/ 118 PetscErrorCode MatCreateAIJViennaCL(MPI_Comm comm,PetscInt m,PetscInt n,PetscInt M,PetscInt N,PetscInt d_nz,const PetscInt d_nnz[],PetscInt o_nz,const PetscInt o_nnz[],Mat *A) 119 { 120 PetscErrorCode ierr; 121 PetscMPIInt size; 122 123 PetscFunctionBegin; 124 ierr = MatCreate(comm,A);CHKERRQ(ierr); 125 ierr = MatSetSizes(*A,m,n,M,N);CHKERRQ(ierr); 126 ierr = MPI_Comm_size(comm,&size);CHKERRMPI(ierr); 127 if (size > 1) { 128 ierr = MatSetType(*A,MATMPIAIJVIENNACL);CHKERRQ(ierr); 129 ierr = MatMPIAIJSetPreallocation(*A,d_nz,d_nnz,o_nz,o_nnz);CHKERRQ(ierr); 130 } else { 131 ierr = MatSetType(*A,MATSEQAIJVIENNACL);CHKERRQ(ierr); 132 ierr = MatSeqAIJSetPreallocation(*A,d_nz,d_nnz);CHKERRQ(ierr); 133 } 134 PetscFunctionReturn(0); 135 } 136 137 /*MC 138 MATAIJVIENNACL - MATMPIAIJVIENNACL= "aijviennacl" = "mpiaijviennacl" - A matrix type to be used for sparse matrices. 139 140 A matrix type (CSR format) whose data resides on GPUs. 141 All matrix calculations are performed using the ViennaCL library. 142 143 This matrix type is identical to MATSEQAIJVIENNACL when constructed with a single process communicator, 144 and MATMPIAIJVIENNACL otherwise. As a result, for single process communicators, 145 MatSeqAIJSetPreallocation is supported, and similarly MatMPIAIJSetPreallocation is supported 146 for communicators controlling multiple processes. It is recommended that you call both of 147 the above preallocation routines for simplicity. 148 149 Options Database Keys: 150 . -mat_type mpiaijviennacl - sets the matrix type to "mpiaijviennacl" during a call to MatSetFromOptions() 151 152 Level: beginner 153 154 .seealso: MatCreateAIJViennaCL(), MATSEQAIJVIENNACL, MatCreateSeqAIJVIENNACL() 155 M*/ 156