Distributed SuperLU data types and function prototypes. More...

#include "superlu_defs.h"

Include dependency graph for superlu_sdefs.h:

Classes
struct	sScalePermstruct_t

struct	sLocalLU_t

struct	sLUValSubBuf_t

struct	strf3Dpartition_t

struct	sLUstruct_t

struct	psgsmv_comm_t

struct	sSOLVEstruct_t

struct	sscuBufs_t

struct	sdiagFactBufs_t

struct	sxT_struct

struct	slsumBmod_buff_t

Macros
#define	MAX_LOOKAHEADS 50

Typedefs
typedef struct sxT_struct	sxT_struct

typedef struct slsumBmod_buff_t	slsumBmod_buff_t

Functions
int_t	scuStatUpdate (int_t knsupc, HyP_t HyP, SCT_t SCT, SuperLUStat_t *stat)

void	sCreate_CompCol_Matrix_dist (SuperMatrix , int_t, int_t, int_t, float , int_t , int_t , Stype_t, Dtype_t, Mtype_t)

void	sCreate_CompRowLoc_Matrix_dist (SuperMatrix , int_t, int_t, int_t, int_t, int_t, float , int_t , int_t , Stype_t, Dtype_t, Mtype_t)

void	sCompRow_to_CompCol_dist (int_t, int_t, int_t, float , int_t , int_t , float , int_t , int_t *)

void	sCompCol_to_CompRow_dist (int_t m, int_t n, int_t nnz, float a, int_t colptr, int_t rowind, float at, int_t rowptr, int_t *colind)

int	psCompRow_loc_to_CompCol_global (int_t, SuperMatrix , gridinfo_t , SuperMatrix *)
	Gather A from the distributed compressed row format to global A in compressed column format. More...

void	sCopy_CompCol_Matrix_dist (SuperMatrix , SuperMatrix )

void	sCreate_Dense_Matrix_dist (SuperMatrix , int_t, int_t, float , int_t, Stype_t, Dtype_t, Mtype_t)

void	sCreate_SuperNode_Matrix_dist (SuperMatrix , int_t, int_t, int_t, float , int_t , int_t , int_t , int_t , int_t *, Stype_t, Dtype_t, Mtype_t)

void	sCopy_Dense_Matrix_dist (int_t, int_t, float , int_t, float , int_t)

void	sallocateA_dist (int_t, int_t, float , int_t , int_t **)

void	sGenXtrue_dist (int_t, int_t, float *, int_t)

void	sFillRHS_dist (char , int_t, float , int_t, SuperMatrix , float , int_t)
	Let rhs[i] = sum of i-th row of A, so the solution vector is all 1's. More...

int	screate_matrix (SuperMatrix , int, float , int , float *, int , FILE , gridinfo_t )

int	screate_matrix_rb (SuperMatrix , int, float , int , float *, int , FILE , gridinfo_t )

int	screate_matrix_dat (SuperMatrix , int, float , int , float *, int , FILE , gridinfo_t )

int	screate_matrix_postfix (SuperMatrix , int, float , int , float *, int , FILE , char , gridinfo_t *)

void	sScalePermstructInit (const int_t, const int_t, sScalePermstruct_t *)
	Allocate storage in ScalePermstruct. More...

void	sScalePermstructFree (sScalePermstruct_t *)
	Deallocate ScalePermstruct. More...

void	sgsequ_dist (SuperMatrix , float , float , float , float , float , int *)

float	slangs_dist (char , SuperMatrix )

void	slaqgs_dist (SuperMatrix , float , float , float, float, float, char )

void	psgsequ (SuperMatrix , float , float , float , float , float , int , gridinfo_t )

float	pslangs (char , SuperMatrix , gridinfo_t *)

void	pslaqgs (SuperMatrix , float , float , float, float, float, char )

int	psPermute_Dense_Matrix (int_t, int_t, int_t[], int_t[], float[], int, float[], int, int, gridinfo_t *)
	Permute the distributed dense matrix: B <= perm(X). perm[i] = j means the i-th row of X is in the j-th row of B. More...

int	sp_strsv_dist (char , char , char , SuperMatrix , SuperMatrix , float , int *)

int	sp_sgemv_dist (char , float, SuperMatrix , float , int, float, float , int)
	SpGEMV. More...

int	sp_sgemm_dist (char , int, float, SuperMatrix , float , int, float, float , int)

float	sdistribute (superlu_dist_options_t , int_t, SuperMatrix , Glu_freeable_t , sLUstruct_t , gridinfo_t *)

void	psgssvx_ABglobal (superlu_dist_options_t , SuperMatrix , sScalePermstruct_t , float , int, int, gridinfo_t , sLUstruct_t , float , SuperLUStat_t , int *)

float	psdistribute (superlu_dist_options_t , int_t, SuperMatrix , sScalePermstruct_t , Glu_freeable_t , sLUstruct_t , gridinfo_t )

float	psdistribute_allgrid (superlu_dist_options_t options, int_t n, SuperMatrix A, sScalePermstruct_t ScalePermstruct, Glu_freeable_t Glu_freeable, sLUstruct_t LUstruct, gridinfo_t grid, int *supernodeMask)

float	psdistribute_allgrid_index_only (superlu_dist_options_t options, int_t n, SuperMatrix A, sScalePermstruct_t ScalePermstruct, Glu_freeable_t Glu_freeable, sLUstruct_t LUstruct, gridinfo_t grid, int *supernodeMask)

void	psgssvx (superlu_dist_options_t , SuperMatrix , sScalePermstruct_t , float , int, int, gridinfo_t , sLUstruct_t , sSOLVEstruct_t , float , SuperLUStat_t , int )

void	psCompute_Diag_Inv (int_t, sLUstruct_t , gridinfo_t , SuperLUStat_t , int )

int	sSolveInit (superlu_dist_options_t , SuperMatrix , int_t[], int_t[], int_t, sLUstruct_t , gridinfo_t , sSOLVEstruct_t *)
	Initialize the data structure for the solution phase. More...

void	sSolveFinalize (superlu_dist_options_t , sSOLVEstruct_t )
	Release the resources used for the solution phase. More...

void	sDestroy_A3d_gathered_on_2d (sSOLVEstruct_t , gridinfo3d_t )

int_t	psgstrs_init (int_t, int_t, int_t, int_t, int_t[], int_t[], gridinfo_t grid, Glu_persist_t , sSOLVEstruct_t *)

int_t	psgstrs_init_device_lsum_x (superlu_dist_options_t , int_t, int_t, int_t, gridinfo_t , sLUstruct_t , sSOLVEstruct_t , int *)

int_t	psgstrs_delete_device_lsum_x (sSOLVEstruct_t *)

void	pxgstrs_finalize (pxgstrs_comm_t *)

int	sldperm_dist (int, int, int_t, int_t[], int_t[], float[], int_t *, float[], float[])

int	sstatic_schedule (superlu_dist_options_t , int, int, sLUstruct_t , gridinfo_t , SuperLUStat_t , int_t , int_t , int *)

void	sLUstructInit (const int_t, sLUstruct_t *)
	Allocate storage in LUstruct. More...

void	sLUstructFree (sLUstruct_t *)
	Deallocate LUstruct. More...

void	sDestroy_LU (int_t, gridinfo_t , sLUstruct_t )
	Destroy distributed L & U matrices. More...

void	sDestroy_Tree (int_t, gridinfo_t , sLUstruct_t )
	Destroy broadcast and reduction trees used in triangular solve. More...

void	sscatter_l (int ib, int ljb, int nsupc, int_t iukp, int_t xsup, int klst, int nbrow, int_t lptr, int temp_nbrow, int_t usub, int_t lsub, float tempv, int indirect_thread, int indirect2, int_t Lrowind_bc_ptr, float Lnzval_bc_ptr, gridinfo_t *grid)

void	sscatter_u (int ib, int jb, int nsupc, int_t iukp, int_t xsup, int klst, int nbrow, int_t lptr, int temp_nbrow, int_t lsub, int_t usub, float tempv, int_t Ufstnz_br_ptr, float Unzval_br_ptr, gridinfo_t *grid)

int_t	psgstrf (superlu_dist_options_t , int, int, float anorm, sLUstruct_t , gridinfo_t , SuperLUStat_t , int *)

void	psgstrs_Bglobal (superlu_dist_options_t , int_t, sLUstruct_t , gridinfo_t , float , int_t, int, SuperLUStat_t , int )

void	psgstrs (superlu_dist_options_t , int_t, sLUstruct_t , sScalePermstruct_t , gridinfo_t , float , int_t, int_t, int_t, int, sSOLVEstruct_t , SuperLUStat_t , int )

void	psgstrf2_trsm (superlu_dist_options_t options, int_t k0, int_t k, double thresh, Glu_persist_t , gridinfo_t , sLocalLU_t , MPI_Request , int tag_ub, SuperLUStat_t , int *info)

void	psgstrs2_omp (int_t k0, int_t k, Glu_persist_t , gridinfo_t , sLocalLU_t , Ublock_info_t , SuperLUStat_t *)

int_t	psReDistribute_B_to_X (float B, int_t m_loc, int nrhs, int_t ldb, int_t fst_row, int_t ilsum, float x, sScalePermstruct_t , Glu_persist_t , gridinfo_t , sSOLVEstruct_t *)

void	slsum_fmod (float , float , float , float , int, int, int_t, int fmod, int_t, int_t, int_t, int_t , gridinfo_t , sLocalLU_t , MPI_Request[], SuperLUStat_t *)

void	slsum_bmod (float , float , float , int, int_t, int bmod, int_t , Ucb_indptr_t , int_t , int_t , gridinfo_t , sLocalLU_t , MPI_Request[], SuperLUStat_t *)

void	slsum_fmod_inv (float , float , float , float , int, int_t, int fmod, int_t , gridinfo_t , sLocalLU_t , SuperLUStat_t *, int_t , int_t *, int_t, int_t, int_t, int_t, int, int)

void	slsum_fmod_inv_master (float , float , float , float , int, int, int_t, int fmod, int_t, int_t , gridinfo_t , sLocalLU_t , SuperLUStat_t **, int_t, int_t, int_t, int_t, int, int)

void	slsum_bmod_inv (float , float , float , float , int, int_t, int bmod, int_t , Ucb_indptr_t , int_t , int_t , gridinfo_t , sLocalLU_t , SuperLUStat_t , int_t , int_t *, int_t, int_t, int, int)

void	slsum_bmod_inv_master (float , float , float , float , int, int_t, int bmod, int_t , Ucb_indptr_t , int_t , int_t , gridinfo_t , sLocalLU_t , SuperLUStat_t *, int_t, int_t, int, int)

void	sComputeLevelsets (int, int_t, gridinfo_t , Glu_persist_t , sLocalLU_t , int_t )

void	psconvertU (superlu_dist_options_t , gridinfo_t , sLUstruct_t , SuperLUStat_t , int)

void	slsum_fmod_inv_gpu_wrap (int, int, int, int, float , float , int, int, int_t, int fmod, C_Tree , C_Tree , int_t , int_t , int64_t , float , int64_t , float , int64_t , int_t , int64_t , int_t , int , gridinfo_t , int_t, uint64_t , uint64_t , float , float , int , int , int , int , int , int , int , int , int , int , int , int , int , int , int , int , int , int *, int)

void	slsum_bmod_inv_gpu_wrap (superlu_dist_options_t , int, int, int, int, float , float , int, int, int_t, int , C_Tree , C_Tree , int_t , int_t , int64_t , int_t , int64_t , int_t , int64_t , float , int64_t , float , int64_t , float , int64_t , int_t , int64_t , int_t , gridinfo_t , int_t, uint64_t , uint64_t , float , float , int , int , int , int , int , int , int , int , int , int , int , int , int , int , int , int , int , int *, int)

void	psgsrfs (superlu_dist_options_t , int_t, SuperMatrix , float, sLUstruct_t , sScalePermstruct_t , gridinfo_t , float[], int_t, float[], int_t, int, sSOLVEstruct_t , float , SuperLUStat_t , int *)

void	psgsrfs3d (superlu_dist_options_t , int_t, SuperMatrix , float, sLUstruct_t , sScalePermstruct_t , gridinfo3d_t , strf3Dpartition_t , float , int_t, float , int_t, int, sSOLVEstruct_t , float , SuperLUStat_t , int )

void	psgsrfs_ABXglobal (superlu_dist_options_t , int_t, SuperMatrix , float, sLUstruct_t , gridinfo_t , float , int_t, float , int_t, int, float , SuperLUStat_t , int *)

int	psgsmv_AXglobal_setup (SuperMatrix , Glu_persist_t , gridinfo_t , int_t , int_t [], float [], int_t *[], int_t[])

int	psgsmv_AXglobal (int_t, int_t[], float[], int_t[], float[], float[])

int	psgsmv_AXglobal_abs (int_t, int_t[], float[], int_t[], float[], float[])

void	psgsmv_init (SuperMatrix , int_t , gridinfo_t , psgsmv_comm_t )

void	psgsmv (int_t, SuperMatrix , gridinfo_t , psgsmv_comm_t *, float x[], float ax[])

void	psgsmv_finalize (psgsmv_comm_t *)

int_t	sinitLsumBmod_buff (int_t ns, int nrhs, slsumBmod_buff_t *lbmod_buf)

int_t	sleafForestBackSolve3d (superlu_dist_options_t options, int_t treeId, int_t n, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float x, float lsum, float recvbuf, MPI_Request send_req, int nrhs, slsumBmod_buff_t lbmod_buf, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

int_t	snonLeafForestBackSolve3d (int_t treeId, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float x, float lsum, sxT_struct xT_s, float recvbuf, MPI_Request send_req, int nrhs, slsumBmod_buff_t lbmod_buf, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

int_t	slasum_bmod_Tree (int_t pTree, int_t cTree, float lsum, float x, sxT_struct xT_s, int nrhs, slsumBmod_buff_t lbmod_buf, sLUstruct_t LUstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, SuperLUStat_t stat)

int_t	slsumForestBsolve (int_t k, int_t treeId, float lsum, float x, sxT_struct xT_s, int nrhs, slsumBmod_buff_t lbmod_buf, sLUstruct_t LUstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, SuperLUStat_t stat)

int_t	sbCastXk2Pck (int_t k, sxT_struct xT_s, int nrhs, sLUstruct_t LUstruct, gridinfo_t grid, xtrsTimer_t xtrsTimer)

int_t	slsumReducePrK (int_t k, float x, float lsum, float recvbuf, int nrhs, sLUstruct_t LUstruct, gridinfo_t grid, xtrsTimer_t xtrsTimer)

int_t	snonLeafForestForwardSolve3d (int_t treeId, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float x, float lsum, sxT_struct xT_s, float recvbuf, float rtemp, MPI_Request send_req, int nrhs, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

int_t	sleafForestForwardSolve3d (superlu_dist_options_t options, int_t treeId, int_t n, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float x, float lsum, float recvbuf, float rtemp, MPI_Request send_req, int nrhs, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

int	strs_compute_communication_structure (superlu_dist_options_t options, int_t n, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, int supernodeMask, gridinfo_t grid, SuperLUStat_t stat)

int_t	sreduceSolvedX_newsolve (int_t treeId, int_t sender, int_t receiver, float x, int nrhs, strf3Dpartition_t trf3Dpartition, sLUstruct_t LUstruct, gridinfo3d_t grid3d, float recvbuf, xtrsTimer_t xtrsTimer)

void	slsum_fmod_leaf (int_t treeId, strf3Dpartition_t trf3Dpartition, float lsum, float x, float xk, float rtemp, int nrhs, int knsupc, int_t k, int fmod, int_t nlb, int_t lptr, int_t luptr, int_t xsup, gridinfo_t grid, sLocalLU_t Llu, MPI_Request send_req[], SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

void	slsum_fmod_leaf_newsolve (strf3Dpartition_t trf3Dpartition, float lsum, float x, float xk, float rtemp, int nrhs, int knsupc, int_t k, int fmod, int_t nlb, int_t lptr, int_t luptr, int_t xsup, gridinfo_t grid, sLocalLU_t Llu, MPI_Request send_req[], SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

void	slsum_bmod_GG (float lsum, float x, float xk, int nrhs, slsumBmod_buff_t lbmod_buf, int_t k, int bmod, int_t Urbs, Ucb_indptr_t Ucb_indptr, int_t Ucb_valptr, int_t xsup, gridinfo_t grid, sLocalLU_t Llu, MPI_Request send_req[], SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

void	slsum_bmod_GG_newsolve (strf3Dpartition_t trf3Dpartition, float lsum, float x, float xk, int nrhs, slsumBmod_buff_t lbmod_buf, int_t k, int bmod, int_t Urbs, Ucb_indptr_t Ucb_indptr, int_t Ucb_valptr, int_t xsup, gridinfo_t grid, sLocalLU_t Llu, MPI_Request send_req[], SuperLUStat_t stat, xtrsTimer_t xtrsTimer)

int_t	psReDistribute3d_B_to_X (float B, int_t m_loc, int nrhs, int_t ldb, int_t fst_row, int_t ilsum, float x, sScalePermstruct_t ScalePermstruct, Glu_persist_t Glu_persist, gridinfo3d_t grid3d, sSOLVEstruct_t *SOLVEstruct)

int_t	psReDistribute3d_X_to_B (int_t n, float B, int_t m_loc, int_t ldb, int_t fst_row, int nrhs, float x, int_t ilsum, sScalePermstruct_t ScalePermstruct, Glu_persist_t Glu_persist, gridinfo3d_t grid3d, sSOLVEstruct_t *SOLVEstruct)

void	psgstrs3d (superlu_dist_options_t , int_t n, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float B, int_t m_loc, int_t fst_row, int_t ldb, int nrhs, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, int *info)

void	psgstrs3d_newsolve (superlu_dist_options_t options, int_t n, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float B, int_t m_loc, int_t fst_row, int_t ldb, int nrhs, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, int *info)

int_t	psgsTrBackSolve3d (superlu_dist_options_t options, int_t n, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float x3d, float lsum3d, sxT_struct xT_s, float recvbuf, MPI_Request send_req, int nrhs, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

int_t	psgsTrForwardSolve3d (superlu_dist_options_t options, int_t n, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float x3d, float lsum3d, sxT_struct xT_s, float recvbuf, MPI_Request send_req, int nrhs, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

int_t	psgsTrForwardSolve3d_newsolve (superlu_dist_options_t options, int_t n, sLUstruct_t LUstruct, sScalePermstruct_t ScalePermstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float x3d, float lsum3d, float recvbuf, MPI_Request send_req, int nrhs, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, xtrsTimer_t xtrsTimer)

int_t	psgsTrBackSolve3d_newsolve (superlu_dist_options_t options, int_t n, sLUstruct_t LUstruct, strf3Dpartition_t trf3Dpartition, gridinfo3d_t grid3d, float x3d, float lsum3d, float recvbuf, MPI_Request send_req, int nrhs, sSOLVEstruct_t SOLVEstruct, SuperLUStat_t stat, xtrsTimer_t *xtrsTimer)

int_t	sbroadcastAncestor3d (strf3Dpartition_t trf3Dpartition, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t SCT)

int_t	slocalSolveXkYk (trtype_t trtype, int_t k, float x, int nrhs, sLUstruct_t LUstruct, gridinfo_t grid, SuperLUStat_t stat)

int_t	siBcastXk2Pck (int_t k, float x, int nrhs, int sendList, MPI_Request send_req, sLUstruct_t LUstruct, gridinfo_t grid, xtrsTimer_t *xtrsTimer)

int_t	strs_B_init3d (int_t nsupers, float x, int nrhs, sLUstruct_t LUstruct, gridinfo3d_t *grid3d)

int_t	strs_X_gather3d (float x, int nrhs, strf3Dpartition_t trf3Dpartition, sLUstruct_t LUstruct, gridinfo3d_t grid3d, xtrsTimer_t *xtrsTimer)

int_t	sfsolveReduceLsum3d (int_t treeId, int_t sender, int_t receiver, float lsum, float recvbuf, int nrhs, strf3Dpartition_t trf3Dpartition, sLUstruct_t LUstruct, gridinfo3d_t grid3d, xtrsTimer_t xtrsTimer)

int_t	sbsolve_Xt_bcast (int_t ilvl, sxT_struct xT_s, int nrhs, strf3Dpartition_t trf3Dpartition, sLUstruct_t LUstruct, gridinfo3d_t grid3d, xtrsTimer_t *xtrsTimer)

int_t	sp2pSolvedX3d (int_t treeId, int_t sender, int_t receiver, float x, int nrhs, strf3Dpartition_t trf3Dpartition, sLUstruct_t LUstruct, gridinfo3d_t grid3d, xtrsTimer_t *xtrsTimer)

float *	floatMalloc_dist (int_t)

float *	floatCalloc_dist (int_t)

void *	suser_malloc_dist (int_t, int_t)

void	suser_free_dist (int_t, int_t)

int_t	sQuerySpace_dist (int_t, sLUstruct_t , gridinfo_t , SuperLUStat_t , superlu_dist_mem_usage_t )

void	sClone_CompRowLoc_Matrix_dist (SuperMatrix , SuperMatrix )

void	sCopy_CompRowLoc_Matrix_dist (SuperMatrix , SuperMatrix )

void	sZero_CompRowLoc_Matrix_dist (SuperMatrix *)
	Sets all entries of a matrix to zero, A_{i,j}=0, for i,j=1,..,n. More...

void	sScaleAddId_CompRowLoc_Matrix_dist (SuperMatrix *, float)
	Scale and add I: scales a matrix and adds an identity. A_{i,j} = c * A_{i,j} + \delta_{i,j} for i,j=1,...,n and \delta_{i,j} is the Kronecker delta. More...

void	sScaleAdd_CompRowLoc_Matrix_dist (SuperMatrix , SuperMatrix , float)
	Scale and add: adds a scalar multiple of one matrix to another. A_{i,j} = c * A_{i,j} + B_{i,j}$ for i,j=1,...,n. More...

void	sZeroLblocks (int, int, gridinfo_t , sLUstruct_t )
	Sets all entries of matrix L to zero. More...

void	sZeroUblocks (int iam, int n, gridinfo_t , sLUstruct_t )
	Sets all entries of matrix U to zero. More...

float	sMaxAbsLij (int iam, int n, Glu_persist_t , sLUstruct_t , gridinfo_t *)
	Find max(abs(L(i,j))) More...

float	sMaxAbsUij (int iam, int n, Glu_persist_t , sLUstruct_t , gridinfo_t *)
	Find max(abs(U(i,j))) More...

void	sfill_dist (float *, int_t, float)
	Fills a float precision array with a given value. More...

void	sinf_norm_error_dist (int_t, int_t, float , int_t, float , int_t, gridinfo_t *)
	Check the inf-norm of the error vector. More...

void	psinf_norm_error (int, int_t, int_t, float[], int_t, float[], int_t, MPI_Comm)
	Check the inf-norm of the error vector. More...

void	sreadhb_dist (int, FILE , int_t , int_t , int_t , float , int_t , int_t **)

void	sreadtriple_dist (FILE , int_t , int_t , int_t , float , int_t , int_t **)

void	sreadtriple_noheader (FILE , int_t , int_t , int_t , float , int_t , int_t **)

void	sreadrb_dist (int, FILE , int_t , int_t , int_t , float , int_t , int_t **)

void	sreadMM_dist (FILE , int_t , int_t , int_t , float , int_t , int_t **)

int	sread_binary (FILE , int_t , int_t , int_t , float , int_t , int_t **)

void	validateInput_psgssvx3d (superlu_dist_options_t , SuperMatrix A, int ldb, int nrhs, gridinfo3d_t , int info)
	Validates the input parameters for a given problem. More...

void	sallocScalePermstruct_RC (sScalePermstruct_t *, int_t m, int_t n)

void	sscaleMatrixDiagonally (fact_t Fact, sScalePermstruct_t , SuperMatrix , SuperLUStat_t , gridinfo_t , int rowequ, int colequ, int *iinfo)

void	sperform_row_permutation (superlu_dist_options_t , fact_t Fact, sScalePermstruct_t , sLUstruct_t LUstruct, int_t m, int_t n, gridinfo_t , SuperMatrix A, SuperMatrix GA, SuperLUStat_t , int job, int Equil, int rowequ, int colequ, int iinfo)

float	scomputeA_Norm (int notran, SuperMatrix , gridinfo_t )
	This function computes the norm of a matrix A. More...

float	sdist_psymbtonum (superlu_dist_options_t , int_t, SuperMatrix , sScalePermstruct_t , Pslu_freeable_t , sLUstruct_t , gridinfo_t )

void	psGetDiagU (int_t, sLUstruct_t , gridinfo_t , float *)

int	s_c2cpp_GetHWPM (SuperMatrix , gridinfo_t , sScalePermstruct_t *)

void	sPrintLblocks (int, int_t, gridinfo_t , Glu_persist_t , sLocalLU_t *)
	Print the blocks in the factored matrix L. More...

void	sPrintUblocks (int, int_t, gridinfo_t , Glu_persist_t , sLocalLU_t *)
	Print the blocks in the factored matrix U. More...

void	sPrint_CompCol_Matrix_dist (SuperMatrix *)

void	sPrint_Dense_Matrix_dist (SuperMatrix *)

int	sPrint_CompRowLoc_Matrix_dist (SuperMatrix *)

int	file_sPrint_CompRowLoc_Matrix_dist (FILE fp, SuperMatrix A)

void	Printfloat5 (char , int_t, float )

int	file_Printfloat5 (FILE , char , int_t, float *)

void	sGenCOOLblocks (int, int_t, gridinfo_t , Glu_persist_t , sLocalLU_t , int_t , int_t , float , int_t , int_t *)

void	sGenCSCLblocks (int, int_t, gridinfo_t , Glu_persist_t , sLocalLU_t , float , int_t , int_t , int_t , int_t *)

void	sGenCSRLblocks (int, int_t, gridinfo_t , Glu_persist_t , sLocalLU_t , float , int_t , int_t , int_t , int_t *)

void	nv_init_wrapper (MPI_Comm)

void	sprepare_multiGPU_buffers (int, int, int, int, int, int)

void	sdelete_multiGPU_buffers ()

int	sgemm_ (const char , const char , const int , const int , const int , const float , const float , const int , const float , const int , const float , float , const int *)

int	strsv_ (char , char , char , int , float , int , float , int )

int	strsm_ (const char , const char , const char , const char , const int , const int , const float , const float , const int , float , const int *)

void	sgemv_ (const char , const int , const int , const float , const float a, const int , const float , const int , const float , float , const int *)

void	sger_ (const int , const int , const float , const float , const int , const float , const int , float , const int *)

int	sscal_ (const int n, const float alpha, float dx, const int incx)

int	saxpy_ (const int n, const float alpha, const float x, const int incx, float y, const int incy)

int	superlu_sgemm (const char transa, const char transb, int m, int n, int k, float alpha, float a, int lda, float b, int ldb, float beta, float *c, int ldc)

int	superlu_strsm (const char sideRL, const char uplo, const char transa, const char diag, const int m, const int n, const float alpha, const float a, const int lda, float b, const int ldb)

int	superlu_sger (const int m, const int n, const float alpha, const float x, const int incx, const float y, const int incy, float *a, const int lda)

int	superlu_sscal (const int n, const float alpha, float *x, const int incx)

int	superlu_saxpy (const int n, const float alpha, const float x, const int incx, float y, const int incy)

int	superlu_sgemv (const char trans, const int m, const int n, const float alpha, const float a, const int lda, const float x, const int incx, const float beta, float y, const int incy)

int	superlu_strsv (char uplo, char trans, char diag, int n, float a, int lda, float *x, int incx)

void	strtri_ (char , char , int , float , int , int )

int	screate_matrix3d (SuperMatrix A, int nrhs, float rhs, int ldb, float *x, int ldx, FILE fp, gridinfo3d_t grid3d)

int	screate_matrix_postfix3d (SuperMatrix A, int nrhs, float rhs, int ldb, float *x, int ldx, FILE fp, char postfix, gridinfo3d_t *grid3d)

int	screate_block_diag_3d (SuperMatrix A, int batchCount, int nrhs, float rhs, int ldb, float *x, int ldx, FILE fp, char postfix, gridinfo3d_t *grid3d)

int	screate_batch_systems (handle_t SparseMatrix_handles, int batchCount, int nrhs, float rhs, int ldb, float *x, int ldx, FILE fp, char postfix, gridinfo3d_t *grid3d)

void	sGatherNRformat_loc3d (fact_t Fact, NRformat_loc A, float B, int ldb, int nrhs, gridinfo3d_t grid3d, NRformat_loc3d *)

void	sGatherNRformat_loc3d_allgrid (fact_t Fact, NRformat_loc A, float B, int ldb, int nrhs, gridinfo3d_t grid3d, NRformat_loc3d *)

int	sScatter_B3d (NRformat_loc3d A3d, gridinfo3d_t grid3d)

void	psgssvx3d (superlu_dist_options_t , SuperMatrix , sScalePermstruct_t , float B[], int ldb, int nrhs, gridinfo3d_t , sLUstruct_t , sSOLVEstruct_t , float berr, SuperLUStat_t , int *info)

int_t	psgstrf3d (superlu_dist_options_t , int m, int n, float anorm, strf3Dpartition_t , SCT_t , sLUstruct_t , gridinfo3d_t , SuperLUStat_t , int *)

void	sInit_HyP (superlu_dist_options_t , HyP_t HyP, sLocalLU_t *Llu, int_t mcb, int_t mrb)

void	Free_HyP (HyP_t *HyP)

int	updateDirtyBit (int_t k0, HyP_t HyP, gridinfo_t grid)

void	sblock_gemm_scatter (int_t lb, int_t j, Ublock_info_t Ublock_info, Remain_info_t Remain_info, float L_mat, int ldl, float U_mat, int ldu, float bigV, int_t knsupc, int_t klst, int_t lsub, int_t usub, int_t ldt, int_t thread_id, int indirect, int indirect2, int_t Lrowind_bc_ptr, float Lnzval_bc_ptr, int_t Ufstnz_br_ptr, float Unzval_br_ptr, int_t xsup, gridinfo_t , SuperLUStat_t )

int_t	sblock_gemm_scatterTopLeft (int_t lb, int_t j, float bigV, int_t knsupc, int_t klst, int_t lsub, int_t usub, int_t ldt, int indirect, int indirect2, HyP_t HyP, sLUstruct_t , gridinfo_t , SCT_t SCT, SuperLUStat_t )

int_t	sblock_gemm_scatterTopRight (int_t lb, int_t j, float bigV, int_t knsupc, int_t klst, int_t lsub, int_t usub, int_t ldt, int indirect, int indirect2, HyP_t HyP, sLUstruct_t , gridinfo_t , SCT_t SCT, SuperLUStat_t )

int_t	sblock_gemm_scatterBottomLeft (int_t lb, int_t j, float bigV, int_t knsupc, int_t klst, int_t lsub, int_t usub, int_t ldt, int indirect, int indirect2, HyP_t HyP, sLUstruct_t , gridinfo_t , SCT_t SCT, SuperLUStat_t )

int_t	sblock_gemm_scatterBottomRight (int_t lb, int_t j, float bigV, int_t knsupc, int_t klst, int_t lsub, int_t usub, int_t ldt, int indirect, int indirect2, HyP_t HyP, sLUstruct_t , gridinfo_t , SCT_t SCT, SuperLUStat_t )

void	sgather_u (int_t num_u_blks, Ublock_info_t Ublock_info, int_t usub, float uval, float bigU, int_t ldu, int_t *xsup, int_t klst)

void	sgather_l (int_t num_LBlk, int_t knsupc, Remain_info_t L_info, float lval, int_t LD_lval, float *L_buff)

void	sRgather_L (int_t k, int_t lsub, float lusup, gEtreeInfo_t , Glu_persist_t , gridinfo_t , HyP_t , int_t myIperm, int_t iperm_c_supno)

void	sRgather_U (int_t k, int_t jj0, int_t usub, float uval, float bigU, gEtreeInfo_t , Glu_persist_t , gridinfo_t , HyP_t , int_t myIperm, int_t iperm_c_supno, int_t perm_u)

void	sbcastPermutedSparseA (SuperMatrix A, sScalePermstruct_t ScalePermstruct, Glu_freeable_t Glu_freeable, sLUstruct_t LUstruct, gridinfo3d_t *grid3d)

void	snewTrfPartitionInit (int_t nsupers, sLUstruct_t LUstruct, gridinfo3d_t grid3d)

strf3Dpartition_t *	sinitTrf3Dpartition (int_t nsupers, superlu_dist_options_t options, sLUstruct_t LUstruct, gridinfo3d_t *grid3d)

strf3Dpartition_t *	sinitTrf3Dpartition_allgrid (int_t n, superlu_dist_options_t options, sLUstruct_t LUstruct, gridinfo3d_t *grid3d)

strf3Dpartition_t *	sinitTrf3DpartitionLUstructgrid0 (int_t n, superlu_dist_options_t options, sLUstruct_t LUstruct, gridinfo3d_t *grid3d)

void	sDestroy_trf3Dpartition (strf3Dpartition_t *trf3Dpartition)

void	s3D_printMemUse (strf3Dpartition_t trf3Dpartition, sLUstruct_t LUstruct, gridinfo3d_t *grid3d)

void	sinit3DLUstructForest (int_t myTreeIdxs, int_t myZeroTrIdxs, sForest_t *sForests, sLUstruct_t LUstruct, gridinfo3d_t *grid3d)

int_t	sgatherAllFactoredLUFr (int_t myZeroTrIdxs, sForest_t sForests, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t *SCT)

int_t	sLpanelUpdate (int_t off0, int_t nsupc, float ublk_ptr, int_t ld_ujrow, float lusup, int_t nsupr, SCT_t *)

void	sgstrf2 (int_t k, float diagBlk, int_t LDA, float BlockUfactor, int_t LDU, double thresh, int_t xsup, superlu_dist_options_t options, SuperLUStat_t stat, int info)

void	Local_Sgstrf2 (superlu_dist_options_t options, int_t k, double thresh, float BlockUFactor, Glu_persist_t , gridinfo_t , sLocalLU_t , SuperLUStat_t , int info, SCT_t )

int_t	sTrs2_GatherU (int_t iukp, int_t rukp, int_t klst, int_t nsupc, int_t ldu, int_t usub, float uval, float *tempv)

int_t	sTrs2_ScatterU (int_t iukp, int_t rukp, int_t klst, int_t nsupc, int_t ldu, int_t usub, float uval, float *tempv)

int_t	sTrs2_GatherTrsmScatter (int_t klst, int_t iukp, int_t rukp, int_t usub, float uval, float tempv, int_t knsupc, int nsupr, float lusup, Glu_persist_t *Glu_persist)

void	psgstrs2 (int_t m, int_t k0, int_t k, Glu_persist_t Glu_persist, gridinfo_t grid, sLocalLU_t Llu, SuperLUStat_t stat)

void	psgstrf2 (superlu_dist_options_t , int_t nsupers, int_t k0, int_t k, double thresh, Glu_persist_t , gridinfo_t , sLocalLU_t , MPI_Request , int, SuperLUStat_t , int *)

int_t	sAllocLlu_3d (int_t nsupers, sLUstruct_t LUstruct, gridinfo3d_t grid3d)

int_t	sp3dScatter (int_t n, sLUstruct_t LUstruct, gridinfo3d_t grid3d, int *supernodeMask)

int_t	sscatter3dLPanels (int_t nsupers, sLUstruct_t LUstruct, gridinfo3d_t grid3d, int *supernodeMask)

int_t	sscatter3dUPanels (int_t nsupers, sLUstruct_t LUstruct, gridinfo3d_t grid3d, int *supernodeMask)

int_t	scollect3dLpanels (int_t layer, int_t nsupers, sLUstruct_t LUstruct, gridinfo3d_t grid3d)

int_t	scollect3dUpanels (int_t layer, int_t nsupers, sLUstruct_t LUstruct, gridinfo3d_t grid3d)

int_t	sp3dCollect (int_t layer, int_t n, sLUstruct_t LUstruct, gridinfo3d_t grid3d)

int_t	szeroSetLU (int_t nnodes, int_t nodeList, sLUstruct_t , gridinfo3d_t *)

int	sAllocGlu_3d (int_t n, int_t nsupers, sLUstruct_t *)

int	sDeAllocLlu_3d (int_t n, sLUstruct_t , gridinfo3d_t )

int	sDeAllocGlu_3d (sLUstruct_t *)

int_t	sreduceAncestors3d (int_t sender, int_t receiver, int_t nnodes, int_t nodeList, float Lval_buf, float Uval_buf, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t SCT)

int	sreduceAllAncestors3d (int_t ilvl, int_t myNodeCount, int_t treePerm, sLUValSubBuf_t LUvsb, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t *SCT)

int_t	sgatherFactoredLU (int_t sender, int_t receiver, int_t nnodes, int_t nodeList, sLUValSubBuf_t LUvsb, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t *SCT)

int_t	sgatherAllFactoredLU (strf3Dpartition_t trf3Dpartition, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t SCT)

int_t	sinit3DLUstruct (int_t myTreeIdxs, int_t myZeroTrIdxs, int_t nodeCount, int_t nodeList, sLUstruct_t LUstruct, gridinfo3d_t *grid3d)

int_t	szSendLPanel (int_t k, int_t receiver, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t *SCT)

int_t	szRecvLPanel (int_t k, int_t sender, float alpha, float beta, float Lval_buf, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t SCT)

int_t	szSendUPanel (int_t k, int_t receiver, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t *SCT)

int_t	szRecvUPanel (int_t k, int_t sender, float alpha, float beta, float Uval_buf, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SCT_t SCT)

int_t	sIBcast_LPanel (int_t k, int_t k0, int_t lsub, float lusup, gridinfo_t , int msgcnt, MPI_Request , int ToSendR, int_t xsup, int)

int_t	sBcast_LPanel (int_t k, int_t k0, int_t lsub, float lusup, gridinfo_t , int msgcnt, int *ToSendR, int_t xsup, SCT_t *, int)

int_t	sIBcast_UPanel (int_t k, int_t k0, int_t usub, float uval, gridinfo_t , int msgcnt, MPI_Request , int ToSendD, int)

int_t	sBcast_UPanel (int_t k, int_t k0, int_t usub, float uval, gridinfo_t , int msgcnt, int ToSendD, SCT_t , int)

int_t	sIrecv_LPanel (int_t k, int_t k0, int_t Lsub_buf, float Lval_buf, gridinfo_t , MPI_Request , sLocalLU_t *, int)

int_t	sIrecv_UPanel (int_t k, int_t k0, int_t Usub_buf, float , sLocalLU_t , gridinfo_t , MPI_Request *, int)

int_t	sWait_URecv (MPI_Request , int msgcnt, SCT_t *)

int_t	sWait_LRecv (MPI_Request , int msgcnt, int msgcntsU, gridinfo_t , SCT_t *)

int_t	sISend_UDiagBlock (int_t k0, float ublk_ptr, int_t size, MPI_Request , gridinfo_t *, int)

int_t	sRecv_UDiagBlock (int_t k0, float ublk_ptr, int_t size, int_t src, gridinfo_t , SCT_t *, int)

int_t	sPackLBlock (int_t k, float Dest, Glu_persist_t , gridinfo_t , sLocalLU_t )

int_t	sISend_LDiagBlock (int_t k0, float lblk_ptr, int_t size, MPI_Request , gridinfo_t *, int)

int_t	sIRecv_UDiagBlock (int_t k0, float ublk_ptr, int_t size, int_t src, MPI_Request , gridinfo_t , SCT_t , int)

int_t	sIRecv_LDiagBlock (int_t k0, float L_blk_ptr, int_t size, int_t src, MPI_Request , gridinfo_t , SCT_t , int)

int_t	sUDiagBlockRecvWait (int_t k, int IrecvPlcd_D, int factored_L, MPI_Request , gridinfo_t , sLUstruct_t , SCT_t )

int_t	sDiagFactIBCast (int_t k, int_t k0, float BlockUFactor, float BlockLFactor, int IrecvPlcd_D, MPI_Request , MPI_Request , MPI_Request , MPI_Request , gridinfo_t , superlu_dist_options_t , double thresh, sLUstruct_t LUstruct, SuperLUStat_t , int info, SCT_t *, int tag_ub)

int_t	sUPanelTrSolve (int_t k, float BlockLFactor, float bigV, int_t ldt, Ublock_info_t , gridinfo_t , sLUstruct_t , SuperLUStat_t , SCT_t *)

int_t	sLPanelUpdate (int_t k, int IrecvPlcd_D, int factored_L, MPI_Request , float BlockUFactor, gridinfo_t , sLUstruct_t , SCT_t *)

int_t	sUPanelUpdate (int_t k, int factored_U, MPI_Request , float BlockLFactor, float bigV, int_t ldt, Ublock_info_t , gridinfo_t , sLUstruct_t , SuperLUStat_t , SCT_t *)

int_t	sIBcastRecvLPanel (int_t k, int_t k0, int msgcnt, MPI_Request , MPI_Request , int_t Lsub_buf, float Lval_buf, int factored, gridinfo_t , sLUstruct_t , SCT_t *, int tag_ub)

int_t	sIBcastRecvUPanel (int_t k, int_t k0, int msgcnt, MPI_Request , MPI_Request , int_t Usub_buf, float Uval_buf, gridinfo_t , sLUstruct_t , SCT_t , int tag_ub)

int_t	sWaitL (int_t k, int msgcnt, int msgcntU, MPI_Request , MPI_Request , gridinfo_t , sLUstruct_t , SCT_t *)

int_t	sWaitU (int_t k, int msgcnt, MPI_Request , MPI_Request , gridinfo_t , sLUstruct_t , SCT_t )

int_t	sLPanelTrSolve (int_t k, int factored_L, float BlockUFactor, gridinfo_t , sLUstruct_t )

int	getNsupers (int, Glu_persist_t *)

int_t	initPackLUInfo (int_t nsupers, packLUInfo_t *packLUInfo)

int	freePackLUInfo (packLUInfo_t *packLUInfo)

int_t	sSchurComplementSetup (int_t k, int msgcnt, Ublock_info_t , Remain_info_t , uPanelInfo_t , lPanelInfo_t , int_t , int_t , int_t , float bigU, int_t Lsub_buf, float Lval_buf, int_t Usub_buf, float Uval_buf, gridinfo_t , sLUstruct_t *)

int_t	sSchurComplementSetupGPU (int_t k, msgs_t msgs, packLUInfo_t , int_t , int_t , int_t , gEtreeInfo_t , factNodelists_t , sscuBufs_t , sLUValSubBuf_t LUvsb, gridinfo_t , sLUstruct_t , HyP_t )

float *	sgetBigV (int_t, int_t)

float *	sgetBigU (superlu_dist_options_t , int_t, gridinfo_t , sLUstruct_t *)

int_t	sLluBufInit (sLUValSubBuf_t , sLUstruct_t )

int_t	sinitScuBufs (superlu_dist_options_t , int_t ldt, int_t num_threads, int_t nsupers, sscuBufs_t , sLUstruct_t , gridinfo_t )

int	sfreeScuBufs (sscuBufs_t *scuBufs)

int_t	ssparseTreeFactor (int_t nnodes, int_t perm_c_supno, treeTopoInfo_t treeTopoInfo, commRequests_t comReqs, sscuBufs_t scuBufs, packLUInfo_t packLUInfo, msgs_t msgs, sLUValSubBuf_t LUvsb, sdiagFactBufs_t dFBuf, factStat_t factStat, factNodelists_t fNlists, superlu_dist_options_t options, int_t gIperm_c_supno, int_t ldt, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SuperLUStat_t stat, double thresh, SCT_t SCT, int *info)

int_t	sdenseTreeFactor (int_t nnnodes, int_t perm_c_supno, commRequests_t comReqs, sscuBufs_t scuBufs, packLUInfo_t packLUInfo, msgs_t msgs, sLUValSubBuf_t LUvsb, sdiagFactBufs_t dFBuf, factStat_t factStat, factNodelists_t fNlists, superlu_dist_options_t options, int_t gIperm_c_supno, int_t ldt, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SuperLUStat_t stat, double thresh, SCT_t SCT, int tag_ub, int info)

int_t	ssparseTreeFactor_ASYNC (sForest_t sforest, commRequests_t comReqss, sscuBufs_t scuBufs, packLUInfo_t packLUInfo, msgs_t msgss, sLUValSubBuf_t LUvsbs, sdiagFactBufs_t dFBufs, factStat_t factStat, factNodelists_t fNlists, gEtreeInfo_t gEtreeInfo, superlu_dist_options_t options, int_t gIperm_c_supno, int_t ldt, HyP_t HyP, sLUstruct_t LUstruct, gridinfo3d_t grid3d, SuperLUStat_t stat, double thresh, SCT_t SCT, int tag_ub, int info)

sLUValSubBuf_t **	sLluBufInitArr (int_t numLA, sLUstruct_t *LUstruct)

int	sLluBufFreeArr (int_t numLA, sLUValSubBuf_t **LUvsbs)

sdiagFactBufs_t **	sinitDiagFactBufsArr (int mxLeafNode, int ldt, gridinfo_t *grid)

sdiagFactBufs_t **	sinitDiagFactBufsArrMod (int mxLeafNode, int ldts, gridinfo_t grid)

int	sfreeDiagFactBufsArr (int mxLeafNode, sdiagFactBufs_t **dFBufs)

int	sinitDiagFactBufs (int ldt, sdiagFactBufs_t *dFBuf)

int_t	checkRecvUDiag (int_t k, commRequests_t comReqs, gridinfo_t grid, SCT_t *SCT)

int_t	checkRecvLDiag (int_t k, commRequests_t comReqs, gridinfo_t , SCT_t *)

int	psflatten_LDATA (superlu_dist_options_t options, int_t n, sLUstruct_t LUstruct, gridinfo_t grid, SuperLUStat_t stat)

void	psconvert_flatten_skyline2UROWDATA (superlu_dist_options_t , gridinfo_t , sLUstruct_t , SuperLUStat_t , int n)

void	psconvertUROWDATA2skyline (superlu_dist_options_t , gridinfo_t , sLUstruct_t , SuperLUStat_t , int n)

int_t	sReDistribute_A (SuperMatrix A, sScalePermstruct_t ScalePermstruct, Glu_freeable_t Glu_freeable, int_t xsup, int_t supno, gridinfo_t grid, int_t colptr[], int_t rowind[], float *a[])

float	psdistribute3d_Yang (superlu_dist_options_t options, int_t n, SuperMatrix A, sScalePermstruct_t ScalePermstruct, Glu_freeable_t Glu_freeable, sLUstruct_t LUstruct, gridinfo3d_t grid3d)

int	psgssvx3d_csc_batch (superlu_dist_options_t , int batchCount, int m, int n, int nnz, int nrhs, handle_t , float *RHSptr, int ldRHS, float ReqPtr, float CeqPtr, int RpivPtr, int CpivPtr, DiagScale_t DiagScale, handle_t F, float *Xptr, int ldX, float *Berrs, gridinfo3d_t grid3d, SuperLUStat_t stat, int info)
	Solve a batch of linear systems Ai * Xi = Bi with direct method, computing the LU factorization of each matrix Ai; This is the fixed-size interface: all the input matrices have the same sparsity structure. More...

int	sequil_batch (superlu_dist_options_t , int batchCount, int m, int n, handle_t , float ReqPtr, float CeqPtr, DiagScale_t *)
	Equilibrate the systems using the LAPACK-style algorithm. More...

int	spivot_batch (superlu_dist_options_t , int batchCount, int m, int n, handle_t , float ReqPtr, float CeqPtr, DiagScale_t , int *RpivPtr)
	Compute row pivotings for each matrix, for numerical stability. More...

Variables
float *	sready_x

float *	sready_lsum

Detailed Description

Distributed SuperLU data types and function prototypes.

The source code is distributed under BSD license, see the file License.txt at the top-level directory.

-- Distributed SuperLU routine (version 9.0) --
Lawrence Berkeley National Lab, Univ. of California Berkeley,
Georgia Institute of Technology
November 1, 2007
April 5, 2015
September 18, 2018  version 6.0
February 8, 2019  version 6.1.1
May 10, 2019 version 7.0.0

Macro Definition Documentation

◆ MAX_LOOKAHEADS

#define MAX_LOOKAHEADS 50

Typedef Documentation

◆ slsumBmod_buff_t

typedef struct slsumBmod_buff_t slsumBmod_buff_t

◆ sxT_struct

typedef struct sxT_struct sxT_struct

Function Documentation

◆ checkRecvLDiag()

int_t checkRecvLDiag	(	int_t	k,
		commRequests_t *	comReqs,
		gridinfo_t *	grid,
		SCT_t *	SCT
	)

◆ checkRecvUDiag()

int_t checkRecvUDiag	(	int_t	k,
		commRequests_t *	comReqs,
		gridinfo_t *	grid,
		SCT_t *	SCT
	)

◆ file_Printfloat5()

int file_Printfloat5	(	FILE *	fp,
		char *	name,
		int_t	len,
		float *	x
	)

◆ file_sPrint_CompRowLoc_Matrix_dist()

int file_sPrint_CompRowLoc_Matrix_dist	(	FILE *	fp,
		SuperMatrix *	A
	)

◆ floatCalloc_dist()

float * floatCalloc_dist ( int_t n )

Here is the caller graph for this function:

◆ floatMalloc_dist()

float * floatMalloc_dist ( int_t n )

◆ Free_HyP()

void Free_HyP ( HyP_t * HyP )

◆ freePackLUInfo()

int freePackLUInfo ( packLUInfo_t * packLUInfo )

◆ getNsupers()

int getNsupers	(	int	n,
		Glu_persist_t *	Glu_persist
	)

◆ initPackLUInfo()

int_t initPackLUInfo	(	int_t	nsupers,
		packLUInfo_t *	packLUInfo
	)

◆ Local_Sgstrf2()

void Local_Sgstrf2	(	superlu_dist_options_t *	options,
		int_t	k,
		double	thresh,
		float *	BlockUFactor,
		Glu_persist_t *	,
		gridinfo_t *	,
		sLocalLU_t *	,
		SuperLUStat_t *	,
		int *	info,
		SCT_t *
	)

Here is the caller graph for this function:

◆ nv_init_wrapper()

void nv_init_wrapper ( MPI_Comm )

◆ Printfloat5()

void Printfloat5	(	char *	name,
		int_t	len,
		float *	x
	)

Here is the caller graph for this function:

◆ psCompRow_loc_to_CompCol_global()

int psCompRow_loc_to_CompCol_global	(	int_t	need_value,
		SuperMatrix *	A,
		gridinfo_t *	grid,
		SuperMatrix *	GA
	)

Gather A from the distributed compressed row format to global A in compressed column format.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psCompute_Diag_Inv()

void psCompute_Diag_Inv	(	int_t	n,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		SuperLUStat_t *	stat,
		int *	info
	)

Purpose
=======
  Compute the inverse of the diagonal blocks of the L and U
  triangular matrices.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psconvert_flatten_skyline2UROWDATA()

void psconvert_flatten_skyline2UROWDATA	(	superlu_dist_options_t *	options,
		gridinfo_t *	grid,
		sLUstruct_t *	LUstruct,
		SuperLUStat_t *	stat,
		int	n
	)

Here is the call graph for this function:

◆ psconvertU()

void psconvertU	(	superlu_dist_options_t *	,
		gridinfo_t *	,
		sLUstruct_t *	,
		SuperLUStat_t *	,
		int
	)

Here is the caller graph for this function:

◆ psconvertUROWDATA2skyline()

void psconvertUROWDATA2skyline	(	superlu_dist_options_t *	options,
		gridinfo_t *	grid,
		sLUstruct_t *	LUstruct,
		SuperLUStat_t *	stat,
		int	n
	)

Here is the call graph for this function:

◆ psdistribute()

float psdistribute	(	superlu_dist_options_t *	options,
		int_t	n,
		SuperMatrix *	A,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_freeable_t *	Glu_freeable,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psdistribute3d_Yang()

float psdistribute3d_Yang	(	superlu_dist_options_t *	options,
		int_t	n,
		SuperMatrix *	A,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_freeable_t *	Glu_freeable,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psdistribute_allgrid()

float psdistribute_allgrid	(	superlu_dist_options_t *	options,
		int_t	n,
		SuperMatrix *	A,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_freeable_t *	Glu_freeable,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		int *	supernodeMask
	)

Here is the call graph for this function:

◆ psdistribute_allgrid_index_only()

float psdistribute_allgrid_index_only	(	superlu_dist_options_t *	options,
		int_t	n,
		SuperMatrix *	A,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_freeable_t *	Glu_freeable,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		int *	supernodeMask
	)

Here is the call graph for this function:

◆ psflatten_LDATA()

int psflatten_LDATA	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		SuperLUStat_t *	stat
	)

Here is the caller graph for this function:

◆ psGetDiagU()

void psGetDiagU	(	int_t	n,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		float *	diagU
	)

Purpose
=======

GetDiagU extracts the main diagonal of matrix U of the LU factorization.

Arguments
=========

n        (input) int
         Dimension of the matrix.

LUstruct (input) sLUstruct_t*
         The data structures to store the distributed L and U factors.
         see superlu_ddefs.h for its definition.

grid     (input) gridinfo_t*
         The 2D process mesh. It contains the MPI communicator, the number
         of process rows (NPROW), the number of process columns (NPCOL),
         and my process rank. It is an input argument to all the
         parallel routines.

diagU    (output) double*, dimension (n)
         The main diagonal of matrix U.
         On exit, it is available on all processes.


Note
====

The diagonal blocks of the L and U matrices are stored in the L
data structures, and are on the diagonal processes of the
2D process grid.

This routine is modified from gather_diag_to_all() in psgstrs_Bglobal.c.

Here is the call graph for this function:

◆ psgsequ()

void psgsequ	(	SuperMatrix *	A,
		float *	r,
		float *	c,
		float *	rowcnd,
		float *	colcnd,
		float *	amax,
		int *	info,
		gridinfo_t *	grid
	)

    Purpose
    =======

    PSGSEQU computes row and column scalings intended to equilibrate an
    M-by-N sparse matrix A and reduce its condition number. R returns the row
    scale factors and C the column scale factors, chosen to try to make
    the largest element in each row and column of the matrix B with
    elements B(i,j)=R(i)*A(i,j)*C(j) have absolute value 1.

    R(i) and C(j) are restricted to be between SMLNUM = smallest safe
    number and BIGNUM = largest safe number.  Use of these scaling
    factors is not guaranteed to reduce the condition number of A but
    works well in practice.

    See supermatrix.h for the definition of 'SuperMatrix' structure.

    Arguments
    =========

    A       (input) SuperMatrix*
            The matrix of dimension (A->nrow, A->ncol) whose equilibration
            factors are to be computed. The type of A can be:
            Stype = SLU_NR_loc; Dtype = SLU_S; Mtype = SLU_GE.

    R       (output) float*, size A->nrow
            If INFO = 0 or INFO > M, R contains the row scale factors
            for A.

    C       (output) float*, size A->ncol
            If INFO = 0,  C contains the column scale factors for A.

    ROWCND  (output) float*
            If INFO = 0 or INFO > M, ROWCND contains the ratio of the
            smallest R(i) to the largest R(i).  If ROWCND >= 0.1 and
            AMAX is neither too large nor too small, it is not worth
            scaling by R.

    COLCND  (output) float*
            If INFO = 0, COLCND contains the ratio of the smallest
            C(i) to the largest C(i).  If COLCND >= 0.1, it is not
            worth scaling by C.

    AMAX    (output) float*
            Absolute value of largest matrix element.  If AMAX is very
            close to overflow or very close to underflow, the matrix
            should be scaled.

    INFO    (output) int*
            = 0:  successful exit
            < 0:  if INFO = -i, the i-th argument had an illegal value
            > 0:  if INFO = i,  and i is
                  <= M:  the i-th row of A is exactly zero
                  >  M:  the (i-M)-th column of A is exactly zero

    GRID    (input) gridinof_t*
            The 2D process mesh.
    =====================================================================

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgsmv()

void psgsmv	(	int_t	abs,
		SuperMatrix *	A_internal,
		gridinfo_t *	grid,
		psgsmv_comm_t *	gsmv_comm,
		float	x[],
		float	ax[]
	)

Here is the caller graph for this function:

◆ psgsmv_AXglobal()

int psgsmv_AXglobal	(	int_t	m,
		int_t	update[],
		float	val[],
		int_t	bindx[],
		float	X[],
		float	ax[]
	)

Performs sparse matrix-vector multiplication.
val/bindx stores the distributed MSR matrix A
X is global
ax product is distributed the same way as A

Here is the caller graph for this function:

◆ psgsmv_AXglobal_abs()

int psgsmv_AXglobal_abs	(	int_t	m,
		int_t	update[],
		float	val[],
		int_t	bindx[],
		float	X[],
		float	ax[]
	)

Here is the caller graph for this function:

◆ psgsmv_AXglobal_setup()

int psgsmv_AXglobal_setup	(	SuperMatrix *	,
		Glu_persist_t *	,
		gridinfo_t *	,
		int_t *	,
		int_t *	[],
		float *	[],
		int_t *	[],
		int_t	[]
	)

Here is the caller graph for this function:

◆ psgsmv_finalize()

void psgsmv_finalize ( psgsmv_comm_t * gsmv_comm )

Here is the caller graph for this function:

◆ psgsmv_init()

void psgsmv_init	(	SuperMatrix *	A,
		int_t *	row_to_proc,
		gridinfo_t *	grid,
		psgsmv_comm_t *	gsmv_comm
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgsrfs()

void psgsrfs	(	superlu_dist_options_t *	,
		int_t	,
		SuperMatrix *	,
		float	,
		sLUstruct_t *	,
		sScalePermstruct_t *	,
		gridinfo_t *	,
		float	[],
		int_t	,
		float	[],
		int_t	,
		int	,
		sSOLVEstruct_t *	,
		float *	,
		SuperLUStat_t *	,
		int *
	)

Here is the caller graph for this function:

◆ psgsrfs3d()

void psgsrfs3d	(	superlu_dist_options_t *	options,
		int_t	n,
		SuperMatrix *	A,
		float	anorm,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		gridinfo3d_t *	grid3d,
		strf3Dpartition_t *	trf3Dpartition,
		float *	B,
		int_t	ldb,
		float *	X,
		int_t	ldx,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		float *	berr,
		SuperLUStat_t *	stat,
		int *	info
	)

Purpose
=======

PSGSRFS3D improves the computed solution to a system of linear
equations and provides error bounds and backward error estimates
for the solution.

Arguments
=========

options (input) superlu_dist_options_t* (global)
        The structure defines the input parameters to control
        how the LU decomposition and triangular solve are performed.

n      (input) int (global)
       The order of the system of linear equations.

A      (input) SuperMatrix*
   The original matrix A, or the scaled A if equilibration was done.
       A is also permuted into diag(R)*A*diag(C)*Pc'. The type of A can be:
       Stype = SLU_NR_loc; Dtype = SLU_S; Mtype = SLU_GE.

anorm  (input) double
       The norm of the original matrix A, or the scaled A if
       equilibration was done.

LUstruct (input) sLUstruct_t*
       The distributed data structures storing L and U factors.
       The L and U factors are obtained from pdgstrf for
       the possibly scaled and permuted matrix A.
       See superlu_sdefs.h for the definition of 'sLUstruct_t'.

ScalePermstruct (input) sScalePermstruct_t* (global)
        The data structure to store the scaling and permutation vectors
        describing the transformations performed to the matrix A.

grid   (input) gridinfo_t*
       The 2D process mesh. It contains the MPI communicator, the number
       of process rows (NPROW), the number of process columns (NPCOL),
       and my process rank. It is an input argument to all the
       parallel routines.
       Grid can be initialized by subroutine SUPERLU_GRIDINIT.
       See superlu_defs.h for the definition of 'gridinfo_t'.

B      (input) float* (local)
       The m_loc-by-NRHS right-hand side matrix of the possibly
       equilibrated system. That is, B may be overwritten by diag(R)*B.

ldb    (input) int (local)
       Leading dimension of matrix B.

X      (input/output) float* (local)
       On entry, the solution matrix Y, as computed by PSGSTRS, of the
           transformed system A1*Y = Pc*Pr*B. where
           A1 = Pc*Pr*diag(R)*A*diag(C)*Pc' and Y = Pc*diag(C)^(-1)*X.
       On exit, the improved solution matrix Y.

       In order to obtain the solution X to the original system,
       Y should be permutated by Pc^T, and premultiplied by diag(C)
       if DiagScale = COL or BOTH.
       This must be done after this routine is called.

ldx    (input) int (local)
       Leading dimension of matrix X.

nrhs   (input) int
       Number of right-hand sides.

SOLVEstruct (output) sSOLVEstruct_t* (global)
       Contains the information for the communication during the
       solution phase.

berr   (output) float*, dimension (nrhs)
        The componentwise relative backward error of each solution
        vector X(j) (i.e., the smallest relative change in
        any element of A or B that makes X(j) an exact solution).

stat   (output) SuperLUStat_t*
       Record the statistics about the refinement steps.
       See util.h for the definition of SuperLUStat_t.

info   (output) int*
       = 0: successful exit
       < 0: if info = -i, the i-th argument had an illegal value

Internal Parameters
===================

ITMAX is the maximum number of steps of iterative refinement.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgsrfs_ABXglobal()

void psgsrfs_ABXglobal	(	superlu_dist_options_t *	options,
		int_t	n,
		SuperMatrix *	A,
		float	anorm,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		float *	B,
		int_t	ldb,
		float *	X,
		int_t	ldx,
		int	nrhs,
		float *	berr,
		SuperLUStat_t *	stat,
		int *	info
	)

Purpose
=======

psgsrfs_ABXglobal improves the computed solution to a system of linear
equations and provides error bounds and backward error estimates
for the solution.

Arguments
=========

n      (input) int (global)
       The order of the system of linear equations.

A      (input) SuperMatrix*
   The original matrix A, or the scaled A if equilibration was done.
       A is also permuted into the form Pc*Pr*A*Pc', where Pr and Pc
       are permutation matrices. The type of A can be:
       Stype = SLU_NCP; Dtype = SLU_S; Mtype = SLU_GE.

       NOTE: Currently, A must reside in all processes when calling
             this routine.

anorm  (input) double
       The norm of the original matrix A, or the scaled A if
       equilibration was done.

LUstruct (input) sLUstruct_t*
       The distributed data structures storing L and U factors.
       The L and U factors are obtained from psgstrf for
       the possibly scaled and permuted matrix A.
       See superlu_ddefs.h for the definition of 'sLUstruct_t'.

grid   (input) gridinfo_t*
       The 2D process mesh. It contains the MPI communicator, the number
       of process rows (NPROW), the number of process columns (NPCOL),
       and my process rank. It is an input argument to all the
       parallel routines.
       Grid can be initialized by subroutine SUPERLU_GRIDINIT.
       See superlu_ddefs.h for the definition of 'gridinfo_t'.

B      (input) float* (global)
       The N-by-NRHS right-hand side matrix of the possibly equilibrated
       and row permuted system.

       NOTE: Currently, B must reside on all processes when calling
             this routine.

ldb    (input) int (global)
       Leading dimension of matrix B.

X      (input/output) float* (global)
       On entry, the solution matrix X, as computed by PSGSTRS.
       On exit, the improved solution matrix X.
       If DiagScale = COL or BOTH, X should be premultiplied by diag(C)
       in order to obtain the solution to the original system.

       NOTE: Currently, X must reside on all processes when calling
             this routine.

ldx    (input) int (global)
       Leading dimension of matrix X.

nrhs   (input) int
       Number of right-hand sides.

berr   (output) double*, dimension (nrhs)
        The componentwise relative backward error of each solution
        vector X(j) (i.e., the smallest relative change in
        any element of A or B that makes X(j) an exact solution).

stat   (output) SuperLUStat_t*
       Record the statistics about the refinement steps.
       See util.h for the definition of SuperLUStat_t.

info   (output) int*
       = 0: successful exit
       < 0: if info = -i, the i-th argument had an illegal value

Internal Parameters
===================

ITMAX is the maximum number of steps of iterative refinement.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgssvx()

void psgssvx	(	superlu_dist_options_t *	,
		SuperMatrix *	,
		sScalePermstruct_t *	,
		float *	,
		int	,
		int	,
		gridinfo_t *	,
		sLUstruct_t *	,
		sSOLVEstruct_t *	,
		float *	,
		SuperLUStat_t *	,
		int *
	)

Here is the caller graph for this function:

◆ psgssvx3d()

void psgssvx3d	(	superlu_dist_options_t *	options,
		SuperMatrix *	A,
		sScalePermstruct_t *	ScalePermstruct,
		float	B[],
		int	ldb,
		int	nrhs,
		gridinfo3d_t *	grid3d,
		sLUstruct_t *	LUstruct,
		sSOLVEstruct_t *	SOLVEstruct,
		float *	berr,
		SuperLUStat_t *	stat,
		int *	info
	)

Here is the caller graph for this function:

◆ psgssvx3d_csc_batch()

int psgssvx3d_csc_batch	(	superlu_dist_options_t *	options,
		int	batchCount,
		int	m,
		int	n,
		int	nnz,
		int	nrhs,
		handle_t *	SparseMatrix_handles,
		float **	RHSptr,
		int *	ldRHS,
		float **	ReqPtr,
		float **	CeqPtr,
		int **	RpivPtr,
		int **	CpivPtr,
		DiagScale_t *	DiagScale,
		handle_t *	F,
		float **	Xptr,
		int *	ldX,
		float **	Berrs,
		gridinfo3d_t *	grid3d,
		SuperLUStat_t *	stat,
		int *	info
	)

Solve a batch of linear systems Ai * Xi = Bi with direct method, computing the LU factorization of each matrix Ai;
This is the fixed-size interface: all the input matrices have the same sparsity structure.

Parameters

  
    [in] options solver options

    [in] batchCount number of matrices in the batch

    [in] m row dimension of the matrices

    [in] n column dimension of the matrices

    [in] nnz number of non-zero entries in each matrix

    [in] nrhs number of right-hand-sides

    [in,out] SparseMatrix_handles  array of sparse matrix handles, of size 'batchCount', each pointing to the actual storage in CSC format, see 'NCformat' in SuperMatix structure
     Each A is overwritten by row/col scaling R*A*C

    [in,out] RHSptr  array of pointers to dense storage of right-hand sides B
     Each B is overwritten by row/col scaling R*B*C

    [in] ldRHS array of leading dimensions of RHS

    [in,out] ReqPtr array of pointers to diagonal row scaling vectors R, each of size m
   ReqPtr[] are allocated internally if equilibration is asked for

    [in,out] CeqPtr array of pointers to diagonal colum scaling vectors C, each of size n
   CeqPtr[] are allocated internally if equilibration is asked for

    [in,out] RpivPtr array of pointers to row permutation vectors, each of size m

    [in,out] CpivPtr array of pointers to column permutation vectors, each of size n

    [in,out] DiagScale array of indicators how equilibration is done for each matrix

    [out] F array of handles pointing to the factored matrices

    [out] Xptr array of pointers to dense storage of solution

    [in] ldX array of leading dimensions of X

    [out] Berrs array of poiniters to backward errors

    [in] 
  
  

     grid3d contains MPI communicator
Parameters

  
    [out] stat records algorithms statistics such as runtime, memory usage, etc.

    [out] info flags the errors on return

!!! CHECK SETTING: TO BE SURE TO USE GPU VERSIONS !!!! gpu3dVersion superlu_acc_offload

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgssvx_ABglobal()

void psgssvx_ABglobal	(	superlu_dist_options_t *	,
		SuperMatrix *	,
		sScalePermstruct_t *	,
		float *	,
		int	,
		int	,
		gridinfo_t *	,
		sLUstruct_t *	,
		float *	,
		SuperLUStat_t *	,
		int *
	)

Here is the caller graph for this function:

◆ psgsTrBackSolve3d()

int_t psgsTrBackSolve3d	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	x3d,
		float *	lsum3d,
		sxT_struct *	xT_s,
		float *	recvbuf,
		MPI_Request *	send_req,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Loop over all the levels from root to leaf

Adding dlsumBmod_buff_t* lbmod_buf

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgsTrBackSolve3d_newsolve()

int_t psgsTrBackSolve3d_newsolve	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	x3d,
		float *	lsum3d,
		float *	recvbuf,
		MPI_Request *	send_req,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Adding dlsumBmod_buff_t* lbmod_buf

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgstrf()

int_t psgstrf	(	superlu_dist_options_t *	options,
		int	m,
		int	n,
		float	anorm,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		SuperLUStat_t *	stat,
		int *	info
	)

Purpose
=======

PSGSTRF performs the LU factorization in parallel.

Arguments
=========

options (input) superlu_dist_options_t*
        The structure defines the input parameters to control
        how the LU decomposition will be performed.
        The following field should be defined:
        o ReplaceTinyPivot (yes_no_t)
          = NO:  do not modify pivots
          = YES: replace tiny pivots by sqrt(epsilon)*norm(A) during
                 LU factorization.

m      (input) int
       Number of rows in the matrix.

n      (input) int
       Number of columns in the matrix.

anorm  (input) float
       The norm of the original matrix A, or the scaled A if
       equilibration was done.

LUstruct (input/output) sLUstruct_t*
        The data structures to store the distributed L and U factors.
        The following fields should be defined:

        o Glu_persist (input) Glu_persist_t*
          Global data structure (xsup, supno) replicated on all processes,
          describing the supernode partition in the factored matrices
          L and U:
        xsup[s] is the leading column of the s-th supernode,
            supno[i] is the supernode number to which column i belongs.

        o Llu (input/output) sLocalLU_t*
          The distributed data structures to store L and U factors.
          See superlu_sdefs.h for the definition of 'sLocalLU_t'.

grid   (input) gridinfo_t*
       The 2D process mesh. It contains the MPI communicator, the number
       of process rows (NPROW), the number of process columns (NPCOL),
       and my process rank. It is an input argument to all the
       parallel routines.
       Grid can be initialized by subroutine SUPERLU_GRIDINIT.
       See superlu_ddefs.h for the definition of 'gridinfo_t'.

stat   (output) SuperLUStat_t*
       Record the statistics on runtime and floating-point operation count.
       See util.h for the definition of 'SuperLUStat_t'.

info   (output) int*
       = 0: successful exit
       < 0: if info = -i, the i-th argument had an illegal value
       > 0: if info = i, U(i,i) is exactly zero. The factorization has
            been completed, but the factor U is exactly singular,
            and division by zero will occur if it is used to solve a
            system of equations.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgstrf2()

void psgstrf2	(	superlu_dist_options_t *	,
		int_t	nsupers,
		int_t	k0,
		int_t	k,
		double	thresh,
		Glu_persist_t *	,
		gridinfo_t *	,
		sLocalLU_t *	,
		MPI_Request *	,
		int	,
		SuperLUStat_t *	,
		int *
	)

◆ psgstrf2_trsm()

void psgstrf2_trsm	(	superlu_dist_options_t *	options,
		int_t	k0,
		int_t	k,
		double	thresh,
		Glu_persist_t *	Glu_persist,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		MPI_Request *	U_diag_blk_send_req,
		int	tag_ub,
		SuperLUStat_t *	stat,
		int *	info
	)

Purpose
=======
  Panel factorization -- block column k

  Factor diagonal and subdiagonal blocks and test for exact singularity.
  Only the column processes that own block column *k* participate
  in the work.

Arguments
=========
options (input) superlu_dist_options_t* (global)
        The structure defines the input parameters to control
        how the LU decomposition will be performed.

k0     (input) int (global)
       Counter of the next supernode to be factorized.

k      (input) int (global)
       The column number of the block column to be factorized.

thresh (input) double (global)
       The threshold value = s_eps * anorm.

Glu_persist (input) Glu_persist_t*
       Global data structures (xsup, supno) replicated on all processes.

grid   (input) gridinfo_t*
       The 2D process mesh.

Llu    (input/output) sLocalLU_t*
       Local data structures to store distributed L and U matrices.

U_diag_blk_send_req (input/output) MPI_Request*
       List of send requests to send down the diagonal block of U.

tag_ub (input) int
       Upper bound of MPI tag values.

stat   (output) SuperLUStat_t*
       Record the statistics about the factorization.
       See SuperLUStat_t structure defined in util.h.

info   (output) int*
       = 0: successful exit
       < 0: if info = -i, the i-th argument had an illegal value
       > 0: if info = i, U(i,i) is exactly zero. The factorization has
            been completed, but the factor U is exactly singular,
            and division by zero will occur if it is used to solve a
            system of equations.

ALWAYS SEND TO ALL OTHERS - TO FIX

Here is the call graph for this function:

◆ psgstrf3d()

int_t psgstrf3d	(	superlu_dist_options_t *	options,
		int	m,
		int	n,
		float	anorm,
		strf3Dpartition_t *	trf3Dpartition,
		SCT_t *	SCT,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SuperLUStat_t *	stat,
		int *	info
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgsTrForwardSolve3d()

int_t psgsTrForwardSolve3d	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	x3d,
		float *	lsum3d,
		sxT_struct *	xT_s,
		float *	recvbuf,
		MPI_Request *	send_req,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Loop over all the levels from root to leaf

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgsTrForwardSolve3d_newsolve()

int_t psgsTrForwardSolve3d_newsolve	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	x3d,
		float *	lsum3d,
		float *	recvbuf,
		MPI_Request *	send_req,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgstrs()

void psgstrs	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		gridinfo_t *	grid,
		float *	B,
		int_t	m_loc,
		int_t	fst_row,
		int_t	ldb,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		int *	info
	)

Purpose
=======

PSGSTRS solves a system of distributed linear equations
A*X = B with a general N-by-N matrix A using the LU factorization
computed by PSGSTRF.
If the equilibration, and row and column permutations were performed,
the LU factorization was performed for A1 where
    A1 = Pc*Pr*diag(R)*A*diag(C)*Pc^T = L*U
and the linear system solved is
    A1 * Y = Pc*Pr*B1, where B was overwritten by B1 = diag(R)*B, and
the permutation to B1 by Pc*Pr is applied internally in this routine.

Arguments
=========

options (input) superlu_dist_options_t*
        The structure defines the input parameters to control
        how the LU decomposition and triangular solve are performed.

n      (input) int (global)
       The order of the system of linear equations.

LUstruct (input) sLUstruct_t*
       The distributed data structures storing L and U factors.
       The L and U factors are obtained from PSGSTRF for
       the possibly scaled and permuted matrix A.
       See superlu_sdefs.h for the definition of 'sLUstruct_t'.
       A may be scaled and permuted into A1, so that
       A1 = Pc*Pr*diag(R)*A*diag(C)*Pc^T = L*U

grid   (input) gridinfo_t*
       The 2D process mesh. It contains the MPI communicator, the number
       of process rows (NPROW), the number of process columns (NPCOL),
       and my process rank. It is an input argument to all the
       parallel routines.
       Grid can be initialized by subroutine SUPERLU_GRIDINIT.
       See superlu_defs.h for the definition of 'gridinfo_t'.

B      (input/output) float*
       On entry, the distributed right-hand side matrix of the possibly
       equilibrated system. That is, B may be overwritten by diag(R)*B.
       On exit, the distributed solution matrix Y of the possibly
       equilibrated system if info = 0, where Y = Pc*diag(C)^(-1)*X,
       and X is the solution of the original system.

m_loc  (input) int (local)
       The local row dimension of matrix B.

fst_row (input) int (global)
       The row number of B's first row in the global matrix.

ldb    (input) int (local)
       The leading dimension of matrix B.

nrhs   (input) int (global)
       Number of right-hand sides.

SOLVEstruct (input) sSOLVEstruct_t* (global)
       Contains the information for the communication during the
       solution phase.

stat   (output) SuperLUStat_t*
       Record the statistics about the triangular solves.
       See util.h for the definition of 'SuperLUStat_t'.

info   (output) int*
       = 0: successful exit
    < 0: if info = -i, the i-th argument had an illegal value

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgstrs2()

void psgstrs2	(	int_t	m,
		int_t	k0,
		int_t	k,
		Glu_persist_t *	Glu_persist,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		SuperLUStat_t *	stat
	)

◆ psgstrs2_omp()

void psgstrs2_omp	(	int_t	k0,
		int_t	k,
		Glu_persist_t *	Glu_persist,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		Ublock_info_t *	Ublock_info,
		SuperLUStat_t *	stat
	)

4/19/2019

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgstrs3d()

void psgstrs3d	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	B,
		int_t	m_loc,
		int_t	fst_row,
		int_t	ldb,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		int *	info
	)

Purpose


PSGSTRS solves a system of distributed linear equations
A*X = B with a general N-by-N matrix A using the LU factorization
computed by PSGSTRF.
If the equilibration, and row and column permutations were performed,
the LU factorization was performed for A1 where
    A1 = Pc*Pr*diag(R)*A*diag(C)*Pc^T = L*U
and the linear system solved is
    A1 * Y = Pc*Pr*B1, where B was overwritten by B1 = diag(R)*B, and
the permutation to B1 by Pc*Pr is applied internally in this routine.

Arguments


n      (input) int (global)
       The order of the system of linear equations.

LUstruct (input) sLUstruct_t*
       The distributed data structures storing L and U factors.
       The L and U factors are obtained from PSGSTRF for
       the possibly scaled and permuted matrix A.
       See superlu_ddefs.h for the definition of 'sLUstruct_t'.
       A may be scaled and permuted into A1, so that
       A1 = Pc*Pr*diag(R)*A*diag(C)*Pc^T = L*U

grid   (input) gridinfo_t*
       The 2D process mesh. It contains the MPI communicator, the number
       of process rows (NPROW), the number of process columns (NPCOL),
       and my process rank. It is an input argument to all the
       parallel routines.
       Grid can be initialized by subroutine SUPERLU_GRIDINIT.
       See superlu_defs.h for the definition of 'gridinfo_t'.

B      (input/output) float*
       On entry, the distributed right-hand side matrix of the possibly
       equilibrated system. That is, B may be overwritten by diag(R)*B.
       On exit, the distributed solution matrix Y of the possibly
       equilibrated system if info = 0, where Y = Pc*diag(C)^(-1)*X,
       and X is the solution of the original system.

m_loc  (input) int (local)
       The local row dimension of matrix B.

fst_row (input) int (global)
       The row number of B's first row in the global matrix.

ldb    (input) int (local)
       The leading dimension of matrix B.

nrhs   (input) int (global)
       Number of right-hand sides.

SOLVEstruct (input) sSOLVEstruct_t* (global)
       Contains the information for the communication during the
       solution phase.

stat   (output) SuperLUStat_t*
       Record the statistics about the triangular solves.
       See util.h for the definition of 'SuperLUStat_t'.

info   (output) int*
    = 0: successful exit
    < 0: if info = -i, the i-th argument had an illegal value

Initializing xT

Setup the headers for xT

Reduce the Solve flops from all the grids to grid zero

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgstrs3d_newsolve()

void psgstrs3d_newsolve	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	B,
		int_t	m_loc,
		int_t	fst_row,
		int_t	ldb,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		int *	info
	)

Reduce the Solve flops from all the grids to grid zero

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgstrs_Bglobal()

void psgstrs_Bglobal	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		float *	B,
		int_t	ldb,
		int	nrhs,
		SuperLUStat_t *	stat,
		int *	info
	)

Purpose
=======

psgstrs_Bglobal solves a system of distributed linear equations
A*X = B with a general N-by-N matrix A using the LU factorization
computed by psgstrf.

Arguments
=========

options (input) superlu_dist_options_t*
        The structure defines the input parameters to control
        how the LU decomposition and triangular solve are performed.

n      (input) int (global)
       The order of the system of linear equations.

LUstruct (input) sLUstruct_t*
       The distributed data structures storing L and U factors.
       The L and U factors are obtained from psgstrf for
       the possibly scaled and permuted matrix A.
       See superlu_ddefs.h for the definition of 'sLUstruct_t'.

grid   (input) gridinfo_t*
       The 2D process mesh. It contains the MPI communicator, the number
       of process rows (NPROW), the number of process columns (NPCOL),
       and my process rank. It is an input argument to all the
       parallel routines.
       Grid can be initialized by subroutine SUPERLU_GRIDINIT.
       See superlu_ddefs.h for the definition of 'gridinfo_t'.

B      (input/output) float*
       On entry, the right-hand side matrix of the possibly equilibrated
       and row permuted system.
       On exit, the solution matrix of the possibly equilibrated
       and row permuted system if info = 0;

       NOTE: Currently, the N-by-NRHS  matrix B must reside on all
             processes when calling this routine.

ldb    (input) int (global)
       Leading dimension of matrix B.

nrhs   (input) int (global)
       Number of right-hand sides.

stat   (output) SuperLUStat_t*
       Record the statistics about the triangular solves.
       See util.h for the definition of 'SuperLUStat_t'.

info   (output) int*
       = 0: successful exit
    < 0: if info = -i, the i-th argument had an illegal value

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psgstrs_delete_device_lsum_x()

int_t psgstrs_delete_device_lsum_x ( sSOLVEstruct_t * SOLVEstruct )

Here is the caller graph for this function:

◆ psgstrs_init()

int_t psgstrs_init	(	int_t	n,
		int_t	m_loc,
		int_t	nrhs,
		int_t	fst_row,
		int_t	perm_r[],
		int_t	perm_c[],
		gridinfo_t *	grid,
		Glu_persist_t *	Glu_persist,
		sSOLVEstruct_t *	SOLVEstruct
	)

Purpose
=======
  Set up the communication pattern for redistribution between B and X
  in the triangular solution.

Arguments
=========

n      (input) int (global)
       The dimension of the linear system.

m_loc  (input) int (local)
       The local row dimension of the distributed input matrix.

nrhs   (input) int (global)
       Number of right-hand sides.

fst_row (input) int (global)
       The row number of matrix B's first row in the global matrix.

perm_r (input) int* (global)
       The row permutation vector.

perm_c (input) int* (global)
       The column permutation vector.

grid   (input) gridinfo_t*
       The 2D process mesh.

Here is the caller graph for this function:

◆ psgstrs_init_device_lsum_x()

int_t psgstrs_init_device_lsum_x	(	superlu_dist_options_t *	options,
		int_t	n,
		int_t	m_loc,
		int_t	nrhs,
		gridinfo_t *	grid,
		sLUstruct_t *	LUstruct,
		sSOLVEstruct_t *	SOLVEstruct,
		int *	supernodeMask
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psinf_norm_error()

void psinf_norm_error	(	int	iam,
		int_t	n,
		int_t	nrhs,
		float	x[],
		int_t	ldx,
		float	xtrue[],
		int_t	ldxtrue,
		MPI_Comm	slucomm
	)

Check the inf-norm of the error vector.

Here is the caller graph for this function:

◆ pslangs()

float pslangs	(	char *	norm,
		SuperMatrix *	A,
		gridinfo_t *	grid
	)

    Purpose
    =======

    PSLANGS returns the value of the one norm, or the Frobenius norm, or
    the infinity norm, or the element of largest absolute value of a
    real matrix A.

    Description
    ===========

    PSLANGE returns the value

       PSLANGE = ( max(abs(A(i,j))), NORM = 'M' or 'm'
                 (
                 ( norm1(A),         NORM = '1', 'O' or 'o'
                 (
                 ( normI(A),         NORM = 'I' or 'i'
                 (
                 ( normF(A),         NORM = 'F', 'f', 'E' or 'e'

    where  norm1  denotes the  one norm of a matrix (maximum column sum),
    normI  denotes the  infinity norm  of a matrix  (maximum row sum) and
    normF  denotes the  Frobenius norm of a matrix (square root of sum of
    squares).  Note that  max(abs(A(i,j)))  is not a  matrix norm.

    Arguments
    =========

    NORM    (input) CHARACTER*1
            Specifies the value to be returned in DLANGE as described above.
    A       (input) SuperMatrix*
            The M by N sparse matrix A.
    GRID    (input) gridinof_t*
            The 2D process mesh.
   =====================================================================

Here is the call graph for this function:

Here is the caller graph for this function:

◆ pslaqgs()

void pslaqgs	(	SuperMatrix *	A,
		float *	r,
		float *	c,
		float	rowcnd,
		float	colcnd,
		float	amax,
		char *	equed
	)

    Purpose
    =======

    PSLAQGS equilibrates a general sparse M by N matrix A using the row
    and column scaling factors in the vectors R and C.

    See supermatrix.h for the definition of 'SuperMatrix' structure.

    Arguments
    =========

    A       (input/output) SuperMatrix*
            On exit, the equilibrated matrix.  See EQUED for the form of
            the equilibrated matrix. The type of A can be:
        Stype = SLU_NR_loc; Dtype = SLU_S; Mtype = SLU_GE.

    R       (input) float*, dimension (A->nrow)
            The row scale factors for A.

    C       (input) float*, dimension (A->ncol)
            The column scale factors for A.

    ROWCND  (input) float
            Ratio of the smallest R(i) to the largest R(i).

    COLCND  (input) float
            Ratio of the smallest C(i) to the largest C(i).

    AMAX    (input) float
            Absolute value of largest matrix entry.

    EQUED   (output) char*
            Specifies the form of equilibration that was done.
            = 'N':  No equilibration
            = 'R':  Row equilibration, i.e., A has been premultiplied by
                    diag(R).
            = 'C':  Column equilibration, i.e., A has been postmultiplied
                    by diag(C).
            = 'B':  Both row and column equilibration, i.e., A has been
                    replaced by diag(R) * A * diag(C).

    Internal Parameters
    ===================

    THRESH is a threshold value used to decide if row or column scaling
    should be done based on the ratio of the row or column scaling
    factors.  If ROWCND < THRESH, row scaling is done, and if
    COLCND < THRESH, column scaling is done.

    LARGE and SMALL are threshold values used to decide if row scaling
    should be done based on the absolute size of the largest matrix
    element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.

    =====================================================================

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psPermute_Dense_Matrix()

int psPermute_Dense_Matrix	(	int_t	fst_row,
		int_t	m_loc,
		int_t	row_to_proc[],
		int_t	perm[],
		float	X[],
		int	ldx,
		float	B[],
		int	ldb,
		int	nrhs,
		gridinfo_t *	grid
	)

Permute the distributed dense matrix: B <= perm(X). perm[i] = j means the i-th row of X is in the j-th row of B.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psReDistribute3d_B_to_X()

int_t psReDistribute3d_B_to_X	(	float *	B,
		int_t	m_loc,
		int	nrhs,
		int_t	ldb,
		int_t	fst_row,
		int_t *	ilsum,
		float *	x,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_persist_t *	Glu_persist,
		gridinfo3d_t *	grid3d,
		sSOLVEstruct_t *	SOLVEstruct
	)

Purpose

  Re-distribute B on the diagonal processes of the 2D process mesh (only on grid 0).

Note

  This routine can only be called after the routine pxgstrs_init(),
  in which the structures of the send and receive buffers are set up.

Arguments


B      (input) float*
       The distributed right-hand side matrix of the possibly
       equilibrated system.

m_loc  (input) int (local)
       The local row dimension of matrix B.

nrhs   (input) int (global)
       Number of right-hand sides.

ldb    (input) int (local)
       Leading dimension of matrix B.

fst_row (input) int (global)
       The row number of B's first row in the global matrix.

ilsum  (input) int* (global)
       Starting position of each supernode in a full array.

x      (output) float*
       The solution vector. It is valid only on the diagonal processes.

ScalePermstruct (input) dScalePermstruct_t*
       The data structure to store the scaling and permutation vectors
       describing the transformations performed to the original matrix A.

grid   (input) gridinfo_t*
       The 2D process mesh.

SOLVEstruct (input) dSOLVEstruct_t*
       Contains the information for the communication during the
       solution phase.

Return value

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psReDistribute3d_X_to_B()

int_t psReDistribute3d_X_to_B	(	int_t	n,
		float *	B,
		int_t	m_loc,
		int_t	ldb,
		int_t	fst_row,
		int	nrhs,
		float *	x,
		int_t *	ilsum,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_persist_t *	Glu_persist,
		gridinfo3d_t *	grid3d,
		sSOLVEstruct_t *	SOLVEstruct
	)

Purpose

  Re-distribute X on the diagonal processes to B distributed on all
  the processes (only on grid 0)

Note

  This routine can only be called after the routine pxgstrs_init(),
  in which the structures of the send and receive buffers are set up.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ psReDistribute_B_to_X()

int_t psReDistribute_B_to_X	(	float *	B,
		int_t	m_loc,
		int	nrhs,
		int_t	ldb,
		int_t	fst_row,
		int_t *	ilsum,
		float *	x,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_persist_t *	Glu_persist,
		gridinfo_t *	grid,
		sSOLVEstruct_t *	SOLVEstruct
	)

Purpose
=======
  Re-distribute B on the diagonal processes of the 2D process mesh.

Note
====
  This routine can only be called after the routine psgstrs_init(),
  in which the structures of the send and receive buffers are set up.

Arguments
=========

B      (input) float*
       The distributed right-hand side matrix of the possibly
       equilibrated system.

m_loc  (input) int (local)
       The local row dimension of matrix B.

nrhs   (input) int (global)
       Number of right-hand sides.

ldb    (input) int (local)
       Leading dimension of matrix B.

fst_row (input) int (global)
       The row number of B's first row in the global matrix.

ilsum  (input) int* (global)
       Starting position of each supernode in a full array.

x      (output) float*
       The solution vector. It is valid only on the diagonal processes.

ScalePermstruct (input) sScalePermstruct_t*
       The data structure to store the scaling and permutation vectors
       describing the transformations performed to the original matrix A.

grid   (input) gridinfo_t*
       The 2D process mesh.

SOLVEstruct (input) sSOLVEstruct_t*
       Contains the information for the communication during the
       solution phase.

Return value
============

Here is the call graph for this function:

Here is the caller graph for this function:

◆ pxgstrs_finalize()

void pxgstrs_finalize ( pxgstrs_comm_t * gstrs_comm )

◆ s3D_printMemUse()

void s3D_printMemUse	(	strf3Dpartition_t *	trf3Dpartition,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ s_c2cpp_GetHWPM()

int s_c2cpp_GetHWPM	(	SuperMatrix *	A,
		gridinfo_t *	grid,
		sScalePermstruct_t *	ScalePermstruct
	)

Purpose
=======

Get heavy-weight perfect matching (HWPM).

Reference:


Arguments
=========

A      (input) SuperMatrix*
       The distributed input matrix A of dimension (A->nrow, A->ncol).
       The type of A can be: Stype = SLU_NR_loc; Dtype = SLU_D; Mtype = SLU_GE.

grid   (input) gridinfo_t*
       SuperLU's 2D process mesh.

ScalePermstruct (output) sScalePermstruct_t*
       ScalePermstruct->perm_r stores the permutation obtained from HWPM.

Here is the caller graph for this function:

◆ sallocateA_dist()

void sallocateA_dist	(	int_t	n,
		int_t	nnz,
		float **	a,
		int_t **	asub,
		int_t **	xa
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sAllocGlu_3d()

int sAllocGlu_3d	(	int_t	n,
		int_t	nsupers,
		sLUstruct_t *	LUstruct
	)

Here is the call graph for this function:

◆ sAllocLlu_3d()

int_t sAllocLlu_3d	(	int_t	nsupers,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ sallocScalePermstruct_RC()

void sallocScalePermstruct_RC	(	sScalePermstruct_t *	ScalePermstruct,
		int_t	m,
		int_t	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ saxpy_()

int saxpy_	(	const int *	n,
		const float *	alpha,
		const float *	x,
		const int *	incx,
		float *	y,
		const int *	incy
	)

◆ sBcast_LPanel()

int_t sBcast_LPanel	(	int_t	k,
		int_t	k0,
		int_t *	lsub,
		float *	lusup,
		gridinfo_t *	,
		int *	msgcnt,
		int **	ToSendR,
		int_t *	xsup,
		SCT_t *	,
		int
	)

◆ sBcast_UPanel()

int_t sBcast_UPanel	(	int_t	k,
		int_t	k0,
		int_t *	usub,
		float *	uval,
		gridinfo_t *	,
		int *	msgcnt,
		int *	ToSendD,
		SCT_t *	,
		int
	)

◆ sbcastPermutedSparseA()

void sbcastPermutedSparseA	(	SuperMatrix *	A,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_freeable_t *	Glu_freeable,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sbCastXk2Pck()

int_t sbCastXk2Pck	(	int_t	k,
		sxT_struct *	xT_s,
		int	nrhs,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		xtrsTimer_t *	xtrsTimer
	)

Here is the caller graph for this function:

◆ sblock_gemm_scatter()

void sblock_gemm_scatter	(	int_t	lb,
		int_t	j,
		Ublock_info_t *	Ublock_info,
		Remain_info_t *	Remain_info,
		float *	L_mat,
		int	ldl,
		float *	U_mat,
		int	ldu,
		float *	bigV,
		int_t	knsupc,
		int_t	klst,
		int_t *	lsub,
		int_t *	usub,
		int_t	ldt,
		int_t	thread_id,
		int *	indirect,
		int *	indirect2,
		int_t **	Lrowind_bc_ptr,
		float **	Lnzval_bc_ptr,
		int_t **	Ufstnz_br_ptr,
		float **	Unzval_br_ptr,
		int_t *	xsup,
		gridinfo_t *	,
		SuperLUStat_t *
	)

◆ sblock_gemm_scatterBottomLeft()

int_t sblock_gemm_scatterBottomLeft	(	int_t	lb,
		int_t	j,
		float *	bigV,
		int_t	knsupc,
		int_t	klst,
		int_t *	lsub,
		int_t *	usub,
		int_t	ldt,
		int *	indirect,
		int *	indirect2,
		HyP_t *	HyP,
		sLUstruct_t *	,
		gridinfo_t *	,
		SCT_t *	SCT,
		SuperLUStat_t *
	)

◆ sblock_gemm_scatterBottomRight()

int_t sblock_gemm_scatterBottomRight	(	int_t	lb,
		int_t	j,
		float *	bigV,
		int_t	knsupc,
		int_t	klst,
		int_t *	lsub,
		int_t *	usub,
		int_t	ldt,
		int *	indirect,
		int *	indirect2,
		HyP_t *	HyP,
		sLUstruct_t *	,
		gridinfo_t *	,
		SCT_t *	SCT,
		SuperLUStat_t *
	)

◆ sblock_gemm_scatterTopLeft()

int_t sblock_gemm_scatterTopLeft	(	int_t	lb,
		int_t	j,
		float *	bigV,
		int_t	knsupc,
		int_t	klst,
		int_t *	lsub,
		int_t *	usub,
		int_t	ldt,
		int *	indirect,
		int *	indirect2,
		HyP_t *	HyP,
		sLUstruct_t *	,
		gridinfo_t *	,
		SCT_t *	SCT,
		SuperLUStat_t *
	)

◆ sblock_gemm_scatterTopRight()

int_t sblock_gemm_scatterTopRight	(	int_t	lb,
		int_t	j,
		float *	bigV,
		int_t	knsupc,
		int_t	klst,
		int_t *	lsub,
		int_t *	usub,
		int_t	ldt,
		int *	indirect,
		int *	indirect2,
		HyP_t *	HyP,
		sLUstruct_t *	,
		gridinfo_t *	,
		SCT_t *	SCT,
		SuperLUStat_t *
	)

◆ sbroadcastAncestor3d()

int_t sbroadcastAncestor3d	(	strf3Dpartition_t *	trf3Dpartition,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

Here is the caller graph for this function:

◆ sbsolve_Xt_bcast()

int_t sbsolve_Xt_bcast	(	int_t	ilvl,
		sxT_struct *	xT_s,
		int	nrhs,
		strf3Dpartition_t *	trf3Dpartition,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sClone_CompRowLoc_Matrix_dist()

void sClone_CompRowLoc_Matrix_dist	(	SuperMatrix *	,
		SuperMatrix *
	)

◆ scollect3dLpanels()

int_t scollect3dLpanels	(	int_t	layer,
		int_t	nsupers,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ scollect3dUpanels()

int_t scollect3dUpanels	(	int_t	layer,
		int_t	nsupers,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ sCompCol_to_CompRow_dist()

void sCompCol_to_CompRow_dist	(	int_t	m,
		int_t	n,
		int_t	nnz,
		float *	a,
		int_t *	colptr,
		int_t *	rowind,
		float **	at,
		int_t **	rowptr,
		int_t **	colind
	)

Here is the caller graph for this function:

◆ sCompRow_to_CompCol_dist()

void sCompRow_to_CompCol_dist	(	int_t	,
		int_t	,
		int_t	,
		float *	,
		int_t *	,
		int_t *	,
		float **	,
		int_t **	,
		int_t **
	)

Here is the caller graph for this function:

◆ scomputeA_Norm()

float scomputeA_Norm	(	int	notran,
		SuperMatrix *	A,
		gridinfo_t *	grid
	)

This function computes the norm of a matrix A.

Parameters

notran	A flag which determines the norm type to be calculated.
A	The input matrix for which the norm is computed.
grid	The gridinfo_t object that contains the information of the grid.

Returns: Returns the computed norm of the matrix A.

the iam process is the root (iam=0), it prints the computed norm to the standard output.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sComputeLevelsets()

void sComputeLevelsets	(	int	,
		int_t	,
		gridinfo_t *	,
		Glu_persist_t *	,
		sLocalLU_t *	,
		int_t *
	)

◆ sCopy_CompCol_Matrix_dist()

void sCopy_CompCol_Matrix_dist	(	SuperMatrix *	,
		SuperMatrix *
	)

◆ sCopy_CompRowLoc_Matrix_dist()

void sCopy_CompRowLoc_Matrix_dist	(	SuperMatrix *	A,
		SuperMatrix *	B
	)

◆ sCopy_Dense_Matrix_dist()

void sCopy_Dense_Matrix_dist	(	int_t	,
		int_t	,
		float *	,
		int_t	,
		float *	,
		int_t
	)

Here is the caller graph for this function:

◆ screate_batch_systems()

int screate_batch_systems	(	handle_t *	SparseMatrix_handles,
		int	batchCount,
		int	nrhs,
		float **	rhs,
		int *	ldb,
		float **	x,
		int *	ldx,
		FILE *	fp,
		char *	postfix,
		gridinfo3d_t *	grid3d
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ screate_block_diag_3d()

int screate_block_diag_3d	(	SuperMatrix *	A,
		int	batchCount,
		int	nrhs,
		float **	rhs,
		int *	ldb,
		float **	x,
		int *	ldx,
		FILE *	fp,
		char *	postfix,
		gridinfo3d_t *	grid3d
	)

◆ sCreate_CompCol_Matrix_dist()

void sCreate_CompCol_Matrix_dist	(	SuperMatrix *	,
		int_t	,
		int_t	,
		int_t	,
		float *	,
		int_t *	,
		int_t *	,
		Stype_t	,
		Dtype_t	,
		Mtype_t
	)

Here is the caller graph for this function:

◆ sCreate_CompRowLoc_Matrix_dist()

void sCreate_CompRowLoc_Matrix_dist	(	SuperMatrix *	,
		int_t	,
		int_t	,
		int_t	,
		int_t	,
		int_t	,
		float *	,
		int_t *	,
		int_t *	,
		Stype_t	,
		Dtype_t	,
		Mtype_t
	)

Here is the caller graph for this function:

◆ sCreate_Dense_Matrix_dist()

void sCreate_Dense_Matrix_dist	(	SuperMatrix *	,
		int_t	,
		int_t	,
		float *	,
		int_t	,
		Stype_t	,
		Dtype_t	,
		Mtype_t
	)

◆ screate_matrix()

int screate_matrix	(	SuperMatrix *	A,
		int	nrhs,
		float **	rhs,
		int *	ldb,
		float **	x,
		int *	ldx,
		FILE *	fp,
		gridinfo_t *	grid
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ screate_matrix3d()

int screate_matrix3d	(	SuperMatrix *	A,
		int	nrhs,
		float **	rhs,
		int *	ldb,
		float **	x,
		int *	ldx,
		FILE *	fp,
		gridinfo3d_t *	grid3d
	)

◆ screate_matrix_dat()

int screate_matrix_dat	(	SuperMatrix *	,
		int	,
		float **	,
		int *	,
		float **	,
		int *	,
		FILE *	,
		gridinfo_t *
	)

◆ screate_matrix_postfix()

int screate_matrix_postfix	(	SuperMatrix *	A,
		int	nrhs,
		float **	rhs,
		int *	ldb,
		float **	x,
		int *	ldx,
		FILE *	fp,
		char *	postfix,
		gridinfo_t *	grid
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ screate_matrix_postfix3d()

int screate_matrix_postfix3d	(	SuperMatrix *	A,
		int	nrhs,
		float **	rhs,
		int *	ldb,
		float **	x,
		int *	ldx,
		FILE *	fp,
		char *	postfix,
		gridinfo3d_t *	grid3d
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ screate_matrix_rb()

int screate_matrix_rb	(	SuperMatrix *	,
		int	,
		float **	,
		int *	,
		float **	,
		int *	,
		FILE *	,
		gridinfo_t *
	)

◆ sCreate_SuperNode_Matrix_dist()

void sCreate_SuperNode_Matrix_dist	(	SuperMatrix *	,
		int_t	,
		int_t	,
		int_t	,
		float *	,
		int_t *	,
		int_t *	,
		int_t *	,
		int_t *	,
		int_t *	,
		Stype_t	,
		Dtype_t	,
		Mtype_t
	)

◆ scuStatUpdate()

int_t scuStatUpdate	(	int_t	knsupc,
		HyP_t *	HyP,
		SCT_t *	SCT,
		SuperLUStat_t *	stat
	)

◆ sDeAllocGlu_3d()

int sDeAllocGlu_3d ( sLUstruct_t * LUstruct )

◆ sDeAllocLlu_3d()

int sDeAllocLlu_3d	(	int_t	n,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ sdelete_multiGPU_buffers()

void sdelete_multiGPU_buffers ( )

Here is the caller graph for this function:

◆ sdenseTreeFactor()

int_t sdenseTreeFactor	(	int_t	nnnodes,
		int_t *	perm_c_supno,
		commRequests_t *	comReqs,
		sscuBufs_t *	scuBufs,
		packLUInfo_t *	packLUInfo,
		msgs_t *	msgs,
		sLUValSubBuf_t *	LUvsb,
		sdiagFactBufs_t *	dFBuf,
		factStat_t *	factStat,
		factNodelists_t *	fNlists,
		superlu_dist_options_t *	options,
		int_t *	gIperm_c_supno,
		int_t	ldt,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SuperLUStat_t *	stat,
		double	thresh,
		SCT_t *	SCT,
		int	tag_ub,
		int *	info
	)

◆ sDestroy_A3d_gathered_on_2d()

void sDestroy_A3d_gathered_on_2d	(	sSOLVEstruct_t *	SOLVEstruct,
		gridinfo3d_t *	grid3d
	)

Here is the caller graph for this function:

◆ sDestroy_LU()

void sDestroy_LU	(	int_t	n,
		gridinfo_t *	grid,
		sLUstruct_t *	LUstruct
	)

Destroy distributed L & U matrices.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sDestroy_Tree()

void sDestroy_Tree	(	int_t	n,
		gridinfo_t *	grid,
		sLUstruct_t *	LUstruct
	)

Destroy broadcast and reduction trees used in triangular solve.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sDestroy_trf3Dpartition()

void sDestroy_trf3Dpartition ( strf3Dpartition_t * trf3Dpartition )

Here is the caller graph for this function:

◆ sDiagFactIBCast()

int_t sDiagFactIBCast	(	int_t	k,
		int_t	k0,
		float *	BlockUFactor,
		float *	BlockLFactor,
		int *	IrecvPlcd_D,
		MPI_Request *	,
		MPI_Request *	,
		MPI_Request *	,
		MPI_Request *	,
		gridinfo_t *	,
		superlu_dist_options_t *	,
		double	thresh,
		sLUstruct_t *	LUstruct,
		SuperLUStat_t *	,
		int *	info,
		SCT_t *	,
		int	tag_ub
	)

◆ sdist_psymbtonum()

float sdist_psymbtonum	(	superlu_dist_options_t *	options,
		int_t	n,
		SuperMatrix *	A,
		sScalePermstruct_t *	ScalePermstruct,
		Pslu_freeable_t *	Pslu_freeable,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid
	)

Purpose
=======
  Distribute the input matrix onto the 2D process mesh.

Arguments
=========

options (input) superlu_dist_options_t*
        The structure defines the input parameters to control
        how the LU decomposition and triangular solve are performed.
        options->Fact specifies whether or not the L and U structures
        will be re-used:
          = SamePattern_SameRowPerm: L and U structures are input, and
                                     unchanged on exit.
            This routine should not be called for this case, an error
            is generated.  Instead, pddistribute routine should be called.
          = DOFACT or SamePattern: L and U structures are computed and output.

n      (Input) int
       Dimension of the matrix.

A      (Input) SuperMatrix*
   The distributed input matrix A of dimension (A->nrow, A->ncol).
       A may be overwritten by diag(R)*A*diag(C)*Pc^T.
       The type of A can be: Stype = NR; Dtype = SLU_D; Mtype = GE.

ScalePermstruct (Input) sScalePermstruct_t*
       The data structure to store the scaling and permutation vectors
       describing the transformations performed to the original matrix A.

Glu_freeable (Input) *Glu_freeable_t
       The global structure describing the graph of L and U.

LUstruct (Input) sLUstruct_t*
       Data structures for L and U factors.

grid   (Input) gridinfo_t*
       The 2D process mesh.

Return value
============
  < 0, number of bytes allocated on return from the dist_symbLU
  > 0, number of bytes allocated for performing the distribution
      of the data, when out of memory.
       (an approximation).

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sdistribute()

float sdistribute	(	superlu_dist_options_t *	options,
		int_t	n,
		SuperMatrix *	A,
		Glu_freeable_t *	Glu_freeable,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid
	)

Purpose
=======
  Distribute the matrix onto the 2D process mesh.

Arguments
=========

options (input) superlu_dist_options_t *
       options->Fact specifies whether or not the L and U structures will be re-used.
       = SamePattern_SameRowPerm: L and U structures are input, and
                                  unchanged on exit.
       = DOFACT or SamePattern: L and U structures are computed and output.

n      (input) int
       Dimension of the matrix.

A      (input) SuperMatrix*
   The original matrix A, permuted by columns, of dimension
       (A->nrow, A->ncol). The type of A can be:
       Stype = SLU_NCP; Dtype = SLU_S; Mtype = SLU_GE.

LUstruct (input) sLUstruct_t*
       Data structures for L and U factors.

grid   (input) gridinfo_t*
       The 2D process mesh.

Return value
============
  > 0, working storage (in bytes) required to perform redistribution.
       (excluding LU factor size)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sequil_batch()

int sequil_batch	(	superlu_dist_options_t *	options,
		int	batchCount,
		int	m,
		int	n,
		handle_t *	SparseMatrix_handles,
		float **	ReqPtr,
		float **	CeqPtr,
		DiagScale_t *	DiagScale
	)

Equilibrate the systems using the LAPACK-style algorithm.

Parameters

[in]	options	solver options
[in]	batchCount	number of matrices in the batch
[in]	m	row dimension of the matrices
[in]	n	column dimension of the matrices
[in,out]	SparseMatrix_handles	pointers to the matrices in the batch, each pointing to the actual stoage in CSC format On entry, the original matrices On exit, each matrix may be overwritten by diag(R)Adiag(C)
[out]	ReqPtr	pointers to row scaling vectors (allocated internally)
[out]	CeqPtr	pointers to column scaling vectors (allocated internally)
[in,out]	DiagScale	arrays indicating how each system is equilibrated: {ROW, COL, BOTH}

Return value i: = 0: successful exit > 0: indicates the first matrix in the batch has zero row or column if i <= m: the i-th row of A is exactly zero if i > m: the (i-m)-th column of A is exactly zero

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sfill_dist()

void sfill_dist	(	float *	a,
		int_t	alen,
		float	dval
	)

Fills a float precision array with a given value.

◆ sFillRHS_dist()

void sFillRHS_dist	(	char *	trans,
		int_t	nrhs,
		float *	x,
		int_t	ldx,
		SuperMatrix *	A,
		float *	rhs,
		int_t	ldb
	)

Let rhs[i] = sum of i-th row of A, so the solution vector is all 1's.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sfreeDiagFactBufsArr()

int sfreeDiagFactBufsArr	(	int	mxLeafNode,
		sdiagFactBufs_t **	dFBufs
	)

Here is the caller graph for this function:

◆ sfreeScuBufs()

int sfreeScuBufs ( sscuBufs_t * scuBufs )

Here is the caller graph for this function:

◆ sfsolveReduceLsum3d()

int_t sfsolveReduceLsum3d	(	int_t	treeId,
		int_t	sender,
		int_t	receiver,
		float *	lsum,
		float *	recvbuf,
		int	nrhs,
		strf3Dpartition_t *	trf3Dpartition,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		xtrsTimer_t *	xtrsTimer
	)

Here is the caller graph for this function:

◆ sgather_l()

void sgather_l	(	int_t	num_LBlk,
		int_t	knsupc,
		Remain_info_t *	L_info,
		float *	lval,
		int_t	LD_lval,
		float *	L_buff
	)

◆ sgather_u()

void sgather_u	(	int_t	num_u_blks,
		Ublock_info_t *	Ublock_info,
		int_t *	usub,
		float *	uval,
		float *	bigU,
		int_t	ldu,
		int_t *	xsup,
		int_t	klst
	)

◆ sgatherAllFactoredLU()

int_t sgatherAllFactoredLU	(	strf3Dpartition_t *	trf3Dpartition,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

Here is the caller graph for this function:

◆ sgatherAllFactoredLUFr()

int_t sgatherAllFactoredLUFr	(	int_t *	myZeroTrIdxs,
		sForest_t *	sForests,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

◆ sgatherFactoredLU()

int_t sgatherFactoredLU	(	int_t	sender,
		int_t	receiver,
		int_t	nnodes,
		int_t *	nodeList,
		sLUValSubBuf_t *	LUvsb,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

◆ sGatherNRformat_loc3d()

void sGatherNRformat_loc3d	(	fact_t	Fact,
		NRformat_loc *	A,
		float *	B,
		int	ldb,
		int	nrhs,
		gridinfo3d_t *	grid3d,
		NRformat_loc3d **
	)

◆ sGatherNRformat_loc3d_allgrid()

void sGatherNRformat_loc3d_allgrid	(	fact_t	Fact,
		NRformat_loc *	A,
		float *	B,
		int	ldb,
		int	nrhs,
		gridinfo3d_t *	grid3d,
		NRformat_loc3d **
	)

Here is the caller graph for this function:

◆ sgemm_()

int sgemm_	(	const char *	,
		const char *	,
		const int *	,
		const int *	,
		const int *	,
		const float *	,
		const float *	,
		const int *	,
		const float *	,
		const int *	,
		const float *	,
		float *	,
		const int *
	)

Here is the caller graph for this function:

◆ sgemv_()

void sgemv_	(	const char *	,
		const int *	,
		const int *	,
		const float *	,
		const float *	a,
		const int *	,
		const float *	,
		const int *	,
		const float *	,
		float *	,
		const int *
	)

Here is the caller graph for this function:

◆ sGenCOOLblocks()

void sGenCOOLblocks	(	int	,
		int_t	,
		gridinfo_t *	,
		Glu_persist_t *	,
		sLocalLU_t *	,
		int_t **	,
		int_t **	,
		float **	,
		int_t *	,
		int_t *
	)

◆ sGenCSCLblocks()

void sGenCSCLblocks	(	int	,
		int_t	,
		gridinfo_t *	,
		Glu_persist_t *	,
		sLocalLU_t *	,
		float **	,
		int_t **	,
		int_t **	,
		int_t *	,
		int_t *
	)

◆ sGenCSRLblocks()

void sGenCSRLblocks	(	int	,
		int_t	,
		gridinfo_t *	,
		Glu_persist_t *	,
		sLocalLU_t *	,
		float **	,
		int_t **	,
		int_t **	,
		int_t *	,
		int_t *
	)

◆ sGenXtrue_dist()

void sGenXtrue_dist	(	int_t	n,
		int_t	nrhs,
		float *	x,
		int_t	ldx
	)

Here is the caller graph for this function:

◆ sger_()

void sger_	(	const int *	,
		const int *	,
		const float *	,
		const float *	,
		const int *	,
		const float *	,
		const int *	,
		float *	,
		const int *
	)

Here is the caller graph for this function:

◆ sgetBigU()

float * sgetBigU	(	superlu_dist_options_t *	,
		int_t	,
		gridinfo_t *	,
		sLUstruct_t *
	)

◆ sgetBigV()

float * sgetBigV	(	int_t	,
		int_t
	)

◆ sgsequ_dist()

void sgsequ_dist	(	SuperMatrix *	A,
		float *	r,
		float *	c,
		float *	rowcnd,
		float *	colcnd,
		float *	amax,
		int *	info
	)

Purpose
  =======

  SGSEQU_DIST computes row and column scalings intended to equilibrate an
  M-by-N sparse matrix A and reduce its condition number. R returns the row
  scale factors and C the column scale factors, chosen to try to make
  the largest element in each row and column of the matrix B with
  elements B(i,j)=R(i)*A(i,j)*C(j) have absolute value 1.

  R(i) and C(j) are restricted to be between SMLNUM = smallest safe
  number and BIGNUM = largest safe number.  Use of these scaling
  factors is not guaranteed to reduce the condition number of A but
  works well in practice.

  See supermatrix.h for the definition of 'SuperMatrix' structure.

  Arguments
  =========

  A       (input) SuperMatrix*
          The matrix of dimension (A->nrow, A->ncol) whose equilibration
          factors are to be computed. The type of A can be:
          Stype = SLU_NC; Dtype = SLU_S; Mtype = SLU_GE.

  R       (output) float*, size A->nrow
          If INFO = 0 or INFO > M, R contains the row scale factors
          for A.

  C       (output) float*, size A->ncol
          If INFO = 0,  C contains the column scale factors for A.

  ROWCND  (output) float*
          If INFO = 0 or INFO > M, ROWCND contains the ratio of the
          smallest R(i) to the largest R(i).  If ROWCND >= 0.1 and
          AMAX is neither too large nor too small, it is not worth
          scaling by R.

  COLCND  (output) float*
          If INFO = 0, COLCND contains the ratio of the smallest
          C(i) to the largest C(i).  If COLCND >= 0.1, it is not
          worth scaling by C.

  AMAX    (output) float*
          Absolute value of largest matrix element.  If AMAX is very
          close to overflow or very close to underflow, the matrix
          should be scaled.

  INFO    (output) int*
          = 0:  successful exit
          < 0:  if INFO = -i, the i-th argument had an illegal value
          > 0:  if INFO = i,  and i is
                <= A->nrow:  the i-th row of A is exactly zero
                >  A->ncol:  the (i-M)-th column of A is exactly zero

  =====================================================================

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sgstrf2()

void sgstrf2	(	int_t	k,
		float *	diagBlk,
		int_t	LDA,
		float *	BlockUfactor,
		int_t	LDU,
		double	thresh,
		int_t *	xsup,
		superlu_dist_options_t *	options,
		SuperLUStat_t *	stat,
		int *	info
	)

Here is the call graph for this function:

◆ sIBcast_LPanel()

int_t sIBcast_LPanel	(	int_t	k,
		int_t	k0,
		int_t *	lsub,
		float *	lusup,
		gridinfo_t *	,
		int *	msgcnt,
		MPI_Request *	,
		int **	ToSendR,
		int_t *	xsup,
		int
	)

◆ sIBcast_UPanel()

int_t sIBcast_UPanel	(	int_t	k,
		int_t	k0,
		int_t *	usub,
		float *	uval,
		gridinfo_t *	,
		int *	msgcnt,
		MPI_Request *	,
		int *	ToSendD,
		int
	)

◆ sIBcastRecvLPanel()

int_t sIBcastRecvLPanel	(	int_t	k,
		int_t	k0,
		int *	msgcnt,
		MPI_Request *	,
		MPI_Request *	,
		int_t *	Lsub_buf,
		float *	Lval_buf,
		int *	factored,
		gridinfo_t *	,
		sLUstruct_t *	,
		SCT_t *	,
		int	tag_ub
	)

◆ sIBcastRecvUPanel()

int_t sIBcastRecvUPanel	(	int_t	k,
		int_t	k0,
		int *	msgcnt,
		MPI_Request *	,
		MPI_Request *	,
		int_t *	Usub_buf,
		float *	Uval_buf,
		gridinfo_t *	,
		sLUstruct_t *	,
		SCT_t *	,
		int	tag_ub
	)

◆ siBcastXk2Pck()

int_t siBcastXk2Pck	(	int_t	k,
		float *	x,
		int	nrhs,
		int **	sendList,
		MPI_Request *	send_req,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		xtrsTimer_t *	xtrsTimer
	)

Here is the caller graph for this function:

◆ sinf_norm_error_dist()

void sinf_norm_error_dist	(	int_t	n,
		int_t	nrhs,
		float *	x,
		int_t	ldx,
		float *	xtrue,
		int_t	ldxtrue,
		gridinfo_t *	grid
	)

Check the inf-norm of the error vector.

Here is the caller graph for this function:

◆ sinit3DLUstruct()

int_t sinit3DLUstruct	(	int_t *	myTreeIdxs,
		int_t *	myZeroTrIdxs,
		int_t *	nodeCount,
		int_t **	nodeList,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ sinit3DLUstructForest()

void sinit3DLUstructForest	(	int_t *	myTreeIdxs,
		int_t *	myZeroTrIdxs,
		sForest_t **	sForests,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

Here is the caller graph for this function:

◆ sInit_HyP()

void sInit_HyP	(	superlu_dist_options_t *	,
		HyP_t *	HyP,
		sLocalLU_t *	Llu,
		int_t	mcb,
		int_t	mrb
	)

Here is the caller graph for this function:

◆ sinitDiagFactBufs()

int sinitDiagFactBufs	(	int	ldt,
		sdiagFactBufs_t *	dFBuf
	)

Here is the caller graph for this function:

◆ sinitDiagFactBufsArr()

sdiagFactBufs_t ** sinitDiagFactBufsArr	(	int	mxLeafNode,
		int	ldt,
		gridinfo_t *	grid
	)

◆ sinitDiagFactBufsArrMod()

sdiagFactBufs_t ** sinitDiagFactBufsArrMod	(	int	mxLeafNode,
		int *	ldts,
		gridinfo_t *	grid
	)

Here is the caller graph for this function:

◆ sinitLsumBmod_buff()

int_t sinitLsumBmod_buff	(	int_t	ns,
		int	nrhs,
		slsumBmod_buff_t *	lbmod_buf
	)

Here is the caller graph for this function:

◆ sinitScuBufs()

int_t sinitScuBufs	(	superlu_dist_options_t *	,
		int_t	ldt,
		int_t	num_threads,
		int_t	nsupers,
		sscuBufs_t *	,
		sLUstruct_t *	,
		gridinfo_t *
	)

Here is the caller graph for this function:

◆ sinitTrf3Dpartition()

strf3Dpartition_t * sinitTrf3Dpartition	(	int_t	nsupers,
		superlu_dist_options_t *	options,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ sinitTrf3Dpartition_allgrid()

strf3Dpartition_t * sinitTrf3Dpartition_allgrid	(	int_t	n,
		superlu_dist_options_t *	options,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ sinitTrf3DpartitionLUstructgrid0()

strf3Dpartition_t * sinitTrf3DpartitionLUstructgrid0	(	int_t	n,
		superlu_dist_options_t *	options,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ sIRecv_LDiagBlock()

int_t sIRecv_LDiagBlock	(	int_t	k0,
		float *	L_blk_ptr,
		int_t	size,
		int_t	src,
		MPI_Request *	,
		gridinfo_t *	,
		SCT_t *	,
		int
	)

◆ sIrecv_LPanel()

int_t sIrecv_LPanel	(	int_t	k,
		int_t	k0,
		int_t *	Lsub_buf,
		float *	Lval_buf,
		gridinfo_t *	,
		MPI_Request *	,
		sLocalLU_t *	,
		int
	)

◆ sIRecv_UDiagBlock()

int_t sIRecv_UDiagBlock	(	int_t	k0,
		float *	ublk_ptr,
		int_t	size,
		int_t	src,
		MPI_Request *	,
		gridinfo_t *	,
		SCT_t *	,
		int
	)

◆ sIrecv_UPanel()

int_t sIrecv_UPanel	(	int_t	k,
		int_t	k0,
		int_t *	Usub_buf,
		float *	,
		sLocalLU_t *	,
		gridinfo_t *	,
		MPI_Request *	,
		int
	)

◆ sISend_LDiagBlock()

int_t sISend_LDiagBlock	(	int_t	k0,
		float *	lblk_ptr,
		int_t	size,
		MPI_Request *	,
		gridinfo_t *	,
		int
	)

◆ sISend_UDiagBlock()

int_t sISend_UDiagBlock	(	int_t	k0,
		float *	ublk_ptr,
		int_t	size,
		MPI_Request *	,
		gridinfo_t *	,
		int
	)

Here is the caller graph for this function:

◆ slangs_dist()

float slangs_dist	(	char *	norm,
		SuperMatrix *	A
	)

Purpose
  =======

  SLANGS_DIST returns the value of the one norm, or the Frobenius norm, or
  the infinity norm, or the element of largest absolute value of a
  real matrix A.

  Description
  ===========

  SLANGE returns the value

     SLANGE = ( max(abs(A(i,j))), NORM = 'M' or 'm'
              (
              ( norm1(A),         NORM = '1', 'O' or 'o'
              (
              ( normI(A),         NORM = 'I' or 'i'
              (
              ( normF(A),         NORM = 'F', 'f', 'E' or 'e'

  where  norm1  denotes the  one norm of a matrix (maximum column sum),
  normI  denotes the  infinity norm  of a matrix  (maximum row sum) and
  normF  denotes the  Frobenius norm of a matrix (square root of sum of
  squares).  Note that  max(abs(A(i,j)))  is not a  matrix norm.

  Arguments
  =========

  NORM    (input) CHARACTER*1
          Specifies the value to be returned in SLANGE as described above.
  A       (input) SuperMatrix*
          The M by N sparse matrix A.

 =====================================================================

Here is the caller graph for this function:

◆ slaqgs_dist()

void slaqgs_dist	(	SuperMatrix *	A,
		float *	r,
		float *	c,
		float	rowcnd,
		float	colcnd,
		float	amax,
		char *	equed
	)

  Purpose
  =======

  SLAQGS_DIST equilibrates a general sparse M by N matrix A using the row and
  scaling factors in the vectors R and C.

  See supermatrix.h for the definition of 'SuperMatrix' structure.

  Arguments
  =========

  A       (input/output) SuperMatrix*
          On exit, the equilibrated matrix.  See EQUED for the form of
          the equilibrated matrix. The type of A can be:
     Stype = NC; Dtype = SLU_S; Mtype = GE.

  R       (input) float*, dimension (A->nrow)
          The row scale factors for A.

  C       (input) float*, dimension (A->ncol)
          The column scale factors for A.

  ROWCND  (input) float
          Ratio of the smallest R(i) to the largest R(i).

  COLCND  (input) float
          Ratio of the smallest C(i) to the largest C(i).

  AMAX    (input) float
          Absolute value of largest matrix entry.

  EQUED   (output) char*
          Specifies the form of equilibration that was done.
          = 'N':  No equilibration
          = 'R':  Row equilibration, i.e., A has been premultiplied by
                  diag(R).
          = 'C':  Column equilibration, i.e., A has been postmultiplied
                  by diag(C).
          = 'B':  Both row and column equilibration, i.e., A has been
                  replaced by diag(R) * A * diag(C).

  Internal Parameters
  ===================

  THRESH is a threshold value used to decide if row or column scaling
  should be done based on the ratio of the row or column scaling
  factors.  If ROWCND < THRESH, row scaling is done, and if
  COLCND < THRESH, column scaling is done.

  LARGE and SMALL are threshold values used to decide if row scaling
  should be done based on the absolute size of the largest matrix
  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.

  =====================================================================

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slasum_bmod_Tree()

int_t slasum_bmod_Tree	(	int_t	pTree,
		int_t	cTree,
		float *	lsum,
		float *	x,
		sxT_struct *	xT_s,
		int	nrhs,
		slsumBmod_buff_t *	lbmod_buf,
		sLUstruct_t *	LUstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		SuperLUStat_t *	stat
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sldperm_dist()

int sldperm_dist	(	int	job,
		int	n,
		int_t	nnz,
		int_t	colptr[],
		int_t	adjncy[],
		float	nzval[],
		int_t *	perm,
		float	u[],
		float	v[]
	)

Purpose
=======

  SLDPERM finds a row permutation so that the matrix has large
  entries on the diagonal.

Arguments
=========

job    (input) int
       Control the action. Possible values for JOB are:
       = 1 : Compute a row permutation of the matrix so that the
             permuted matrix has as many entries on its diagonal as
             possible. The values on the diagonal are of arbitrary size.
             HSL subroutine MC21A/AD is used for this.
       = 2 : Compute a row permutation of the matrix so that the smallest
             value on the diagonal of the permuted matrix is maximized.
       = 3 : Compute a row permutation of the matrix so that the smallest
             value on the diagonal of the permuted matrix is maximized.
             The algorithm differs from the one used for JOB = 2 and may
             have quite a different performance.
       = 4 : Compute a row permutation of the matrix so that the sum
             of the diagonal entries of the permuted matrix is maximized.
       = 5 : Compute a row permutation of the matrix so that the product
             of the diagonal entries of the permuted matrix is maximized
             and vectors to scale the matrix so that the nonzero diagonal
             entries of the permuted matrix are one in absolute value and
             all the off-diagonal entries are less than or equal to one in
             absolute value.
       Restriction: 1 <= JOB <= 5.

n      (input) int
       The order of the matrix.

nnz    (input) int
       The number of nonzeros in the matrix.

adjncy (input) int*, of size nnz
       The adjacency structure of the matrix, which contains the row
       indices of the nonzeros.

colptr (input) int*, of size n+1
       The pointers to the beginning of each column in ADJNCY.

nzval  (input) float*, of size nnz
       The nonzero values of the matrix. nzval[k] is the value of
       the entry corresponding to adjncy[k].
       It is not used if job = 1.

perm   (output) int*, of size n
       The permutation vector. perm[i] = j means row i in the
       original matrix is in row j of the permuted matrix.

u      (output) float*, of size n
       If job = 5, the natural logarithms of the row scaling factors.

v      (output) float*, of size n
       If job = 5, the natural logarithms of the column scaling factors.
       The scaled matrix B has entries b_ij = a_ij * exp(u_i + v_j).

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sleafForestBackSolve3d()

int_t sleafForestBackSolve3d	(	superlu_dist_options_t *	options,
		int_t	treeId,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	x,
		float *	lsum,
		float *	recvbuf,
		MPI_Request *	send_req,
		int	nrhs,
		slsumBmod_buff_t *	lbmod_buf,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sleafForestForwardSolve3d()

int_t sleafForestForwardSolve3d	(	superlu_dist_options_t *	options,
		int_t	treeId,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	x,
		float *	lsum,
		float *	recvbuf,
		float *	rtemp,
		MPI_Request *	send_req,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sLluBufFreeArr()

int sLluBufFreeArr	(	int_t	numLA,
		sLUValSubBuf_t **	LUvsbs
	)

Here is the caller graph for this function:

◆ sLluBufInit()

int_t sLluBufInit	(	sLUValSubBuf_t *	,
		sLUstruct_t *
	)

Here is the caller graph for this function:

◆ sLluBufInitArr()

sLUValSubBuf_t ** sLluBufInitArr	(	int_t	numLA,
		sLUstruct_t *	LUstruct
	)

Here is the caller graph for this function:

◆ slocalSolveXkYk()

int_t slocalSolveXkYk	(	trtype_t	trtype,
		int_t	k,
		float *	x,
		int	nrhs,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		SuperLUStat_t *	stat
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sLPanelTrSolve()

int_t sLPanelTrSolve	(	int_t	k,
		int *	factored_L,
		float *	BlockUFactor,
		gridinfo_t *	,
		sLUstruct_t *
	)

◆ sLPanelUpdate()

int_t sLPanelUpdate	(	int_t	k,
		int *	IrecvPlcd_D,
		int *	factored_L,
		MPI_Request *	,
		float *	BlockUFactor,
		gridinfo_t *	,
		sLUstruct_t *	,
		SCT_t *
	)

◆ sLpanelUpdate()

int_t sLpanelUpdate	(	int_t	off0,
		int_t	nsupc,
		float *	ublk_ptr,
		int_t	ld_ujrow,
		float *	lusup,
		int_t	nsupr,
		SCT_t *
	)

◆ slsum_bmod()

void slsum_bmod	(	float *	lsum,
		float *	x,
		float *	xk,
		int	nrhs,
		int_t	k,
		int *	bmod,
		int_t *	Urbs,
		Ucb_indptr_t **	Ucb_indptr,
		int_t **	Ucb_valptr,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		MPI_Request	send_req[],
		SuperLUStat_t *	stat
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_bmod_GG()

void slsum_bmod_GG	(	float *	lsum,
		float *	x,
		float *	xk,
		int	nrhs,
		slsumBmod_buff_t *	lbmod_buf,
		int_t	k,
		int *	bmod,
		int_t *	Urbs,
		Ucb_indptr_t **	Ucb_indptr,
		int_t **	Ucb_valptr,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		MPI_Request	send_req[],
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_bmod_GG_newsolve()

void slsum_bmod_GG_newsolve	(	strf3Dpartition_t *	trf3Dpartition,
		float *	lsum,
		float *	x,
		float *	xk,
		int	nrhs,
		slsumBmod_buff_t *	lbmod_buf,
		int_t	k,
		int *	bmod,
		int_t *	Urbs,
		Ucb_indptr_t **	Ucb_indptr,
		int_t **	Ucb_valptr,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		MPI_Request	send_req[],
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_bmod_inv()

void slsum_bmod_inv	(	float *	lsum,
		float *	x,
		float *	xk,
		float *	rtemp,
		int	nrhs,
		int_t	k,
		int *	bmod,
		int_t *	Urbs,
		Ucb_indptr_t **	Ucb_indptr,
		int_t **	Ucb_valptr,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		SuperLUStat_t **	stat,
		int_t *	root_send,
		int_t *	nroot_send,
		int_t	sizelsum,
		int_t	sizertemp,
		int	thread_id,
		int	num_thread
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_bmod_inv_gpu_wrap()

void slsum_bmod_inv_gpu_wrap	(	superlu_dist_options_t *	,
		int	,
		int	,
		int	,
		int	,
		float *	,
		float *	,
		int	,
		int	,
		int_t	,
		int *	,
		C_Tree *	,
		C_Tree *	,
		int_t *	,
		int_t *	,
		int64_t *	,
		int_t *	,
		int64_t *	,
		int_t *	,
		int64_t *	,
		float *	,
		int64_t *	,
		float *	,
		int64_t *	,
		float *	,
		int64_t *	,
		int_t *	,
		int64_t *	,
		int_t *	,
		gridinfo_t *	,
		int_t	,
		uint64_t *	,
		uint64_t *	,
		float *	,
		float *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int
	)

Here is the caller graph for this function:

◆ slsum_bmod_inv_master()

void slsum_bmod_inv_master	(	float *	lsum,
		float *	x,
		float *	xk,
		float *	rtemp,
		int	nrhs,
		int_t	k,
		int *	bmod,
		int_t *	Urbs,
		Ucb_indptr_t **	Ucb_indptr,
		int_t **	Ucb_valptr,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		SuperLUStat_t **	stat,
		int_t	sizelsum,
		int_t	sizertemp,
		int	thread_id,
		int	num_thread
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_fmod()

void slsum_fmod	(	float *	lsum,
		float *	x,
		float *	xk,
		float *	rtemp,
		int	nrhs,
		int	knsupc,
		int_t	k,
		int *	fmod,
		int_t	nlb,
		int_t	lptr,
		int_t	luptr,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		MPI_Request	send_req[],
		SuperLUStat_t *	stat
	)

Purpose
=======
  Perform local block modifications: lsum[i] -= L_i,k * X[k].

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_fmod_inv()

void slsum_fmod_inv	(	float *	lsum,
		float *	x,
		float *	xk,
		float *	rtemp,
		int	nrhs,
		int_t	k,
		int *	fmod,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		SuperLUStat_t **	stat,
		int_t *	leaf_send,
		int_t *	nleaf_send,
		int_t	sizelsum,
		int_t	sizertemp,
		int_t	recurlevel,
		int_t	maxsuper,
		int	thread_id,
		int	num_thread
	)

Purpose
=======
  Perform local block modifications: lsum[i] -= L_i,k * X[k].

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_fmod_inv_gpu_wrap()

void slsum_fmod_inv_gpu_wrap	(	int	,
		int	,
		int	,
		int	,
		float *	,
		float *	,
		int	,
		int	,
		int_t	,
		int *	fmod,
		C_Tree *	,
		C_Tree *	,
		int_t *	,
		int_t *	,
		int64_t *	,
		float *	,
		int64_t *	,
		float *	,
		int64_t *	,
		int_t *	,
		int64_t *	,
		int_t *	,
		int *	,
		gridinfo_t *	,
		int_t	,
		uint64_t *	,
		uint64_t *	,
		float *	,
		float *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int *	,
		int
	)

Here is the caller graph for this function:

◆ slsum_fmod_inv_master()

void slsum_fmod_inv_master	(	float *	lsum,
		float *	x,
		float *	xk,
		float *	rtemp,
		int	nrhs,
		int	knsupc,
		int_t	k,
		int *	fmod,
		int_t	nlb,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		SuperLUStat_t **	stat,
		int_t	sizelsum,
		int_t	sizertemp,
		int_t	recurlevel,
		int_t	maxsuper,
		int	thread_id,
		int	num_thread
	)

Purpose
=======
  Perform local block modifications: lsum[i] -= L_i,k * X[k].

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_fmod_leaf()

void slsum_fmod_leaf	(	int_t	treeId,
		strf3Dpartition_t *	trf3Dpartition,
		float *	lsum,
		float *	x,
		float *	xk,
		float *	rtemp,
		int	nrhs,
		int	knsupc,
		int_t	k,
		int *	fmod,
		int_t	nlb,
		int_t	lptr,
		int_t	luptr,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		MPI_Request	send_req[],
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsum_fmod_leaf_newsolve()

void slsum_fmod_leaf_newsolve	(	strf3Dpartition_t *	trf3Dpartition,
		float *	lsum,
		float *	x,
		float *	xk,
		float *	rtemp,
		int	nrhs,
		int	knsupc,
		int_t	k,
		int *	fmod,
		int_t	nlb,
		int_t	lptr,
		int_t	luptr,
		int_t *	xsup,
		gridinfo_t *	grid,
		sLocalLU_t *	Llu,
		MPI_Request	send_req[],
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsumForestBsolve()

int_t slsumForestBsolve	(	int_t	k,
		int_t	treeId,
		float *	lsum,
		float *	x,
		sxT_struct *	xT_s,
		int	nrhs,
		slsumBmod_buff_t *	lbmod_buf,
		sLUstruct_t *	LUstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		SuperLUStat_t *	stat
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ slsumReducePrK()

int_t slsumReducePrK	(	int_t	k,
		float *	x,
		float *	lsum,
		float *	recvbuf,
		int	nrhs,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		xtrsTimer_t *	xtrsTimer
	)

Here is the caller graph for this function:

◆ sLUstructFree()

void sLUstructFree ( sLUstruct_t * LUstruct )

Deallocate LUstruct.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sLUstructInit()

void sLUstructInit	(	const	int_t,
		sLUstruct_t *	LUstruct
	)

Allocate storage in LUstruct.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sMaxAbsLij()

float sMaxAbsLij	(	int	iam,
		int	n,
		Glu_persist_t *	Glu_persist,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid
	)

Find max(abs(L(i,j)))

Here is the caller graph for this function:

◆ sMaxAbsUij()

float sMaxAbsUij	(	int	iam,
		int	n,
		Glu_persist_t *	Glu_persist,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid
	)

Find max(abs(U(i,j)))

Here is the caller graph for this function:

◆ snewTrfPartitionInit()

void snewTrfPartitionInit	(	int_t	nsupers,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ snonLeafForestBackSolve3d()

int_t snonLeafForestBackSolve3d	(	int_t	treeId,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	x,
		float *	lsum,
		sxT_struct *	xT_s,
		float *	recvbuf,
		MPI_Request *	send_req,
		int	nrhs,
		slsumBmod_buff_t *	lbmod_buf,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Pkk(Yk) = sumOver_PrK (Yk)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ snonLeafForestForwardSolve3d()

int_t snonLeafForestForwardSolve3d	(	int_t	treeId,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		strf3Dpartition_t *	trf3Dpartition,
		gridinfo3d_t *	grid3d,
		float *	x,
		float *	lsum,
		sxT_struct *	xT_s,
		float *	recvbuf,
		float *	rtemp,
		MPI_Request *	send_req,
		int	nrhs,
		sSOLVEstruct_t *	SOLVEstruct,
		SuperLUStat_t *	stat,
		xtrsTimer_t *	xtrsTimer
	)

Pkk(Yk) = sumOver_PrK (Yk)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sp2pSolvedX3d()

int_t sp2pSolvedX3d	(	int_t	treeId,
		int_t	sender,
		int_t	receiver,
		float *	x,
		int	nrhs,
		strf3Dpartition_t *	trf3Dpartition,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		xtrsTimer_t *	xtrsTimer
	)

Here is the caller graph for this function:

◆ sp3dCollect()

int_t sp3dCollect	(	int_t	layer,
		int_t	n,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

◆ sp3dScatter()

int_t sp3dScatter	(	int_t	n,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		int *	supernodeMask
	)

◆ sp_sgemm_dist()

int sp_sgemm_dist	(	char *	transa,
		int	n,
		float	alpha,
		SuperMatrix *	A,
		float *	b,
		int	ldb,
		float	beta,
		float *	c,
		int	ldc
	)

  Purpose
    =======

    sp_sgemm_dist performs one of the matrix-matrix operations

       C := alpha*op( A )*op( B ) + beta*C,

    where  op( X ) is one of

       op( X ) = X   or   op( X ) = X'   or   op( X ) = conjg( X' ),

    alpha and beta are scalars, and A, B and C are matrices, with op( A )
    an m by k matrix,  op( B )  a  k by n matrix and  C an m by n matrix.


    Parameters
    ==========

    TRANSA - (input) char*
             On entry, TRANSA specifies the form of op( A ) to be used in
             the matrix multiplication as follows:
                TRANSA = 'N' or 'n',  op( A ) = A.
                TRANSA = 'T' or 't',  op( A ) = A'.
                TRANSA = 'C' or 'c',  op( A ) = conjg( A' ).
             Unchanged on exit.

    TRANSB - (input) char*
             On entry, TRANSB specifies the form of op( B ) to be used in
             the matrix multiplication as follows:
                TRANSB = 'N' or 'n',  op( B ) = B.
                TRANSB = 'T' or 't',  op( B ) = B'.
                TRANSB = 'C' or 'c',  op( B ) = conjg( B' ).
             Unchanged on exit.

    M      - (input) int
             On entry,  M  specifies  the number of rows of the matrix
         op( A ) and of the matrix C.  M must be at least zero.
         Unchanged on exit.

    N      - (input) int
             On entry,  N specifies the number of columns of the matrix
         op( B ) and the number of columns of the matrix C. N must be
         at least zero.
         Unchanged on exit.

    K      - (input) int
             On entry, K specifies the number of columns of the matrix
         op( A ) and the number of rows of the matrix op( B ). K must
         be at least  zero.
             Unchanged on exit.

    ALPHA  - (input) float
             On entry, ALPHA specifies the scalar alpha.

    A      - (input) SuperMatrix*
             Matrix A with a sparse format, of dimension (A->nrow, A->ncol).
             Currently, the type of A can be:
                 Stype = NC or NCP; Dtype = SLU_S; Mtype = GE.
             In the future, more general A can be handled.

    B      - float array of DIMENSION ( LDB, kb ), where kb is
             n when TRANSB = 'N' or 'n',  and is  k otherwise.
             Before entry with  TRANSB = 'N' or 'n',  the leading k by n
             part of the array B must contain the matrix B, otherwise
             the leading n by k part of the array B must contain the
             matrix B.
             Unchanged on exit.

    LDB    - (input) int
             On entry, LDB specifies the first dimension of B as declared
             in the calling (sub) program. LDB must be at least max( 1, n ).
             Unchanged on exit.

    BETA   - (input) float
             On entry, BETA specifies the scalar beta. When BETA is
             supplied as zero then C need not be set on input.

    C      - float array of DIMENSION ( LDC, n ).
             Before entry, the leading m by n part of the array C must
             contain the matrix C,  except when beta is zero, in which
             case C need not be set on entry.
             On exit, the array C is overwritten by the m by n matrix
         ( alpha*op( A )*B + beta*C ).

    LDC    - (input) int
             On entry, LDC specifies the first dimension of C as declared
             in the calling (sub)program. LDC must be at least max(1,m).
             Unchanged on exit.

    ==== Sparse Level 3 Blas routine.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sp_sgemv_dist()

int sp_sgemv_dist	(	char *	trans,
		float	alpha,
		SuperMatrix *	A,
		float *	x,
		int	incx,
		float	beta,
		float *	y,
		int	incy
	)

SpGEMV.

  Purpose
    =======

    sp_sgemv_dist()  performs one of the matrix-vector operations
       y := alpha*A*x + beta*y,   or   y := alpha*A'*x + beta*y,
    where alpha and beta are scalars, x and y are vectors and A is a
    sparse A->nrow by A->ncol matrix.

    Parameters
    ==========

    TRANS  - (input) char*
             On entry, TRANS specifies the operation to be performed as
             follows:
                TRANS = 'N' or 'n'   y := alpha*A*x + beta*y.
                TRANS = 'T' or 't'   y := alpha*A'*x + beta*y.
                TRANS = 'C' or 'c'   y := alpha*A'*x + beta*y.

    ALPHA  - (input) double
             On entry, ALPHA specifies the scalar alpha.

    A      - (input) SuperMatrix*
             Matrix A with a sparse format, of dimension (A->nrow, A->ncol).
             Currently, the type of A can be:
                 Stype = SLU_NC or SLU_NCP; Dtype = SLU_S; Mtype = SLU_GE.
             In the future, more general A can be handled.

    X      - (input) float*, array of DIMENSION at least
             ( 1 + ( n - 1 )*abs( INCX ) ) when TRANS = 'N' or 'n'
             and at least
             ( 1 + ( m - 1 )*abs( INCX ) ) otherwise.
             Before entry, the incremented array X must contain the
             vector x.

    INCX   - (input) int
             On entry, INCX specifies the increment for the elements of
             X. INCX must not be zero.

    BETA   - (input) float
             On entry, BETA specifies the scalar beta. When BETA is
             supplied as zero then Y need not be set on input.

    Y      - (output) float*,  array of DIMENSION at least
             ( 1 + ( m - 1 )*abs( INCY ) ) when TRANS = 'N' or 'n'
             and at least
             ( 1 + ( n - 1 )*abs( INCY ) ) otherwise.
             Before entry with BETA non-zero, the incremented array Y
             must contain the vector y. On exit, Y is overwritten by the
             updated vector y.

    INCY   - (input) int
             On entry, INCY specifies the increment for the elements of
             Y. INCY must not be zero.

    ==== Sparse Level 2 Blas routine.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sp_strsv_dist()

int sp_strsv_dist	(	char *	uplo,
		char *	trans,
		char *	diag,
		SuperMatrix *	L,
		SuperMatrix *	U,
		float *	x,
		int *	info
	)

  Purpose
  =======

  sp_strsv_dist() solves one of the systems of equations
      A*x = b,   or   A'*x = b,
  where b and x are n element vectors and A is a sparse unit , or
  non-unit, upper or lower triangular matrix.
  No test for singularity or near-singularity is included in this
  routine. Such tests must be performed before calling this routine.

  Parameters
  ==========

  uplo   - (input) char*
           On entry, uplo specifies whether the matrix is an upper or
            lower triangular matrix as follows:
               uplo = 'U' or 'u'   A is an upper triangular matrix.
               uplo = 'L' or 'l'   A is a lower triangular matrix.

  trans  - (input) char*
            On entry, trans specifies the equations to be solved as
            follows:
               trans = 'N' or 'n'   A*x = b.
               trans = 'T' or 't'   A'*x = b.
               trans = 'C' or 'c'   A'*x = b.

  diag   - (input) char*
            On entry, diag specifies whether or not A is unit
            triangular as follows:
               diag = 'U' or 'u'   A is assumed to be unit triangular.
               diag = 'N' or 'n'   A is not assumed to be unit
                                   triangular.

  L       - (input) SuperMatrix*
        The factor L from the factorization Pr*A*Pc=L*U. Use
            compressed row subscripts storage for supernodes, i.e.,
            L has types: Stype = SLU_SC, Dtype = SLU_S, Mtype = SLU_TRLU.

  U       - (input) SuperMatrix*
         The factor U from the factorization Pr*A*Pc=L*U.
         U has types: Stype = SLU_NC, Dtype = SLU_S, Mtype = SLU_TRU.

  x       - (input/output) float*
            Before entry, the incremented array X must contain the n
            element right-hand side vector b. On exit, X is overwritten
            with the solution vector x.

  info    - (output) int*
            If *info = -i, the i-th argument had an illegal value.

Here is the call graph for this function:

◆ sPackLBlock()

int_t sPackLBlock	(	int_t	k,
		float *	Dest,
		Glu_persist_t *	,
		gridinfo_t *	,
		sLocalLU_t *
	)

◆ sperform_row_permutation()

void sperform_row_permutation	(	superlu_dist_options_t *	options,
		fact_t	Fact,
		sScalePermstruct_t *	ScalePermstruct,
		sLUstruct_t *	LUstruct,
		int_t	m,
		int_t	n,
		gridinfo_t *	grid,
		SuperMatrix *	A,
		SuperMatrix *	GA,
		SuperLUStat_t *	stat,
		int	job,
		int	Equil,
		int *	rowequ,
		int *	colequ,
		int *	iinfo
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ spivot_batch()

int spivot_batch	(	superlu_dist_options_t *	options,
		int	batchCount,
		int	m,
		int	n,
		handle_t *	SparseMatrix_handles,
		float **	ReqPtr,
		float **	CeqPtr,
		DiagScale_t *	DiagScale,
		int **	RpivPtr
	)

Compute row pivotings for each matrix, for numerical stability.


Parameters

  
    [in] options solver options

    [in] batchCount number of matrices in the batch

    [in] m row dimension of the matrices

    [in] n column dimension of the matrices

    [in,out] SparseMatrix_handles pointers to the matrices in the batch, each pointing to the actual stoage in CSC format
    On entry, the original matrices, may be overwritten by A1 <- diag(R)*A*diag(C) from dequil_batch()
    On exit, each matrix may be A2 <- Pr*A1

    [in,out] ReqPtr pointers to row scaling vectors, maybe overwritten by scaling from MC64

    [in,out] CeqPtr pointers to column scaling vectors, maybe overwritten by scaling from MC64

    [in,out] DiagScale array indicating how each system is equilibrated: {ROW, COL, BOTH}

    [in,out] RpivPtr pointers to row permutation vectors for each matrix, each of size m
    On exit, each RpivPtr[] is applied to each matrix

Return value:
    0,  success
    -1, invalid RowPerm option; an Identity perm_r[] is returned
    d, indicates that the d-th matrix is the first one in the batch encountering error

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sprepare_multiGPU_buffers()

void sprepare_multiGPU_buffers	(	int	,
		int	,
		int	,
		int	,
		int	,
		int
	)

Here is the caller graph for this function:

◆ sPrint_CompCol_Matrix_dist()

void sPrint_CompCol_Matrix_dist ( SuperMatrix * )

Here is the caller graph for this function:

◆ sPrint_CompRowLoc_Matrix_dist()

int sPrint_CompRowLoc_Matrix_dist ( SuperMatrix * )

◆ sPrint_Dense_Matrix_dist()

void sPrint_Dense_Matrix_dist ( SuperMatrix * )

◆ sPrintLblocks()

void sPrintLblocks	(	int	iam,
		int_t	nsupers,
		gridinfo_t *	grid,
		Glu_persist_t *	Glu_persist,
		sLocalLU_t *	Llu
	)

Print the blocks in the factored matrix L.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sPrintUblocks()

void sPrintUblocks	(	int	iam,
		int_t	nsupers,
		gridinfo_t *	grid,
		Glu_persist_t *	Glu_persist,
		sLocalLU_t *	Llu
	)

Print the blocks in the factored matrix U.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sQuerySpace_dist()

int_t sQuerySpace_dist	(	int_t	n,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		SuperLUStat_t *	stat,
		superlu_dist_mem_usage_t *	mem_usage
	)

mem_usage consists of the following fields:
for_lu (float)
     The amount of space used in bytes for the L\U data structures.
total (float)
     The amount of space needed in bytes to perform factorization.
expansions (int)
     Number of memory expansions during the LU factorization.

Here is the caller graph for this function:

◆ sread_binary()

int sread_binary	(	FILE *	fp,
		int_t *	m,
		int_t *	n,
		int_t *	nnz,
		float **	nzval,
		int_t **	rowind,
		int_t **	colptr
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sreadhb_dist()

void sreadhb_dist	(	int	iam,
		FILE *	fp,
		int_t *	nrow,
		int_t *	ncol,
		int_t *	nonz,
		float **	nzval,
		int_t **	rowind,
		int_t **	colptr
	)

Purpose
=======

Read a FLOAT PRECISION matrix stored in Harwell-Boeing format
as described below.

Line 1 (A72,A8)
    Col. 1 - 72   Title (TITLE)
 Col. 73 - 80  Key (KEY)

Line 2 (5I14)
    Col. 1 - 14   Total number of lines excluding header (TOTCRD)
    Col. 15 - 28  Number of lines for pointers (PTRCRD)
    Col. 29 - 42  Number of lines for row (or variable) indices (INDCRD)
    Col. 43 - 56  Number of lines for numerical values (VALCRD)
 Col. 57 - 70  Number of lines for right-hand sides (RHSCRD)
                   (including starting guesses and solution vectors
           if present)
                  (zero indicates no right-hand side data is present)

Line 3 (A3, 11X, 4I14)
    Col. 1 - 3    Matrix type (see below) (MXTYPE)
    Col. 15 - 28  Number of rows (or variables) (NROW)
    Col. 29 - 42  Number of columns (or elements) (NCOL)
 Col. 43 - 56  Number of row (or variable) indices (NNZERO)
               (equal to number of entries for assembled matrices)
    Col. 57 - 70  Number of elemental matrix entries (NELTVL)
               (zero in the case of assembled matrices)
Line 4 (2A16, 2A20)
    Col. 1 - 16   Format for pointers (PTRFMT)
 Col. 17 - 32  Format for row (or variable) indices (INDFMT)
 Col. 33 - 52  Format for numerical values of coefficient matrix (VALFMT)
    Col. 53 - 72 Format for numerical values of right-hand sides (RHSFMT)

Line 5 (A3, 11X, 2I14) Only present if there are right-hand sides present
    Col. 1        Right-hand side type:
              F for full storage or M for same format as matrix
    Col. 2        G if a starting vector(s) (Guess) is supplied. (RHSTYP)
    Col. 3        X if an exact solution vector(s) is supplied.
 Col. 15 - 28  Number of right-hand sides (NRHS)
 Col. 29 - 42  Number of row indices (NRHSIX)
                  (ignored in case of unassembled matrices)

The three character type field on line 3 describes the matrix type.
The following table lists the permitted values for each of the three
characters. As an example of the type field, RSA denotes that the matrix
is real, symmetric, and assembled.

First Character:
 R Real matrix
 C Complex matrix
 P Pattern only (no numerical values supplied)

Second Character:
 S Symmetric
 U Unsymmetric
 H Hermitian
 Z Skew symmetric
 R Rectangular

Third Character:
 A Assembled
 E Elemental matrices (unassembled)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sreadMM_dist()

void sreadMM_dist	(	FILE *	fp,
		int_t *	m,
		int_t *	n,
		int_t *	nonz,
		float **	nzval,
		int_t **	rowind,
		int_t **	colptr
	)

brief

Output parameters
=================
  (nzval, rowind, colptr): (*rowind)[*] contains the row subscripts of
     nonzeros in columns of matrix A; (*nzval)[*] the numerical values;
 column i of A is given by (*nzval)[k], k = (*rowind)[i],...,
     (*rowind)[i+1]-1.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sreadrb_dist()

void sreadrb_dist	(	int	iam,
		FILE *	fp,
		int_t *	nrow,
		int_t *	ncol,
		int_t *	nonz,
		float **	nzval,
		int_t **	rowind,
		int_t **	colptr
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sreadtriple_dist()

void sreadtriple_dist	(	FILE *	fp,
		int_t *	m,
		int_t *	n,
		int_t *	nonz,
		float **	nzval,
		int_t **	rowind,
		int_t **	colptr
	)

brief

Output parameters
=================
  (nzval, rowind, colptr): (*rowind)[*] contains the row subscripts of
     nonzeros in columns of matrix A; (*nzval)[*] the numerical values;
 column i of A is given by (*nzval)[k], k = (*rowind)[i],...,
     (*rowind)[i+1]-1.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sreadtriple_noheader()

void sreadtriple_noheader	(	FILE *	fp,
		int_t *	m,
		int_t *	n,
		int_t *	nonz,
		float **	nzval,
		int_t **	rowind,
		int_t **	colptr
	)

brief

Output parameters
=================
  (nzval, rowind, colptr): (*rowind)[*] contains the row subscripts of
     nonzeros in columns of matrix A; (*nzval)[*] the numerical values;
 column i of A is given by (*nzval)[k], k = (*rowind)[i],...,
     (*rowind)[i+1]-1.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sRecv_UDiagBlock()

int_t sRecv_UDiagBlock	(	int_t	k0,
		float *	ublk_ptr,
		int_t	size,
		int_t	src,
		gridinfo_t *	,
		SCT_t *	,
		int
	)

Here is the caller graph for this function:

◆ sReDistribute_A()

int_t sReDistribute_A	(	SuperMatrix *	A,
		sScalePermstruct_t *	ScalePermstruct,
		Glu_freeable_t *	Glu_freeable,
		int_t *	xsup,
		int_t *	supno,
		gridinfo_t *	grid,
		int_t *	colptr[],
		int_t *	rowind[],
		float *	a[]
	)

Purpose
=======
  Re-distribute A on the 2D process mesh.

Arguments
=========

A      (input) SuperMatrix*
   The distributed input matrix A of dimension (A->nrow, A->ncol).
       A may be overwritten by diag(R)*A*diag(C)*Pc^T.
       The type of A can be: Stype = SLU_NR_loc; Dtype = SLU_S; Mtype = SLU_GE.

ScalePermstruct (input) sScalePermstruct_t*
       The data structure to store the scaling and permutation vectors
       describing the transformations performed to the original matrix A.

Glu_freeable (input) *Glu_freeable_t
       The global structure describing the graph of L and U.

grid   (input) gridinfo_t*
       The 2D process mesh.

colptr (output) int*

rowind (output) int*

a      (output) float*

Return value
============
  > 0, working storage (in bytes) required to perform redistribution.
       (excluding LU factor size)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sreduceAllAncestors3d()

int sreduceAllAncestors3d	(	int_t	ilvl,
		int_t *	myNodeCount,
		int_t **	treePerm,
		sLUValSubBuf_t *	LUvsb,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

Here is the caller graph for this function:

◆ sreduceAncestors3d()

int_t sreduceAncestors3d	(	int_t	sender,
		int_t	receiver,
		int_t	nnodes,
		int_t *	nodeList,
		float *	Lval_buf,
		float *	Uval_buf,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

◆ sreduceSolvedX_newsolve()

int_t sreduceSolvedX_newsolve	(	int_t	treeId,
		int_t	sender,
		int_t	receiver,
		float *	x,
		int	nrhs,
		strf3Dpartition_t *	trf3Dpartition,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		float *	recvbuf,
		xtrsTimer_t *	xtrsTimer
	)

Here is the caller graph for this function:

◆ sRgather_L()

void sRgather_L	(	int_t	k,
		int_t *	lsub,
		float *	lusup,
		gEtreeInfo_t *	,
		Glu_persist_t *	,
		gridinfo_t *	,
		HyP_t *	,
		int_t *	myIperm,
		int_t *	iperm_c_supno
	)

◆ sRgather_U()

void sRgather_U	(	int_t	k,
		int_t	jj0,
		int_t *	usub,
		float *	uval,
		float *	bigU,
		gEtreeInfo_t *	,
		Glu_persist_t *	,
		gridinfo_t *	,
		HyP_t *	,
		int_t *	myIperm,
		int_t *	iperm_c_supno,
		int_t *	perm_u
	)

◆ sscal_()

int sscal_	(	const int *	n,
		const float *	alpha,
		float *	dx,
		const int *	incx
	)

◆ sScaleAdd_CompRowLoc_Matrix_dist()

void sScaleAdd_CompRowLoc_Matrix_dist	(	SuperMatrix *	A,
		SuperMatrix *	B,
		float	c
	)

Scale and add: adds a scalar multiple of one matrix to another. A_{i,j} = c * A_{i,j} + B_{i,j}$ for i,j=1,...,n.

◆ sScaleAddId_CompRowLoc_Matrix_dist()

void sScaleAddId_CompRowLoc_Matrix_dist	(	SuperMatrix *	A,
		float	c
	)

Scale and add I: scales a matrix and adds an identity. A_{i,j} = c * A_{i,j} + \delta_{i,j} for i,j=1,...,n and \delta_{i,j} is the Kronecker delta.

Here is the call graph for this function:

◆ sscaleMatrixDiagonally()

void sscaleMatrixDiagonally	(	fact_t	Fact,
		sScalePermstruct_t *	ScalePermstruct,
		SuperMatrix *	A,
		SuperLUStat_t *	stat,
		gridinfo_t *	grid,
		int *	rowequ,
		int *	colequ,
		int *	iinfo
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sScalePermstructFree()

void sScalePermstructFree ( sScalePermstruct_t * ScalePermstruct )

Deallocate ScalePermstruct.

Here is the caller graph for this function:

◆ sScalePermstructInit()

void sScalePermstructInit	(	const	int_t,
		const	int_t,
		sScalePermstruct_t *	ScalePermstruct
	)

Allocate storage in ScalePermstruct.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sscatter3dLPanels()

int_t sscatter3dLPanels	(	int_t	nsupers,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		int *	supernodeMask
	)

◆ sscatter3dUPanels()

int_t sscatter3dUPanels	(	int_t	nsupers,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		int *	supernodeMask
	)

◆ sScatter_B3d()

int sScatter_B3d	(	NRformat_loc3d *	A3d,
		gridinfo3d_t *	grid3d
	)

Here is the caller graph for this function:

◆ sscatter_l()

void sscatter_l	(	int	ib,
		int	ljb,
		int	nsupc,
		int_t	iukp,
		int_t *	xsup,
		int	klst,
		int	nbrow,
		int_t	lptr,
		int	temp_nbrow,
		int_t *	usub,
		int_t *	lsub,
		float *	tempv,
		int *	indirect_thread,
		int *	indirect2,
		int_t **	Lrowind_bc_ptr,
		float **	Lnzval_bc_ptr,
		gridinfo_t *	grid
	)

◆ sscatter_u()

void sscatter_u	(	int	ib,
		int	jb,
		int	nsupc,
		int_t	iukp,
		int_t *	xsup,
		int	klst,
		int	nbrow,
		int_t	lptr,
		int	temp_nbrow,
		int_t *	lsub,
		int_t *	usub,
		float *	tempv,
		int_t **	Ufstnz_br_ptr,
		float **	Unzval_br_ptr,
		gridinfo_t *	grid
	)

◆ sSchurComplementSetup()

int_t sSchurComplementSetup	(	int_t	k,
		int *	msgcnt,
		Ublock_info_t *	,
		Remain_info_t *	,
		uPanelInfo_t *	,
		lPanelInfo_t *	,
		int_t *	,
		int_t *	,
		int_t *	,
		float *	bigU,
		int_t *	Lsub_buf,
		float *	Lval_buf,
		int_t *	Usub_buf,
		float *	Uval_buf,
		gridinfo_t *	,
		sLUstruct_t *
	)

◆ sSchurComplementSetupGPU()

int_t sSchurComplementSetupGPU	(	int_t	k,
		msgs_t *	msgs,
		packLUInfo_t *	,
		int_t *	,
		int_t *	,
		int_t *	,
		gEtreeInfo_t *	,
		factNodelists_t *	,
		sscuBufs_t *	,
		sLUValSubBuf_t *	LUvsb,
		gridinfo_t *	,
		sLUstruct_t *	,
		HyP_t *
	)

◆ sSolveFinalize()

void sSolveFinalize	(	superlu_dist_options_t *	options,
		sSOLVEstruct_t *	SOLVEstruct
	)

Release the resources used for the solution phase.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sSolveInit()

int sSolveInit	(	superlu_dist_options_t *	options,
		SuperMatrix *	A,
		int_t	perm_r[],
		int_t	perm_c[],
		int_t	nrhs,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		sSOLVEstruct_t *	SOLVEstruct
	)

Initialize the data structure for the solution phase.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ ssparseTreeFactor()

int_t ssparseTreeFactor	(	int_t	nnodes,
		int_t *	perm_c_supno,
		treeTopoInfo_t *	treeTopoInfo,
		commRequests_t *	comReqs,
		sscuBufs_t *	scuBufs,
		packLUInfo_t *	packLUInfo,
		msgs_t *	msgs,
		sLUValSubBuf_t *	LUvsb,
		sdiagFactBufs_t *	dFBuf,
		factStat_t *	factStat,
		factNodelists_t *	fNlists,
		superlu_dist_options_t *	options,
		int_t *	gIperm_c_supno,
		int_t	ldt,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SuperLUStat_t *	stat,
		double	thresh,
		SCT_t *	SCT,
		int *	info
	)

◆ ssparseTreeFactor_ASYNC()

int_t ssparseTreeFactor_ASYNC	(	sForest_t *	sforest,
		commRequests_t **	comReqss,
		sscuBufs_t *	scuBufs,
		packLUInfo_t *	packLUInfo,
		msgs_t **	msgss,
		sLUValSubBuf_t **	LUvsbs,
		sdiagFactBufs_t **	dFBufs,
		factStat_t *	factStat,
		factNodelists_t *	fNlists,
		gEtreeInfo_t *	gEtreeInfo,
		superlu_dist_options_t *	options,
		int_t *	gIperm_c_supno,
		int_t	ldt,
		HyP_t *	HyP,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SuperLUStat_t *	stat,
		double	thresh,
		SCT_t *	SCT,
		int	tag_ub,
		int *	info
	)

Here is the caller graph for this function:

◆ sstatic_schedule()

int sstatic_schedule	(	superlu_dist_options_t *	options,
		int	m,
		int	n,
		sLUstruct_t *	LUstruct,
		gridinfo_t *	grid,
		SuperLUStat_t *	stat,
		int_t *	perm_c_supno,
		int_t *	iperm_c_supno,
		int *	info
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sTrs2_GatherTrsmScatter()

int_t sTrs2_GatherTrsmScatter	(	int_t	klst,
		int_t	iukp,
		int_t	rukp,
		int_t *	usub,
		float *	uval,
		float *	tempv,
		int_t	knsupc,
		int	nsupr,
		float *	lusup,
		Glu_persist_t *	Glu_persist
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sTrs2_GatherU()

int_t sTrs2_GatherU	(	int_t	iukp,
		int_t	rukp,
		int_t	klst,
		int_t	nsupc,
		int_t	ldu,
		int_t *	usub,
		float *	uval,
		float *	tempv
	)

Here is the caller graph for this function:

◆ sTrs2_ScatterU()

int_t sTrs2_ScatterU	(	int_t	iukp,
		int_t	rukp,
		int_t	klst,
		int_t	nsupc,
		int_t	ldu,
		int_t *	usub,
		float *	uval,
		float *	tempv
	)

Here is the caller graph for this function:

◆ strs_B_init3d()

int_t strs_B_init3d	(	int_t	nsupers,
		float *	x,
		int	nrhs,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d
	)

Here is the caller graph for this function:

◆ strs_compute_communication_structure()

int strs_compute_communication_structure	(	superlu_dist_options_t *	options,
		int_t	n,
		sLUstruct_t *	LUstruct,
		sScalePermstruct_t *	ScalePermstruct,
		int *	supernodeMask,
		gridinfo_t *	grid,
		SuperLUStat_t *	stat
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ strs_X_gather3d()

int_t strs_X_gather3d	(	float *	x,
		int	nrhs,
		strf3Dpartition_t *	trf3Dpartition,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		xtrsTimer_t *	xtrsTimer
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ strsm_()

int strsm_	(	const char *	,
		const char *	,
		const char *	,
		const char *	,
		const int *	,
		const int *	,
		const float *	,
		const float *	,
		const int *	,
		float *	,
		const int *
	)

Here is the caller graph for this function:

◆ strsv_()

int strsv_	(	char *	,
		char *	,
		char *	,
		int *	,
		float *	,
		int *	,
		float *	,
		int *
	)

Here is the caller graph for this function:

◆ strtri_()

void strtri_	(	char *	,
		char *	,
		int *	,
		float *	,
		int *	,
		int *
	)

Here is the caller graph for this function:

◆ sUDiagBlockRecvWait()

int_t sUDiagBlockRecvWait	(	int_t	k,
		int *	IrecvPlcd_D,
		int *	factored_L,
		MPI_Request *	,
		gridinfo_t *	,
		sLUstruct_t *	,
		SCT_t *
	)

◆ sUPanelTrSolve()

int_t sUPanelTrSolve	(	int_t	k,
		float *	BlockLFactor,
		float *	bigV,
		int_t	ldt,
		Ublock_info_t *	,
		gridinfo_t *	,
		sLUstruct_t *	,
		SuperLUStat_t *	,
		SCT_t *
	)

◆ sUPanelUpdate()

int_t sUPanelUpdate	(	int_t	k,
		int *	factored_U,
		MPI_Request *	,
		float *	BlockLFactor,
		float *	bigV,
		int_t	ldt,
		Ublock_info_t *	,
		gridinfo_t *	,
		sLUstruct_t *	,
		SuperLUStat_t *	,
		SCT_t *
	)

◆ superlu_saxpy()

int superlu_saxpy	(	const int	n,
		const float	alpha,
		const float *	x,
		const int	incx,
		float *	y,
		const int	incy
	)

Here is the caller graph for this function:

◆ superlu_sgemm()

int superlu_sgemm	(	const char *	transa,
		const char *	transb,
		int	m,
		int	n,
		int	k,
		float	alpha,
		float *	a,
		int	lda,
		float *	b,
		int	ldb,
		float	beta,
		float *	c,
		int	ldc
	)

Here is the caller graph for this function:

◆ superlu_sgemv()

int superlu_sgemv	(	const char *	trans,
		const int	m,
		const int	n,
		const float	alpha,
		const float *	a,
		const int	lda,
		const float *	x,
		const int	incx,
		const float	beta,
		float *	y,
		const int	incy
	)

◆ superlu_sger()

int superlu_sger	(	const int	m,
		const int	n,
		const float	alpha,
		const float *	x,
		const int	incx,
		const float *	y,
		const int	incy,
		float *	a,
		const int	lda
	)

Here is the caller graph for this function:

◆ superlu_sscal()

int superlu_sscal	(	const int	n,
		const float	alpha,
		float *	x,
		const int	incx
	)

Here is the caller graph for this function:

◆ superlu_strsm()

int superlu_strsm	(	const char *	sideRL,
		const char *	uplo,
		const char *	transa,
		const char *	diag,
		const int	m,
		const int	n,
		const float	alpha,
		const float *	a,
		const int	lda,
		float *	b,
		const int	ldb
	)

Here is the caller graph for this function:

◆ superlu_strsv()

int superlu_strsv	(	char *	uplo,
		char *	trans,
		char *	diag,
		int	n,
		float *	a,
		int	lda,
		float *	x,
		int	incx
	)

◆ suser_free_dist()

void suser_free_dist	(	int_t	bytes,
		int_t	which_end
	)

◆ suser_malloc_dist()

void * suser_malloc_dist	(	int_t	bytes,
		int_t	which_end
	)

◆ sWait_LRecv()

int_t sWait_LRecv	(	MPI_Request *	,
		int *	msgcnt,
		int *	msgcntsU,
		gridinfo_t *	,
		SCT_t *
	)

◆ sWait_URecv()

int_t sWait_URecv	(	MPI_Request *	,
		int *	msgcnt,
		SCT_t *
	)

◆ sWaitL()

int_t sWaitL	(	int_t	k,
		int *	msgcnt,
		int *	msgcntU,
		MPI_Request *	,
		MPI_Request *	,
		gridinfo_t *	,
		sLUstruct_t *	,
		SCT_t *
	)

◆ sWaitU()

int_t sWaitU	(	int_t	k,
		int *	msgcnt,
		MPI_Request *	,
		MPI_Request *	,
		gridinfo_t *	,
		sLUstruct_t *	,
		SCT_t *
	)

◆ sZero_CompRowLoc_Matrix_dist()

void sZero_CompRowLoc_Matrix_dist ( SuperMatrix * A )

Sets all entries of a matrix to zero, A_{i,j}=0, for i,j=1,..,n.

Here is the call graph for this function:

◆ sZeroLblocks()

void sZeroLblocks	(	int	iam,
		int	n,
		gridinfo_t *	grid,
		sLUstruct_t *	LUstruct
	)

Sets all entries of matrix L to zero.

Here is the caller graph for this function:

◆ szeroSetLU()

int_t szeroSetLU	(	int_t	nnodes,
		int_t *	nodeList,
		sLUstruct_t *	,
		gridinfo3d_t *
	)

◆ sZeroUblocks()

void sZeroUblocks	(	int	iam,
		int	n,
		gridinfo_t *	grid,
		sLUstruct_t *	LUstruct
	)

Sets all entries of matrix U to zero.

Here is the caller graph for this function:

◆ szRecvLPanel()

int_t szRecvLPanel	(	int_t	k,
		int_t	sender,
		float	alpha,
		float	beta,
		float *	Lval_buf,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

◆ szRecvUPanel()

int_t szRecvUPanel	(	int_t	k,
		int_t	sender,
		float	alpha,
		float	beta,
		float *	Uval_buf,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

◆ szSendLPanel()

int_t szSendLPanel	(	int_t	k,
		int_t	receiver,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

◆ szSendUPanel()

int_t szSendUPanel	(	int_t	k,
		int_t	receiver,
		sLUstruct_t *	LUstruct,
		gridinfo3d_t *	grid3d,
		SCT_t *	SCT
	)

◆ updateDirtyBit()

int updateDirtyBit	(	int_t	k0,
		HyP_t *	HyP,
		gridinfo_t *	grid
	)

◆ validateInput_psgssvx3d()

void validateInput_psgssvx3d	(	superlu_dist_options_t *	options,
		SuperMatrix *	A,
		int	ldb,
		int	nrhs,
		gridinfo3d_t *	grid3d,
		int *	info
	)

Validates the input parameters for a given problem.

This function checks the input parameters for a given problem and sets the error code in the 'info' variable accordingly. If there is an error, it prints an error message and returns.

Parameters

[in]	options	Pointer to the options structure containing Fact, RowPerm, ColPerm, and IterRefine values.
[in]	A	Pointer to the matrix A structure containing nrow, ncol, Stype, Dtype, and Mtype values.
[in]	ldb	The leading dimension of the array B.
[in]	nrhs	The number of right-hand sides.
[in]	grid	Pointer to the grid structure.
[out]	info	Pointer to an integer variable that stores the error code.

Here is the call graph for this function:

Here is the caller graph for this function:

Variable Documentation

◆ sready_lsum

float* sready_lsum

extern

◆ sready_x

float* sready_x

extern

[out]	stat	records algorithms statistics such as runtime, memory usage, etc.
[out]	info	flags the errors on return

Classes

Macros

Typedefs

Functions

Variables

Detailed Description

Macro Definition Documentation

◆ MAX_LOOKAHEADS

Typedef Documentation

◆ slsumBmod_buff_t

◆ sxT_struct

Function Documentation

◆ checkRecvLDiag()

◆ checkRecvUDiag()

◆ file_Printfloat5()

◆ file_sPrint_CompRowLoc_Matrix_dist()

◆ floatCalloc_dist()

◆ floatMalloc_dist()

◆ Free_HyP()

◆ freePackLUInfo()

◆ getNsupers()

◆ initPackLUInfo()

◆ Local_Sgstrf2()

◆ nv_init_wrapper()

◆ Printfloat5()

◆ psCompRow_loc_to_CompCol_global()

◆ psCompute_Diag_Inv()

◆ psconvert_flatten_skyline2UROWDATA()

◆ psconvertU()

◆ psconvertUROWDATA2skyline()

◆ psdistribute()

◆ psdistribute3d_Yang()

◆ psdistribute_allgrid()

◆ psdistribute_allgrid_index_only()

◆ psflatten_LDATA()

◆ psGetDiagU()

◆ psgsequ()

◆ psgsmv()

◆ psgsmv_AXglobal()

◆ psgsmv_AXglobal_abs()

◆ psgsmv_AXglobal_setup()

◆ psgsmv_finalize()

◆ psgsmv_init()

◆ psgsrfs()

◆ psgsrfs3d()

◆ psgsrfs_ABXglobal()

◆ psgssvx()

◆ psgssvx3d()

◆ psgssvx3d_csc_batch()

◆ psgssvx_ABglobal()

◆ psgsTrBackSolve3d()

◆ psgsTrBackSolve3d_newsolve()

◆ psgstrf()

◆ psgstrf2()

◆ psgstrf2_trsm()

◆ psgstrf3d()

◆ psgsTrForwardSolve3d()

◆ psgsTrForwardSolve3d_newsolve()

◆ psgstrs()

◆ psgstrs2()

◆ psgstrs2_omp()

◆ psgstrs3d()

◆ psgstrs3d_newsolve()

◆ psgstrs_Bglobal()

◆ psgstrs_delete_device_lsum_x()

◆ psgstrs_init()

◆ psgstrs_init_device_lsum_x()

◆ psinf_norm_error()

◆ pslangs()

◆ pslaqgs()

◆ psPermute_Dense_Matrix()

◆ psReDistribute3d_B_to_X()

◆ psReDistribute3d_X_to_B()

◆ psReDistribute_B_to_X()

◆ pxgstrs_finalize()

◆ s3D_printMemUse()

◆ s_c2cpp_GetHWPM()

◆ sallocateA_dist()

◆ sAllocGlu_3d()

◆ sAllocLlu_3d()