core_chetrfsp.c File Reference

#include "common.h"
#include "cblas.h"
#include "blend/solver.h"
#include "pastix_ccores.h"
#include "kernels_trace.h"
#include <lapacke.h>

Go to the source code of this file.

Functions
static void	core_chetf2sp (pastix_int_t n, pastix_complex32_t *A, pastix_int_t lda, pastix_int_t *nbpivots, float criterion)
	Compute the sequential static pivoting factorization of the hermitian matrix n-by-n A such that A = L * D * conjf(L^t). More...
void	core_chetrfsp (pastix_int_t n, pastix_complex32_t *A, pastix_int_t lda, pastix_int_t *nbpivots, float criterion)
	Compute the block static pivoting factorization of the hermitian matrix n-by-n A such that A = L * D * conjf(L^t). More...
int	cpucblk_chetrfsp1d_hetrf (SolverMatrix *solvmtx, SolverCblk *cblk, void *dataL)
	Computes the LDL^h factorization of the diagonal block in a panel. More...
void	core_chetrfsp1d_gemm (const SolverCblk *cblk, const SolverBlok *blok, SolverCblk *fcblk, const pastix_complex32_t *L, pastix_complex32_t *C, pastix_complex32_t *work)
int	cpucblk_chetrfsp1d_panel (SolverMatrix *solvmtx, SolverCblk *cblk, void *L, void *DLh)
	Compute the LDL^h factorization of one panel. More...
int	cpucblk_chetrfsp1d (SolverMatrix *solvmtx, SolverCblk *cblk, pastix_complex32_t *DLh, pastix_complex32_t *work, pastix_int_t lwork)
	Perform the LDL^h factorization of a given panel and apply all its updates. More...
int	cpucblk_chetrfsp1dplus (SolverMatrix *solvmtx, SolverCblk *cblk)
	Perform the LDL^h factorization of a given panel and submit tasks for the subsequent updates. More...
void	cpucblk_chetrfsp1dplus_update (SolverMatrix *solvmtx, SolverBlok *blok, pastix_complex32_t *work)
	Apply the updates of the LDL^h factorisation of a given panel. More...

Detailed Description

PaStiX kernel routines for LDL^h factorization.

Copyright

2011-2023 Bordeaux INP, CNRS (LaBRI UMR 5800), Inria, Univ. Bordeaux. All rights reserved.

Version

6.3.2

Author

Mathieu Faverge

Pierre Ramet

Xavier Lacoste

Gregoire Pichon

Alycia Lisito

Nolan Bredel

Date

2023-12-11

Generated arithmetic file

from /builds/solverstack/pastix/kernels/core_zhetrfsp.c, normal z -> c, Wed Dec 13 12:09:11 2023

Definition in file core_chetrfsp.c.

Function Documentation

core_chetf2sp()

static void core_chetf2sp ( pastix_int_t  n, pastix_complex32_t *  A, pastix_int_t  lda, pastix_int_t *  nbpivots, float  criterion  )

inlinestatic

Compute the sequential static pivoting factorization of the hermitian matrix n-by-n A such that A = L * D * conjf(L^t).

Parameters

[in]	n	The number of rows and columns of the matrix A.
[in,out]	A	The matrix A to factorize with LDL^h factorization. The matrix is of size lda -by- n.
[in]	lda	The leading dimension of the matrix A.
[in,out]	nbpivots	Pointer to the number of piovting operations made during factorization. It is updated during this call
[in]	criterion	Threshold use for static pivoting. If diagonal value is under this threshold, its value is replaced by the threshold and the number of pivots is incremented.

Definition at line 66 of file core_chetrfsp.c.

References pastix_int_t.

Referenced by core_chetrfsp().

core_chetrfsp()

void core_chetrfsp	(	pastix_int_t	n,
		pastix_complex32_t *	A,
		pastix_int_t	lda,
		pastix_int_t *	nbpivots,
		float	criterion
	)

Compute the block static pivoting factorization of the hermitian matrix n-by-n A such that A = L * D * conjf(L^t).

Parameters

[in]	n	The number of rows and columns of the matrix A.
[in,out]	A	The matrix A to factorize with LDL^h factorization. The matrix is of size lda -by- n.
[in]	lda	The leading dimension of the matrix A.
[in,out]	nbpivots	Pointer to the number of piovting operations made during factorization. It is updated during this call
[in]	criterion	Threshold use for static pivoting. If diagonal value is under this threshold, its value is replaced by the threshold and the nu,ber of pivots is incremented.

Definition at line 147 of file core_chetrfsp.c.

References core_chetf2sp(), and pastix_int_t.

cpucblk_chetrfsp1d_hetrf()

int cpucblk_chetrfsp1d_hetrf	(	SolverMatrix *	solvmtx,
		SolverCblk *	cblk,
		void *	dataL
	)

Computes the LDL^h factorization of the diagonal block in a panel.

Parameters

[in]	solvmtx	Solver Matrix structure of the problem
[in]	cblk	Pointer to the structure representing the panel to factorize in the cblktab array. Next column blok must be accessible through cblk[1].
[in,out]	dataL	The pointer to the correct representation of lower part of the data. coeftab if the block is in full rank. Must be of size cblk.stride -by- cblk.width. pastix_lr_block if the block is compressed.

Returns

The number of static pivoting performed during the diagonal block factorization.

Definition at line 241 of file core_chetrfsp.c.

References solver_cblk_s::cblktype, solver_matrix_s::diagthreshold, solver_cblk_s::fcolnum, kernel_trace_start(), solver_cblk_s::lcolnum, pastix_int_t, PastixKernelHETRF, pastix_lrblock_s::rk, pastix_lrblock_s::rkmax, solver_cblk_s::stride, and pastix_lrblock_s::u.

Referenced by cpucblk_chetrfsp1d_panel(), and fct_blok_chetrfsp_cpu().

core_chetrfsp1d_gemm()

void core_chetrfsp1d_gemm	(	const SolverCblk *	cblk,
		const SolverBlok *	blok,
		SolverCblk *	fcblk,
		const pastix_complex32_t *	L,
		pastix_complex32_t *	C,
		pastix_complex32_t *	work
	)

core_chetrfsp1d_gemm - Computes the LDL^h factorization of one panel and apply all the trsm updates to this panel.

Parameters

[in]	cblk	The pointer to the data structure that describes the panel from which we compute the contributions. Next column blok must be accessible through cblk[1].
[in]	blok	The pointer to the data structure that describes the blok from which we compute the contributions.
[in]	fcblk	The pointer to the data structure that describes the panel on which we compute the contributions. Next column blok must be accessible through fcblk[1].
[in,out]	L	The pointer to the matrix storing the coefficients of the panel. Must be of size cblk.stride -by- cblk.width
[in,out]	C	The pointer to the matrix storing the coefficients of the target.
[in,out]	work	Temporary buffer used in core_cgemdm().

Definition at line 323 of file core_chetrfsp.c.

References blok_rownbr(), cblk_colnbr(), solver_cblk_s::cblktype, solver_blok_s::coefind, core_cgemdm(), solver_cblk_s::fblokptr, solver_cblk_s::fcolnum, solver_blok_s::frownum, is_block_inside_fblock(), pastix_int_t, PASTIX_SUCCESS, PastixConjTrans, PastixNoTrans, and solver_cblk_s::stride.

Referenced by cpucblk_chetrfsp1d(), and cpucblk_chetrfsp1dplus_update().

cpucblk_chetrfsp1d_panel()

int cpucblk_chetrfsp1d_panel	(	SolverMatrix *	solvmtx,
		SolverCblk *	cblk,
		void *	L,
		void *	DLh
	)

Compute the LDL^h factorization of one panel.

Parameters

[in]	solvmtx	Solver Matrix structure of the problem
[in]	cblk	Pointer to the structure representing the panel to factorize in the cblktab array. Next column blok must be accessible through cblk[1].
[in,out]	L	The pointer to the correct representation of lower part of the data. coeftab if the block is in full rank. Must be of size cblk.stride -by- cblk.width. pastix_lr_block if the block is compressed.
[in,out]	DLh	The pointer to the correct representation of DLh matrix (stored in the upper part by default). coeftab if the block is in full rank. Must be of size cblk.stride -by- cblk.width. pastix_lr_block if the block is compressed.

Returns

The number of static pivoting during factorization of the diagonal block.

Definition at line 439 of file core_chetrfsp.c.

References solver_cblk_s::cblktype, cpucblk_chetrfsp1d_hetrf(), cpucblk_cscalo(), cpucblk_ctrsmsp(), solver_matrix_s::lowrank, pastix_int_t, PastixConjTrans, PastixNonUnit, PastixNoTrans, PastixRight, and PastixUpper.

Referenced by cpucblk_chetrfsp1d(), and cpucblk_chetrfsp1dplus().

cpucblk_chetrfsp1d()

int cpucblk_chetrfsp1d	(	SolverMatrix *	solvmtx,
		SolverCblk *	cblk,
		pastix_complex32_t *	DLh,
		pastix_complex32_t *	work,
		pastix_int_t	lwork
	)

Perform the LDL^h factorization of a given panel and apply all its updates.

Parameters

[in]	solvmtx	Solver Matrix structure of the problem
[in]	cblk	Pointer to the structure representing the panel to factorize in the cblktab array. Next column blok must be accessible through cblk[1].
[in]	DLh	Temporary memory buffer to store the conjugate transpose of DLh.
[in]	work	Temporary memory buffer.
[in]	lwork	Temporary workspace dimension.

Returns

The number of static pivoting during factorization of the diagonal block.

Definition at line 494 of file core_chetrfsp.c.

References cblk_getdataL(), cblk_getdataU(), solver_matrix_s::cblktab, solver_cblk_s::cblktype, core_chetrfsp1d_gemm(), cpucblk_calloc(), cpucblk_calloc_lrws(), cpucblk_cgemmsp(), cpucblk_chetrfsp1d_panel(), cpucblk_crelease_deps(), solver_cblk_s::fblokptr, solver_blok_s::fcblknm, solver_matrix_s::lowrank, pastix_int_t, PastixLCoef, and PastixTrans.

Referenced by sequential_chetrf(), thread_chetrf_dynamic(), and thread_chetrf_static().

cpucblk_chetrfsp1dplus()

int cpucblk_chetrfsp1dplus	(	SolverMatrix *	solvmtx,
		SolverCblk *	cblk
	)

Perform the LDL^h factorization of a given panel and submit tasks for the subsequent updates.

Parameters

[in]	solvmtx	Solver Matrix structure of the problem
[in]	cblk	Pointer to the structure representing the panel to factorize in the cblktab array. Next column blok must be accessible through cblk[1].

Returns

The number of static pivoting during factorization of the diagonal block.

Definition at line 573 of file core_chetrfsp.c.

References solver_matrix_s::bloktab, cblk_getdataL(), solver_matrix_s::cblktab, solver_cblk_s::cblktype, cpucblk_chetrfsp1d_panel(), solver_cblk_s::fblokptr, solver_blok_s::fcblknm, pastix_int_t, pqueuePush1(), solver_cblk_s::priority, and solver_cblk_s::threadid.

Referenced by thread_chetrf_dynamic().

cpucblk_chetrfsp1dplus_update()

void cpucblk_chetrfsp1dplus_update	(	SolverMatrix *	solvmtx,
		SolverBlok *	blok,
		pastix_complex32_t *	work
	)

Apply the updates of the LDL^h factorisation of a given panel.

Parameters

[in]	solvmtx	Solver Matrix structure of the problem
[in]	blok	Pointer to the blok where the update start.
[in]	work	Temporary memory buffer.
[in]	lwork	Temporary workspace dimension.

Definition at line 626 of file core_chetrfsp.c.

References cblk_getdataL(), solver_matrix_s::cblktab, solver_cblk_s::cblktype, core_chetrfsp1d_gemm(), cpucblk_calloc(), cpucblk_crelease_deps(), solver_cblk_s::fblokptr, solver_blok_s::fcblknm, solver_blok_s::lcblknm, solver_cblk_s::lcoeftab, and PastixLCoef.

Referenced by thread_chetrf_dynamic().