Refinement internal function documentation

TODO. More...

Typedefs
typedef pastix_int_t(*	refine_fct_t) (pastix_data_t *, pastix_rhs_t, pastix_rhs_t)
	Select the refinement function to call depending on the matrix type and the precision.

Functions
void	d_refine_init (struct d_solver *solver, pastix_data_t *pastix_data)
	Initiate functions pointers to define basic operations. More...
pastix_int_t	d_gmres_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	d_grad_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	d_pivot_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	d_bicgstab_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
void	c_refine_init (struct c_solver *solver, pastix_data_t *pastix_data)
	Initiate functions pointers to define basic operations. More...
pastix_int_t	c_gmres_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	c_grad_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	c_pivot_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	c_bicgstab_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
void	s_refine_init (struct s_solver *solver, pastix_data_t *pastix_data)
	Initiate functions pointers to define basic operations. More...
pastix_int_t	s_gmres_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	s_grad_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	s_pivot_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	s_bicgstab_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
void	z_refine_init (struct z_solver *solver, pastix_data_t *pastix_data)
	Initiate functions pointers to define basic operations. More...
pastix_int_t	z_gmres_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	z_grad_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	z_pivot_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
pastix_int_t	z_bicgstab_smp (pastix_data_t *pastix_data, pastix_rhs_t xp, pastix_rhs_t bp)
void	z_refine_output_oneiter (pastix_fixdbl_t t0, pastix_fixdbl_t tf, double err, pastix_int_t nb_iters)
	Print statistics about one iteration. More...
void	z_refine_output_final (pastix_data_t *pastix_data, pastix_complex64_t err, pastix_int_t nb_iters, pastix_fixdbl_t tf, void *x, pastix_complex64_t *gmresx)
	Final output. More...
void	c_refine_output_oneiter (pastix_fixdbl_t t0, pastix_fixdbl_t tf, float err, pastix_int_t nb_iters)
	Print statistics about one iteration. More...
void	c_refine_output_final (pastix_data_t *pastix_data, pastix_complex32_t err, pastix_int_t nb_iters, pastix_fixdbl_t tf, void *x, pastix_complex32_t *gmresx)
	Final output. More...
void	d_refine_output_oneiter (pastix_fixdbl_t t0, pastix_fixdbl_t tf, double err, pastix_int_t nb_iters)
	Print statistics about one iteration. More...
void	d_refine_output_final (pastix_data_t *pastix_data, double err, pastix_int_t nb_iters, pastix_fixdbl_t tf, void *x, double *gmresx)
	Final output. More...
void	s_refine_output_oneiter (pastix_fixdbl_t t0, pastix_fixdbl_t tf, float err, pastix_int_t nb_iters)
	Print statistics about one iteration. More...
void	s_refine_output_final (pastix_data_t *pastix_data, float err, pastix_int_t nb_iters, pastix_fixdbl_t tf, void *x, float *gmresx)
	Final output. More...

Detailed Description

TODO.

Function Documentation

d_refine_init()

void d_refine_init	(	struct d_solver *	solver,
		pastix_data_t *	pastix_data
	)

Initiate functions pointers to define basic operations.

Parameters

[out]	solver	The structure to be filled
[in]	pastix_data	TODO

Definition at line 121 of file d_refine_functions.c.

References bcsc_dspmv(), bcsc_dspsv(), bvec_daxpy_seq(), bvec_daxpy_smp(), bvec_dcopy_seq(), bvec_dcopy_smp(), bvec_ddot_seq(), bvec_ddot_smp(), bvec_dgemv_seq(), bvec_dgemv_smp(), bvec_dnrm2_seq(), bvec_dnrm2_smp(), bvec_dscal_seq(), bvec_dscal_smp(), bvec_free(), bvec_malloc(), d_refine_output_final(), d_refine_output_oneiter(), pastix_data_s::iparm, IPARM_SCHEDULER, and PastixSchedSequential.

Referenced by d_bicgstab_smp(), d_gmres_smp(), d_grad_smp(), and d_pivot_smp().

d_gmres_smp()

pastix_int_t d_gmres_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

d_gmres_smp - Function computing GMRES iterative refinement.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

H stores the h_{i,j} elements of the upper hessenberg matrix H (See Alg. 9.5 p 270) V stores the v_{i} vectors W stores the M^{-1} v_{i} vectors to avoid the application of the preconditioner on the output result (See line 11 of Alg 9.5)

If no preconditioner is applied, or the user wants to save memory, W stores only temporarily one vector for the Ax product (ldw is set to 0 to reuse the same vector at each iteration)

Algorithm from Iterative Methods for Sparse Linear systems, Y. Saad, Second Ed. p267-273

The version implemented is the Right preconditioned algorithm.

Compute x_m = x_0 + M^{-1} V_m y_m = x_0 + W_m y_m

Since we saved memory, we do not have (M^{-1} V_m) stored, thus we compute: w = V_m y_m w = M^{-1} (V_m y_m) x = x0 + (M^{-1} (V_m y_m))

Since we did not saved memory, we do have (M^{-1} V_m) stored in W_m if precond is true, thus we compute: x = x0 + W_m y_m, if precond x = x0 + V_m y_m, if not precond

Exit only if maximum number of iteration is reached. Exit on residual if checked at the beginning of the outer loop to be sure that the final residual of Ax-b is equal to the estimator computed within the inner loop.

Definition at line 49 of file d_refine_gmres.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, d_refine_init(), pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_GMRES_IM, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, and pastix_data_s::steps.

d_grad_smp()

pastix_int_t d_grad_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

d_grad_smp - Refine the solution using conjugate gradian method.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 49 of file d_refine_grad.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, d_refine_init(), pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, and pastix_data_s::steps.

d_pivot_smp()

pastix_int_t d_pivot_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

d_pivot_smp - Refine the solution using static pivoting method.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 48 of file d_refine_pivot.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, d_refine_init(), pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, and pastix_data_s::steps.

d_bicgstab_smp()

pastix_int_t d_bicgstab_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

d_bicgstab_smp - Function computing bicgstab iterative refinement.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 49 of file d_refine_bicgstab.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, d_refine_init(), pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, and pastix_data_s::steps.

c_refine_init()

void c_refine_init	(	struct c_solver *	solver,
		pastix_data_t *	pastix_data
	)

Initiate functions pointers to define basic operations.

Parameters

[out]	solver	The structure to be filled
[in]	pastix_data	TODO

Definition at line 121 of file c_refine_functions.c.

References bcsc_cspmv(), bcsc_cspsv(), bvec_caxpy_seq(), bvec_caxpy_smp(), bvec_ccopy_seq(), bvec_ccopy_smp(), bvec_cgemv_seq(), bvec_cgemv_smp(), bvec_cnrm2_seq(), bvec_cnrm2_smp(), bvec_cscal_seq(), bvec_cscal_smp(), bvec_free(), bvec_malloc(), c_refine_output_final(), c_refine_output_oneiter(), pastix_data_s::iparm, IPARM_SCHEDULER, and PastixSchedSequential.

Referenced by c_bicgstab_smp(), c_gmres_smp(), c_grad_smp(), and c_pivot_smp().

c_gmres_smp()

pastix_int_t c_gmres_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

c_gmres_smp - Function computing GMRES iterative refinement.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

H stores the h_{i,j} elements of the upper hessenberg matrix H (See Alg. 9.5 p 270) V stores the v_{i} vectors W stores the M^{-1} v_{i} vectors to avoid the application of the preconditioner on the output result (See line 11 of Alg 9.5)

If no preconditioner is applied, or the user wants to save memory, W stores only temporarily one vector for the Ax product (ldw is set to 0 to reuse the same vector at each iteration)

Algorithm from Iterative Methods for Sparse Linear systems, Y. Saad, Second Ed. p267-273

The version implemented is the Right preconditioned algorithm.

Compute x_m = x_0 + M^{-1} V_m y_m = x_0 + W_m y_m

Since we saved memory, we do not have (M^{-1} V_m) stored, thus we compute: w = V_m y_m w = M^{-1} (V_m y_m) x = x0 + (M^{-1} (V_m y_m))

Since we did not saved memory, we do have (M^{-1} V_m) stored in W_m if precond is true, thus we compute: x = x0 + W_m y_m, if precond x = x0 + V_m y_m, if not precond

Exit only if maximum number of iteration is reached. Exit on residual if checked at the beginning of the outer loop to be sure that the final residual of Ax-b is equal to the estimator computed within the inner loop.

Definition at line 49 of file c_refine_gmres.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, c_refine_init(), pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_GMRES_IM, IPARM_ITERMAX, IPARM_MIXED, pastix_int_t, PastixNoTrans, and pastix_data_s::steps.

c_grad_smp()

pastix_int_t c_grad_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

c_grad_smp - Refine the solution using conjugate gradian method.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 49 of file c_refine_grad.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, c_refine_init(), pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, and pastix_data_s::steps.

c_pivot_smp()

pastix_int_t c_pivot_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

c_pivot_smp - Refine the solution using static pivoting method.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 48 of file c_refine_pivot.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, c_refine_init(), pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, and pastix_data_s::steps.

c_bicgstab_smp()

pastix_int_t c_bicgstab_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

c_bicgstab_smp - Function computing bicgstab iterative refinement.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 49 of file c_refine_bicgstab.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, c_refine_init(), pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, and pastix_data_s::steps.

s_refine_init()

void s_refine_init	(	struct s_solver *	solver,
		pastix_data_t *	pastix_data
	)

Initiate functions pointers to define basic operations.

Parameters

[out]	solver	The structure to be filled
[in]	pastix_data	TODO

Definition at line 121 of file s_refine_functions.c.

References bcsc_sspmv(), bcsc_sspsv(), bvec_free(), bvec_malloc(), bvec_saxpy_seq(), bvec_saxpy_smp(), bvec_scopy_seq(), bvec_scopy_smp(), bvec_sdot_seq(), bvec_sdot_smp(), bvec_sgemv_seq(), bvec_sgemv_smp(), bvec_snrm2_seq(), bvec_snrm2_smp(), bvec_sscal_seq(), bvec_sscal_smp(), pastix_data_s::iparm, IPARM_SCHEDULER, PastixSchedSequential, s_refine_output_final(), and s_refine_output_oneiter().

Referenced by s_bicgstab_smp(), s_gmres_smp(), s_grad_smp(), and s_pivot_smp().

s_gmres_smp()

pastix_int_t s_gmres_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

s_gmres_smp - Function computing GMRES iterative refinement.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

H stores the h_{i,j} elements of the upper hessenberg matrix H (See Alg. 9.5 p 270) V stores the v_{i} vectors W stores the M^{-1} v_{i} vectors to avoid the application of the preconditioner on the output result (See line 11 of Alg 9.5)

If no preconditioner is applied, or the user wants to save memory, W stores only temporarily one vector for the Ax product (ldw is set to 0 to reuse the same vector at each iteration)

Algorithm from Iterative Methods for Sparse Linear systems, Y. Saad, Second Ed. p267-273

The version implemented is the Right preconditioned algorithm.

Compute x_m = x_0 + M^{-1} V_m y_m = x_0 + W_m y_m

Since we saved memory, we do not have (M^{-1} V_m) stored, thus we compute: w = V_m y_m w = M^{-1} (V_m y_m) x = x0 + (M^{-1} (V_m y_m))

Since we did not saved memory, we do have (M^{-1} V_m) stored in W_m if precond is true, thus we compute: x = x0 + W_m y_m, if precond x = x0 + V_m y_m, if not precond

Exit only if maximum number of iteration is reached. Exit on residual if checked at the beginning of the outer loop to be sure that the final residual of Ax-b is equal to the estimator computed within the inner loop.

Definition at line 49 of file s_refine_gmres.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_GMRES_IM, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, s_refine_init(), and pastix_data_s::steps.

s_grad_smp()

pastix_int_t s_grad_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

s_grad_smp - Refine the solution using conjugate gradian method.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 49 of file s_refine_grad.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, s_refine_init(), and pastix_data_s::steps.

s_pivot_smp()

pastix_int_t s_pivot_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

s_pivot_smp - Refine the solution using static pivoting method.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 48 of file s_refine_pivot.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, s_refine_init(), and pastix_data_s::steps.

s_bicgstab_smp()

pastix_int_t s_bicgstab_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

s_bicgstab_smp - Function computing bicgstab iterative refinement.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 49 of file s_refine_bicgstab.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, s_refine_init(), and pastix_data_s::steps.

z_refine_init()

void z_refine_init	(	struct z_solver *	solver,
		pastix_data_t *	pastix_data
	)

Initiate functions pointers to define basic operations.

Parameters

[out]	solver	The structure to be filled
[in]	pastix_data	TODO

Definition at line 121 of file z_refine_functions.c.

References bcsc_zspmv(), bcsc_zspsv(), bvec_free(), bvec_malloc(), bvec_zaxpy_seq(), bvec_zaxpy_smp(), bvec_zcopy_seq(), bvec_zcopy_smp(), bvec_zgemv_seq(), bvec_zgemv_smp(), bvec_znrm2_seq(), bvec_znrm2_smp(), bvec_zscal_seq(), bvec_zscal_smp(), pastix_data_s::iparm, IPARM_SCHEDULER, PastixSchedSequential, z_refine_output_final(), and z_refine_output_oneiter().

Referenced by z_bicgstab_smp(), z_gmres_smp(), z_grad_smp(), and z_pivot_smp().

z_gmres_smp()

pastix_int_t z_gmres_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

z_gmres_smp - Function computing GMRES iterative refinement.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

H stores the h_{i,j} elements of the upper hessenberg matrix H (See Alg. 9.5 p 270) V stores the v_{i} vectors W stores the M^{-1} v_{i} vectors to avoid the application of the preconditioner on the output result (See line 11 of Alg 9.5)

If no preconditioner is applied, or the user wants to save memory, W stores only temporarily one vector for the Ax product (ldw is set to 0 to reuse the same vector at each iteration)

Algorithm from Iterative Methods for Sparse Linear systems, Y. Saad, Second Ed. p267-273

The version implemented is the Right preconditioned algorithm.

Compute x_m = x_0 + M^{-1} V_m y_m = x_0 + W_m y_m

Since we saved memory, we do not have (M^{-1} V_m) stored, thus we compute: w = V_m y_m w = M^{-1} (V_m y_m) x = x0 + (M^{-1} (V_m y_m))

Since we did not saved memory, we do have (M^{-1} V_m) stored in W_m if precond is true, thus we compute: x = x0 + W_m y_m, if precond x = x0 + V_m y_m, if not precond

Exit only if maximum number of iteration is reached. Exit on residual if checked at the beginning of the outer loop to be sure that the final residual of Ax-b is equal to the estimator computed within the inner loop.

Definition at line 49 of file z_refine_gmres.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_GMRES_IM, IPARM_ITERMAX, IPARM_MIXED, pastix_int_t, PastixNoTrans, pastix_data_s::steps, and z_refine_init().

z_grad_smp()

pastix_int_t z_grad_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

z_grad_smp - Refine the solution using conjugate gradian method.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 49 of file z_refine_grad.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, pastix_data_s::steps, and z_refine_init().

z_pivot_smp()

pastix_int_t z_pivot_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

z_pivot_smp - Refine the solution using static pivoting method.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 48 of file z_refine_pivot.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, pastix_data_s::steps, and z_refine_init().

z_bicgstab_smp()

pastix_int_t z_bicgstab_smp	(	pastix_data_t *	pastix_data,
		pastix_rhs_t	xp,
		pastix_rhs_t	bp
	)

z_bicgstab_smp - Function computing bicgstab iterative refinement.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[out]	xp	The solution vector.
[in]	bp	The right hand side member (only one).

Returns

Number of iterations

Definition at line 49 of file z_refine_bicgstab.c.

References pastix_rhs_s::b, pastix_data_s::bcsc, pastix_data_s::dparm, DPARM_EPSILON_REFINEMENT, pastix_data_s::iparm, IPARM_ITERMAX, IPARM_MIXED, IPARM_VERBOSE, pastix_int_t, PastixNoTrans, PastixVerboseNot, pastix_data_s::procnum, pastix_data_s::steps, and z_refine_init().

z_refine_output_oneiter()

void z_refine_output_oneiter	(	pastix_fixdbl_t	t0,
		pastix_fixdbl_t	tf,
		double	err,
		pastix_int_t	nb_iters
	)

Print statistics about one iteration.

Parameters

[in]	t0	The clock value at the beginning of the iteration
[in]	tf	The clock value at the end of the iteration
[in]	err	The backward error after the iteration
[in]	nb_iters	Current number of refinement iterations

Definition at line 52 of file z_refine_functions.c.

Referenced by z_refine_init().

z_refine_output_final()

void z_refine_output_final	(	pastix_data_t *	pastix_data,
		pastix_complex64_t	err,
		pastix_int_t	nb_iters,
		pastix_fixdbl_t	tf,
		void *	x,
		pastix_complex64_t *	gmresx
	)

Final output.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[in]	err	The final backward error
[in]	nb_iters	The final number of iterations
[in]	tf	The final clock value
[in,out]	x	The vector that is to be overwritten by gmresx
[in]	gmresx	The final solution

Definition at line 90 of file z_refine_functions.c.

Referenced by z_refine_init().

c_refine_output_oneiter()

void c_refine_output_oneiter	(	pastix_fixdbl_t	t0,
		pastix_fixdbl_t	tf,
		float	err,
		pastix_int_t	nb_iters
	)

Print statistics about one iteration.

Parameters

[in]	t0	The clock value at the beginning of the iteration
[in]	tf	The clock value at the end of the iteration
[in]	err	The backward error after the iteration
[in]	nb_iters	Current number of refinement iterations

Definition at line 52 of file c_refine_functions.c.

Referenced by c_refine_init().

c_refine_output_final()

void c_refine_output_final	(	pastix_data_t *	pastix_data,
		pastix_complex32_t	err,
		pastix_int_t	nb_iters,
		pastix_fixdbl_t	tf,
		void *	x,
		pastix_complex32_t *	gmresx
	)

Final output.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[in]	err	The final backward error
[in]	nb_iters	The final number of iterations
[in]	tf	The final clock value
[in,out]	x	The vector that is to be overwritten by gmresx
[in]	gmresx	The final solution

Definition at line 90 of file c_refine_functions.c.

Referenced by c_refine_init().

d_refine_output_oneiter()

void d_refine_output_oneiter	(	pastix_fixdbl_t	t0,
		pastix_fixdbl_t	tf,
		double	err,
		pastix_int_t	nb_iters
	)

Print statistics about one iteration.

Parameters

[in]	t0	The clock value at the beginning of the iteration
[in]	tf	The clock value at the end of the iteration
[in]	err	The backward error after the iteration
[in]	nb_iters	Current number of refinement iterations

Definition at line 52 of file d_refine_functions.c.

Referenced by d_refine_init().

d_refine_output_final()

void d_refine_output_final	(	pastix_data_t *	pastix_data,
		double	err,
		pastix_int_t	nb_iters,
		pastix_fixdbl_t	tf,
		void *	x,
		double *	gmresx
	)

Final output.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[in]	err	The final backward error
[in]	nb_iters	The final number of iterations
[in]	tf	The final clock value
[in,out]	x	The vector that is to be overwritten by gmresx
[in]	gmresx	The final solution

Definition at line 90 of file d_refine_functions.c.

Referenced by d_refine_init().

s_refine_output_oneiter()

void s_refine_output_oneiter	(	pastix_fixdbl_t	t0,
		pastix_fixdbl_t	tf,
		float	err,
		pastix_int_t	nb_iters
	)

Print statistics about one iteration.

Parameters

[in]	t0	The clock value at the beginning of the iteration
[in]	tf	The clock value at the end of the iteration
[in]	err	The backward error after the iteration
[in]	nb_iters	Current number of refinement iterations

Definition at line 52 of file s_refine_functions.c.

Referenced by s_refine_init().

s_refine_output_final()

void s_refine_output_final	(	pastix_data_t *	pastix_data,
		float	err,
		pastix_int_t	nb_iters,
		pastix_fixdbl_t	tf,
		void *	x,
		float *	gmresx
	)

Final output.

Parameters

[in]	pastix_data	The PaStiX data structure that describes the solver instance.
[in]	err	The final backward error
[in]	nb_iters	The final number of iterations
[in]	tf	The final clock value
[in,out]	x	The vector that is to be overwritten by gmresx
[in]	gmresx	The final solution

Definition at line 90 of file s_refine_functions.c.

Referenced by s_refine_init().