Auxilliary Routines Parameters APIs#

LAMCH#

group lamch

Functions

template<typename T>
T lamch(char *cmach)#

LAMCH determines single precision machine parameters.

Purpose:

   LAMCH determines single precision machine parameters.

Parameters:

CMACH – [in]

CMACH is CHARACTER*1

Specifies the value to be returned by SLAMCH:

= ‘E’ or ‘e’, SLAMCH := eps

= ‘S’ or ‘s , SLAMCH := sfmin

= ‘B’ or ‘b’, SLAMCH := base

= ‘P’ or ‘p’, SLAMCH := eps*base

= ‘N’ or ‘n’, SLAMCH := t

= ‘R’ or ‘r’, SLAMCH := rnd

= ‘M’ or ‘m’, SLAMCH := emin

= ‘U’ or ‘u’, SLAMCH := rmin

= ‘L’ or ‘l’, SLAMCH := emax

= ‘O’ or ‘o’, SLAMCH := rmax

where

eps = relative machine precision

sfmin = safe minimum, such that 1/sfmin does not overflow

base = base of the machine

prec = eps*base

t = number of (base) digits in the mantissa

rnd = 1.0 when rounding occurs in addition, 0.0 otherwise

emin = minimum exponent before (gradual) underflow

rmin = underflow threshold - base**(emin-1)

emax = largest exponent before overflow

rmax = overflow threshold - (base**emax)*(1-eps)

Returns:

T Template based return value of the function.

LABAD#

group labad

Functions

template<typename T>
void labad(T *small_val, T *large)#

LABAD takes as input the values computed by LAMCH for underflow and

overflow, and returns the square root of each of these values.

Purpose:

 LABAD takes as input the values computed by SLAMCH for underflow and
 overflow, and   returns the square root of each of these values if the
 log of LARGE is sufficiently large.  This subroutine is intended to
 identify machines with a large exponent range, such as the Crays, and
 redefine the underflow and overflow limits to be the square roots of
 the values computed by SLAMCH.  This subroutine is needed because
 SLAMCH does not compensate for poor arithmetic in the upper half of
 the exponent range, as is found on a Cray.

Parameters:
  • SMALL_VAL – [inout]

    SMALL_VAL is REAL

    On entry, the underflow threshold as computed by SLAMCH. On exit, if LOG10(LARGE) is sufficiently large, the square root of SMALL, otherwise unchanged.

  • LARGE – [inout]

    LARGE is REAL

    On entry, the overflow threshold as computed by SLAMCH. On exit, if LOG10(LARGE) is sufficiently large, the square root of LARGE, otherwise unchanged.

ILAVER#

group ilaver

Functions

inline void ilaver(integer *vers_major, integer *vers_minor, integer *vers_patch__)#

ILAVER returns the LAPACK version.

Purpose:

 This subroutine   returns the LAPACK version.

Parameters:
  • VERS_MAJOR – [out]

    VERS_MAJOR is INTEGER

    return the lapack major version

  • VERS_MINOR – [out]

    VERS_MINOR is INTEGER

    return the lapack minor version from the major version

  • VERS_PATCH – [out]

    VERS_PATCH is INTEGER

    return the lapack patch version from the minor version

IPARAM2STAGE#

group iparam2stage

Functions

inline integer iparam2stage(integer *ispec, char *name, char *opts, integer *ni, integer *nbi, integer *ibi, integer *nxi)#

IPARAM2STAGE program sets problem and machine dependent parameters.

Purpose:

    This program sets problem and machine dependent parameters
    useful for xHETRD_2STAGE, xHETRD_HE2HB, xHETRD_HB2ST,
    xGEBRD_2STAGE, xGEBRD_GE2GB, xGEBRD_GB2BD
    and related subroutines for eigenvalue problems.
    It is called whenever ILAENV is called with 17 <= ISPEC <= 21.
    It is called whenever ILAENV2STAGE is called with 1 <= ISPEC <= 5
    with a direct conversion ISPEC + 16.

Parameters:
  • ISPEC – [in]

    ISPEC is integer scalar

    ISPEC specifies which tunable parameter IPARAM2STAGE should return.

    ISPEC=17: the optimal blocksize nb for the reduction to BAND

    ISPEC=18: the optimal blocksize ib for the eigenvectors singular vectors update routine

    ISPEC=19: The length of the array that store the Housholder representation for the second stage Band to Tridiagonal or Bidiagonal

    ISPEC=20: The workspace needed for the routine in input.

    ISPEC=21: For future release.

  • NAME – [in]

    NAME is character string

    Name of the calling subroutine

  • OPTS – [in]

    OPTS is CHARACTER*(*)

    The character options to the subroutine NAME, concatenated into a single character string. For example, UPLO = ‘U’, TRANS = ‘T’, and DIAG = ‘N’ for a triangular routine would be specified as OPTS = ‘UTN’.

  • NI – [in] NI is INTEGER which is the size of the matrix

  • NBI – [in]

    NBI is INTEGER which is the used in the reduciton,

    (e.g., the size of the band), needed to compute workspace and LHOUS2.

  • IBI – [in] IBI is INTEGER which represent the IB of the reduciton, needed to compute workspace and LHOUS2.

  • NXI – [in] NXI is INTEGER needed in the future release.

Returns:

INTEGER Return value of the function.

ILADIAG#

group iladiag

Functions

inline integer iladiag(char *diag)#

ILADIAG translated from a character string specifying if a matrix

has unit diagonal or not to the relevant BLAST-specified integer constant.

Purpose:

 This subroutine translated from a character string specifying if a
 matrix has unit diagonal or not to the relevant BLAST-specified
 integer constant.

 ILADIAG   returns an INTEGER. If ILADIAG < 0, then the input is not a
 character indicating a unit or non-unit diagonal. Otherwise ILADIAG
 returns the constant value corresponding to DIAG.

Parameters:

diag – [in]

DIAG is character pointer.

= N means non-unit diagonal.

= U means unit diagonal.

Returns:

INTEGER Return value of the function.

ILATRANS#

group ilatrans

Functions

inline integer ilatrans(char *trans)#

ILATRANS translates from a character string specifying a transposition

operation to the relevant BLAST-specified integer constant.

Purpose:

 This subroutine translates from a character string specifying a
 transposition operation to the relevant BLAST-specified integer
 constant.

 ILATRANS   returns an INTEGER.  If ILATRANS < 0, then the input is not
 a character indicating a transposition operator.  Otherwise ILATRANS
   returns the constant value corresponding to TRANS

Parameters:

trans – [in]

trans is charater pointer.

= N means no transposition

= T means transposition

= C means conjugate transposition

Returns:

INTEGER Return value of the function.

ILAPREC#

group ilaprec

Functions

inline integer ilaprec(char *prec)#

ILAPREC translated from a character string specifying an intermediate

precision to the relevant BLAST-specified integer constant.

Purpose:

 This subroutine translated from a character string specifying an
 intermediate precision to the relevant BLAST-specified integer
 constant.

 ILAPREC   returns an INTEGER.  If ILAPREC < 0, then the input is not a
 character indicating a supported intermediate precision.  Otherwise
 ILAPREC   returns the constant value corresponding to PREC.

Parameters:

prec – [in]

PREC is character pointer.

= S means single precision

= D means double precision

= I means indigenous precision

= X or E means extra precision.

Other values returns invalid result.

Returns:

INTEGER Return value of the function.

ILAENV2STAGE#

group ilaenv2stage

Functions

inline integer ilaenv2stage(integer *ispec, char *name, char *opts, integer *n1, integer *n2, integer *n3, integer *n4)#

ILAENV2STAGE is called from the LAPACK routines to choose problem-dependent

parameters for the local environment.

Purpose:

   ILAENV2STAGE is called from the LAPACK routines to choose problem-dependent
   parameters for the local environment.  See ISPEC for a description of
   the parameters.
   It sets problem and machine dependent parameters useful for *_2STAGE and
   related subroutines.

   ILAENV2STAGE   returns an INTEGER
   if ILAENV2STAGE >= 0: ILAENV2STAGE   returns the value of the parameter
                         specified by ISPEC
   if ILAENV2STAGE < 0:  if ILAENV2STAGE = -k, the k-th argument had an
                         illegal value.

   This version provides a set of parameters which should give good,
   but not optimal, performance on many of the currently available
   computers for the 2-stage solvers. Users are encouraged to modify this
   subroutine to set the tuning parameters for their particular machine using
   the option and problem size information in the arguments.

   This routine will not function correctly if it is converted to all
   lower case.  Converting it to all upper case is allowed.

Parameters:
  • ISPEC – [in]

    ISPEC is INTEGER

    Specifies the parameter to be returned as the value of ILAENV2STAGE.

    = 1: the optimal blocksize nb for the reduction to BAND

    = 2: the optimal blocksize ib for the eigenvectors singular vectors update routine

    = 3: The length of the array that store the Housholder representation for the second stage Band to Tridiagonal or Bidiagonal

    = 4: The workspace needed for the routine in input.

    = 5: For future release.

  • NAME – [in]

    NAME is CHARACTER*(*)

    The name of the calling subroutine, in either upper case or lower case.

  • OPTS – [in]

    OPTS is CHARACTER*(*)

    The character options to the subroutine NAME, concatenated into a single character string. For example, UPLO = ‘U’, TRANS = ‘T’, and DIAG = ‘N’ for a triangular routine would be specified as OPTS = ‘UTN’.

  • N1 – [in] N1 is INTEGER

  • N2 – [in] N2 is INTEGER

  • N3 – [in] N3 is INTEGER

  • N4 – [in]

    N4 is INTEGER

    Problem dimensions for the subroutine NAME; these may not all be required.

Returns:

INTEGER Return value of the function.

ILAUPLO#

group ilauplo

Functions

inline integer ilauplo(char *uplo)#

ILAUPLO translated from a character string specifying a upper- or

lower-triangular matrix to the relevant BLAST-specified integer constant.

Purpose:

 This subroutine translated from a character string specifying a
 upper- or lower-triangular matrix to the relevant BLAST-specified
 integer constant.

 ILAUPLO   returns an INTEGER.  If ILAUPLO < 0, then the input is not
 a character indicating an upper- or lower-triangular matrix.
 Otherwise ILAUPLO   returns the constant value corresponding to UPLO

Parameters:

uplo – [in]

uplo is charater pointer.

= ‘L’ means lower triangular matrix

= ‘U’ means upper triangular matrix

Other values, returns invalid result.

Returns:

INTEGER Return value of the function.

LATRANS#

group latrans

Functions

inline void hla_transtype(char *ret_val, integer *trans)#

CHLA_TRANSTYPE subroutine translates from a BLAST-specified integer constant

to the character string specifying a transposition operation.

Purpose:

 This subroutine translates from a BLAST-specified integer constant to
 the character string specifying a transposition operation.

 CHLA_TRANSTYPE   returns an CHARACTER*1.  If CHLA_TRANSTYPE is 'X',
 then input is not an integer indicating a transposition operator.
 Otherwise CHLA_TRANSTYPE   returns the constant value corresponding to
 TRANS.