Macros and constants, which are type specific.
Here reside declarations related to supported matrix numerical types, and other declarations according to the build time options.
If you wish to use this library with different matrix numerical types, you shall regenerate the library source code accordingly; see the README file how to do this.
Only a small part of these declarations is needed to the user (see matrix_type_symbols_section).
Therefore, only the declarations which are commented are actually meant to be used in functions; please regard the remaining ones as internal. More...

Macros
#define	RSB_LIBRSB_VER_STRING "1.2.0"
	Library version string. More...

#define	RSB_HEADER_VERSION_STRING "librsb version 1.2.0.11 - 202112161631"
	Library header version string. More...

#define	RSB_LIBRSB_VER_MAJOR 1
	Major version. More...

#define	RSB_LIBRSB_VER_MINOR 2
	Minor version. More...

#define	RSB_LIBRSB_VER_PATCH 0
	Patch version. More...

#define	RSB_LIBRSB_VER 10201
	Version number. More...

#define	RSB_LIBRSB_VER_DATE 202112161631
	Version release date. More...

#define	RSB_HAVE_TYPE_DOUBLE 1
	Type double is supported, so RSB_HAVE_TYPE_DOUBLE is defined . More...

#define	RSB_HAVE_TYPE_FLOAT 1
	Type float is supported, so RSB_HAVE_TYPE_FLOAT is defined . More...

#define	RSB_HAVE_TYPE_FLOAT_COMPLEX 1
	Type float complex is supported, so RSB_HAVE_TYPE_FLOAT_COMPLEX is defined . More...

#define	RSB_HAVE_TYPE_DOUBLE_COMPLEX 1
	Type double complex is supported, so RSB_HAVE_TYPE_DOUBLE_COMPLEX is defined . More...

#define	RSB_DEFAULT_TYPE double
	The default numerical matrix type (can be used for declarations), used in the example programs. More...

#define	RSB_DEFAULT_POSSIBLY_INTEGER_TYPE double
	The default, integer if possible , numerical type (can be used for declarations). More...

#define	RSB_DEFAULT_POSSIBLY_FIRST_BLAS_TYPE float
	The default, blas if possible , numerical type (can be used for declarations). More...

#define	RSB_DEFAULT_TYPE_STRING "double"
	A string specifying the name of the default type. More...

#define	RSB_DEFAULT_POSSIBLY_INTEGER_TYPE_STRING "double"
	A string specifying the name of the default possibly integer type. More...

#define	RSB_DEFAULT_SYMMETRY RSB_SYMMETRY_U
	The default symmetry flag. More...

#define	RSB_DEFAULT_TRANSPOSITION RSB_TRANSPOSITION_N
	The default transposition flag (no transposition). More...

#define	RSB_ROWS_TRANSPOSITIONS_ARRAY {RSB_TRANSPOSITION_N, RSB_TRANSPOSITION_T, RSB_TRANSPOSITION_C, RSB_INVALID_TRANS }
	An array with transposition constants. More...

#define	RSB_TYPE_INDEX_DOUBLE 0

#define	RSB_TYPE_INDEX_FLOAT 1

#define	RSB_TYPE_INDEX_FLOAT_COMPLEX 2

#define	RSB_TYPE_INDEX_DOUBLE_COMPLEX 3

Values for valid matrix transposition flags.
The Hermitian flag will act as simple transposed, for non complex types.
#define	RSB_TRANSPOSITION_N 0x4E
	N: Non transposed flag, valid for rsb_trans_t typed variables. More...

#define	RSB_TRANSPOSITION_T 0x54
	T: Transposed flag value, valid for rsb_trans_t valued variables. More...

#define	RSB_TRANSPOSITION_C 0x43
	C: Conjugated transpose flag, valid for rsb_trans_t typed variables. More...

Valid symbol values for matrix numerical type specification -- type codes -- (type \see #rsb_type_t).

#define	RSB_NUMERICAL_TYPE_SAME_TYPE 1
	a bogus type flag for specifying no type conversion More...

#define	RSB_NUMERICAL_TYPE_DOUBLE 'D'
	Character code for type double. More...

#define	RSB_NUMERICAL_TYPE_SAME_TYPE 1
	a bogus type flag for specifying no type conversion More...

#define	RSB_NUMERICAL_TYPE_FLOAT 'S'
	Character code for type float. More...

#define	RSB_NUMERICAL_TYPE_SAME_TYPE 1
	a bogus type flag for specifying no type conversion More...

#define	RSB_NUMERICAL_TYPE_FLOAT_COMPLEX 'C'
	Character code for type float complex. More...

#define	RSB_NUMERICAL_TYPE_SAME_TYPE 1
	a bogus type flag for specifying no type conversion More...

#define	RSB_NUMERICAL_TYPE_DOUBLE_COMPLEX 'Z'
	Character code for type double complex. More...

#define	RSB_NUMERICAL_TYPE_FORTRAN_SAME_TYPE 1
	a bogus type flag for specifying no type conversion More...

#define	RSB_NUMERICAL_TYPE_FORTRAN_INT ICHAR('I')
	Character code for type int, to be used (only) from Fortran. More...

#define	RSB_NUMERICAL_TYPE_FORTRAN_DOUBLE ICHAR('D')
	Character code for type double, to be used (only) from Fortran. More...

#define	RSB_NUMERICAL_TYPE_FORTRAN_FLOAT ICHAR('S')
	Character code for type float, to be used (only) from Fortran. More...

#define	RSB_NUMERICAL_TYPE_FORTRAN_FLOAT_COMPLEX ICHAR('C')
	Character code for type float complex, to be used (only) from Fortran. More...

#define	RSB_NUMERICAL_TYPE_FORTRAN_DOUBLE_COMPLEX ICHAR('Z')
	Character code for type double complex, to be used (only) from Fortran. More...

#define	RSB_NUMERICAL_TYPE_DEFAULT RSB_NUMERICAL_TYPE_DOUBLE
	A default numerical matrix type. More...

#define	RSB_NUMERICAL_TYPE_DEFAULT_INTEGER RSB_NUMERICAL_TYPE_DOUBLE
	A default numerical matrix type; if possible, an integer one. More...

#define	RSB_NUMERICAL_TYPE_INVALID_TYPE '?'
	By definition, an invalid type code. More...

#define	RSB_NUMERICAL_TYPE_FIRST_BLAS RSB_NUMERICAL_TYPE_FLOAT
	A default numerical matrix type; if possible, not integer one. If no such type is configured in, then the invalid type. More...

#define	RSB_CHAR_AS_TRANSPOSITION(TRANSC)
	Get the right transposition flag out of either n, c, t chars. More...

Miscellaneous constants.
#define	RSB_CONST_MAX_TUNING_ROUNDS 16
	Maximal count of tuning rounds in one invocation of (rsb_tune_spmm/rsb_tune_spsm). More...

Detailed Description

Macros and constants, which are type specific.
Here reside declarations related to supported matrix numerical types, and other declarations according to the build time options.
If you wish to use this library with different matrix numerical types, you shall regenerate the library source code accordingly; see the README file how to do this.
Only a small part of these declarations is needed to the user (see matrix_type_symbols_section).
Therefore, only the declarations which are commented are actually meant to be used in functions; please regard the remaining ones as internal.

Macro Definition Documentation

◆ RSB_CHAR_AS_TRANSPOSITION

#define RSB_CHAR_AS_TRANSPOSITION ( TRANSC )

Value:

(                                                                                                               \
                (TRANSC) == ('N') ? (RSB_TRANSPOSITION_N) :             \
                (TRANSC) == ('n') ? (RSB_TRANSPOSITION_N) :             \
                (TRANSC) == ('T') ? (RSB_TRANSPOSITION_T) :             \
                (TRANSC) == ('t') ? (RSB_TRANSPOSITION_T) :             \
                (TRANSC) == ('C') ? (RSB_TRANSPOSITION_C) :             \
                (TRANSC) == ('c') ? (RSB_TRANSPOSITION_C) :             \
                '?'                                                                                             \
)

Get the right transposition flag out of either n, c, t chars.

◆ RSB_CONST_MAX_TUNING_ROUNDS

#define RSB_CONST_MAX_TUNING_ROUNDS 16

Maximal count of tuning rounds in one invocation of (rsb_tune_spmm/rsb_tune_spsm).

◆ RSB_DEFAULT_POSSIBLY_FIRST_BLAS_TYPE

#define RSB_DEFAULT_POSSIBLY_FIRST_BLAS_TYPE float

The default, blas if possible , numerical type (can be used for declarations).

◆ RSB_DEFAULT_POSSIBLY_INTEGER_TYPE

#define RSB_DEFAULT_POSSIBLY_INTEGER_TYPE double

The default, integer if possible , numerical type (can be used for declarations).

◆ RSB_DEFAULT_POSSIBLY_INTEGER_TYPE_STRING

#define RSB_DEFAULT_POSSIBLY_INTEGER_TYPE_STRING "double"

A string specifying the name of the default possibly integer type.

◆ RSB_DEFAULT_SYMMETRY

#define RSB_DEFAULT_SYMMETRY RSB_SYMMETRY_U

The default symmetry flag.

◆ RSB_DEFAULT_TRANSPOSITION

#define RSB_DEFAULT_TRANSPOSITION RSB_TRANSPOSITION_N

The default transposition flag (no transposition).

◆ RSB_DEFAULT_TYPE

#define RSB_DEFAULT_TYPE double

The default numerical matrix type (can be used for declarations), used in the example programs.

◆ RSB_DEFAULT_TYPE_STRING

#define RSB_DEFAULT_TYPE_STRING "double"

A string specifying the name of the default type.

◆ RSB_HAVE_TYPE_DOUBLE

#define RSB_HAVE_TYPE_DOUBLE 1

Type double is supported, so RSB_HAVE_TYPE_DOUBLE is defined .

◆ RSB_HAVE_TYPE_DOUBLE_COMPLEX

#define RSB_HAVE_TYPE_DOUBLE_COMPLEX 1

Type double complex is supported, so RSB_HAVE_TYPE_DOUBLE_COMPLEX is defined .

◆ RSB_HAVE_TYPE_FLOAT

#define RSB_HAVE_TYPE_FLOAT 1

Type float is supported, so RSB_HAVE_TYPE_FLOAT is defined .

◆ RSB_HAVE_TYPE_FLOAT_COMPLEX

#define RSB_HAVE_TYPE_FLOAT_COMPLEX 1

Type float complex is supported, so RSB_HAVE_TYPE_FLOAT_COMPLEX is defined .

◆ RSB_HEADER_VERSION_STRING

#define RSB_HEADER_VERSION_STRING "librsb version 1.2.0.11 - 202112161631"

Library header version string.

◆ RSB_LIBRSB_VER

#define RSB_LIBRSB_VER 10201

Version number.

◆ RSB_LIBRSB_VER_DATE

#define RSB_LIBRSB_VER_DATE 202112161631

Version release date.

◆ RSB_LIBRSB_VER_MAJOR

#define RSB_LIBRSB_VER_MAJOR 1

Major version.

◆ RSB_LIBRSB_VER_MINOR

#define RSB_LIBRSB_VER_MINOR 2

Minor version.

◆ RSB_LIBRSB_VER_PATCH

#define RSB_LIBRSB_VER_PATCH 0

Patch version.

◆ RSB_LIBRSB_VER_STRING

#define RSB_LIBRSB_VER_STRING "1.2.0"

Library version string.

◆ RSB_NUMERICAL_TYPE_DEFAULT

#define RSB_NUMERICAL_TYPE_DEFAULT RSB_NUMERICAL_TYPE_DOUBLE

A default numerical matrix type.

◆ RSB_NUMERICAL_TYPE_DEFAULT_INTEGER

#define RSB_NUMERICAL_TYPE_DEFAULT_INTEGER RSB_NUMERICAL_TYPE_DOUBLE

A default numerical matrix type; if possible, an integer one.

◆ RSB_NUMERICAL_TYPE_DOUBLE

#define RSB_NUMERICAL_TYPE_DOUBLE 'D'

Character code for type double.

◆ RSB_NUMERICAL_TYPE_DOUBLE_COMPLEX

#define RSB_NUMERICAL_TYPE_DOUBLE_COMPLEX 'Z'

Character code for type double complex.

◆ RSB_NUMERICAL_TYPE_FIRST_BLAS

#define RSB_NUMERICAL_TYPE_FIRST_BLAS RSB_NUMERICAL_TYPE_FLOAT

A default numerical matrix type; if possible, not integer one. If no such type is configured in, then the invalid type.

◆ RSB_NUMERICAL_TYPE_FLOAT

#define RSB_NUMERICAL_TYPE_FLOAT 'S'

Character code for type float.

◆ RSB_NUMERICAL_TYPE_FLOAT_COMPLEX

#define RSB_NUMERICAL_TYPE_FLOAT_COMPLEX 'C'

Character code for type float complex.

◆ RSB_NUMERICAL_TYPE_FORTRAN_DOUBLE

#define RSB_NUMERICAL_TYPE_FORTRAN_DOUBLE ICHAR('D')

Character code for type double, to be used (only) from Fortran.

◆ RSB_NUMERICAL_TYPE_FORTRAN_DOUBLE_COMPLEX

#define RSB_NUMERICAL_TYPE_FORTRAN_DOUBLE_COMPLEX ICHAR('Z')

Character code for type double complex, to be used (only) from Fortran.

◆ RSB_NUMERICAL_TYPE_FORTRAN_FLOAT

#define RSB_NUMERICAL_TYPE_FORTRAN_FLOAT ICHAR('S')

Character code for type float, to be used (only) from Fortran.

◆ RSB_NUMERICAL_TYPE_FORTRAN_FLOAT_COMPLEX

#define RSB_NUMERICAL_TYPE_FORTRAN_FLOAT_COMPLEX ICHAR('C')

Character code for type float complex, to be used (only) from Fortran.

◆ RSB_NUMERICAL_TYPE_FORTRAN_INT

#define RSB_NUMERICAL_TYPE_FORTRAN_INT ICHAR('I')

Character code for type int, to be used (only) from Fortran.