libsim Versione 7.2.6
|
◆ fp_s
single precision floating point (4 byte IEEE) Definizione alla linea 258 del file kinds.F90. 258! Copyright (C) 2010 ARPA-SIM <urpsim@smr.arpa.emr.it>
259! authors:
260! Davide Cesari <dcesari@arpa.emr.it>
261! Paolo Patruno <ppatruno@arpa.emr.it>
262
263! This program is free software; you can redistribute it and/or
264! modify it under the terms of the GNU General Public License as
265! published by the Free Software Foundation; either version 2 of
266! the License, or (at your option) any later version.
267
268! This program is distributed in the hope that it will be useful,
269! but WITHOUT ANY WARRANTY; without even the implied warranty of
270! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
271! GNU General Public License for more details.
272
273! You should have received a copy of the GNU General Public License
274! along with this program. If not, see <http://www.gnu.org/licenses/>.
275#include "config.h"
276!> \defgroup base Libsim package, base library.
277!! The libsim base library defines modules and classes of general
278!! use for scientifical applications in Fortran 90. In order to
279!! compile and link programs using this library, you have to insert
280!! the required \c USE statements in the program units involved,
281!! specify the location of module files when compiling (tipically \c
282!! -I/usr/lib/gfortran/modules or \c -I/usr/lib64/gfortran/modules or
283!! \c -I/usr/include) and indicate the library name \c -lsim_base when
284!! linking, assuming that the library has been installed in a default
285!! location.
286
287!> Definition of constants to be used for declaring variables of a
288!! desired type. This module defines constants that can be portably
289!! used when declaring variables (through the \c KIND attribute) and
290!! when defining constants (through the underscore character \c _ ) in
291!! order to be sure that the desired type is used.
292!!
293!! There is a subtle difference between platform-default single and
294!! double precision real, obtained by declaring a variable of type \c
295!! REAL or \c DOUBLE \c PRECISION respectively, and single and double
296!! precision IEEE (4 and 8 bytes respectively) which are the standard
297!! IEEE data types and which are declared through \c REAL(kind=fp_s)
298!! and \c REAL(kind=fp_d) respectively: these two pairs of types
299!! usually coincide, but it may be not the case on some platforms, so
300!! you should choose one or the other approach depending on situation.
301!!
302!! Example of typical use:
303!! \code
304!! USE kinds
305!! ...
306!! INTEGER(kind=int_b) :: ab, bb
307!! REAL(kind=fp_d) :: dd
308!!
309!! ab = 13_int_b
310!! dd = REAL(ab, kind=fp_d)
311!! ...
312!! \endcode
313!! \ingroup base
315IMPLICIT NONE
316
317INTEGER, PARAMETER :: int_b = selected_int_kind(1) !< 1-byte integer (byte)
318INTEGER, PARAMETER :: int_s = selected_int_kind(4) !< 2-byte integer (short)
319INTEGER, PARAMETER :: int_l = selected_int_kind(8) !< 4-byte integer (long)
320INTEGER, PARAMETER, PRIVATE :: &
321 int_ll_t = selected_int_kind(16)
322!> 8-byte integer (long long) if supported, otherwise 4-byte integer
323INTEGER, PARAMETER :: int_ll = &
324 ( ( ( 1 + sign( 1, int_ll_t ) ) / 2 ) * int_ll_t ) + &
325 ( ( ( 1 - sign( 1, int_ll_t ) ) / 2 ) * int_l )
326
327INTEGER, PARAMETER :: fp_s = selected_real_kind(6) !< single precision floating point (4 byte IEEE)
328INTEGER, PARAMETER :: fp_d = selected_real_kind(15) !< double precision floating point (8 byte IEEE)
329INTEGER, PARAMETER, PRIVATE :: fp_q_t = selected_real_kind(20)
330!> quad precision floating point (16 byte IEEE) if supported, otherwise double precision floating point
331INTEGER, PARAMETER :: fp_q = &
332 ( ( ( 1 + sign( 1, fp_q_t ) ) / 2 ) * fp_q_t ) + &
333 ( ( ( 1 - sign( 1, fp_q_t ) ) / 2 ) * fp_d )
334
335INTEGER, PARAMETER :: ptr_c = sizeof_ptr_c !< kind for an integer having the same size of a C pointer
336
Definition of constants to be used for declaring variables of a desired type. Definition kinds.F90:245 |