Open MPI logo

MPI_Type_create_f90_real(3) man page (version 1.4.5)

  |   Home   |   Support   |   FAQ   |  

« Return to documentation listing

Table of Contents


MPI_Type_create_f90_real - Returns a bounded MPI real datatype


C Syntax

#include <mpi.h>
int MPI_Type_create_f90_real(int p, int r, MPI_Datatype *newtype)

Fortran Syntax

INCLUDE ’mpif.h’

C++ Syntax

#include <mpi.h>
static MPI::Datatype MPI::Datatype::Create_f90_real(int p, int r)

Input Parameters

Precision, in decimal digits (integer).
Decimal exponent range (integer).

Output Parameters

New data type (handle).
Fortran only: Error status (integer).


This function provides a way to declare KIND-parameterized REAL MPI datatypes. The arguments are interpreted in a similar fashion to the F90 function SELECTED_REAL_KIND. The parameters p and r must be scalar integers. The argument p represents the required level of numerical precision, in decimal digits. The r parameter indicates the range of exponents desired: the returned datatype will have at least one exponent between +r and -r (inclusive).

Either p or r, but not both, may be omitted from calls to SELECTED_REAL_KIND. Similarly, either argument to MPI_Type_create_f90_real may be set to MPI_UNDEFINED.


It is erroneous to supply values for p and r not supported by the compiler.

The Fortran function SELECTED_REAL_KIND maps a large number of (p,r) pairs to a much smaller number of KIND parameters supported by the compiler. KIND parameters are not specified by the language and are not portable. From the point of view of the language, variables of the same base type and KIND parameter are equivalent, even if their KIND parameters were generated by different (p,r) arguments to SELECTED_REAL_KIND. However, to help facilitate interoperability in a heterogeneous environment, equivalency is more strictly defined for datatypes returned by MPI_Type_create_f90_real. Two MPI datatypes, each generated by this function, will match if and only if they have identical values for both p and r.

The interaction between the datatypes returned by this function and the external32 data representation - used by MPI_Pack_external, MPI_Unpack_external and many MPI_File functions - is subtle. The external32 representation of returned datatypes is as follows.

    if (p > 33) and/or (r > 4931):
        external32 size = n/a (undefined)
    else if (p > 15) and/or (r > 307):
        external32 size = 16
    else if (p > 6) and/or (r > 37):
        external32 size = 8
        external32 size = 4

If the external32 representation of a datatype is undefined, so are the results of using that datatype in operations that require the external32 format. Care should be taken not to use incompatible datatypes indirectly, e.g., as part of another datatype or through a duplicated datatype, in these functions.

If a variable is declared specifying a nondefault KIND value that was not obtained with SELECTED_REAL_KIND (i.e., p and/or r are unknown), the only way to obtain a matching MPI datatype is to use the functions MPI_Sizeof and MPI_Type_match_size.


Almost all MPI routines return an error value; C routines as the value of the function and Fortran routines in the last argument. C++ functions do not return errors. If the default error handler is set to MPI::ERRORS_THROW_EXCEPTIONS, then on error the C++ exception mechanism will be used to throw an MPI:Exception object.

Before the error value is returned, the current MPI error handler is called. By default, this error handler aborts the MPI job, except for I/O function errors. The error handler may be changed with MPI_Comm_set_errhandler; the predefined error handler MPI_ERRORS_RETURN may be used to cause error values to be returned. Note that MPI does not guarantee that an MPI program can continue past an error.

See the MPI man page for a full list of MPI error codes.

See Also


« Return to documentation listing