The following program is formally valid though the use of of C_CHAR as INTEGER kind is doubtful and will not work on all systems. (C_CHAR is for CHARACTERS; there is C_SIGNED_CHAR for INTEGERS, which may or may not have the same kind number; in gfortran is has.)
However, it is simply a PARAMETER (of value 1) and as an integer of kind=1 exists, it may not be rejected according to the standard. gfortran currently rejects it. It should print a big warning but it is formally correct. See also: http://groups.google.com/group/comp.lang.fortran/browse_thread/thread/026a820711ab1e70 use iso_c_binding implicit none Character(kind=c_char) :: OK_1 Integer( kind=c_signed_char) :: OK_2 Integer( kind=c_char) :: doubtful ! Formally OK, but bad usage print *, 'c_char = ', c_char ! is 1 print *, 'c_signed_char = ', c_signed_char ! is 1 end -- Summary: Bind(C): gfortran rejects the questionable but formally correct INTEGER(C_CHAR) Product: gcc Version: unknown Status: UNCONFIRMED Keywords: rejects-valid Severity: normal Priority: P3 Component: fortran AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: burnus at gcc dot gnu dot org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40110
