https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89431
--- Comment #4 from MarkEggleston <mark.eggleston at codethink dot com> --- (In reply to Dominique d'Humieres from comment #3) > program test > use iso_fortran_env > implicit none > write(*,*) integer_kinds > #ifdef __GFC_INT_1__ > write(*,*) "__GFC_INT_1__ defined" > #else > write(*,*) "__GFC_INT_1__ undefined" > #endif > #ifdef __GFC_INT_2__ > write(*,*) "__GFC_INT_2__ defined" > #else > write(*,*) "__GFC_INT_2__ undefined" > #endif > #ifdef __GFC_INT_8__ > write(*,*) "__GFC_INT_8__ defined" > #else > write(*,*) "__GFC_INT_8__ undefined" > #endif > #ifdef __GFC_INT_16__ > write(*,*) "__GFC_INT_16__ defined" > #else > write(*,*) "__GFC_INT_16__ undefined" > #endif > write(*,*) real_kinds > #ifdef __GFC_REAL_10__ > write(*,*) "__GFC_REAL_10__ defined" > #else > write(*,*) "__GFC_REAL_10__ undefined" > #endif > #ifdef __GFC_REAL_10__ > write(*,*) "__GFC_REAL_16__ defined" > #else > write(*,*) "__GFC_REAL_16__ undefined" > #endif > end program > > gives > > 1 2 4 8 16 > __GFC_INT_1__ defined > __GFC_INT_2__ defined > __GFC_INT_8__ defined > __GFC_INT_16__ defined > 4 8 10 16 > __GFC_REAL_10__ defined > __GFC_REAL_16__ defined > > So either the doc has to be adjusted to gcc/fortran/cpp.c or > gcc/fortran/cpp.c has to be adjusted to the doc. Since Fortran uses the name integer for its type rather than int I suggest that the code should be changed.
