------- Comment #2 from burnus at gcc dot gnu dot org 2007-04-04 15:47 ------- Note: NAG f95 has the error: Error: foo.f90, line 7: IMPLICIT setting for letter C changed after use in TEST Errors in declarations, no further processing for TEST
I think strictly speaking, NAG f95 is right. With g95, ifort and sunf95, C is REAL(4). Thus they ignore the implicit Real*8 (C) for the parameter "C". (Of cause, if one places the implicit before the parameter statment, all compiler do the same: C is a Real*8) I would argue that an error message make sense. While the error message in NAG f95 is clearer, the one of gfortran is already pretty good. What does the original reporter expected -- C being REAL(4) or REAL(8)? I think most source code readers expect the latter while the three compilers which accept it have all REAL(4). If you decide to allow this, please add a default warning about chaning the implicit type. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31465
