[Bug fortran/107659] New: C procedure with no global scope is seen as global

[Bug fortran/107659] New: C procedure with no global scope is seen as global

[urbanjost at comcast dot net]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107659

            Bug ID: 107659
           Summary: C procedure with no global scope is seen as global
           Product: gcc
           Version: 10.3.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: urbanjost at comcast dot net
  Target Milestone: ---

Created attachment 53886
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53886&action=edit
single file source appears to be required to reproduce the problem

This is reduced from a much larger (>100 000, line) code and is
very sensitive to nearly any change to the remaining file, and
requires everything being in a single file as well to reproduce
reliably. I have workarounds now, but it might be a reproducer
for something more generic. The "remove" C function is required,
which is on Linux/Unix/... machines but not sure if that exists
on MSWindows.

gfortran -c xbug.f90
xbug.f90:42:27:

   13 |    function c_remove(c_path) bind(c,name="remove") result(c_err)
      |   2                        
......
   42 |       call remove(self%key)
      |                           1
Error: Global binding name ‘remove’ at (1) is already being used as a FUNCTION
at (2)

module m_system
use,intrinsic :: iso_c_binding,   only : c_int, c_char
use,intrinsic :: iso_fortran_env, only : int32
implicit none
private
public :: CALLC
contains
elemental impure function CALLC(path) result(err)
character(*),intent(in) :: path
integer(c_int)          :: err
character(kind=c_char,len=1),allocatable :: temp(:)
interface
   function c_remove(c_path) bind(c,name="remove") result(c_err)
      import c_char,c_int
      character(kind=c_char,len=1),intent(in) :: c_path(*)
      integer(c_int)                          :: c_err
   end function
end interface
   err= c_remove(temp)
end function CALLC
end module m_system
module m_list
implicit none
private
public remove
interface remove
   module procedure remove_char
end interface
public dictionary
type dictionary
   character(len=:),allocatable :: key(:)
end type dictionary
contains
subroutine remove_char(list)
character(len=:),allocatable :: list(:)
    list=['aaaaa']
end subroutine remove_char

subroutine dict_delete(self,key)
class(dictionary),intent(in) :: self
character(len=*),intent(in)  :: key
      call remove(self%key)
end subroutine dict_delete
end module m_list

[Bug fortran/107659] C procedure with no global scope is seen as global

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107659

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2022-11-16
     Ever confirmed|0                           |1

--- Comment #1 from anlauf at gcc dot gnu.org ---
Confirmed.

NAG accepts the code just fine.

[Bug fortran/107659] C procedure with no global scope is seen as global

[anlauf at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107659

--- Comment #2 from anlauf at gcc dot gnu.org ---
I was trying a fix that regressed on binding_label_tests_34.f90,
but looking into that it appears that this test is not correct,
as well as the comment at the top of it.  The fix for pr94737 was
likely incomplete.

F2018, 18.9.2, paragraph 2 actually says:

! (2) If a variable or common block has the BIND attribute with the NAME=
! specifier and the value of its expression, after discarding leading and
! trailing blanks, has nonzero length, the variable or common block has
! this as its binding label. The case of letters in the binding label is
! significant. ...

Since NAG and Intel accepts the testcase, it will need updating...