[Bug fortran/64958] Warn if INTENT(IN) is changed by passing to no-intent argument

[Bug fortran/64958] Warn if INTENT(IN) is changed by passing to no-intent argument

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64958

Thomas Koenig <tkoenig at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Version|unknown                     |5.0
           Severity|normal                      |enhancement

[Bug fortran/64958] New: Warn if INTENT(IN) is changed by passing to no-intent argument

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64958

            Bug ID: 64958
           Summary: Warn if INTENT(IN) is changed by passing to no-intent
                    argument
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: tkoenig at gcc dot gnu.org

Consider the following code:

module foo
  implicit none
contains
  subroutine bar(n)
    integer, intent(in) :: n
    print *,"bar before dusty", n
    call dusty(n)
    print *,"bar after dusty", n
  end subroutine bar
end module foo

program main
  use foo
  implicit none
  integer :: n
  n = 5
  call bar(n)
end program main

subroutine dusty(n)
  integer n
  n = n - 1
end

What "dusty" does is illegal, but it may not be easy for the programmer
to catch this (especially if "dusty" is in a library).

So, it would be nice if -fcheck=all would contain a check which checked
if intent(in) dummy arguments are changed by being passed to
intent(none).

[Bug fortran/64958] Warn if INTENT(IN) is changed by passing to no-intent argument

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64958

--- Comment #1 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
If this mechanism is in place, it could also be used to catch code like

call foo(a+1)

...


subroutine foo(a)
   a = a + 1
end subroutine

[Bug fortran/64958] Warn if INTENT(IN) is changed by passing to no-intent argument

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64958

Tobias Burnus <burnus at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |burnus at gcc dot gnu.org

--- Comment #2 from Tobias Burnus <burnus at gcc dot gnu.org> ---
How is this supposed to work? I think in general, it is not that simple to test
for it. Or do you want to restrict it to scalar numeric intrinsic types - and
create a shadow variable? Even that might not work with optimization as we tell
the middle end that the variable doesn't change - and it might then in some
cases simply optimize the check away.

[Bug fortran/64958] Warn if INTENT(IN) is changed by passing to no-intent argument

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64958

--- Comment #3 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
If it is forbidden to modify an actual argument, we could make
a tempoary copy of that argument, pass that to a procedure where
it maches an intent(unknown) dummy argument, then do the comparison.

So, we would replace

subroutine bar(n)
  integer, intent(in) :: n
  call dusty(n)
end subroutine bar

with

subroutine bar(n)
  integer, intent(in) :: n
  integer :: shadow_n

  shadow_n = n
  call dusty(shadow_n)
  if (n /= shadow_n) call _gfortran_runtime_error("bletchful")
end subroutine bar

where dusty has no explicit interface or the first dummy argument
of dusty has unknown intent.

It would probably make most sense for scalar intrinsic types inculding
character variables.  Arrays of intrinsic types might also be a possibility,
but with decreasing return on investment, so to speak.

[Bug fortran/64958] Warn if INTENT(IN) is changed by passing to no-intent argument

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64958

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2015-03-06
     Ever confirmed|0                           |1

--- Comment #4 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
Confirmed.