[Bug fortran/77504] [9/10/11/12 Regression] "is used uninitialized" with allocatable string and array constructors

[mailling-lists-bd at posteo dot de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=77504

Baptiste Demoulin <mailling-lists-bd at posteo dot de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mailling-lists-bd at posteo dot de

--- Comment #27 from Baptiste Demoulin <mailling-lists-bd at posteo dot de> ---
I don't know if it is related to this bug, but I found that this program:

program test_allocatable
  implicit none

  integer, allocatable  :: array(:)
  character(len=:), allocatable :: string

  integer :: i
  character(len=64) :: buf

  allocate(array(4))
  array = [1, 2, 3, 4]

  do i=1, size(array) - 1
     write(buf, '(I0)') array(i)
     string = string//trim(buf)//','
  end do

  write(buf, '(I0)') array(size(array))
  string = string//trim(buf)

  write(*, '(A)') string

end program test_allocatable

yields:

gfortran -Wuninitialized allocatable_string.f90
allocatable_string.f90:20:28:

   20 |   string = string//trim(buf)
      |                            ^
Warning: ‘.__var_2_realloc_string’ may be used uninitialized
[-Wmaybe-uninitialized]

If the line "string = string//trim(buf)" after the loop is commented, then the
warning disppears.

The test was run on Fedora, where "gfortran --version" yields:

GNU Fortran (GCC) 11.2.1 20211203 (Red Hat 11.2.1-7)
Copyright (C) 2021 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.