[Bug fortran/66575] New: Endless compilation on missing end interface

[Bug fortran/66575] New: Endless compilation on missing end interface

[gerhard.steinmetz.fortran@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66575

            Bug ID: 66575
           Summary: Endless compilation on missing end interface
           Product: gcc
           Version: 5.1.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gerhard.steinmetz.fortran@t-online.de
  Target Milestone: ---

If the first "end interface" is missing :

$ cat zlcti_1.f90
   module m
      interface operator (.f.)
         procedure f
      !end interface
      interface operator (.g.)
         procedure g
      end interface
   contains
      integer function f(x)
         integer, intent(in) :: x
         f = x
      end function
      integer function g(x)
         integer, intent(in) :: x
         g = x
      end function
   end


or wrong for whatever reason, e.g. :

$ cat zlcti_2.f90
   module m
      interface operator (.f.)
         procedure f
      end !interface
      interface operator (.g.)
         procedure g
      end interface
   contains
      integer function f(x)
         integer, intent(in) :: x
         f = x
      end function
      integer function g(x)
         integer, intent(in) :: x
         g = x
      end function
   end


then compilation will not end, even after minutes.

$ time gfortran -c zlcti_1.f90
# ^C after a few minutes

real    5m10.508s
user    0m0.000s
sys     0m0.001s

Tested with gfortran 5.1.1, 4.9.0, 4.8.3 (on SUSE Linux 13.2, 64 bit).
Tested with several combinations of options and several flags.

[Bug fortran/66575] Endless compilation on missing end interface

[gerhard.steinmetz.fortran@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66575

--- Comment #1 from Gerhard Steinmetz <gerhard.steinmetz.fortran@t-online.de> ---
The above sources can be reduced further. The part below "contains"
(inclusive contains) is not needed to produce the effect :


$ cat zlcti_4.f90
   module m
      interface operator (.f.)
         procedure f
      interface operator (.g.)
         procedure g
      end interface
   end


$ cat zlcti_6.f90
   module m
   interface operator (.f.)
      procedure f
   end
   interface operator (.g.)
      procedure g
   end

[Bug fortran/66575] Endless compilation on missing end interface

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66575

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |WAITING
   Last reconfirmed|                            |2015-06-27
     Ever confirmed|0                           |1

--- Comment #2 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
All the tests give an error in a fraction of a second on x86_64-apple-darwin14.

[Bug fortran/66575] Endless compilation on missing end interface

[gerhard.steinmetz.fortran@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66575

--- Comment #3 from Gerhard Steinmetz <gerhard.steinmetz.fortran@t-online.de> ---
Playing around, I've found other cases (wrong code with procedure)
causing endless compilation on my environment, e.g.


$ cat zlctp_1.f90
program p
   procedure(g) :: g
   procedure(g) :: g
end


$ cat zlctp_2.f90
program p
   procedure(g) :: g
   procedure(g) !! anything
end


$ time gfortran -c zlctp_2.f90      # with 5.1.1, 4.9.0, 4.8.3
# ^C after a few minutes

real    3m46.187s
user    0m0.000s
sys     0m0.002s

[Bug fortran/66575] Endless compilation on missing end interface

[dominiq at lps dot ens.fr]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66575

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW

--- Comment #4 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
I confirm the endless compilation for the tests in comment 3 with 4.8 up to
trunk (6.0).