[Bug fortran/18737] New: ICE on invalid use of external keyword

[Bug fortran/18737] New: ICE on invalid use of external keyword

[coudert at clipper dot ens dot fr]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 736 bytes --]

$ cat a.f90 
program test
  implicit none
  real(8) :: x
  external bug

  x = 2
  print *, bug(x)
  
end program test


$ gfortran -c a.f90
a.f90: In function ‘MAIN__’:
a.f90:7: internal compiler error: in gfc_typenode_for_spec, at
fortran/trans-types.c:613

-- 
           Summary: ICE on invalid use of external keyword
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: coudert at clipper dot ens dot fr
                CC: gcc-bugs at gcc dot gnu dot org
  GCC host triplet: i686-unkown-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18737

[Bug fortran/18737] ICE on invalid use of external keyword

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-11-30 14:03 -------
Confirmed.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
           Keywords|                            |ice-on-invalid-code
   Last reconfirmed|0000-00-00 00:00:00         |2004-11-30 14:03:39
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18737

[Bug fortran/18737] ICE on invalid use of external keyword

[tobi at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
OtherBugsDependingO|                            |19292
              nThis|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18737

[Bug fortran/18737] ICE on invalid use of external keyword

[tobi at gcc dot gnu dot org]

------- Additional Comments From tobi at gcc dot gnu dot org  2005-03-14 20:34 -------
Actually, this slightly modified valid testcase fails with the same error:
program test
!  implicit none
  real(8) :: x
  external bug

  x = 2
  print *, bug(x)
  
end program test




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18737

[Bug fortran/18737] ICE on invalid use of external keyword

[tobi at gcc dot gnu dot org]

------- Additional Comments From tobi at gcc dot gnu dot org  2005-03-14 20:37 -------
Excuse, the testcase with "implicit none" commented out passes.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2005-03-01 14:09:30         |2005-03-14 20:37:30
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18737

[Bug fortran/18737] ICE on invalid use of external keyword

[tobi at gcc dot gnu dot org]

------- Additional Comments From tobi at gcc dot gnu dot org  2005-03-14 20:52 -------
The bug is caused by the call to gfc_set_default_type in resolve.c:3961 passing
0 as second argument, thereby suppressing the error that should be issued there.
 Unfortunately, passing 1 for the argument breaks code which calls intrinsic
functions.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2005-03-14 20:37:30         |2005-03-14 20:52:30
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18737

[Bug fortran/18737] ICE on invalid use of external keyword

[paulthomas2 at wanadoo dot fr]

------- Additional Comments From paulthomas2 at wanadoo dot fr  2005-08-03 11:42 -------
(In reply to comment #4)
> The bug is caused by the call to gfc_set_default_type in resolve.c:3961 
passing
> 0 as second argument, thereby suppressing the error that should be issued 
there.
>  Unfortunately, passing 1 for the argument breaks code which calls intrinsic
> functions.

This sounds familiar!  I can add this to the pr16940 fix; ie. before calling 
gfc_set_default_type, we look for an external declaration, as well as a generic 
interface.  I'll put it on the TODO list.

Paul T


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18737