[Bug fortran/39861] New: Internal Error with INTRINSIC in module

[Bug fortran/39861] New: Internal Error with INTRINSIC in module

[dominiq at lps dot ens dot fr]

At revision 146606 the following code

module vector_calculus
  intrinsic :: dot_product, sqrt

contains

  function len(r)
    real, dimension(:), intent(in) :: r
    real :: len
    len = sqrt(dot_product(r,r))
  end function len

end module vector_calculus
end

gives

pr36192_mod_red.f90:12.26:

end module vector_calculus
                          1
Internal Error at (1):
write_symbol(): bad module symbol 'x'

This code has been extracted from a larger one that did not give this error at
revision 146282.
I suspect revision 146554.


-- 
           Summary: Internal Error with INTRINSIC in module
           Product: gcc
           Version: 4.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: dominiq at lps dot ens dot fr


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39861

[Bug fortran/39861] [4.5 Regression] ICE with INTRINSIC in module: write_symbol(): bad module symbol

[burnus at gcc dot gnu dot org]

-- 

burnus at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
           Keywords|                            |ice-on-valid-code
   Last reconfirmed|0000-00-00 00:00:00         |2009-04-23 09:30:47
               date|                            |
            Summary|Internal Error with         |[4.5 Regression] ICE with
                   |INTRINSIC in module         |INTRINSIC in module:
                   |                            |write_symbol(): bad module
                   |                            |symbol
   Target Milestone|---                         |4.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39861

[Bug fortran/39861] [4.5 Regression] ICE with INTRINSIC in module: write_symbol(): bad module symbol

[pault at gcc dot gnu dot org]

------- Comment #1 from pault at gcc dot gnu dot org  2009-04-23 11:10 -------
(In reply to comment #0)

Hi Dominique,

Could I ask a Bear-of-Little-Brain question here?

> write_symbol(): bad module symbol 'x'

Where does the symbol 'x' come from?

Cheers

Paul


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39861

[Bug fortran/39861] [4.5 Regression] ICE with INTRINSIC in module: write_symbol(): bad module symbol

[dominiq at lps dot ens dot fr]

------- Comment #2 from dominiq at lps dot ens dot fr  2009-04-23 11:25 -------
> Where does the symbol 'x' come from?

Good question! The code generate a file vector_calculus.mod0 containing:

GFORTRAN module version '0' created from pr36192_mod_red.f90 on Thu Apr 23
13:17:45 2009
MD5:00000000000000000000000000000000 -- If you edit this, you'll get what you
deserve.

(() () () () () () () () () () () () () () () () () () () () () () () ()
() () ())

()

()

()

()

(2 'dot_product' '(intrinsic)' 'dot_product' 1 ((PROCEDURE
UNKNOWN-INTENT UNKNOWN-PROC DECL UNKNOWN INTRINSIC FUNCTION) (REAL 4 0 0
REAL ()) 3 0 (4 5) () 2 () () () 0 0)
6 'len' 'vector_calculus' 'len' 1 ((PROCEDURE UNKNOWN-INTENT MODULE-PROC
DECL UNKNOWN FUNCTION ALWAYS_EXPLICIT) (REAL 4 0 0 REAL ()) 7 0 (8) () 6
() () () 0 0)
9 'sqrt' '(intrinsic)' 'sqrt' 1 ((PROCEDURE UNKNOWN-INTENT UNKNOWN-PROC
DECL UNKNOWN INTRINSIC FUNCTION) (REAL 4 0 0 REAL ()) 10 0 (11) () 9 ()
() () 0 0)
12 'vector_calculus' 'vector_calculus' 'vector_calculus' 1 ((MODULE
UNKNOWN-INTENT UNKNOWN-PROC UNKNOWN UNKNOWN) (UNKNOWN 0 0 0 UNKNOWN ())
0 0 () () 0 () () () 0 0)
8 'r' '' 'r' 7 ((VARIABLE IN UNKNOWN-PROC UNKNOWN UNKNOWN DIMENSION
DUMMY) (REAL 4 0 0 REAL ()) 0 0 () (1 ASSUMED_SHAPE (CONSTANT (INTEGER 4
0 0 INTEGER ()) 0 '1') ()) 0 () () () 0 0)

the only 'x' being in "ALWAYS_EXPLICIT".


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39861

[Bug fortran/39861] [4.5 Regression] ICE with INTRINSIC in module: write_symbol(): bad module symbol

[janus at gcc dot gnu dot org]

------- Comment #3 from janus at gcc dot gnu dot org  2009-04-23 12:24 -------

> > write_symbol(): bad module symbol 'x'
> 
> Where does the symbol 'x' come from?

The 'x' here apparently is the formal argument of the sqrt() function!

I think Dominique was right in suspecting my r146554, which had the effect that
formal arguments from intrinsic functions are now not only kept in the isym,
but are also copied to the gfc_symbol of the intrinsic function (see
'resolve_intrinsic').

The problem seems to be that for those formal args copied from isym the flavor
is not set:

  if (sym->attr.flavor == FL_UNKNOWN || sym->attr.flavor == FL_LABEL)
    gfc_internal_error ("write_symbol(): bad module symbol '%s'", sym->name);

I will take care of this soon. Thanks for the report.


-- 

janus at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |janus at gcc dot gnu dot org
                   |dot org                     |
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2009-04-23 09:30:47         |2009-04-23 12:24:16
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39861

[Bug fortran/39861] [4.5 Regression] ICE with INTRINSIC in module: write_symbol(): bad module symbol

[janus at gcc dot gnu dot org]

------- Comment #4 from janus at gcc dot gnu dot org  2009-04-24 08:34 -------
Subject: Bug 39861

Author: janus
Date: Fri Apr 24 08:34:14 2009
New Revision: 146677

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=146677
Log:
2009-04-24  Janus Weil  <janus@gcc.gnu.org>

        PR fortran/39861
        PR fortran/39864
        * symbol.c (gfc_copy_formal_args_intr): Set attr.flavor and attr.dummy
        for the formal arguments.


2009-04-24  Janus Weil  <janus@gcc.gnu.org>

        PR fortran/39861
        PR fortran/39864
        * gfortran.dg/intrinsic_1.f90: New.


Added:
    trunk/gcc/testsuite/gfortran.dg/intrinsic_1.f90
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/symbol.c
    trunk/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39861

[Bug fortran/39861] [4.5 Regression] ICE with INTRINSIC in module: write_symbol(): bad module symbol

[janus at gcc dot gnu dot org]

------- Comment #5 from janus at gcc dot gnu dot org  2009-04-24 08:39 -------
Fixed by r146677. Closing.


-- 

janus at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39861