[Bug fortran/32933] New: ICE in simplify_subreg with -fdefault-integer-8

[Bug fortran/32933] New: ICE in simplify_subreg with -fdefault-integer-8

[tkoenig at gcc dot gnu dot org]

Simplification of the ICE with nan_1.f90:

$ !cat
cat nan_6.f90 
program test
  real :: a
  if (min(a, 3.0, 2.0) /= 2.0) call abort
end program test
$ gfortran -fdefault-integer-8 nan_6.f90 
nan_6.f90: In function 'MAIN__':
nan_6.f90:3: internal compiler error: in simplify_subreg, at
simplify-rtx.c:4676
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

This goes away if -ffast-math is specified, so it looks as if
the call to __builtin_isnan is causing the trouble.


-- 
           Summary: ICE in simplify_subreg with -fdefault-integer-8
           Product: gcc
           Version: 4.3.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: tkoenig at gcc dot gnu dot org
OtherBugsDependingO 32770
             nThis:


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[fxcoudert at gcc dot gnu dot org]

------- Comment #1 from fxcoudert at gcc dot gnu dot org  2007-07-31 20:54 -------
(In reply to comment #0)
> $ gfortran -fdefault-integer-8 nan_6.f90 

But there's no integer *at all* in the testcase! How come?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[pinskia at gcc dot gnu dot org]

------- Comment #2 from pinskia at gcc dot gnu dot org  2007-07-31 20:58 -------
(In reply to comment #1)
> But there's no integer *at all* in the testcase! How come?

But there is, internally with __builtin_isnan.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[kargl at gcc dot gnu dot org]

------- Comment #3 from kargl at gcc dot gnu dot org  2007-07-31 21:20 -------
The testcase is illegal code because a is never assigned a value.
I also do not see where __builtin_isnan is supposely coming into
play here?


-- 

kargl at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kargl at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[tkoenig at gcc dot gnu dot org]

------- Comment #4 from tkoenig at gcc dot gnu dot org  2007-08-02 20:38 -------
Same place to ICE for bounds_check_5.f90.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[pinskia at gcc dot gnu dot org]

------- Comment #5 from pinskia at gcc dot gnu dot org  2007-08-06 17:35 -------
*** Bug 33006 has been marked as a duplicate of this bug. ***


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |michael dot a dot richmond
                   |                            |at nasa dot gov


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[fxcoudert at gcc dot gnu dot org]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1046 bytes --]



------- Comment #6 from fxcoudert at gcc dot gnu dot org  2007-08-08 15:24 -------
I reduced bounds_check_5.f90 to the following:

$ cat o2.f90
  INTEGER, POINTER :: x(:)
  logical :: l
  l = associated(x,x)
end
$ gfortran -fdefault-integer-8 -m32 o2.f90
o2.f90: In function ‘MAIN__’:
o2.f90:3: internal compiler error: in simplify_subreg, at simplify-rtx.c:4676

The tree dump is very simple:

  logical8 l;
  struct array1_int8 x;

  x.data = 0B;
  l = _gfortran_associated (&x, &x) && x.dim[0].stride != 0;

I guess we are generating mismatched types somewhere in the && operator.


-- 

fxcoudert at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2007-08-08 15:24:37
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[fxcoudert at gcc dot gnu dot org]

------- Comment #7 from fxcoudert at gcc dot gnu dot org  2007-08-08 17:26 -------
*** Bug 32943 has been marked as a duplicate of this bug. ***


-- 

fxcoudert at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dominiq at lps dot ens dot
                   |                            |fr


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[fxcoudert at gcc dot gnu dot org]

------- Comment #8 from fxcoudert at gcc dot gnu dot org  2007-08-08 17:27 -------
See PR32943 for other examples of testcases that exhibit the same ICE. When
this one is fixed, we should make sure all these have gone.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[fxcoudert at gcc dot gnu dot org]

------- Comment #9 from fxcoudert at gcc dot gnu dot org  2007-08-10 10:31 -------
(In reply to comment #2)
> But there is, internally with __builtin_isnan.

And that's where I got it wrong. In the Fortran front-end, boolean_type_node is
an integer of the default logical kind, while I thought it was coming from the
middle-end (as an int) and hence did not depend on Fortran kinds.

I'm testing patches for both this and bounds_check_5.f90, by adding conversions
at the right places.


-- 

fxcoudert at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |fxcoudert at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2007-08-08 15:24:37         |2007-08-10 10:31:34
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[fxcoudert at gcc dot gnu dot org]

------- Comment #10 from fxcoudert at gcc dot gnu dot org  2007-08-10 13:21 -------
Subject: Bug 32933

Author: fxcoudert
Date: Fri Aug 10 13:20:46 2007
New Revision: 127334

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=127334
Log:
        PR fortran/32933

        * trans-decl.c (gfc_build_builtin_function_decls): Change
        prototype for associated.
        * trans-intrinsic.c (gfc_conv_intrinsic_minmax): Convert the
        result of __builtin_isnan into a boolean.
        (gfc_conv_intrinsic_strcmp): Cleanup.
        (gfc_conv_associated): Convert the result of the associated
        function into a boolean.

        * intrinsics/associated.c: Change return type of associated into
        a C int.

Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/trans-decl.c
    trunk/gcc/fortran/trans-intrinsic.c
    trunk/libgfortran/ChangeLog
    trunk/libgfortran/intrinsics/associated.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933

[Bug fortran/32933] ICE in simplify_subreg with -fdefault-integer-8

[fxcoudert at gcc dot gnu dot org]

------- Comment #11 from fxcoudert at gcc dot gnu dot org  2007-08-10 13:25 -------
Fixed.


-- 

fxcoudert at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32933