[Bug fortran/55852] New: internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[Bug fortran/55852] New: internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[latlon90180+gcc_bugzilla at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

             Bug #: 55852
           Summary: internal compiler error: in gfc_build_intrinsic_call,
                    at fortran/expr.c:4647
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: latlon90180+gcc_bugzilla@gmail.com


Compilation of the following program causes internal compiler error:

!-----------------------------------
program bug
  implicit none

  Real, allocatable:: a(:)

  allocate(a(1:3))

  print*, ubound((a), 1) ! Bad
! print*, ubound(a, 1)   ! OK
! print*, lbound((a), 1) ! OK
! print*, lbound(a, 1)   ! OK

  stop
end program bug
!------------------------------------

$ gfortran-mp-4.8 -Wall -Wextra bug.F90
f951: internal compiler error: in gfc_build_intrinsic_call, at
fortran/expr.c:4647

f951: internal compiler error: Abort trap
gfortran-mp-4.8: internal compiler error: Abort trap (program f951)
Abort trap

$ gfortran-mp-4.8 -v
Using built-in specs.
COLLECT_GCC=gfortran-mp-4.8
COLLECT_LTO_WRAPPER=/opt/local/libexec/gcc/x86_64-apple-darwin10/4.8.0/lto-wrapper
Target: x86_64-apple-darwin10
Configured with: ../gcc-4.8-20121223/configure --prefix=/opt/local
--build=x86_64-apple-darwin10
--enable-languages=c,c++,objc,obj-c++,fortran,java
--libdir=/opt/local/lib/gcc48 --includedir=/opt/local/include/gcc48
--infodir=/opt/local/share/info --mandir=/opt/local/share/man
--datarootdir=/opt/local/share/gcc-4.8 --with-local-prefix=/opt/local
--with-system-zlib --disable-nls --program-suffix=-mp-4.8
--with-gxx-include-dir=/opt/local/include/gcc48/c++/ --with-gmp=/opt/local
--with-mpfr=/opt/local --with-mpc=/opt/local --with-ppl=/opt/local
--with-cloog=/opt/local --enable-cloog-backend=isl
--disable-cloog-version-check --enable-stage1-checking --disable-multilib
--enable-lto --enable-libstdcxx-time --with-as=/opt/local/bin/as
--with-ld=/opt/local/bin/ld --with-ar=/opt/local/bin/ar
--with-bugurl=https://trac.macports.org/newticket --with-pkgversion='MacPorts
gcc48 4.8-20121223_0'
Thread model: posix
gcc version 4.8.0 20121223 (experimental) (MacPorts gcc48 4.8-20121223_0)

[Bug fortran/55852] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

janus at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2013-01-03
                 CC|                            |janus at gcc dot gnu.org
     Ever Confirmed|0                           |1

--- Comment #1 from janus at gcc dot gnu.org 2013-01-03 09:39:32 UTC ---
Confirmed. Fails for me with all of 4.6, 4.7 and trunk, while it works with
4.3.

The backtrace on trunk is:

f951: internal compiler error: in gfc_build_intrinsic_call, at
fortran/expr.c:4646
0x57a1fd gfc_build_intrinsic_call(char const*, locus, unsigned int, ...)
        /home/jweil/gcc48/trunk/gcc/fortran/expr.c:4644
0x6054f4 gfc_simplify_size(gfc_expr*, gfc_expr*, gfc_expr*)
        /home/jweil/gcc48/trunk/gcc/fortran/simplify.c:5590
0x5ffe82 simplify_bound_dim
        /home/jweil/gcc48/trunk/gcc/fortran/simplify.c:3251
0x600580 simplify_bound
        /home/jweil/gcc48/trunk/gcc/fortran/simplify.c:3467
0x606f36 gfc_simplify_ubound(gfc_expr*, gfc_expr*, gfc_expr*)
        /home/jweil/gcc48/trunk/gcc/fortran/simplify.c:6266
0x595427 do_simplify
        /home/jweil/gcc48/trunk/gcc/fortran/intrinsic.c:3811
0x595d30 gfc_intrinsic_func_interface(gfc_expr*, int)
        /home/jweil/gcc48/trunk/gcc/fortran/intrinsic.c:4154
0x5d8cba resolve_unknown_f
        /home/jweil/gcc48/trunk/gcc/fortran/resolve.c:2610
0x5da08c resolve_function
        /home/jweil/gcc48/trunk/gcc/fortran/resolve.c:3211
0x5e167f gfc_resolve_expr(gfc_expr*)
        /home/jweil/gcc48/trunk/gcc/fortran/resolve.c:6546
0x5e91e5 resolve_code
        /home/jweil/gcc48/trunk/gcc/fortran/resolve.c:10091
0x5e7bcd gfc_resolve_blocks(gfc_code*, gfc_namespace*)
        /home/jweil/gcc48/trunk/gcc/fortran/resolve.c:9411
0x5e91a6 resolve_code
        /home/jweil/gcc48/trunk/gcc/fortran/resolve.c:10081
0x5f4b02 resolve_codes
        /home/jweil/gcc48/trunk/gcc/fortran/resolve.c:14908
0x5f4c0b gfc_resolve(gfc_namespace*)
        /home/jweil/gcc48/trunk/gcc/fortran/resolve.c:14936
0x5cc2ba resolve_all_program_units
        /home/jweil/gcc48/trunk/gcc/fortran/parse.c:4397
0x5ccada gfc_parse_file()
        /home/jweil/gcc48/trunk/gcc/fortran/parse.c:4664
0x619381 gfc_be_parse_file
        /home/jweil/gcc48/trunk/gcc/fortran/f95-lang.c:191

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

janus at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|internal compiler error: in |[4.6/4.7/4.8 regression]
                   |gfc_build_intrinsic_call,   |internal compiler error: in
                   |at fortran/expr.c:4647      |gfc_build_intrinsic_call,
                   |                            |at fortran/expr.c:4647

--- Comment #2 from janus at gcc dot gnu.org 2013-01-03 09:42:20 UTC ---
Works also with 4.5, which means it's a regression in 4.6.

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #3 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2013-01-03 11:01:18 UTC ---
Revision 162456 (2010-07-23) is OK, revision 1635293 (2010-08-24) gives a
segmentation fault.

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

Tobias Burnus <burnus at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |ice-on-valid-code
                 CC|                            |burnus at gcc dot gnu.org
   Target Milestone|---                         |4.6.4

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #4 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2013-01-03 11:13:19 UTC ---
gfc_build_intrinsic_call has been introduced at revision 162648.

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P4
                 CC|                            |jakub at gcc dot gnu.org

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[tkoenig at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

Thomas Koenig <tkoenig at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tkoenig at gcc dot gnu.org

--- Comment #5 from Thomas Koenig <tkoenig at gcc dot gnu.org> 2013-01-06 20:56:20 UTC ---
The problem is that gfc_get_sym_tree is asked to
get "size", but that is not available in the current
namespace.

The ICE goes away if "intrinsic:: size" is added.

We could simply remove the assert, but that would
be a bad idea in case the user specifies something
called "size" for something else.

So, we need to generate a different name, like in the
frontend optimization pass, where we use __internal_foo
for an intrinsic named foo.

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[tkoenig at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #6 from Thomas Koenig <tkoenig at gcc dot gnu.org> 2013-01-06 21:59:13 UTC ---
This patch works (not regression-tested yet), but the method
using the state variable seems hackish and error-prone.

What do you think?

Index: expr.c
===================================================================
--- expr.c      (Revision 194850)
+++ expr.c      (Arbeitskopie)
@@ -4623,7 +4623,8 @@
    want to add arguments but with a NULL-expression.  */

 gfc_expr*
-gfc_build_intrinsic_call (const char* name, locus where, unsigned numarg, ...)
+gfc_build_intrinsic_call (const char* name, const char *symtree_name,
+                         locus where, unsigned numarg, ...)
 {
   gfc_expr* result;
   gfc_actual_arglist* atail;
@@ -4641,11 +4642,17 @@
   result->value.function.name = name;
   result->value.function.isym = isym;

-  result->symtree = gfc_find_symtree (gfc_current_ns->sym_root, name);
-  gcc_assert (result->symtree
-             && (result->symtree->n.sym->attr.flavor == FL_PROCEDURE
-                 || result->symtree->n.sym->attr.flavor == FL_UNKNOWN));
+  if (symtree_name == NULL)
+    {
+      result->symtree = gfc_find_symtree (gfc_current_ns->sym_root, name);

+      gcc_assert (result->symtree &&
+                 (result->symtree->n.sym->attr.flavor == FL_PROCEDURE
+                  || result->symtree->n.sym->attr.flavor == FL_UNKNOWN));
+    }
+  else
+    gfc_get_sym_tree (symtree_name, gfc_current_ns, &result->symtree, true);
+
   va_start (ap, numarg);
   atail = NULL;
   for (i = 0; i < numarg; ++i)
Index: simplify.c
===================================================================
--- simplify.c  (Revision 194850)
+++ simplify.c  (Arbeitskopie)
@@ -33,6 +33,7 @@

 gfc_expr gfc_bad_expr;

+bool artificial_call = false;

 /* Note that 'simplification' is not just transforming expressions.
    For functions that are not simplified at compile time, range
@@ -3248,7 +3249,10 @@
          gfc_expr* dim = result;
          mpz_set_si (dim->value.integer, d);

+         artificial_call = true;
          result = gfc_simplify_size (array, dim, kind);
+         artificial_call = false;
+
          gfc_free_expr (dim);
          if (!result)
            goto returnNull;
@@ -5512,7 +5516,10 @@
        {
          mpz_set_ui (e->value.integer, n + 1);

+         artificial_call = true;
          f = gfc_simplify_size (source, e, NULL);
+         artificial_call = false;
+
          gfc_free_expr (e);
          if (f == NULL)
            {
@@ -5584,11 +5591,18 @@
       /* Otherwise, we build a new SIZE call.  This is hopefully at least
         simpler than the original one.  */
       if (!simplified)
-       simplified = gfc_build_intrinsic_call ("size", array->where, 3,
-                                              gfc_copy_expr (replacement),
-                                              gfc_copy_expr (dim),
-                                              gfc_copy_expr (kind));
+       {
+         const char *symtree_name;
+         if (artificial_call)
+           symtree_name = "__internal_size";
+         else
+           symtree_name = NULL;

+         simplified = gfc_build_intrinsic_call ("size", symtree_name,
array->where, 3,
+                                                gfc_copy_expr (replacement),
+                                                gfc_copy_expr (dim),
+                                                gfc_copy_expr (kind));
+       }
       return simplified;
     }

Index: gfortran.h
===================================================================
--- gfortran.h  (Revision 194850)
+++ gfortran.h  (Arbeitskopie)
@@ -2797,7 +2797,8 @@
 bool gfc_has_ultimate_allocatable (gfc_expr *);
 bool gfc_has_ultimate_pointer (gfc_expr *);

-gfc_expr* gfc_build_intrinsic_call (const char*, locus, unsigned, ...);
+gfc_expr* gfc_build_intrinsic_call (const char*, const char *, locus,
+                                   unsigned, ...);
 gfc_try gfc_check_vardef_context (gfc_expr*, bool, bool, bool, const char*);

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #7 from Tobias Burnus <burnus at gcc dot gnu.org> 2013-01-07 11:11:01 UTC ---
Author: burnus
Date: Mon Jan  7 11:10:53 2013
New Revision: 194966

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=194966
Log:
2013-01-07  Tobias Burnus  <burnus@net-b.de>
            Thomas Koenig  <tkoenig@gcc.gnu.org>

        PR fortran/55852
        * expr.c (gfc_build_intrinsic_call): Avoid clashes
        with user's procedures.
        * gfortran.h (gfc_build_intrinsic_call): Update prototype.
        * simplify.c (gfc_simplify_size): Update call.
        * class.c (finalization_scalarizer, finalization_get_offset,
        finalizer_insert_packed_call, generate_finalization_wrapper):
        Clean up by using gfc_build_intrinsic_call.

2013-01-07  Tobias Burnus  <burnus@net-b.de>

        PR fortran/55852
        * gfortran.dg/intrinsic_size_3.f90: New.


Added:
    trunk/gcc/testsuite/gfortran.dg/intrinsic_size_3.f90
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/class.c
    trunk/gcc/fortran/expr.c
    trunk/gcc/fortran/gfortran.h
    trunk/gcc/fortran/simplify.c
    trunk/gcc/testsuite/ChangeLog

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

Tobias Burnus <burnus at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |slayoo at staszic dot
                   |                            |waw.pl

--- Comment #8 from Tobias Burnus <burnus at gcc dot gnu.org> 2013-01-07 15:23:51 UTC ---
*** Bug 54940 has been marked as a duplicate of this bug. ***

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #9 from Tobias Burnus <burnus at gcc dot gnu.org> 2013-01-07 19:11:13 UTC ---
Author: burnus
Date: Mon Jan  7 19:11:04 2013
New Revision: 194995

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=194995
Log:
2013-01-07  Tobias Burnus  <burnus@net-b.de>
            Thomas Koenig  <tkoenig@gcc.gnu.org>
            Jakub Jelinek  <jakub@redhat.com>

        PR fortran/55852
        * expr.c (gfc_build_intrinsic_call): Avoid clashes
        with user's procedures.
        * gfortran.h (gfc_build_intrinsic_call): Update prototype.
        (GFC_PREFIX): Define.
        * simplify.c (gfc_simplify_size): Update call.

2013-01-07  Tobias Burnus  <burnus@net-b.de>

        PR fortran/55852
        * gfortran.dg/intrinsic_size_3.f90: New.


Added:
    branches/gcc-4_7-branch/gcc/testsuite/gfortran.dg/intrinsic_size_3.f90
Modified:
    branches/gcc-4_7-branch/gcc/fortran/ChangeLog
    branches/gcc-4_7-branch/gcc/fortran/expr.c
    branches/gcc-4_7-branch/gcc/fortran/gfortran.h
    branches/gcc-4_7-branch/gcc/fortran/simplify.c
    branches/gcc-4_7-branch/gcc/testsuite/ChangeLog

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[schwab@linux-m68k.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #10 from Andreas Schwab <schwab@linux-m68k.org> 2013-01-08 16:55:58 UTC ---
The test case fails because the match is too strict.

$ grep iszs intrinsic_size_3.f90.003t.original 
  integer(kind=2) iszs;
          iszs = (integer(kind=2)) MAX_EXPR <(D.854->dim[0].ubound -
D.854->dim[0].lbound) + 1, 0>;

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #11 from Tobias Burnus <burnus at gcc dot gnu.org> 2013-02-13 14:38:49 UTC ---
(In reply to comment #10)
> The test case fails because the match is too strict.

Should be fixed - for the the trunk only - since 2012-01-09 via
http://gcc.gnu.org/viewcvs?view=revision&revision=195064

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #12 from janus at gcc dot gnu.org 2013-02-13 14:44:13 UTC ---
(In reply to comment #11)
> Should be fixed - for the the trunk only - since 2012-01-09

Ok, so can we close this guy? Or is there need for a backport?

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #13 from Tobias Burnus <burnus at gcc dot gnu.org> 2013-02-13 15:09:06 UTC ---
Author: burnus
Date: Wed Feb 13 15:08:59 2013
New Revision: 196016

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=196016
Log:
2012-02-13   Tobias Burnus  <burnus@net-b.de>

        Backport from mainline
        2012-01-09  Uros Bizjak  <ubizjak@gmail.com>

        PR fortran/55852
        * gfortran.dg/intrinsic_size_3.f90: Make scan-tree-dump-times
        number matching more robust.


Modified:
    branches/gcc-4_7-branch/gcc/testsuite/ChangeLog
    branches/gcc-4_7-branch/gcc/testsuite/gfortran.dg/intrinsic_size_3.f90

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #14 from Tobias Burnus <burnus at gcc dot gnu.org> 2013-02-13 17:51:17 UTC ---
Author: burnus
Date: Wed Feb 13 17:51:11 2013
New Revision: 196020

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=196020
Log:
2013-02-13  Tobias Burnus  <burnus@net-b.de>

        Backported from mainline
        2013-01-07  Tobias Burnus  <burnus@net-b.de>
                    Thomas Koenig  <tkoenig@gcc.gnu.org>
                    Jakub Jelinek  <jakub@redhat.com>

        PR fortran/55852
        * expr.c (gfc_build_intrinsic_call): Avoid clashes
        with user's procedures.
        * gfortran.h (gfc_build_intrinsic_call): Update prototype.
        (GFC_PREFIX): Define.
        * simplify.c (gfc_simplify_size): Update call.

2013-02-13  Tobias Burnus  <burnus@net-b.de>

        Backported from mainline
        2013-01-07  Tobias Burnus  <burnus@net-b.de>
                    Uros Bizjak  <ubizjak@gmail.com>

        PR fortran/55852
        * gfortran.dg/intrinsic_size_3.f90: New.


Added:
    branches/gcc-4_6-branch/gcc/testsuite/gfortran.dg/intrinsic_size_3.f90
Modified:
    branches/gcc-4_6-branch/gcc/fortran/ChangeLog
    branches/gcc-4_6-branch/gcc/fortran/expr.c
    branches/gcc-4_6-branch/gcc/fortran/gfortran.h
    branches/gcc-4_6-branch/gcc/fortran/simplify.c
    branches/gcc-4_6-branch/gcc/testsuite/ChangeLog

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

Tobias Burnus <burnus at gcc dot gnu.org> changed:

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

--- Comment #15 from Tobias Burnus <burnus at gcc dot gnu.org> 2013-02-13 17:52:27 UTC ---
FIXED on the trunk (4.8) and on the 4.6/4.7 branches.

Thanks for the report - and sorry for the regression.

[Bug fortran/55852] [4.6/4.7/4.8 regression] internal compiler error: in gfc_build_intrinsic_call, at fortran/expr.c:4647

[latlon90180+gcc_bugzilla at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55852

--- Comment #16 from A. Kasahara <latlon90180+gcc_bugzilla at gmail dot com> 2013-02-13 19:21:33 UTC ---
Thank you for the fix!