[Bug fortran/66041] New: [6.0 Regression] Matmul ICE

[Bug fortran/66041] New: [6.0 Regression] Matmul ICE

[Joost.VandeVondele at mat dot ethz.ch]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66041

            Bug ID: 66041
           Summary: [6.0 Regression] Matmul ICE
           Product: gcc
           Version: 6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: Joost.VandeVondele at mat dot ethz.ch
  Target Milestone: ---

> cat test.f90
REAL*8, ALLOCATABLE :: a(:,:), b(:,:), v(:)

ALLOCATE(a(N,N),b(N,N),v(N))

DO i=1,N
v = MATMUL(a,b(:,i))
ENDDO

END

> gfortran -O2 test.f90
test.f90:6:0:

 v = MATMUL(a,b(:,i))
 1
internal compiler error: in gfc_walk_array_ref, at fortran/trans-array.c:8900
0x69057b gfc_walk_array_ref(gfc_ss*, gfc_expr*, gfc_ref*)
        ../../gcc/gcc/fortran/trans-array.c:8900
0x690b90 gfc_walk_expr(gfc_expr*)
        ../../gcc/gcc/fortran/trans-array.c:9264
0x69287c gfc_conv_expr_descriptor(gfc_se*, gfc_expr*)
        ../../gcc/gcc/fortran/trans-array.c:6568
0x6cfb03 gfc_conv_intrinsic_bound
        ../../gcc/gcc/fortran/trans-intrinsic.c:1860
0x6da3de gfc_conv_intrinsic_function(gfc_se*, gfc_expr*)
        ../../gcc/gcc/fortran/trans-intrinsic.c:8048
0x6b767e gfc_conv_expr_op
        ../../gcc/gcc/fortran/trans-expr.c:3264
0x6b5e33 gfc_conv_expr_val(gfc_se*, gfc_expr*)
        ../../gcc/gcc/fortran/trans-expr.c:7414
0x6b5ed8 gfc_conv_expr_type(gfc_se*, gfc_expr*, tree_node*)
        ../../gcc/gcc/fortran/trans-expr.c:7428
0x686ddd gfc_conv_array_ref(gfc_se*, gfc_array_ref*, gfc_expr*, locus*)
        ../../gcc/gcc/fortran/trans-array.c:3306
0x6b651c gfc_conv_variable
        ../../gcc/gcc/fortran/trans-expr.c:2575
0x6b767e gfc_conv_expr_op
        ../../gcc/gcc/fortran/trans-expr.c:3264
0x6b767e gfc_conv_expr_op
        ../../gcc/gcc/fortran/trans-expr.c:3264
0x6b4777 gfc_trans_assignment_1
        ../../gcc/gcc/fortran/trans-expr.c:9097
0x67ddc5 trans_code
        ../../gcc/gcc/fortran/trans.c:1682
0x6fa4a4 gfc_trans_simple_do
        ../../gcc/gcc/fortran/trans-stmt.c:1680
0x6fa4a4 gfc_trans_do(gfc_code*, tree_node*)
        ../../gcc/gcc/fortran/trans-stmt.c:1843
0x67dbda trans_code
        ../../gcc/gcc/fortran/trans.c:1791
0x6fa4a4 gfc_trans_simple_do
        ../../gcc/gcc/fortran/trans-stmt.c:1680
0x6fa4a4 gfc_trans_do(gfc_code*, tree_node*)
        ../../gcc/gcc/fortran/trans-stmt.c:1843
0x67dbda trans_code
        ../../gcc/gcc/fortran/trans.c:1791

[Bug fortran/66041] [6 Regression] Matmul ICE

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66041

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |6.0
            Summary|[6.0 Regression] Matmul ICE |[6 Regression] Matmul ICE

[Bug fortran/66041] [6 Regression] Matmul ICE

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66041

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

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

--- Comment #1 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
Mine.

[Bug fortran/66041] [6 Regression] Matmul ICE

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66041

--- Comment #3 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
Created attachment 35491
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=35491&action=edit
Patch

When calling lbound for the lower bound of an array section, a reference to the
full array was built.  This still contained a reference to ar->start, which hit
an assert.

This rather obvious patch corrects this.

[Bug fortran/66041] [6 Regression] Matmul ICE

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66041

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #35491|0                           |1
        is obsolete|                            |

--- Comment #4 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
Created attachment 35504
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=35504&action=edit
Better patch

[Bug fortran/66041] [6 Regression] Matmul ICE

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66041

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

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

--- Comment #6 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
Fixed, closing.

[Bug fortran/66041] [6 Regression] Matmul ICE

[tkoenig at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66041

--- Comment #7 from Thomas Koenig <tkoenig at gcc dot gnu.org> ---
Author: tkoenig
Date: Tue May 12 06:37:43 2015
New Revision: 223031

URL: https://gcc.gnu.org/viewcvs?rev=223031&root=gcc&view=rev
Log:
2015-05-12  Thomas Koenig  <tkoenig@gcc.gnu.org>

        PR fortran/66041
        PR fortran/37131
        * gfortran.h (gfc_array_spec):  Add field resolved.
        * array.c (gfc_resolve_array_spec):  Resolve array spec
        only once.


Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/array.c
    trunk/gcc/fortran/gfortran.h