[Bug c/16409] New: [3.5 regression] ICE in size_binop, at fold-const.c

["reichelt at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>]

Richard, your patch
  http://gcc.gnu.org/ml/gcc-cvs/2004-07/msg00205.html
causes a regression on mainline: Compiling

=====================================================
const int i = 1;
void foo() { char a[][i] = {""}; }
=====================================================

or 

=====================================================
void foo(int i) { char a[][i] = {""}; }
=====================================================

with the C frontend causes an ICE:

bug.c: In function `foo':
bug.c:2: internal compiler error: in size_binop, at fold-const.c:1598
Please submit a full bug report, [etc.]

Note that all released compiler versions crash on these examples,
but since the tree-ssa merge the code could be compiled on mainline.

Btw, we really should add a testcase for this, since it is not the first
time code like this crashes (see PR 10089, PR 12207).

-- 
           Summary: [3.5 regression] ICE in size_binop, at fold-const.c
           Product: gcc
           Version: 3.5.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code, monitored
          Severity: normal
          Priority: P2
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: reichelt at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org,kenner at gcc dot gnu
                    dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16409