[Bug fortran/107397] [10/11/12/13 Regression] ICE in gfc_arith_plus, at fortran/arith.cc:654

["kargl at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107397

kargl at gcc dot gnu.org changed:

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

--- Comment #3 from kargl at gcc dot gnu.org ---
Interesting bug.  The BOZ on the rhs is eventually rejected,
but if any error/warning that might be queued is lost during
a pass through  add_init_expr_to_sym().  Eventually, the rhs
expr is freed and the pointer to the initializer is wrong.

This patch fixes the ICE and issues an error.  It has passed
regression testing.

diff --git a/gcc/fortran/decl.cc b/gcc/fortran/decl.cc
index 0f9b2ced4c2..1562dc22bc6 100644
--- a/gcc/fortran/decl.cc
+++ b/gcc/fortran/decl.cc
@@ -2221,6 +2221,14 @@ add_init_expr_to_sym (const char *name, gfc_expr
**initp, locus *var_locus)
            sym->ts.f90_type = init->ts.f90_type;
        }

+      /* Catch the case:  type(t), parameter :: x = z'1'.  */
+      if (sym->ts.type == BT_DERIVED && init->ts.type == BT_BOZ)
+       {
+         gfc_error ("Entity %qs at %L is incompatible with a BOZ "
+                    "literal constant", name, &sym->declared_at);
+         return false;
+       }
+
       /* Add initializer.  Make sure we keep the ranks sane.  */
       if (sym->attr.dimension && init->rank == 0)
        {