[Bug fortran/95053] [11 regression] ICE in f951: gfc_divide()

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

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95053

anlauf at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2020-05-11
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW

--- Comment #6 from anlauf at gcc dot gnu.org ---
It gets actually really weird during parsing.

The following (invalid) code shows that the parser is still in an
early phase where it has not yet decided that it is a FORMAT statement,
but rather could be something else:

  format('x') = x
end

This gives:

    1 |   format('x') = x
      |  1
Error: The function result on the lhs of the assignment at (1) must have the
pointer attribute.

while e.g. Intel detects:

foo.f90(1): error #6072: A dummy argument of a statement function must be a
scalar identifier.   ['x']
  format('x') = x
---------^

The simplest solution is to defer error detection and recovery by restoring
the previous behavior when the basic type of operand 2 to gfc_divide is
non-numeric:

diff --git a/gcc/fortran/arith.c b/gcc/fortran/arith.c
index 1cd0867a941..dd72f44d377 100644
--- a/gcc/fortran/arith.c
+++ b/gcc/fortran/arith.c
@@ -1828,7 +1828,8 @@ gfc_divide (gfc_expr *op1, gfc_expr *op2)
            rc = ARITH_DIV0;
          break;
        default:
-         gfc_internal_error ("gfc_divide(): Bad basic type");
+         /* basic type is non-numeric, handle this elsewhere.  */
+         break;
        }
       if (rc == ARITH_DIV0)
        {