* floatformat.c (floatformat_ibm_long_double_is_valid): Fix
	compiler warnings.
	(floatformat_ibm_long_double): Use
	floatformat_ibm_long_double_is_valid.

From-SVN: r130131