Fixed type_float failures in --ps-protocol mode introduced by the test case for bug #27483.

The reason for the failures was bug #21205 (fixed in 6.0 by dtoa, but still present in 5.0/5.1).

mysql-test/r/type_float.result:
  Modified the failing test case so that bug #21205 does not occur.
mysql-test/t/type_float.test:
  Modified the failing test case so that bug #21205 does not occur.

mysql-test/r/type_float.result

@@ -398,11 +398,11 @@ insert into t1(d) values (9.2233720368547777e+18),
 (9.22337203685479e18),
 (1.84e19);
 update t1 set u = d;
-select * from t1;
-d	u
-9.22337203685478e+18	9223372036854775808
-9.22337203685478e+18	9223372036854779904
-9.22337203685479e+18	9223372036854790144
-1.84e+19	18400000000000000000
+select u from t1;
+u
+9223372036854775808
+9223372036854779904
+9223372036854790144
+18400000000000000000
 drop table t1;
 End of 5.0 tests

mysql-test/t/type_float.test

@@ -265,7 +265,7 @@ insert into t1(d) values (9.2233720368547777e+18),
                          (1.84e19);
 
 update t1 set u = d;
-select * from t1;
+select u from t1;
 
 drop table t1;