There is a test for this bug under
http://hg.openjdk.java.net/jdk9/dev/jdk/rev/0f6be80ba02c.

If you modify the supplied TestEnum.java file and add a few more enum values
(entered below) then
the resulting schema1.xsd has wrong content as the sequence of the enum
values has been changed illegally.


-----------------------------
TestEnumType.java:

import javax.xml.bind.annotation.XmlEnum;



@XmlEnum(String.class)

public enum TestEnumType {

     FIRST, ONE, TWO, THREE, FOUR, FIVE, SIX, LAST


---------------------------------

run schemagen on the java file


Resulting schema1.xsd:

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<xs:schema version="1.0" xmlns:xs="http://www.w3.org/2001/XMLSchema">

  <xs:simpleType name="testEnumType">
    <xs:restriction base="xs:string">
      <xs:enumeration value="FIVE"/>
      <xs:enumeration value="FIRST"/>
      <xs:enumeration value="LAST"/>
      <xs:enumeration value="FOUR"/>
      <xs:enumeration value="SIX"/>
      <xs:enumeration value="TWO"/>
      <xs:enumeration value="THREE"/>
      <xs:enumeration value="ONE"/>
    </xs:restriction>
  </xs:simpleType>
</xs:schema>


The contents of the schema1.xsd file is as expected if running the test with
JDK 7.