Bug 901174 (JBPAPP6-1158)
| Summary: | BasicIIOPInvocationTestCase fails on ibm jdk due to failure in server startup for failed parsing | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [JBoss] JBoss Enterprise Application Platform 6 | Reporter: | Madhumita Sadhukhan <msadhukh> | ||||
| Component: | IIOP | Assignee: | baranowb <bbaranow> | ||||
| Status: | CLOSED WONTFIX | QA Contact: | |||||
| Severity: | high | Docs Contact: | |||||
| Priority: | high | ||||||
| Version: | 6.0.0 | CC: | bbaranow, msadhukh, pslavice | ||||
| Target Milestone: | --- | ||||||
| Target Release: | EAP 6.0.1 | ||||||
| Hardware: | Unspecified | ||||||
| OS: | Unspecified | ||||||
| URL: | http://jira.jboss.org/jira/browse/JBPAPP6-1158 | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | Bug Fix | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2012-11-29 15:39:35 UTC | Type: | Bug | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Attachments: |
|
||||||
|
Description
Madhumita Sadhukhan
2012-11-07 10:19:20 UTC
I have some problems with JDK switch and eclipse exploding with OOM but...
The XML file has something like
{noformat}
<subsystem xmlns="urn:jboss:domain:jacorb:1.2">
<orb>
<initializers xmlns="" xmlns:t="urn:jboss:domain:transactions:1.2" xmlns:j="urn:jboss:domain:jacorb:1.2" transactions="on" security="on"/>
</orb>
</subsystem>
{noformat}
Trick is that initializers sets the namespace(default) to "". If the source I got for woodstock(wstx?) is correct, this is properly picked up on IBM JDK, while it is ignored on oracle.
So it seems that running this test on IBM JDK produces valid parsing failure.
If you dig a bit further, the jarcorb.xml which is a template for jacorb conf for standalone-full.xml looks as follows:
{noformat}
<config>
<extension-module>org.jboss.as.jacorb</extension-module>
<subsystem xmlns="urn:jboss:domain:jacorb:1.2">
<orb socket-binding="jacorb" ssl-socket-binding="jacorb-ssl">
<initializers transactions="spec" security="on"/>
</orb>
</subsystem>
<socket-binding name="jacorb" interface="unsecure" port="3528"/>
<socket-binding name="jacorb-ssl" interface="unsecure" port="3529"/>
</config>
{noformat}
No empty namespace, at first glance, looks like "generate-server-config" ANT task puts bad NS. Or fiel is created from different src( tooo many files and scripting is a killer)
Further investigation: ts.config-as.add-jts <- this target is a culprit.
{noformat}
<xsl:template match="//j:subsystem/j:orb">
<xsl:copy>
<initializers transactions="on" security="on"/>
</xsl:copy>
</xsl:template>
{noformat}
Oracle JDK:
{noformat}
<subsystem xmlns="urn:jboss:domain:jacorb:1.2">
<orb>
<initializers xmlns:j="urn:jboss:domain:jacorb:1.2" xmlns:t="urn:jboss:domain:transactions:1.2" transactions="on" security="on"/>
</orb>
</subsystem>
{noformat}
IBM JDK
{noformat}
<subsystem xmlns="urn:jboss:domain:jacorb:1.2">
<orb>
<initializers xmlns="" xmlns:t="urn:jboss:domain:transactions:1.2" xmlns:j="urn:jboss:domain:jacorb:1.2" transactions="on" security="on"/>
</orb>
</subsystem>
{noformat}
Without transformation:
{noformat}
<subsystem xmlns="urn:jboss:domain:jacorb:1.2">
<orb socket-binding="jacorb" ssl-socket-binding="jacorb-ssl">
<initializers transactions="spec" security="on"/>
</orb>
</subsystem>
{noformat}
This possibly should be filed as bug in IBM JDK. The "xslt" task use the javax.xml.transform.TransformerFactory. If the "javax.xml.transform.TransformerFactory" is not set to transformer factory, it defaults to "com.sun.org.apache.xalan.internal.xsltc.trax.TransformerFactoryImpl". IBM JDK hijacks this. It has dumb implementation of this class which extends some IBM factory class. This is what causes wrong translation. Hack to make test pass. Attachment: Added: JBPAPP-10361-hack.patch Attached hack fix. But since this is a JVM issue, this is a "WontFix" (as has been agreed with Madhumita) Docs QE Status: Removed: NEW Failure is due to issue in IBM JDK hence closing |