[ 
https://issues.apache.org/jira/browse/FELIX-3160?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13148431#comment-13148431
 ] 

Rob Walker commented on FELIX-3160:
-----------------------------------

Just as I was about to give up - I think I got it!

Adding 1 more bundle to the startup set seemed to trigger it (http)

This is the bundle set I load:

===
felix.auto.install.1="${bundle.root}/lib/ext/osgi.compendium.jar"

org.osgi.framework.startlevel.beginning=3

felix.auto.start.1="${bundle.root}/lib/ext/felix.log.jar"

felix.auto.start.2="${bundle.root}/lib/ext/httpjetty.jar" \
                   "${bundle.root}/lib/vt/watchdog.jar"

felix.auto.start.3="${bundle.root}/lib/ext/shell.jar" \
                   "${bundle.root}/lib/ext/shelltui.jar"

# we'll start this transient so we have something to clean up
test.bundle=file:c:/mnt/data/vtmp/lib/shell.gui.jar
===

All the above bundles are standard from felix SVN, except my watchdog bundle. 
Will attach source for that.


Steps I take are:

1 - start cold with the above bundle set

2 -  from shell TUI start the shell.gui.jar transient (my activator code looks 
for this symbolic name on restart) e.g.

     -> install file:c:/mnt/data/vtmp/lib/shell.gui.jar
     Bundle ID: 7
     -> start -t 7

3 - kill the framework (I used Ctrl-C in fact, rather than clean shutdown)

4 - start the framework warm this time. NPE occurs after the auto uninstall of 
the shell GUI jar:

     C:\mnt\Data\runvt>vtmp
     Removing transient bundle: org.apache.felix.shell.gui [7]
     -> Framework error, bundle: org.apache.felix.shell.gui [7]
     java.lang.NullPointerException
             at 
org.apache.felix.framework.BundleRevisionImpl.close(BundleRevisionImpl.java:642)
             at 
org.apache.felix.framework.BundleImpl.closeRevisions(BundleImpl.java:154)
             at 
org.apache.felix.framework.BundleImpl.closeAndDelete(BundleImpl.java:138)
             at 
org.apache.felix.framework.Felix$RefreshHelper.refreshOrRemove(Felix.java:4640)
             at 
org.apache.felix.framework.Felix.refreshPackages(Felix.java:3952)
             at 
org.apache.felix.framework.FrameworkWiringImpl.run(FrameworkWiringImpl.java:172)
             at java.lang.Thread.run(Thread.java:662)



===

Remember to comment out the workaround checked in to SVN to ensure the NPE 
appears:

    synchronized void close()
    {
        try
        {
            resolve(null);
        }
        catch (Exception ex)
        {
            ((BundleImpl) m_bundle).getFramework().getLogger().log(
                Logger.LOG_ERROR, "Error releasing revision: " + 
ex.getMessage(), ex);
        }
        
        //if (m_content != null)
        //{
            m_content.close();
        //}
        
        m_content = null;
        for (int i = 0; (m_contentPath != null) && (i < m_contentPath.size()); 
i++)
        {
            m_contentPath.get(i).close();
        }
        m_contentPath = null;
    }


                
> NPE in BundleRevisionImpl.close() when uninstalling a bundle
> ------------------------------------------------------------
>
>                 Key: FELIX-3160
>                 URL: https://issues.apache.org/jira/browse/FELIX-3160
>             Project: Felix
>          Issue Type: Bug
>          Components: Framework
>    Affects Versions: framework-4.0.0
>            Reporter: Rob Walker
>            Assignee: Rob Walker
>            Priority: Minor
>             Fix For: framework-4.2.0
>
>         Attachments: Activator.java
>
>
> There seems to be a case where m_content can be null when uninstalling a 
> bundle, causing the following code to throw an NPE:
>     synchronized void close()
>     {
>         try
>         {
>             resolve(null);
>         }
>         catch (Exception ex)
>         {
>             ((BundleImpl) m_bundle).getFramework().getLogger().log(
>                 Logger.LOG_ERROR, "Error releasing revision: " + 
> ex.getMessage(), ex);
>         }
>         
>        m_content.close();
> I've added a simple check for null around the close in my version to avoid 
> the exception, which I'll check in.
> I'm not sure of the scenarios where this can legally be null at this point, 
> and whether there's some nastier underlying circumstance that needs 
> investigating. We see it under the following circumstances:
> * we register a listener that looks for framework STARTED
> * when triggered, this iterates all bundles and performs an uninstall on ones 
> which we determine are zombies leftover from a previous run
> Aside from the fact we're calling uninstall from inside an event handling 
> method, there doesn't anything else unusual about this code.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Reply via email to