Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.

Bug 343199

Summary: Threadlock when creating new JAXB project, changing facet version, using JRE library
Product: [WebTools] Dali JPA Tools Reporter: Paul Fullbright <paul.fullbright>
Component: JAXBAssignee: Paul Fullbright <paul.fullbright>
Status: RESOLVED FIXED QA Contact:
Severity: major    
Priority: P3 CC: david_williams, neil.hauge
Version: 3.0Flags: paul.fullbright: pmc_approved?
paul.fullbright: pmc_approved? (raghunathan.srinivasan)
paul.fullbright: pmc_approved? (naci.dai)
paul.fullbright: pmc_approved? (deboer)
paul.fullbright: pmc_approved? (neil.hauge)
paul.fullbright: pmc_approved? (kaloyan)
paul.fullbright: pmc_approved? (cbridgha)
Target Milestone: 3.0 M7   
Hardware: PC   
OS: Windows Vista   
Whiteboard: PMC
Bug Depends on: 343197    
Bug Blocks:    
Attachments:
Description Flags
stack trace
none
hackish patch none

Description Paul Fullbright CLA 2011-04-18 15:51:34 EDT
Created attachment 193524 [details]
stack trace

Create a new JAXB project, choose a Java 6 as the runtime, select facet version 2.1, and choose to use the JRE as the library provider on the JAXB facet page.  A threadlock occurs.  (see attachment)
Comment 1 Paul Fullbright CLA 2011-04-18 16:14:21 EDT
Created attachment 193528 [details]
hackish patch
Comment 2 Paul Fullbright CLA 2011-04-18 16:15:16 EDT
Problem is caused to a large extent by how the data model interacts with the faceted project working copy committing/reverting changes and how the data model wizard page interacts with the data model.

I have a hackish patch (see attachment) that addresses this, but don't think it's a long-term solution.
Comment 3 Paul Fullbright CLA 2011-05-04 17:12:22 EDT
    * Explain why you believe this is a stop-ship defect. Or, if it is a "hotbug" (requested by an adopter) please document it as such. 

     This is a major bug, resulting in workspace deadlock.


    * Is there a work-around? If so, why do you believe the work-around is insufficient? 

     There is no known work-around.


    * How has the fix been tested? Is there a test case attached to the bugzilla record? Has a JUnit Test been added? 

     This patch, while not ideal, has been tested by myself and by an additional developer.  No side effects have been discovered.  Long-term patch is out of scope for this release.


    * Give a brief technical overview. Who has reviewed this fix? 

     I have reviewed this fix.  The patch simply avoids validation by orphan facet configuration pages.  


    * What is the risk associated with this fix?

     Minimal risk.  The patch is very small and very targeted in scope.
Comment 4 David Williams CLA 2011-05-04 23:29:13 EDT
Ahhh ... nothing like a hackish patch for RC1! :) 

But, looks pretty safe and isolated as you say, so would be better than a deadlock. 

I'm not sure what ideas you've had about a long term fix, but be sure to open a bug (if not already) and reference it here so we will have long term cross-references to track back and forth. 

Much thanks.
Comment 5 Paul Fullbright CLA 2011-05-05 10:31:03 EDT
Well, actually we've determined the patch got into M7 somehow (re-targeting to reflect this).  Probably not the worst thing in the world.

But yes, I'll log another bug to make sure it gets taken care of long-term.
Comment 6 Paul Fullbright CLA 2011-05-05 10:35:14 EDT
bug 344846 has been opened to track a long-term fix for this issue.