RFR:7184195 - java.util.logging.Logger.getGlobal().info() doesn't log without configuration

[Jim Gish]

Here's an updated webrev for this fix: 
http://cr.openjdk.java.net/~jgish/Bug7184195-global-logger-failure.2/ 
<http://cr.openjdk.java.net/%7Ejgish/Bug7184195-global-logger-failure.2/>

The changes from the previous one are:
- breaking apart the two tests so that logger initialization in the 
basic getGlobal()/handler test doesn't cloud the deadlock testing issue.
- replacing the references in LogManager to global with getGlobal() 
(basically to get rid of deprecated usage).

Thanks,
     Jim

On 04/24/2013 05:23 PM, Jim Gish wrote:
> Please review 
> http://cr.openjdk.java.net/~jgish/Bug7184195-global-logger-failure.1/ 
> <http://cr.openjdk.java.net/%7Ejgish/Bug7184195-global-logger-failure.1/>
>
> This is a simple fix that removes a long-standing bug in acquiring a 
> using the global Logger in which Logger.getGlobal().info() (for 
> example), would not log without additional configuration that would 
> trigger LogManager initialization.  I have also added a test to 
> reassure us that I have not introduced a deadlock between LogManager 
> initialization and getting the global logger.
>
> In addition, I've introduced a utility class I developed during a 
> recent fix to a logging deadlock, which helps us detect and report on 
> the details of a deadlock.
>
> Thanks,
>    Jim
>

-- 
Jim Gish | Consulting Member of Technical Staff | +1.781.442.0304
Oracle Java Platform Group | Core Libraries Team
35 Network Drive
Burlington, MA 01803
jim.gish at oracle.com