RFR JDK-8195703: BasicJDWPConnectionTest.java: 'App exited unexpectedly with 2'

Alex Menkov alexey.menkov at oracle.com
Thu Oct 11 17:00:06 UTC 2018


Hi Serguei,

I run the test itself and jdk/com/sun/jdi multiple times.

--alex

On 10/10/2018 17:17, serguei.spitsyn at oracle.com wrote:
> Hi Alex,
> 
> It looks good to me.
> How did you test it?
> 
> Thanks,
> Serguei
> 
> 
> On 10/10/18 16:25, Alex Menkov wrote:
>> Hi all,
>>
>> please review a fix for
>> https://bugs.openjdk.java.net/browse/JDK-8195703
>> webrev:
>> http://cr.openjdk.java.net/~amenkov/BasicJDWPConn/webrev.01/
>>
>> I was not able to reproduce the issue, but accordingly the logs in 
>> jira root cause is a transport initialization error "Address already 
>> in use".
>> The test uses Utils.getFreePort() to select some free port, but it can 
>> be race condition when some other app (or other test) uses the port 
>> selected before debuggee application starts to listen on it.
>> The fix uses dynamic port allocation and then parses it from the 
>> debuggee output.
>> Other changes:
>> - dropped catching exceptions and calling System.exit() - this causes 
>> SecurityException in JTReg harness which makes error analysis much 
>> harder;
>> - dropped using of Utils.getFreePort() from jdi/DoubleAgentTest.java 
>> test;
>>   jdi/BadHandshakeTest.java also uses Utils.getFreePort(), but it 
>> handles "Already in use" error re-peeking other free port and 
>> restarting debuggee, so I keep it as is.
>>
>> --alex
>>
> 


More information about the serviceability-dev mailing list